untitled

Driving crime down

Denying criminals the use of the road

October 2004

Police Standards Unit

PA: delivering value

PA Consulting Group is a leading management, systems and technology

consulting firm. Operating worldwide in more than 35 countries, PA draws

on the knowledge and experience of 3,000 people, whose skills span the

initial generation of ideas, insights and solutions all the way through to

detailed implementation.

PA focuses on creating benefits for clients rather than merely proposing them.

Our work is founded on powerful insights into our clients’ issues, and in the

private sector in particular, on the need to deliver superior shareholder returns.

We help accelerate business growth by developing innovative products for

our clients and by the application of emerging technology. We deliver major

transformation programmes, mobilize human resources, and manage complex

IT and technically-challenging programmes.

PA’s results-focused approach is founded on a unique commitment to

excellence, value and independence:

Excellence. We are committed to unremitting excellence and quality in

every aspect of our work: in our relationships with our clients, in the client

assignments we deliver, and in the people we recruit and develop, who enjoy

exciting and rewarding careers at PA.

Value. PA’s consultants bring an intense focus on delivering value through

deep industry insight, the development and application of technology, and

our culture of respect, collaboration and flexibility in working with clients.

Independence. PA is independent of outsourcing, software, hardware

and audit providers. We are the pre-eminent client-side adviser on IT and

outsourcing projects, and deliver, in partnership with our clients, business

solutions tailored to our clients’ needs, rather than solutions pre-determined

by commercial alliances. As an employee-owned company, we are answerable

to our clients and to ourselves only.

“I am pleased to be able to present this

evaluation of the pilot of automatic number

plate recognition intercept teams. I am very

grateful to the police forces that have taken part,

and thank the officers involved for their high levels

of commitment and effort that have been shown to

have delivered such impressive results.

The outcomes from the pilot are impressive:

• Between June 2003 and June 2004, ANPR

teams across 23 forces produced nine/ten times

the national average arrest rate per officer,

totalling more than 13,000 arrests

• Over the same period, officers recovered

property and drugs worth well in excess of

£8 million.

• The average ANPR-intercept officer is

responsible for 33 offences brought to justice

each year – three times the rate for other

forms of policing.

• The Home Office estimates that national roll-out

of ANPR would lead to approximately 26,400

additional offences being brought to justice each

year – a significant contribution of around 15%

towards meeting the Governement’s target for

offences brought to justice.

Although it is only one policing tool, ANPR has uses

in a range of areas, including tackling volume crime,

serious & organised crime, counterterrorism, and

in intelligence gathering. It has also proven a great

asset in tackling the ‘underclass’ of vehicles that

are incorrectly registered, untaxed and uninsured.

In recognition of this, ANPR is integral to delivering

the Home Office’s policy objectives as set out in

Confident Communities in a Secure Britain, the

Home Office strategic plan for the next five years.

It is also a crucial element of the joint Home Office,

Department for Constitutional Affairs and Crown

Prosecution Service strategy for reforming the

criminal justice system: Cutting Crime, Delivering

Justice. The experience gained in the pilot,

highlighted by the evaluation work, is likely to

lead to the introduction of ANPR enabling

legislation as soon as Parliamentary time allows.

The recent Greenaway Report on uninsured

driving also included recommendations to

maximise the effectiveness of ANPR. DfT are

currently planning measures to implement these.

The achievements and good practice established

during the pilot provide an outstanding foundation

for rolling out the concept of ANPR nationally.

This, together with further development suggested

by the evaluation, and stronger partnership

working, brings us closer to our ultimate aim

of denying criminals use of the roads.”

Home Secretary’s

Introduction

Rt Hon David Blunkett MP

Home Secretary

October 2004

This study, commissioned by the Home Office Police Standards Unit (PSU),

would not have been possible without the co-operation of the police forces

involved. We would like to thank the Chief Constables, intercept teams and

support staff of the 23 police forces that took part in this project, namely:

Avon & Somerset Constabulary Cambridgeshire Constabulary

Cheshire Constabulary City of London Police

Cleveland Police Greater Manchester Police

Hampshire Constabulary Hertfordshire Constabulary

Kent Constabulary Lancashire Constabulary

Leicestershire Constabulary Lincolnshire Police

Merseyside Police Metropolitan Police Service

North Wales Police North Yorkshire Police

Northamptonshire Police Northumbria Police

Nottinghamshire Constabulary Staffordshire Police

Warwickshire Police West Midlands Police

West Yorkshire Police

Specific thanks goes to Chief Constable Richard Brunstrom (Head of

Road Policing, ACPO), and Frank Whiteley (Chair, ANPR Steering Group),

Superindendents Alan Ford and Terry Kellaher (formerly of the Home Office

Police Standards Unit) and Robert Ritchie (Home Office, Justice Gap Taskforce).

The views expressed in this report are those of the authors, not

necessarily those of the Home Office Police Standards Unit or the

Association of Chief Police Officers. The consultants who worked on

this project were Charlie Henderson, Panikos Papagapiou, Adrian Gains

and Jim Knox. Any queries in relation to this report should be directed

to [email protected]

Acknowledgements

ANPR is not a new technology, but it was only

recently that the full potential to tackle criminality

was beginning to be realised

In 2002, a number of police forces increased their use of Automatic Number

Plate Recognition (ANPR) systems to include dedicated intercept officers.

These officers were able to intercept and stop vehicles of interest identified by

the ANPR systems and question the driver and/or passengers as appropriate.

The intention was that targeted enforcement would detect, disrupt and deter

criminality. A six-month evaluation of the use of these dedicated intercept

officers (‘Laser 1’) showed the concept to be extremely effective, achieving

arrest rates many times that of conventional policing.

Although these results were encouraging, there was no funding set aside

for the national testing, roll-out and operation of ANPR-enabled intercept

teams. An innovative funding mechanism was, therefore, required. Following

a submission to HM Treasury, conditional approval was given to the Home

Office to test a cost recovery system for dedicated ANPR-enabled intercept

teams. This would allow police to target vehicle documentation offences and

crime in general with the activity part-funded through receipts from the fixed

penalties issued by these teams.

Since 1 June 2003, 23 forces across England and Wales have operated

dedicated intercept officers part-funded under cost recovery (‘Laser 2’).

Executive summary

This report presents the findings of the evaluation of Laser 2 for the period 1

June 2003 to 31 June 2004.

The use of ANPR intercept teams

is aligned with Government policy

The use of ANPR-enabled intercept teams to engage criminality on the road

is clearly aligned with a number of key objectives for the Police Service,

including the National Policing Plan, Strategic Plan for Criminal Justice

2004-08, the Police Service’s National Intelligence Model and the Association

of Chief Police Officers (ACPO) Road Policing strategy. The use of ANPR-

enabled intercept teams also contributes to wider objectives, specifically road

safety (eg enforcing the offences of not wearing seat belts and illegal use of

mobile telephones while driving) and excise collection (eg ensuring that all

vehicles on the road are appropriately taxed).

This also addresses the public’s desire to see more ‘officers on the street’

and more action taken against illegal drivers. Given the link between vehicle

documentation offences (which can be relatively easily identified from national

databases) and wider criminality, it can be shown that the targeting of these

offences through the use of ANPR-enabled intercept teams can make a

significant contribution to wider policy objectives.

During Laser 2, ANPR has been evaluated

in 23 forces for over a year

During the course of Laser 2 project, the total staff input across 23 forces was

368,446 hours – this equates to 192 Full Time Equivalents (FTEs), the majority

of whom were police constables. By the start of the second year of Laser 2

there were approximately 515 officers involved in ANPR related operations.

The majority of ANPR officers’ time (77%) was spent either on intercept duties

or travelling to and from intercept duties. This level of visibility is significantly

higher than a ‘typical’ police officer – a Home Office report identified that on

average a typical police officer spent only 57% of their time away from their

police station. Further, ANPR intercept officers, whether they are travelling

to and from intercept sites or undertaking intercepts can also respond to

incidents as necessary when they occur.

A key aspect to the successful exploitation of ANPR intercept teams was

senior officer commitment to the programme – this ensured that resources

were available as and when required and other officers across the force

provided appropriate intelligence for the ANPR teams to operate on.

The results from Laser 2 have been impressive

The ANPR intercept teams stopped a total of 180,543 vehicles.

From these stops, the intercept officers:

• arrested 13,499 persons, including:

– 2,263 arrests for theft and burglary

– 3,324 arrests for driving offences (for example driving whilst disqualified)

– 1,107 arrests for drugs offences

– 1,386 arrests for auto crime (theft from and of vehicles)

• recovered or seized property, including:

– 1,152 stolen vehicles (valued at over £7.5 million)

– 266 offensive weapons and 13 firearms

– drugs worth over £380,000 from 740 vehicles

– stolen goods worth over £640,000 from 430 vehicles

• issued fixed penalty notices, including:

– 22,825 tickets for failing to display Vehicle Excise Duty (VED)

– 6,299 for no insurance

– 1,496 for no MOT

– 20,290 for a variety of offences, including not wearing a seat belt,

using a mobile telephone whilst driving.

The evaluation also confirmed previous research that had found a correlation

between vehicle documentation offences and volume crime – 3,530 of all

arrests (26%) originated from vehicle stops from No VED or current keeper.

Tracking a sample of these arrests through the criminal justice system, it was

found that an average ANPR full time equivalent will contribute around 31

offences per annum towards to the Government’s Offences Brought to Justice

(OBTJ) target – this is over three times the rate for conventional policing. If an

ANPR intercept team was deployed by each Basic Command Unit this would

contribute 26,400 additional OBTJs per annum towards the target – around

15% of the Government’s target. Since Laser involves redeploying existing

resources more effectively, this represents little incremental costs and hence

good value for money.

The expansion of Laser 1 to Laser 2 has shown that the results achieved

within a small-scale pilot can be achieved across a much wider cross-section

of forces and that these results can be sustained over time.

ANPR has helped pay for itself

Overall the cost recovery process released an additional £1 million in total to

the 23 Laser 2 forces over a nine-month period (to the end of the first financial

year of the pilot). The controls and processes have worked well – while forces

were required to collect additional information and were able to issue new

fixed penalties, there was no evidence to suggest that operational priorities

were distorted – forces achieved comparable arrest rates to Laser 1 where

no cost recovery operated.

Given the focus on recovering monies from Fixed Penalty Notices (FPNs),

the Laser 2 evaluation highlighted the low payment levels associated with

some fines. In particular, the introduction of a £200 fine and 6 penalty points

for no insurance was intended to reduce the burden on courts. However, with

just a 14% payment rate, this has not proved to be the case.

Conclusions

In terms of operation, the use of ANPR intercept teams represents an

innovative approach:

• targeting vehicle documentation enforcement to engage with and

disrupt criminals

• delivered through an intelligence-led piece of technology (an ANPR reader)

• benefiting from officers’ experience (eg observations of vehicle drivers)

• supported by existing policing processes (eg prisoner handling).

On this basis we can conclude that ANPR-enabled intercept teams have

been shown to be an extremely effective means of engaging with criminals.

Laser 2 has built upon the significant success of Laser 1 by proving the

concept across a wider range of forces, over a longer time period and with a

greater level of resource. Using a range of police intelligence and experience,

Laser 2 intercept teams were able to disrupt criminal activity in an efficient and

effective manner, bringing more than three times the number of offences to

justice compared to conventional policing.

While the cost recovery element realised less than 10% of the expenditure

incurred, these monies were important, for example, in helping to improve the

intelligence capability of the ANPR teams and providing part of the administrative

support required. On this basis, we conclude that the cost recovery aspect

contributed to the overall success of Laser 2. The pilot identified a number of

areas where operations could be improved (in particular by having more

accurate data). Once these areas have been addressed, it is expected that

ANPR will be an even more effective policing tool than was shown.

Recommendations

The evaluation highlighted a number of recommendations, including:

• Roll-out of Project Laser – Project Laser has proved that ANPR intercept

teams, if used appropriately, can be an extremely effective police tool in

engaging and dealing with criminality in all its forms. There is a strong case

that Laser is rolled out nationally and this roll-out proceeds as rapidly as

possible to ensure that the benefits to police and society are achieved.

Cost recovery can then be used as a means of supplementing local force

expenditure, in particular in the improvement of intelligence and its handling.

• A review of data used for ANPR – the accuracy of the DVLA database in

particular needs to be investigated. There are also substantial variations in the

quality and accuracy of local intelligence databases that require investigation.

There should be more effective use of intelligence at a national and local

level. Further, the pro-active use of MOT and no insurance databases, planned

in the near future, are an important development and should increase the

productivity of the ANPR intercept teams. These should be fully evaluated

in terms of their strengths and weaknesses.

• A national vehicle intelligence data warehouse – other than for the services

provided by PNC, police forces have to operate with a series of local databases

that are copied and shared between forces. This is a time consuming and

ineffective way of operating and is a further example of the lack of joined up

intelligence management highlighted by the Bichard enquiry report. There is

a need for a national data warehouse to hold all vehicle intelligence to be read

in real time by all ANPR users nationally. In turn, this data warehouse would

also hold ANPR reads and hits as a further source of vehicle intelligence,

providing great benefits to major crime and terrorism enquiries. A means to

fund provision of this data warehouse should be urgently sought by Government.

• Deployment management – currently, most ANPR teams are tasked

and deployed from a central location. This can mean, in some areas,

that considerable time is spent travelling to and from ANPR intercept sites.

Clearly, this is not best use of police time and we suggest that consideration

is given to co-locating ANPR intercept teams with BCUs and roads policing

units, as appropriate. Support systems will need to be put in place to ensure

best practice and intelligence is shared and performance monitored as a whole.

• A review of level of fines and payment rates – there is an apparent

disconnect between the levels of fixed penalties for the more serious offences

and the penalties that are awarded if the case is taken to court – anecdotal

evidence suggests that in some cases penalties are less severe in court,

both in monetary value and the number of points awarded. This could potentially

damage the effectiveness of the fixed penalty scheme and needs to be

urgently reviewed by ACPO and the Department of Constitutional Affairs (DCA).

• Development of a national ANPR strategy – we recommend that the

Home Office and Department for Transport, working with other Government

departments and key stakeholders, develop a detailed strategy and

implementation plan for ANPR for the next few years.

Table of contents

1. Introduction 12

1.1 Context 13

1.2 Evaluation methodology 15

1.3 This report 20

2. Strategic context 22

2.1 Policy context 23

2.2 The drive to make better use of intelligence 26

2.3 The ACPO road policing strategy 29

2.4 The link between vehicle documentation offences and crime 31

3. How Laser 2 ANPR intercept teams operate 36

3.1 Introduction 37

3.2 ANPR deployment 39

3.3 ANPR data sources 41

3.4 When a stop occurs 44

4. Findings: Operational factors 46

4.1 ANPR staff inputs 47

4.2 Team capabilities and support 55

4.3 Location deployment 58

5. Findings: Vehicle stops 60

5.1 ANPR reads, hits and stops 61

5.2 Observation-generated 65

5.3 All vehicle stops (ANPR and observations) 67

6. Findings: Actions taken, property recovered and arrests made 76

6.1 Possible actions taken at a stop site 77

6.2 Vehicle/ person search 78

6.3 Arrests 81

6.4 Other actions 94

7. Findings: Database issues 98

7.1 Context 99

7.2 Data sources 100

8. Findings: Cost recovery 106

8.1 Context 107

8.2 Conditions of cost recovery 108

8.3 Factors affecting the introduction of cost recovery 110

8.4 Fixed penalty notices issued and paid 111

9. Findings: ANPR arrest outcomes 124

9.1 Context 125

9.2 Tracking the outcome of ANPR arrests 125

10. Conclusions and recommendations 132

10.1 Conclusions 132

10.2 Recommendations 138

Appendices 142

Appendix A Acronyms 143

Appendix B: Data collection pro forma 144

Appendix D: Data completeness by field 147

Appendix D: Fixed penalty notices included under cost recovery 149

Appendix E: ANPR case studies 151

Appendix F: National ANPR project board memebership 163

Appendix G: Recorded offence guidance 164

Appendix H: References 166

In 2002, police forces started to use Automatic Number Plate Recognition

(ANPR) systems together with dedicated intercept officers. These officers

were able to intercept and stop vehicles of interest identified by the ANPR

systems and question the driver and/or passengers as appropriate. The

intention was that targeted enforcement would detect, disrupt and deter

criminal use of the roads. A six-month evaluation of the use of these

dedicated intercept officers (‘Laser 1’) showed the concept to be extremely

effective, achieving arrest rates many times that of conventional policing.

Although these results were encouraging, there was no additional funding

available for the national testing, roll-out and operation of ANPR-enabled

intercept teams. An innovative funding mechanism was therefore required.

Following a submission to HM Treasury, conditional approval was given to

the Home Office to test a cost recovery scheme for dedicated ANPR-

enabled intercept teams. This would allow police to target vehicle

documentation offences and crime in general using ANPR-enabled

dedicated intercept teams, with the activity part-funded through receipts

from fixed penalties issued.

Since 1 June 2003, 23 forces across England and Wales have been

operating dedicated intercept officers part-funded under cost recovery

(‘Laser 2’). This report presents the findings of the evaluation of Laser 2

for the 1 June 2003 to 31 June 2004.

Introduction

1.1 Context

1.1.1 Background to the use of Automatic Number Plate Recognition

ANPR is an established technology that enables vehicles observed by

cameras to have their vehicle registration mark (VRM) ‘read’ using pattern

recognition software. When combined with other resources and data,

ANPR can be an extremely powerful tool in:

• Road tolling – for example, the London Congestion Charging Scheme

uses ANPR-enabled cameras to identify vehicles passing in/out of the

congestion charge zone. This information is subsequently used to levy

tolls and to penalise non-payers

• Vehicle tax evasion – for example, the Driver and Vehicle Licensing Agency

(DVLA) uses ANPR as part of a system to ensure that vehicles on the road

have a current Vehicle Excise Duty (VED)

• Congestion warning – for example, Trafficmaster uses a national network of

ANPR cameras to measure speed between cameras and, from this, identify

areas of the road network that are congested. This is then used to provide

information to drivers.

The police have used ANPR systems at strategic points for a number of

years, for example at ports, tunnels and in the ‘ring of steel’ around the City

of London as part of counterterrorism measures. With the improvements in

ANPR technologies (which have led to increased accuracy of read and the

ability to process images more rapidly) and a reduction in costs of ANPR and

camera equipment, the police have begun to look to ANPR as a proactive tool

to help address volume crime.

1.1.2 Laser 1 – developing the concept

Recognising the potential of ANPR, the Home Office provided each police

force in England and Wales with a mobile ANPR unit and back office facility

in 2002. With this equipment, forces came to recognise that one of the most

effective ways of exploiting ANPR was to use it with dedicated intercept teams,

typically comprising around six police officers operating either on motorcycles

or from cars. These officers could then intercept and stop vehicles identified

by the ANPR system as worthy of interest, and were thus called an

‘ANPR-enabled intercept team’.

Given that the use of ANPR-enabled intercept teams represented a significant

development in policing in terms of using technology and intelligence, the

Home Office Police Standards Unit (PSU) and the Association of Chief Police

Officers (ACPO) decided to undertake a small-scale pilot over a six-month

period (30 September 2002 to 30 March 2003) – ‘Laser 1’.

Nine forces were selected to take part in the pilot, reflecting a cross-section

of force types and geographies.

The aim of the pilot was to gather evidence on the operations and impact of

ANPR-enabled intercept teams to inform policy and potential national roll-out.

These teams stopped 39,188 vehicles, arrested over 3,000 persons (of which

only 20% were for driving-related matters) and took a further 45,000 actions.

These included issuing verbal advice or a fixed penalty, or requesting that

vehicle documentation, such as MOT certificate and vehicle insurance, be

presented at a local police station.

An independent evaluation of Laser 1 concluded that:

“ ANPR-enabled intercept teams have shown to be an extremely

effective means of engaging with criminals. Using a range of

police intelligence and experience, intercept teams were able

to disrupt criminal activity in an efficient and effective manner,

achieving arrest rates ten times the national average.”

1.1.3 Laser 2 – testing cost recovery

Although Laser 1 provided encouraging results, there was no additional

funding set aside for the national testing, roll-out and operation of ANPR-

enabled intercept teams. However, following a submission to HM Treasury,

conditional approval was given to the Home Office to test a cost recovery

scheme for dedicated ANPR-enabled intercept teams. This would allow police

to target vehicle documentation offences and crime in general, with the activity

part-funded through receipts from fixed penalties issued.

Following discussions at the National ANPR Project Board (membership listed

in Appendix F), it was decided to undertake a more widespread testing of ANPR,

this time (part) funded by cost recovery – ‘Laser 2.’ The specific objectives of

Laser 2 were as follows:

• to demonstrate whether or not ANPR-enabled intercept teams could continue

to make a significant contribution to the detection of a wide variety of crimes

• to demonstrate that the primary motivation behind the additional activity was

to address criminality, not create revenue

• to inform a policy decision regarding whether or not to introduce primary

legislation to allow for the national roll-out of cost recovery

• to identify and disseminate good practice to maximise the effectiveness

of the teams

Engaging criminality – denying criminals use of the roads, PA Consulting Group (October 2003)

• to demonstrate workable, non-bureaucratic arrangements for recovering the

costs of the intercept teams that did not distort from existing policing priorities

and operations

• to test the rules and guidelines that were prepared for ANPR cost recovery.

All forces were invited by the Home Office and ACPO to participate in Laser 2.

Following submission of cases to the National ANPR Project Board, 23 forces

(including all nine from Laser 1) were accepted onto this further pilot that

started on 1 June 2003.

Part of ANPR enforcement was intended to be funded through receipts from

fixed penalties issued for vehicle documentation offences by the ANPR teams.

The cost recovery element was governed by a number of rules and guidelines,

to which all 23 Laser 2 forces subscribed. The aim of these rules and guidelines

was to ensure the cost recovery element did not distort the way in which ANPR

was used to the detriment of fundamental policy objectives.

The start of Laser 2 also coincided with the introduction of four new fixed

penalties, three of which were particularly relevant to ANPR teams, namely:

• driving without insurance

• driving without MOT certificate (where required)

• not displaying a vehicle excise licence.

Throughout the pilot, the PSU and ACPO supported individual forces by

disseminating good practice and feeding back performance measurement reports.

1.2 Evaluation methodology

1.2.1 Approach

PA Consulting Group (PA) was commissioned by the PSU to undertake an

independent evaluation of the operations of ANPR-enabled intercept teams.

In parallel, a team within the PSU was charged with developing the good

practice guide. In undertaking the evaluation, PA worked closely with this team

to understand practices that worked well and where specific interventions had

been undertaken by PSU. PA also provided information to forces and the PSU

to help identify good practice.

The PSU have developed a good practice guide that addresses key issues

around human resources, technologies and operational practices.

The basis for the evaluation was as follows:

• preparation of a data collection model – this recorded key information on

activities undertaken by the intercept teams and the resource requirements

of these teams. The data collection pro formas used as part of this model

are listed in Appendix B

• collation of this recorded information, data cleansing and validation

• briefings and field visits to each of the participating forces to ensure that

data was collected in a consistent manner and to discuss the operation of

ANPR-enabled intercept teams.

1.2.2 Data collection

Operational information was collected weekly from each Laser 2 force.

For each day of operation, this was:

• total ANPR reads and hits

• officer hours (by rank) for:

– ANPR intercept deployment and non-intercept, eg breaks, travelling time

– prisoner handling up to booking in or handing over

– ANPR admin/spreadsheet data input

• for each vehicle stop:

– day, date, time, location and VRM

– trigger database (or observation) and accuracy of database

– property recovered

– actions taken, including number and type of fixed penalties issued and

arrests made

– crime file reference numbers (to allow for tracking of cases)

where relevant

– ethnicity of vehicle driver and arrested persons.

Further information collated from forces on a quarterly basis included:

• headcount numbers of persons involved in the project

• revenue expenditure, including:

– staff salaries and on-costs (training, national insurance, etc) by rank/grade

– IT and communication systems, including maintenance, associated with

ANPR activity

– vehicle lease, maintenance and running costs (including fuel)

– consumables and ancillary costs

– leased accommodation (including office and IT equipment if applicable)

• agreed capital expenditure

• the number of fixed penalties issued, what these were issued for and

whether they were paid or whether the case has gone to court

• progress and variation against their operational case, highlighting any

significant variations and seeking permission for any change in expenditure.

1.2.3 Data validation

Every effort was made to improve data quality, including making the data

collection pro forma straightforward to use, hosting seminars with ANPR

project managers to discuss data collection issues, building basic checks

into the data entry model, undertaking random checks of data and ensuring

that data was logically consistent.

However, given the scale of collection (180,543 vehicles stopped and with

over 2 million data items recorded by the 23 forces) it was inevitable that there

would be a number of inconsistencies in data collection. The main areas of

inconsistency were:

• Different coding practices. For example, officers recorded the ethnicity of

the vehicle driver using codes reflecting their own force practice rather than

a national standard.

• Recording practices. For example, because of the variety of make-up of

ANPR intercept teams (see Chapter 4 below), there was some inconsistency

between forces in measuring officer hours input.

Part of the data cleansing process involved identifying anomalies and seeking

to address them almost immediately; with the aim of improving the quality of

data during the pilot period. In practice, after the first month data inconsistencies

tended to be isolated rather than routine and procedures were developed to

automate the data validation.

In spite of these issues, the vast majority of the data appears robust and the

large number of records allows some compelling conclusions to be drawn about

the benefits of ANPR-enabled officers compared to conventional policing.

1.2.4 Performance feedback

As part of the data collection cycle the PSU and the 23 forces were provided

with a weekly progress report, one week in arrears. This gave quick feedback

on performance, both relative to other forces and over time, and provided a

means for forces to validate high-level information submitted. The electronic

reports allowed forces to analyse their own data as required – in sufficient time

to make changes to operational deployment. A copy of the headline page of

the electronic report is shown in Figure 1.1 below.

17%

10%

15%

13%

11%

24%

Arrests breakdown in Laser 2

21%

12%

17%

11%

20%

Arrests breakdown for Laser 1

Actions taken

HO/RT1

CLE 2/6 (7)

CLE 2/8 / V62

VDRS / PG9

NEFPN

EFPN

Reported for summons

INTEL log generated

Verbal advice given

No action taken

Arrests breakdown

Robbery

Theft /

burglary

Driving

Drugs

S25

Auto crime

Warrant

Other

1,795

2,611

946

1,136

1,064

1,340

1,596

35,580

3,991

14,177

3,009

31,754

7,593

6,117

19,542

15,104

48,916

FRNs breakdown

Endorsable: No insurance

Endorsable: Other

Non-end. No MOT

Non-end. No VED

Non-end. Other

4,923

2,112

1,166

18,381

13,646

Property recovery

Stolen vehicle

Stolen goods

Firearms

Drugs

Offensive weapon(s)

Other property

Value of stolen vehicle

Value of stolen goods

Value of drugs

Total property recovered

874

333

589

159

282

£5,671,304

£421,453

£381,384

£6,474,140

Overview

Number of stops

Number of arrests

Fixed penalty notice

Pilot arrests per FTE

136,857

10,546

40,228

Staff-days on intercept duty

Staff-days travel to intercept

Staff-days prisoner handling

Staff-days spent on admin

3,657

1,661

414

1,100

Detailed weekly view

Arrests / FTE

Total arrests

FPN targets

FPNs issued

Deployment time

Old summary view

Arrests / FPNs per 100 stops

Arrests / FPNs per 100 hours

Stops

Click here to look how each force

is performing on a weekly basis

Click here for the LASER 2 “Arrests

per full time equivalent” chart

Click here to view comparison of

total arrests and breakdown by force

Click here to see how each force is

performing against its FPNs target

Click here to view all FPNs issued

by force. Breakdown also provided

Click here to see how force

deployment times compare

Click here to see weekly results

in the original format

Click here to see arrests and

FPN performance per 100 stops

Click here for arrests and FPN

performance per 100 intercept hours

Click here to compare the volume

of stops achieved by each force

Robbery

Theft / burglary

Driving

Drugs

S25

Auto crime

Warrant

Other

Figure 1.1: Weekly reporting tool provided to Laser 2 forces (front page only)

For the 14 additional forces joining Project Laser in June 2003, there was a

significant project mobilisation stage. In practice many of these forces were

unable to deploy their intercept teams for the first weeks of the pilot, and

when they were deployed, they were still developing their operational strategy.

Moreover, the data collection process took time to ‘bed-down’.

In recognition, the National ANPR Project Board extended the period covered

by the evaluation to a 13 month period to allow analysis of one year’s good

quality data. Thus in reviewing these findings a number of points are worth noting:

• While ANPR is an established technology, for many forces the use of

ANPR-enabled intercept teams represented a new way of working. As such,

operations changed to reflect feedback from the field. Also most forces used

the pilot as an opportunity to develop the way they used ANPR and varied

the way they deployed intercept teams in response to operational experience.

These evaluation results thus do not cover a ‘steady state’ period – for

example this evaluation report shows that performance improved over

the year for the 14 forces new to ANPR.

• Weeks 30 and 31 covered the Christmas period and operations were

much reduced.

• During January 2004, forces were invited to submit operational cases for June

2004 to May 2005. The majority of forces chose to continue ANPR operations

as before, though some forces re-evaluated and restructured their ANPR

operations to reflect local operational needs (eg Avon and Somerset included

prisoner handling as a core function of the ANPR team, while Leicestershire

devolved ANPR operations to Basic Command Unit level).

1.2.6 Contribution to Narrowing the Justice Gap

Laser 1 had shown that ANPR was a particularly effective tool for targeting

police resources, producing arrest rates many times those normally achieved.

Laser 1, however, had not collected information on the outcome of these

arrests. This information was key to evaluating the potential impact of ANPR

on the justice system and the Government’s target for the number of Offences

Brought To Justice (OBTJ) programme.

Following a presentation to the Narrowing the Justice Gap (NJG) taskforce,

we were asked to review the outcome of these ANPR-generated arrests and

to estimate their contribution to the Government’s target. As part of this exercise,

Laser 2 forces were asked to provide information on the outcomes of the arrests

they made between June 2003 and August 2003.

A number of forces were able to provide this information from their existing

information systems. While the results of this analysis have already been

presented internally to the Narrowing the Justice Gap taskforce, the main

conclusions are also re-presented here for the sake of completeness.

1.3 This report

This report presents the findings of PA’s evaluation of Laser 2. The purpose

of this evaluation was to explore the validity of ANPR-enabled intercept teams,

not to assess relative performance of intercept teams between forces. Results

presented have therefore been aggregated across the 23 forces, though where

appropriate these have been broken down by force. In terms of coverage, the

diagram below sets out the difference between this report and the evaluation

of Laser 1.

This report is set out in nine further chapters as follows:

• chapter 2 gives policy background of ANPR-enabled intercept teams

• chapter 3 provides an overview of Laser 2 ANPR-enabled intercept teams

and how they operate

• chapter 4 outlines operational staff inputs used during the pilot

• chapter 5 identifies ANPR reads, hits and stops, that is the number of VRMs

read by the ANPR units, the number of times these reads led to a match with

an intelligence database, and the number of vehicles of interest stopped by

the intercept teams

• chapter 6 examines database issues

• chapter 7 looks at the actions taken, the property recovered and arrests

Avon and Somerset Constabulary

Cambridgeshire Constabulary

Cheshire Constabulary

City of London Police

Cleveland Police

Greater Manchester Police

Hampshire Constabulary

Hertfordshire Constabulary

Kent Constabulary

Lancashire Constabulary

Leicestershire Constabulary

Lincolnshire Constabulary

Merseyside Police

Metropolitan Police Service

North Wales Police

North Yorkshire Police

Northamptonshire Police

Northumbria Police

Nottinghamshire Constabulary

Staffordshire Police

Warwickshire Police

West Midlands Police

West Yorkshire Police

Avon and Somerset Constabulary

Greater Manchester Police

Kent Constabulary

Metropolitan Police Service

North Wales Police

Northamptonshire Police

Staffordshire Police

West Midlands Police

West Yorkshire Police

Approximately 70 full time equivalents

6 months data

No cost recovery

No evaluation of impact on Justice Gap

No data on ethnicity

9 forces

Approximately 192 full time equivalents

13 months data

Cost recovery

Evaluation of impact on Justice Gap

Data on ethnicity

23 forces

Laser 1 evaluation Laser 2 evaluation

• chapter 8 looks specifically at the cost recovery aspect, including the cost

of ANPR operations and the fine penalties recovered

• chapter 9 examines the outcome of ANPR arrests and presents the potential

contribution that ANPR could make to the Government’s target for Offences

Brought to Justice

• chapter 10 sets out the evaluation conclusions in terms of the original

objectives for the pilot and on this basis makes a number of recommendations.

This report also has eight appendices as follows:

• appendix A lists the acronyms used in this report

• appendix B shows the data collection pro forma as used by the

intercept teams

• appendix C presents a summary of data completeness by field

• appendix D lists all Fixed Penalty Notices that were included within this pilot

• appendix E presents some ANPR case studies from forces as presented on

their websites

• appendix F lists the representation on the National ANPR Project Board

• appendix G provides outline guidance on recorded offences

• appendix H lists documents referenced throughout the report.

In all the graphs and tables in this report, Week 1 refers to the first week of

the evaluation period, ie June 1 2003. Weeks 30 and 31 therefore covered

the Christmas/New Year period.

For some of the analysis column totals may differ slightly from the total

displayed due to rounding.

Officers of the 23 forces involved in Laser 2 have provided the data used to

compile this report (on a weekly basis).

This section of the evaluation sets out the strategic context for the operation

of ANPR-enabled intercept teams. This covers four broad areas.

First, part of the Government’s vision for the criminal justice system is to

bring an additional 150,000 offences to justice in 2008 than is currently the

case and to share information within the system more effectively to reduce

inefficiencies. Within their strategy, the Government identifies ANPR as a

key means to improving police effectiveness [Section 2.1].

Second, the introduction of the National Intelligence Model within the

Police Service and the findings of the recent Bichard inquiry provides a

strong focus for police to ensure information is fully researched, developed

and analysed to provide intelligence for policing and police managers

across forces [Section 2.2].

Thirdly, the ACPO Road Policing Strategy sets out a clear objective of

detecting, disrupting and challenging criminal use of the roads. To achieve

this it is planned that police will make full use of modern technology, in

particular that approximately 2,000 officers will deliver an intercept capability

ANPR. This equates to a police intercept team in every Basic Command

Unit area [Section 2.3].

Strategic context

To support this ACPO have developed an ANPR strategy for the police.

At present, however, there is no complementary strategy for use ANPR

across other Government bodies, including DVLA, DfT, Customs & Excise,

the Highways Agency, VOSA, and the ports authorities [Section 2.3].

Finally, this section identifies that there is substantial evidence that the

pursuit of vehicle documentation offences will lead to more serious crimes

being detected and that relatively little police time is spent undertaking

proactive vehicle documentation checks. As such, there is an opportunity

being lost to address wider criminal issues. If this were addressed by

means of ANPR-enabled intercept teams, this would also meet the public’s

desire to see more ‘officers on the street’ and more action taken against

illegal drivers [Section 2.4].

2.1 Policy context

In the recent strategic plan for the Criminal Justice System

the Government

sets out a vision for the criminal justice system for 2008. This vision is built

around five key objectives:

•

“The public will have confidence that the Criminal Justice

System is effective and that it serves all communities fairly.

• Victims and witnesses will receive a consistently high standard

of service from all criminal justice agencies.

• We will bring more offences to justice through a more modern

and efficient justice process.

• Rigorous enforcement will revolutionise compliance with

sentences and orders of the court.

• Criminal justice will be a joined up, modern and well run service,

and an excellent place to work for people from all backgrounds.”

Cutting Crime, Delivering Justice: A Strategic Plan for Criminal Justice 2004-08, Home

Office/DCA (July 2004)

Ibid (July 2004), p9-10

In terms of the primary objective of an effective criminal justice system,

the strategic plan sets out the Government target of bringing 150,000 more

offences to justice in 2008 and states:

“We will raise the detection rate from 19% to at least 25%, by

improving police effectiveness and deploying new technology,

including enhanced DNA testing and Automatic Number Plate

Recognition systems, across the country to target criminals

more effectively.”

The strategic plan also identifies the need for better intelligence and

information-sharing across the criminal justice system. It highlights that there

is no single data source to identify individuals who may be wanted by a

number of police forces and courts for fine arrears, failing to appear in court

or probation breaches. This means that agencies are often pursuing the same

offender separately for breaches of different types of warrant. The police may

arrest someone and bail them without knowing about outstanding warrants for

them. Equally, unknown to a court, a defendant appearing in front of them may

have failed to answer charges elsewhere or have other outstanding fines or

community punishments. This leads to poorly-informed decisions, frustration

on behalf of the professionals involved and unnecessary costs; it also helps

undermine public confidence in criminal justice.

To address this, the strategic plan sets out a key change in information

sharing, namely:

“We are giving direct access to the Police National Computer to

all Magistrates’ Courts Committees by Autumn 2004. This will

enable warrants to be entered promptly onto the system so

police are aware of and can act on them. We will also link this

into the ANPR system so that offenders wanted on warrants can

be identified when their car is spotted by an ANPR camera.”

Ibid (July 2004), p10

Ibid (July 2004), p42

The Government’s commitment to tackling vehicle crime and, in particular,

addressing the problem of uninsured driving was outlined in the Government’s

response to the publication of the Greenaway report in August 2004

. The Road

Safety Minister David Jamieson announced that the Government will:

• give the police the power to seize and, in appropriate cases, destroy vehicles

that are being driven uninsured

• link the DVLA’s Vehicle Register and the Motor Insurance Databases,

allowing police to know which vehicles on the road are uninsured

• allow fixed penalties for people who ignore reminders that their insurance

has expired.

The DfT also wants to see and is discussing with relevant stakeholders:

• concerted action by insurance companies to continue to improve the Motor

Insurance Database

• simpler and clearer notification procedures so that no one is in any doubt

when their insurance expires

• automatic reminders sent out to those motorists who forget to insure on time.

David Jamieson, Parliamentary Under-Secretary of State in the Department

for Transport said:

“ I very much welcome Professor Greenaway’s report. We know

that lawabiding motorists are fed up with paying the price for the

small, hard core of antisocial motorists who drive uninsured, often

in untaxed or unsafe vehicles.

The Government is determined to tackle head on the menace

of uninsured driving. That is why I have announced today that

we plan to give the police the power to seize and destroy

vehicles that are being driven illegally and to increase police

powers to use new technology to make detection and

enforcement more effective.

We are also working closely with the insurance industry to

improve detection of drivers who fail to insure their vehicles

and to raise awareness of the need for motor insurance.

The message to the small hard core of antisocial motorists

who drive without insurance is clear – uninsured driving

is unacceptable.”

Uninsured Driving in the United Kingdom, Professor David Greenaway (July 2004)

David Jamieson, DfT press release 11 August 2004

Caroline Flint, Parliamentary Under-Secretary of Stete in the Home Office, added:

“ Uninsured driving victimises the law-abiding motorist.

This report gives a sensible way forward to tackling the problem

and across Government we will work hard to take forward its

recommendations. We want to ensure that the police and courts

have the powers they need to tackle offenders and that they can

use them effectively.

We are also working closely with the police to ensure that the

hugely successful Automatic Number Plate Recognition system

is used as effectively as possible to target those who flout the

law and drive without insurance.”

Finding 1. The Government views ANPR as a key tool for bringing

more offences to justice and to identify and pursue the estimated

1 million motorists that drive without insurance as well as those

wanted on warrant.

2.2 The drive to make better use of intelligence

The use of ANPR-enabled intercept teams is an excellent example of an

intelligence-led policing tool. This section looks at the current drivers for

making best use of police intelligence. This applies at both a National and

European level, with ANPR expected to be a key part of policing international

borders and sharing intelligence across European states as part of the

Schengen Acquis.

2.2.1 The National Intelligence Model (NIM)

In the context of the police reform agenda, the NIM is ‘A Model for Policing’

that ensures that information is fully researched, developed and analysed to

provide intelligence that police managers can use to:

• provide strategic direction

• make tactical resourcing decisions about operational policing

• manage risk.

It is important to note that the NIM is not just about crime and not just

about intelligence – it is a model that can be used for most areas of policing.

It offers, for the first time, the realisable goal of integrated intelligence in

which all forces and law enforcement agencies play a part in a system

greater than themselves.

Caroline Flint, DfT press release 11 August 2004

Launched by the National Criminal Intelligence Service (NCIS) and adopted by

the ACPO in 2000, the government placed the NIM at the centre of the police

reform agenda. The model has been designed to impact at three levels of

business: local, cross border and serious and organised crime:

• Level 1 – Local issues – usually the crimes, criminals and other problems

affecting a Basic Command Unit or small force area. The scope of the crimes

will be wide ranging from low value thefts through to serious offences such as

murder. The handling of volume crime will be a particular issue at this level.

• Level 2 – Cross border issues – usually the actions of a criminal or other

specific problems affecting more than one basic command unit. Problems

may affect a group of basic command units, neighbouring forces or a group

of forces. Issues will be capable of resolution by forces, perhaps with support

from the National Crime Squad, HM Customs and Excise, the National

Criminal Intelligence Service or other national resources. Key issues will be

the identification of common problems, the exchange of appropriate data and

the provision of resources for the common good.

• Level 3 – Serious and organised crime – usually operating on a national and

international scale, requiring identification by proactive means and response

primarily through targeting operations by dedicated units and a preventative

response on a national basis.

In the context of ANPR-enabled intercept teams, their primary role is to

address level 1 criminality, though clearly they have a potential role in tackling

level 2 and 3 criminality. For example ANPR units can gather intelligence on

vehicle movements and the deployment of intercept teams on strategic roads

and could potentially detect and disrupt cross border movement of criminals.

PSU have developed a process map of how ANPR contributes to NIM.

This is explained in full in the good practice guide and summarised in

Figure 2.1 overleaf.

2.2.2 Policing Bureaucracy Taskforce recommendations

The Policing Bureaucracy Taskforce (chaired by Sir David O’Dowd, former

Chief Inspector of Constabulary) was established in January 2002 as part of

the Government’s Police Reform programme to seek ways to increase the

presence of uniformed officers in the community by:

• removing the unnecessary burdens borne by front-line staff

• providing adequate support

• revising working practices to enable them to operate more effectively.

The Taskforce report

identified that the public wanted to see more uniformed

police officers in the community and that front-line officers want to dedicate

more of their time to dealing effectively with crime and anti-social conduct and

in bringing offenders to book. It also acknowledged that there was a general

desire for the police and criminal justice professionals to succeed in convicting

the guilty and, in particular, persistent offenders whose activities blight the lives

of whole communities.

ANPR deploy to gather

intelligence on moving criminality

into / within / from area

Add to analysis of

previous ANPR deployment

in area (if any)

Analysis of ‘hits’ to

profile movement of

criminality in vehicles

Intelligence T&CG Process Tactical response

T&CG identifies issue

as priority for rescuing

T&CG consider

intelligence outputs to

inform Tactical Response

‘Semi-speculative’ deployment of

ANPR and Intercept Team as part of a

tactical response (crime analysis only)

Fully intelligence-led deployment of

ANPR and Intercept Team as part of

tactical response (crime and vehicle

movements analysis)

Option 1 Option 2

Intelligence gathered during deployment fed back into system

Figure 2.1: How ANPR contributes to NIM

Policing Bureaucracy Taskforce Report, Home Office (17 September 2002)

In relation to ANPR, the Taskforce recognised that ANPR was an extremely

useful tool; however it could only be fully effective if sufficient intervention

resources, specifically intercept teams, are deployed to respond to a

significant proportion of alerts. It recognised that ANPR:

• increased police performance in crime detection

• lead to higher police visibility and citizen reassurance

• could be partly self-funding through cost recovery

• could reduce time spent on paperwork by increased use of FPNs over

preparing traditional court files for appropriate offences.

The report made a specific recommendation to develop the use of ANPR.

2.2.3 The Bichard inquiry

The independent inquiry arising from the Soham murders chaired by Sir

Michael Bichard investigated a number of issues, including the effectiveness

of the relevant intelligence-based record keeping and information sharing with

other agencies and between forces.

The inquiry report

made a specific recommendation that the Home Office

should lead the development of a national information technology system

for England and Wales to support police intelligence and that it should be

introduced as a matter of urgency. Government has accepted these findings

and recommendations in full.

In this context it is important to note that other than for the services provided

by PNC, police forces are having to operate with a series of local databases

in regard to vehicle intelligence which have to be copied and shared between

forces. This is a time consuming and ineffective way of operating and highlights

the need for a national data warehouse. This could hold all vehicle intelligence

to be read in real time by all ANPR users nationally. In turn, this data warehouse

would also hold ANPR reads and hits as a further source of vehicle intelligence,

providing great benefits to major crime and terrorism enquiries.

2.3 The ACPO road policing strategy

The national ACPO roads policing strategy

presents the use of ANPR

as a core activity for the police to detect and respond to criminal activity

on the roads:

The Bichard Inquiry Report, Sir Michael Bichard, House of Commons (June 2004)

Modern road policing – a manifesto for the future, ACPO (November 2002)

“The police have a duty to tackle criminality, in all its forms,

including contravention of road traffic law much of which is

aimed at poor driver behaviour. We intend to use the police

National Intelligence Model to focus enforcement activity in order

to detect, disrupt and challenge criminal use of the roads. To do

this we will make full use of modern technology, and in particular

Automatic Number Plate Recognition systems that have the

potential to revolutionise road policing.”

To support the development and use of ANPR, ACPO have drafted an ANPR

information, intelligence and technology strategy

. The vision is to roll-out

Laser 2 to all forces such that approximately 2,000 officers are delivering an

intercept capability. This sets out how the Police Service will use ANPR,

specifically to meet its strategic aim of denying criminals the use of the roads

through a national infrastructure of ANPR technology throughout England and

Wales. The intention is to back this up by a police intercept team in every

Basic Command Unit area.

This strategy identified that every force in England and Wales has ANPR

capability and back office facility and shortly this back office facility will enable

ANPR data to be transferred between all forces through the secure and

controlled environment of the Criminal Justice Extranet (CJX). However the

strategy highlighted that while all forces have ANPR equipment, they are using

systems from a variety of suppliers. To address this PSU and ACPO have

recently prepared and circulated a National ANPR Standards document that

details the minimum standards within which police ANPR systems should operate.

This strategy, however, represents the Police Service strategy, and does cover

ANPR across Government, which includes DVLA, Customs & Excise, the

Highways Agency, VOSA, and ports authorities.

ANPR information, intelligence and technology strategy, ACPO (June, 2004)

2.4 The link between vehicle documentation

offences and crime

In the UK there is a high-level of non-compliance with vehicle documentation

requirements, for example:

• There are over 1.76 million vehicles on the road that do not have a valid

vehicle excise licence (approximately 5.5% of all vehicles on the road).

This evasion costs the HM Treasury (HMT) over £190 million per annum.

• DVLA have no registered keeper information for approximately 1.9 million

vehicles on the road. Anecdotal evidence from traffic police suggests that

where registered keeper information exists, the actual keeper is likely to be

different to the registered keeper in at least 10% of cases.

• The Association of British Insurers (ABI) estimates that there are at least one

million persons driving regularly while uninsured, ie about 5% of all drivers.

Accidents involving uninsured motorists cost up to £500 million a year, which

ultimately adds approximately £30 a year to each motorist’s premium.

• While no statistics are collected, it is believed that around 10% of those

vehicles requiring an MOT do not have a current MOT certificate.

Following a nationwide police operation to assess the level of non-compliance

on the roads, the DfT is expected to publish more information on the above in

the autumn of 2004.

2.4.1 There is a correlation between vehicle and traffic offences

and other criminality

Historically, police have not focused on these offences for a number of

reasons. First, the offences themselves were not seen to be as important

as other volume crime. However, evidence suggests that there is a strong

correlation between vehicle crime and other, more serious, crimes – for

example a Home Office study

demonstrated the link between traffic offending

and general criminality. The study found that of those parking illegally in

disabled parking bays:

• 21% of vehicles were of immediate police interest

• 33% of keepers of the vehicles had a criminal record

• 49% of the vehicles had a history of traffic offending

• 18% of vehicles were known or suspected of use in a crime

• 11% of vehicles were in breach of traffic law, eg no VED.

Vehicle Excise Duty Evasion, Department for Transport (2002)

New Research on Uninsured Drivers, Association of British Insurers (March 2004)

Illegal Parking in Disabled Bays: A Means of Offender Targeting, Sylvia Chenery, Chris Henshaw

and Ken Pease, Home Office RDS (1999)

These figures are significantly higher than the ‘average’ vehicle/vehicle driver.

The Home Office has also completed a study of the criminal history of serious

traffic offenders

. The study examined the extent to which anti-social behaviour

on the road was linked to wider criminal activity. It looked specifically at drink

drivers, disqualified drivers and dangerous drivers. A finding was that disqualified

drivers showed a similar offending profile to mainstream criminal offenders.

79% had a criminal record (72% for mainstream offenders), their levels of

previous offending were slightly higher than for mainstream offenders and they

were equally likely to be convicted again within a year (37% were reconvicted).

Importantly, however, police used prior intelligence in only half of all arrests.

This suggested that if police were able to access previous convictions in a

timely fashion, this could help more effectively target resources.

An important point that emerged from the study was the level of non-

specialisation of offence types – those repeatedly committing serious traffic

offences were also likely to commit mainstream offences. The evidence

suggests that these offenders cannot generally be thought of as otherwise

law-abiding members of the public. Even drink drivers (who were less involved

in mainstream crime than other serious traffic offenders) were estimated to be

twice as likely to have a criminal record as members of the general population.

When serious traffic offenders were reconvicted, there was a tendency for

repeat serious traffic offending (especially disqualified driving) although this

was in a context of more generalised criminal offending.

Recent research by the insurance industry evidences the strong link between

serious motoring offences and the one million motorists driving without

insurance

. Specifically, compared to drivers with insurance, uninsured

drivers are:

• ten times more likely to have been convicted of drink driving

• six times more likely to have been convicted of driving a non-roadworthy

vehicle

• three times more likely to have been convicted of driving without due care

and attention.

Finding 2. There is substantial research evidence to suggest that the

pursuit of vehicle documentation offences will lead to more serious

crimes being detected.

The Criminal History of Serious Traffic Offenders, Gerry Rose, Home Office RDS (2000)

New Research on Uninsured Drivers, Association of British Insurers (March 2004)

2.4.2 Scarce police resource is being stretched

A second reason why vehicle documentation enforcement has not been a

police focus has been the significant resource constraints upon traffic

police, ie those officers who would normally undertake vehicle documentation

enforcement. For example, a study published in 2003

estimated that less

than 6% of police personnel are dedicated to traffic and vehicle duties. In spite

of an increase in traffic volume (8% increase between 1997 and 2002)

and

vehicles (13% increase between 1997 and 2002)

, the number of designated

road traffic police fell by 13% between 1997/98 and 2002/03 to approximately

6,900 officers. An analysis of activity undertaken by these traffic police

officers

showed that less than 5% of their time was spent on static vehicle

checks and vehicle documentation checks – this equates to approximately

350 full time officers across England and Wales

Finding 3. These figures suggest that relatively little police time is spent

undertaking proactive vehicle checks and, given the above link between

vehicle documentation offences and more serious crime, this appears to

be an opportunity lost to address wider criminal issues.

Finally, police have not focused on vehicle documentation enforcement due

to the sheer volume of traffic on the road – in the UK there are nearly 30

million vehicles currently registered and over 485 billion vehicle kilometres

driven on the road network per year

Finding 4. The distances travelled on the UK roads presents a huge

logistical problem for police in terms of identifying and filtering out

vehicles worthy of stopping.

With the improvements in ANPR technologies and an overall reduction in IT

costs, it has been proven that ANPR can address these difficulties and become

an effective policing tool

. Criminals, like other citizens, need to use the roads

and, given the potential of ANPR allied with good police intelligence, when

they do so they are susceptible to detection.

Roles and responsibilities review Highways Agency/ACPO, PA Consulting Group (2003)

Road traffic: by type of vehicle: 1992-2002, DfT (2004)

Ibid

Roles and responsibilities review Highways Agency/ACPO, PA Consulting Group (2003)

While vehicle document checks may be undertaken by ordinary officers, no data exists on the

volume undertaken

Ibid

Engaging criminality – denying criminals use of the roads, PA Consulting Group (October 2003)

2.4.3 The public want more action taken against illegal drivers

In terms of public perceptions, surveys of motorists reveal strong support

for action against documentation offences. For example a recent survey

found that:

• three quarters of people surveyed are worried about the number of uninsured

drivers on the road

• more than nine out of ten (97%) people urged the Government to do more

to tackle this problem

• in terms of specific actions against uninsured drivers:

– a third of those questioned would like to see offenders taken off the road

for good with a total driving ban for culprits

– a third favoured confiscation of the vehicle

– a fifth would welcome larger fines

– while a fifth favoured imprisonment.

In terms of industry support, it is interesting to note that in their response to the

current Government review of the uninsured drivers, the ABI recommended:

“We need to see a step-change in enforcement processes, to

improve the actual and perceived chances of uninsured drivers

being caught. A new modern and cost-effective enforcement

process needs to be introduced. . .”

Commissioned by MORE TH>N and conducted by TNS via telephone interview amongst 1,006

GB adults aged 16+ from 11-13 June 2004

Response of The Association of British Insurers on behalf of Motor Conference and the MIB to

The Greenaway Review of Compulsory Motor Insurance and Uninsured Driving, ABI (February

2004)

Finding 5. There is significant public and industry support for a radical

change in the way documentation enforcement takes place, in particular

there is support for enforcement to target uninsured drivers. However, a

legislative change is needed to allow the proactive targeting of vehicles

without insurance (by using intelligence provided by an insurance

database). This would also help to increase the productivity of ANPR

intercept officers.

Finding 6. There is a major drive within policing to make better use of

intelligence, both as a means of targeting resources and to engage with

criminality. In the context of this report, it is clear that as an intelligence-

led policing tool, the effectiveness of ANPR in engaging level 2 and 3

criminality will be limited by the availability of good quality and timely

intelligence.

This section of the evaluation provides more information on how ANPR

intercept teams function. In terms of operation, the use of ANPR intercept

teams represents a radical approach:

• targeting vehicle documentation enforcement to engage with and

disrupt criminals

• delivering through an intelligence-led piece of technology

(an ANPR reader)

• benefiting from officers’ experience (eg observations of vehicle drivers)

• supported by existing policing processes (eg prisoner handling)

[Section 3.1].

ANPR monitoring can be undertaken by a number of means, principally

through fixed infrastructure (CCTV systems), within existing patrol cars

(in-car systems) or as a dedicated mobile unit. No one method of

deployment is significantly more accurate in terms of VRM reads – the

key issue is how the police operate the systems to meet local operational

targets. It is worth noting that ANPR-enabled intercept teams do not rely

solely on ANPR technologies but also use their training, experience and

judgement. Vehicles that are not flagged by the ANPR system but are

being driven suspiciously can also be stopped [Section 3.2].

How Laser 2 APNR

intercept teams operate

In terms of data sources, ANPR can be used with any database that

includes reference to a VRM. Within Laser 2, the principal data sources

were Police National Computer (PNC), local force information systems

and DVLA’s databases of vehicles recorded as not having VED or a

known keeper.

Programmed improvements to existing vehicle databases (DVLA and PNC)

and the development of legislative powers to use other databases proactively

(eg motor insurance database) will provide more and better quality

intelligence to ANPR intercept teams. With the success of ANPR intercept

teams, non-ANPR intercept officers are beginning to supply more vehicle-

based intelligence for the ANPR teams to exploit. Most national vehicle

databases are or will shortly be available for ANPR intercept teams.

While there is still a need to provide this intelligence as part of a national

data warehouse, these will allow ANPR teams to be more effective,

particularly in stopping those vehicles that appear on a number of

databases (‘multiple hits’) [Section 3.4].

3.1 Introduction

ANPR systems read VRMs from digital images, captured either through in-car

systems, closed circuit television camera (CCTV), or a mobile unit (normally

mounted in a vehicle). A key feature of all ANPR systems is their speed and

efficiency of analysis – the systems are capable of checking up to 3,000

number plates per hour of vehicles travelling up to 100 mph. Individual ANPR

units can link up to four cameras and cover several lanes/locations at a time.

Older systems were susceptible to crude manipulation of number plates

(for example using black insulation tape to change an ‘F’ into an ‘E’), and

functioned badly in poor visibility conditions. Newer infrared cameras combine

the latest software, are much more reliable and are able to accurately read

most VRMs – in practice this means ANPR systems are able to correctly

read 95 number plates out of 100.

The conversion of an image of a registration plate into text allows this data

to be used in a variety of ways including cross-referencing with databases.

This process is performed in a fraction of a second. Within a policing context,

ANPR can be used to identify vehicles flagged on the Police National

Computer (PNC), local Force Intelligence Systems (FIS) or other related

databases (eg DVLA or Customs and Excise).

Where there are support resources, action can then be taken immediately

– police know where a vehicle is and the direction in which it is travelling.

Prior to the introduction of ANPR, the volume of traffic helped to conceal those

committing vehicle-related crimes. The use of ANPR and dedicated intercept

teams can thus allow police to actively engage with criminality.

An example of how ANPR can be used with intercept teams is shown in

Figure 3.1. The vehicle passes an ANPR camera (either in-car, CCTV or a

mobile unit). This sends image data to the ANPR system, which ‘reads’ the

VRM and crosschecks it against a database; in this case the PNC and a

Force Intelligence System. Where a match is found, the ANPR operator is

notified and can decide to call for an intercept vehicle.

Figure 3.1: Use of ANPR to direct intercept teams

Stage 1

Vehicle passes

ANPR camera

CCTVIn car system

Stage 3

Decision taken

to stop vehicle

Stage 4

Vehicle stopped in

safe environment

Mobile Unit

Stage 2

ANPR software

checks against

database

W407 GHM

Vehicle:W407 GHM

22/07/04 14:04:03

Confidence: 98%

W407 GHM

Stolen vehicle

Intercept

officers notified

The development and increased use of ANPR technology allows for a more

focused approach than was previously possible. Officers responding to the

ANPR alerts do so in a targeted way using police intelligence with significantly

improved chances of detecting offences and disrupting criminals.

It is worth noting that ANPR-enabled intercept teams do not rely solely on

ANPR technologies but also use their training, experience and judgement.

Vehicles that are not flagged by the ANPR system can also be stopped,

based on the professional judgement of the intercept officer.

Apart from the use of ANPR with intercept teams, all other aspects of

policing and prisoner handling were as per normal force practice. During the

pilot, however, as a result of the effectiveness of ANPR intercept teams, these

practices developed. For example, certain forces (eg West Midlands Police

and Cheshire Constabulary) made use of dedicated prisoner handling units

to support the ANPR intercept team. The intention was to minimise the time

spent by ANPR intercept teams in handling prisoners and to maximise their

time intercepting. However, for the majority of forces, ANPR intercept teams

undertook their own prisoner handling.

Finding 7. Conceptually, the use of ANPR intercept teams represents

a radical approach:

• targeting vehicle documentation enforcement to engage

with and disrupt criminals

• delivered through an intelligence-led piece of technology

(an ANPR reader)

• benefiting from officers’ experience (eg observations of vehicle drivers)

• supported by existing policing processes (eg prisoner handling).

3.2 ANPR deployment

Not all forces used the same equipment or structure of intercept team during

the pilot. In terms of deployment, three approaches were used:

• Mobile ANPR vehicle with intercept capability – The majority of pilot

forces involved in Laser 2 used a static ANPR vehicle, normally a van,

operated in conjunction with dedicated marked mobile police resources,

most usually marked motorcycles. The ANPR van was normally parked at the

side of the road, in a lay-by, verge or central reservation. Motorcyclists were

then deployed approximately 250 metres further down the road to stop

vehicles of interest.

• In-car systems – This form of deployment was based around individual

patrol vehicles fitted with ANPR operating without back-up intercept support

– the officers in the vehicle would stop vehicles of interest. However, when

a vehicle was stopped, the officers operating the ANPR equipment would

be ‘tied up’ with an enquiry and hence the ANPR reader would not be fully

exploited. This method was seen as relatively inefficient means of operation

and hence none of the forces used this as a primary means for deployment

• CCTV – Some forces also used ANPR readers linked to existing public space

CCTV systems and used dedicated intercept teams to follow up on vehicles

of interest. For this deployment, the CCTV control room (situated on local

authority premises for ease of access to the CCTV camera matrix) handles

the incoming video source. Number plate details are then sent via a data

link to the processor unit within the police control room where the relevant

databases are situated so a match can be made. The fixed nature of ANPR

links to CCTV enable it to have a live fast track access to the PNC, allowing

access to the most current information. The police controller is informed

which vehicle is of interest and the intelligence report that has identified it.

An ANPR intercept team is then despatched to vehicles that are identified

in this way.

Finding 8. ANPR deployment can be undertaken by a number of means,

principally through fixed infrastructure (CCTV systems), within existing

patrol cars (in-car systems) or as a dedicated mobile unit. No one method

of deployment is significantly more accurate in terms of VRM reads –

the key issue is how the police operate the systems to meet local

operational targets.

Some forces based their ANPR operations on the force tasking and

co-ordination process in accordance with the National Intelligence Model.

This is evolving, but to ensure that ANPR is deployed appropriately, proper

consultation regarding selection of suitable venues is undertaken with the

team supervisor. Those locations where there are high levels of crime and

high volumes of traffic flow are considered. High volume roads are also

typically high visibility, leading to greater public reassurance, though these

roads do not always make the most appropriate intercept locations.

This is not to say that ANPR cannot be deployed where the volume of

traffic is lighter. Consideration has to be given to the location, the number

of vehicles passing through the area, the number of target vehicles likely to

be encountered and where there is specific intelligence to indicate that its

deployment would be beneficial. The time of day may also influence the

number of vehicles stopped – this is discussed in section 5.3.5 below.

The speed of traffic can make interception more difficult and can determine

whether or not pursuit situations occur. In identifying locations for intercept,

effort is made to ensure smooth traffic flow and consideration should be given

to likely disruption, particularly busy commuter thoroughfares. For example,

fixed CCTV may be more effective for motorway junctions (especially where

they are linked to arterial roads), service areas (where vehicles will come to a

halt) and bridges where the capacity for greater intelligence gathering exists.

The nature of the operation can dictate the size of location for intercept.

For example, smaller operations may utilise roadside lay-bys, whereas larger

operations, involving other agencies, may require a much larger site to operate

safely (such as a shopping centre car park).

3.3 ANPR data sources

ANPR can be used with any database that includes reference to VRM.

In a policing context, the most obvious data source is the PNC. Within

PNC, there are two main indexes, namely Vehicles and Criminal Records.

The Vehicles Index houses 50 million records, containing full descriptive

details of vehicles and, where known, their registered keepers. The vehicles

index includes both:

• Reports, which are based on specific police intelligence. As of the

1 July 2004 there were approximately 936,000 vehicles reported on

PNC as follows:

– 551,767 lost/stolen – 26,633 found

– 41,653 removed – 171,658 destroyed

– 80,654 information – 4,167 seen/checked

– 45,001 corrected

• Markers, which are based on data supplied by third parties, for example

the DVLA. There are approximately eight million markers for a wide range

of possible vehicle documentation offences including ‘no registered keeper’

and ‘no valid vehicle excise duty’.

In terms of ANPR usage, reports are held in the ANPR system for immediate

access, while markers are accessible only when a standard vehicle check is

carried out. For many intercept officers, information on PNC markers will,

therefore, only be available through an enquiry to their control room. For this

reason most forces equip their ANPR systems with information supplied direct

from DVLA on vehicles which have no vehicle excise duty or where there is no

registered keeper as a separate ANPR database.

Police Information Technology Organisation (PITO)

It is worth noting that during March 2004 DVLA had technical problems

producing the ‘No current keeper’ and the ‘No VED’ databases. Forces typically

received updates of these databases monthly, thus during this period, forces

were using increasingly dated data sources that would have been significantly

less accurate – this is reflected in the performance of these databases as can

be seen in section 7. In fact many forces chose not to use these databases

due to these issues.

In addition, most forces also use local intelligence databases with ANPR,

for example the registration marks of those vehicles:

• that have been recorded speeding but have failed to respond the

Notice of Intended Prosecution

• belonging to recently disqualified drivers.

Forces use the data as a means of positive inclusion, that is where a VRM is

matched with a specific marker on the database, then the vehicle is stopped.

Finding 9. The current intelligence databases do not allow ANPR readers

to identify ‘ghost’ VRMs, that is false GB VRMs that have never been

issued by DVLA.

Finding 10. Currently, ANPR teams operate as discrete operations with

no cross-referencing of VRM reads either within or between forces.

3.3.1 Future enhancements

The above finding has been recognised by ACPO and the Home Office and

they are currently seeking to develop a national vehicle intelligence database.

This would allow all forces to work in real time with the same information,

would include all vehicle hits and would be available Nationally rather than

held in individual forces.

In terms of further future developments, enhancements to the driver and

motor insurance database within PNC will also provide patrolling officers

at the roadside with information on drivers’ licence entitlement and their

insurance status. Currently drivers have to produce their documents at a

police station of their choice within seven days (HO/RT/1). In future officers

will able to request a PNC check through the police control room to find out,

in most cases, whether the motorist before them is insured or has the right

licence entitlement to be behind the wheel. Legislation is still needed to

support the pro-active use of this information.

A key issue in relation to any roadside stop is driver identification. There is

currently no requirement for drivers to carry identification or a driving licence.

While many drivers do carry some form of photographic identification, many

do not (and often it is those that do not that are of most interest to the police).

In practice this means that any officer stopping vehicles at the roadside may

have to take an individual to a police station in order to validate their details.

In order to address this, PITO are developing a roadside fingerprint capability

to assist ANPR teams with the identification process.

3.3.2 Continuous registration, MOT and motor insurance databases

In the context of this evaluation it is worth noting that from 1 February 2004

DVLA have been enforcing ‘continuous registration’. Specifically vehicle keepers

who fail to re-licence within eight weeks following the expiry of the old licence

are now being sent a letter stating that they have committed an offence for

which a fine is payable and requesting that the keeper either re-licenses their

vehicle or declares the vehicle off road (Statutory Off-road Notification – SORN).

While it is too early to evaluate the impact of continuous registration, in the

context of this evaluation it is worth noting the potential effects:

• increasing the overall level of VED compliance (through a better process and

associated communication), thereby reducing potential fine revenue for the

ANPR intercept teams

• increased accuracy of the no VED and Current Keeper databases.

The computerisation of the MOT database is due to go live late 2004.

This will give police officers roadside access to information on the MOT

status of a vehicle in a similar way to the drivers and motor insurance

database. In addition, an automatic flagging system will eventually be

introduced to alert police each time a stolen vehicle is taken to a testing

station for an MOT .

It is expected that the drivers and motor insurance database, continuous

registration and the MOT database will dramatically cut the number of

traditional document productions at the police station and reduce the

paperwork burden on the police. Giving officers the tools to check driving

licence, motor insurance and MOT details at the roadside will also make life

easier for the honest motorist who, in most cases, will no longer have to visit

a police station with their papers.

Finding 11. Improvements to existing vehicle databases (DVLA and PNC)

and the development of new databases (MOT and motor insurance) will

provide more and better quality intelligence to ANPR intercept teams.

With the success of ANPR intercept teams, non-ANPR intercept officers

are beginning to supply more vehicle-based intelligence for the ANPR

teams to exploit. These new databases will shortly be provided for ANPR

intercept teams – these will allow ANPR teams to be more effective, in

particular stopping those vehicles that appear on a number of databases

(‘multiple hits’).

3.4 When a stop occurs

Following a vehicle stop, an intercept officer will question the vehicle driver

and/or passenger(s) and where appropriate inspect the vehicle. Following this,

the intercept officer can take a number of actions including:

• Vehicle/person search – an officer may decide that the vehicle or individuals

in it should be searched.

• Recovery of property – the search of a vehicle/person can often lead to the

recovery of stolen goods, drugs or even the vehicle itself.

• Arrest – whereby an officer arrests an individual in relation to an offence.

• Reported for summons – where an individual was reported to appear in

court in relation to minor offences (normally motoring) where a fixed penalty

was not appropriate or the offence was too serious (for example four tyres

with insufficient tread).

• Issuing a fixed penalty – these can be issued for a variety of vehicle/driving

offences, such as contravening directional signs, driving without wearing a

seatbelt or using a mobile phone while driving. The recipient is issued with

a ticket that requires them to pay a fine and, where appropriate, provide their

driving licence for endorsement. The police forces in Laser 2 were able to

use the issue of a small number of tickets (see Appendix D) for cost recovery.

This is covered in section seven of this report.

• Issuing a note requiring follow-up action – these include:

– HO/RT – which requires a driver to present their driving licence and motor

insurance details to a local police station within seven days. This is also

linked to the issue of conditional offers for the offences of driving without

insurance and no MOT .

– CLE2/6 and CLE2/7 – no current excise offence report to DVLA used for

all vehicles.

– CLE2/8 and V62 – no current vehicle excise offence combined with

failing to notify current keeper offence. V62 is application for registration

document only.

– VDRS – Vehicle Defect Rectification Scheme (notice to offender to carry

out repairs to defect within 14 days and have repair certified or face

court proceedings).

– PG9 – vehicle prohibition notice, prohibiting the use of the vehicle on the

road due to its defective state and requiring a full MOT to be undertaken

prior to reuse on road.

• Intelligence log – an officer may decide that during a vehicle stop they

have uncovered information that should be shared with other officers/forces,

for example who was driving a vehicle or who the passengers were. In this

case the officer fills in an intelligence form and sends it to the local

intelligence officer.

• No action taken – where no offence has been committed or the police

consider there is insufficient evidence to prosecute or that an informal

warning may be sufficient.

Figure 3.2 shows the high-level process for how these outcomes are arrived at.

The possible outcomes for an ANPR intercept team vehicle stop are,

in principle, the same as for any vehicle stop. However, as will be shown

in the following Sections, ANPR stops lead to a large volume of arrests

being made and fixed penalties issued.

Figure 3.2: How stops are dealt with

ANPR Trigger

– PNC

– DVLA

– Local database

Observations

– Mobile phone

– No seatbelt

– Vehicle excise duty

– Vehicle defect

– Driving manner

– Known person / vehicle

– Other

Stop vehicle

Has an offence

been comitted

No further action

Establish identity

of driver and

relationship to car

Section 25 arrest

Vehicle / documentation offenceVehicle / documentation offence Arrestable offenceArrestable offence

Does the vehicle / person

need to be searched

Does the vehicle / person

need to be searched

Arrests for

– Robbery

– Theft / burglary

– Auto crime

– Drugs

– Warrant

– Other

Has property

been recovered?

Driving offences

Other offences

Known / wanted

criminals

Arrests for

Driving offences

– Driving whist

disqualified

– Driving under

the influence of

drugs/alcohol

Arrests for

– Robbery

– Theft / burglary

– Auto crime

– Drugs

– Other

Deal with by

fixed penalty

– No insurance

– No MOT

– No Tax

– Other

hypothecated

offences

– Other

non-hypothecated

offence

Revert to traffic

process

– HO / RT1

– CLE 2/6/7/8/9

– VRDS

– PG9

– Report for

summons

YesYes

PossiblyPossibly

NoNo

The level of ANPR intercept activity is broadly a combination of the number

of ANPR intercept teams operating and the size of these teams. This section

of the evaluation looks at issues relating to the resourcing of the ANPR

intercept teams, what skills and support these teams had and how they

decided on where to deploy their resources [Section 4.1].

During the course of the 13 month Laser 2 project, total staff input was

368,446 hours – this equates to 192 Full Time Equivalents (FTEs), the

majority of whom were police constables. By the start of the second year

of Laser 2 there were approximately 515 working as part of ANPR teams

[Section 4.1].

The force returns showed that the majority of ANPR officers’ time (77%)

was spent either on intercept duties or travelling to and from intercept

duties. This is significantly higher than a ‘typical’ police officer – a Home

Office report identified that on average a typical police officer spends only

57% of their time away from their police station. Further, ANPR intercept

officers, whether travelling to and from intercept sites or undertaking

intercepts, are highly visible and can respond to incidents as and when

they occur. Accepting that there are clear differences between the work

undertaken by ANPR intercept officers and conventional policing, the

operation of ANPR-enabled intercept teams provides for an extremely

visible form of policing [Section 4.1].

Findings: Operational factors

All forces adopted similar staffing structures for their intercept teams –

a core team of intercept officers, lead by supervisor/team leader with

appropriate back office support. There was, however, significant variation

in the number of intercept officers operating per team, reflecting different

operational practices. Forces also recognised the need to appropriately

support the operations of ANPR intercept teams, in particular of providing

good central support and intelligence [Section 4.2].

A key aspect to the successful exploitation of ANPR intercept teams was

senior officer commitment – this ensured that resources were available as

and when required and other officers across the force provided appropriate

intelligence for the ANPR teams to operate [Section 4.2].

4.1 ANPR staff inputs

4.1.1 ANPR Staff

The level of ANPR intercept activity is broadly a combination of the number of

ANPR intercept teams operating and the size of these teams. During Laser 2,

the number of teams and staff assigned to these teams was not static, it changed

according to local operational conditions and needs and evolved work practices.

In particular, for those forces that were new to ANPR, it took some time to

recruit and train the teams and set in place the processes to support Laser 2.

Going into the second financial year of Laser 2 in April 2004 (ten months into

Laser 2 and a relatively stable state), there were approximately 450 front-line

staff and 70 back office support staff assigned to 57 teams across the 23

forces, as shown in Figure 4.1.

Figure 4.1 shows that some forces operated quite a large number of teams

but involved relatively few officers in each of them (Cleveland had dedicated

an average of 7 ANPR staff per intercept team) while others operated relatively

few teams, but involved large numbers of officers in these teams (eg the West

Midlands with 17 ANPR staff per intercept team), reflecting different operational

tactics and local staffing issues.

Figure 4.1: Number of ANPR teams and staff operating as part of these teams

4.1.2 Resources directly associated with ANPR activity

To understand the total resource devoted to ANPR and how intercept teams

were spending their time, each force was asked to provide an estimate of the

hours spent on the project per week by grade (Inspector, Sergeant, Constable

or civilian staff) for each week of the pilot under a number of categories:

Force ANPR teams Front-line Support Total Staff Persons per staff

Avon & Somerset 3 27 1 28 9.3

Cambridgeshire 2 18 2 20 10.0

Cheshire 1 14 5 19 19.0

City of London 1 7 3 10 10.0

Cleveland 1 7 2 9 9.0

Greater Manchester 3 21 1.5 22.5 7.5

Hampshire 1 24 2 26 26.0

Hertfordshire 3 22 8 30 10.0

Kent 2 14 3 17 8.5

Lancashire 6 40 3 43 7.2

Leicestershire 2 18 2 20 10.0

Lincolnshire 2 16 6 22 11.0

Merseyside 1 8 1 9 9.0

Metropolitan 10 41 9 50 5.0

North Wales 2 21 6 27 13.5

North Yorkshire 1 8 1 9 9.0

Northamptonshire 4 31 2 33 8.3

Northumbria 1 6 2 8 8.0

Nottinghamshire 1 10 3 13 13.0

Staffordshire 2 20 1 21 10.5

Warwickshire 5 27 1 28 5.6

West Midlands 2 33 3 36 18.0

West Yorkshire 1 12 2 14 14.0

Total / Average 57 445 69.5 514.5 9.0

Note: As supplied by forces as part of their operational cases for April 2004.

• intercept duties – time when officer is on intercept duty. This includes all

those involved in ANPR intercept, eg intercept officers, mobile ANPR

unit/CCTV operators, roadside interviewers, back-office PNC checkers

(such as force intelligence officers) etc

• non-intercept duties – time out on ANPR deployment but not on, or ready

for, intercept duties (eg travelling time to location of operation and break times)

• prisoner handling time – time from point of arrest at the roadside to point of

booking in or handing over to another officer

• administration – time spent on data input and other ANPR-related

administration such as setting up the operation, databases, systems and

resource allocation.

This is not an exhaustive list of the tasks that ANPR staff were involved in.

For example other non-ANPR duties that officers undertook but which were

not incorporated in the above list included:

• general administration

• case preparation/court attendance

• training and leave

• emergency and special operations.

This list, however, does provide some insight into the level of effort that was

required for the different key tasks as well as giving an indication of the kinds

of overheads that were part of an ANPR operation. In week 24 of the pilot,

forces were also asked to record the number of hours of intercept support

that they received from units other than those dedicated to ANPR operations.

This included Armed Response Vehicle (ARV) support and dog handlers.

Finding 12. During the course of the 13 month Laser 2 project, total staff

input was 368,446 hours – this equates to 192 Full Time Equivalents

(FTEs) on the basis of 1,920 operational hours per annum.

This figure is significantly less than the 514.5 quoted in Figure 4.1 above

due to abstractions (ANPR intercept officers and staff undertaking duties not

directly associated with Laser 2). Throughout this report, FTE has been used

as a baseline measure of input. This approach enables us to look at the

relative productivity of forces’ ANPR intercept teams during their active periods.

4.1.3 Resources directly associated with ANPR activity, over time

During Laser 2, ANPR operations took place on 5,010 days across the 23

forces – this equates to approximately 202 days per force per annum during

which an average of 8.3 FTEs were deployed per operational day per force.

Figure 4.2 shows the average number of days the ANPR intercept teams were

working per force area and the total hours worked per week.

Figure 4.2: Number of days ANPR teams were operational/total hours worked by week

Figure 4.2 shows:

• the average number of days worked and hours worked per week were

closely related

• that there was a gradual ramp up of resources in the first month of the pilot

• a significant drop in the average number of days worked during the

Christmas period (weeks 30 and 31) and also towards the end of March –

weeks 40-44. This corresponds with the end of the police leave year and

school Easter holidays

• a fall in the number of hours deployed towards the end of the evaluation.

This also occurred in Laser 1 and was due in part to forces not submitting

data by the end of the evaluation period

• overall, the number of days worked per team was approximately four days

per week – this was broadly similar to Laser 1 during the pilot.

Week

14 464340373431282523191613107 49 52

Days worked Total hours worked ('000)

4.5

3.5

2.5

1.5

0.5

Days worked

Hours worked

4.1.4 Staffing of ANPR teams

Figure 4.3 shows the proportion of time spent on ANPR duties during Laser 2.

The most significant staff input to the pilot by grade was by constables (82%

of resource input and 300,581 hours). These findings are broadly similar to

Laser 1, where 84% of resource input was from Police Constables.

Figure 4.3: Percentage time spent by staff grade

Finding 13. On the basis of the information provided by forces, the average

force staffing for ANPR operations was 0.92 FTE inspectors/sergeants,

6.8 FTE constables and 0.6 FTE civilian staff. On the basis of standard

annualised running costs (including staff overhead costs), the cost of

staffing the pilot was approximately £6.7m over the 13 month period.

4.1.5 Activities undertaken by the ANPR teams

In total, 197,554 hours were spent during the 13 month Laser 2 project

on active intercept duty with an average of 42 hours of intercept time per

operational day (86% of which was constable input, with the remainder

sergeant/inspector input). A further total of 87,717 hours was spent on non-

intercept duties (travelling to/from the intercept site and taking breaks).

This figure varies significantly force by force, with Avon and Somerset and

Lincolnshire spending over 40% of their deployment time travelling to and

from sites, while Cleveland only spent 17% of their deployment time travelling.

This reflects both geography and operational set-up. For example, Avon and

Somerset operate force-wide ANPR teams and therefore need to travel large

distances to some of their most productive areas. Other forces, for example

Cleveland, cover a much smaller area and consequently travel times are

much less.

82%

10%

Constable

Civilian

Inspector

Sergeant

A total of 22,731 hours was booked to Laser 2 to prisoner-handling. During this

time, there were 13,499 arrests by the intercept teams, ie each arrest required

approximately 1.7 hours of processing time (which would include travel time

back to the station).

However, some forces were able to deploy special prisoner handling units,

which were able to hold prisoners at the stop site until there were sufficient

numbers to drive them all to a custody station. For example in the West

Midlands, Operational Command Units (OCUs) bid for the ANPR intercept

teams to operate in their area. In return, they committed their OCU to providing

a minimum of 8 prisoner handlers and custody facilities – this appears to

improve the effectiveness of the ANPR team.

The administrative requirement (which included both civilian and officer time)

for the intercept teams was 60,443 hours – equivalent to 1.4 full time members

of staff per force, focused solely on the administrative duties of ANPR.

Figure 4.4 shows the majority of time was spent by intercept teams in

‘intercepting’, with the remainder being spent on travel to sites/breaks (24%),

administration (17%), and prisoner handling (6%). Civilian support (4% of

deployment effort) was primarily used in the control rooms or mobile ANPR

vans, to run checks on number plates and dispatch officers to intercept

specific vehicles.

Figure 4.4: Percentage time spent by operational area/staff grade

The force returns showed that the majority of ANPR officers’ time is spent

either on intercept duties or travelling to and from intercept duties (77%).

This is significantly higher than a ‘typical’ police officer – a Home Office report

identified that on average a typical police officer spends only 57% of their time

away from their police station. Further, ANPR intercept officers, whether

travelling to and from intercept sites or undertaking intercepts, are highly visible.

53%

24%

17%

Intercept hours

Non intercept hours

Booking in / handover

Administration

Finding 14. Accepting that there are clear differences between the work

undertaken by ANPR intercept officers and conventional policing, the

operation of ANPR-enabled intercept teams provides for an extremely

visible form of policing.

Figure 4.5: Percentage time spent by operational area/staff grade

In terms of activity, the proportion of time spent on intercept duties will be a

factor in determining the number of arrests and actions taken by the intercept

officers. Figure 4.6 shows the average proportion of time spent on intercept

duties across the 23 forces.

100%

80%

60%

40%

20%

Intercept

Non-intercept

Prisoner handling

Administration

Avon and Somerset

Cambridgeshire

Cheshire

City of London

Cleveland

GMP

Hampshire

Hertfordshire

Kent

Lancashire

Leicestershire

Lincolnshire

Merseyside

Metropolitan

North Wales

North Yorkshire

Northamptonshire

Northumbria

Nottinghamshire

Staffordshire

Warwickshire

West Midlands

West Yorkshire

Diary of a Police Officer, PA Consulting Group (2001)

Figure 4.6: Proportion of time spent on intercept duties by week

Figure 4.6 also shows the most and least time spent on intercept duties of any

the 23 pilot forces for each week. No force spent consistently the most or least

time on intercept duties. Overall, the average time spent intercepting stayed

broadly consistent through Laser 2.

Finding 15. In any week, the difference between the most and least time

spent on intercept duties varied considerably (at times over 75%) and

was considerably higher than Laser 1 (where there was relatively little

variation). On average, however, the proportion of time spent on intercept

duties stayed broadly the same throughout the pilot and was similar

across the 23 forces.

Week

14 4643403734312825231916131074952

% time spent on intercept duties

Highest

Average of 23 forces

Lowest

100

4.2 Team capabilities and support

4.2.1 ANPR team make-up

In terms of staff capabilities, ANPR supervisors/managers recognised the need

for a wide skills base within intercept teams:

“Ideally an ANPR team should be made up of officers with a

broad range of skills. Traditionally ANPR has been located

within a Roads Policing environment because the interception

of vehicles requires tactical stopping and pursuit expertise,

together with higher than average standard of police driving

qualifications. However, the nature of the interaction with the

vehicle occupants, once stopped, requires a high level of

investigative training and experience. To this end it is useful if

the team is also made up of officers with a history of criminal

investigation, intelligence handling or search training. This

variety adds to the efficiency and effectiveness of ANPR as a

tactical tool and enhances the operational performance.”

Typically, an ANPR intercept team consists of the following personnel:

• a team supervisor, normally a sergeant, who would manage the ANPR team

and co-ordinate with other ANPR teams and other officers within their force.

They also have a role in liaising with other ANPR forces (both Laser and non-

Laser) in the discussion and adoption of good practice. Supervisor skills include:

– leadership and communication skills, including drive and motivation

– operational and strategic planning

– good knowledge and experience of police powers, roads policing, crime

and general policing duties.

• a team of between four and six experienced constables. Their role would be

to stop vehicles either identified by the ANPR unit or through observation, eg

to identify those vehicles where a driver was using a mobile telephone or not

wearing a seatbelt. They would operate in a highly visible, overt fashion.

Their skills base would include:

– team working

– proactive decision making

– trained in investigative interviewing

– qualified standard/advanced driving

– trained in drugs recognition

– trained in field impairment

– trained in safe search techniques.

Nick Purdie, Northamptonshire Police

• an analyst (in most cases civilian), who would be responsible for providing

back office support to the ANPR team, collating data on the number of ANPR

hits, actions taken and the results, helping prepare files for prosecution, and

collating Fixed Penalty paperwork for transfer to the central ticket office.

This role would allow the intercept officers to maximise their time spent

intercepting. Analyst skills would include:

– database management experience/training

– decision making

– ability to multi-task

– local knowledge

– control room experience.

Finding 16. All forces adopted a similar staffing structure for their

intercept teams – a core team of intercept officers, supported by

administrative support and led by supervisor/team leader. There was,

however, significant variation in the number of intercept officers

operating per team, reflecting different operational practices.

Examples of operational deployments are described in the following case studies.

Case study 1: ANPR deployment by West Midlands Police

The West Midlands ANPR team typically operates using stop sites staffed

by large numbers of people (20+), a strategy that has served the force well

in the past. While this requires considerable commitment from OCUs, good

results (and internal publicity in relation to these results) have helped

increase demand for the centralised ANPR resource. In return for an

OCU’s commitment to providing good crime analysis, a clear objective for

each check, a minimum of eight prisoner handlers and custody facilities,

the ANPR team supplies a sergeant, eight motorcycle-based intercept

officers, a double-crewed pursuit traffic car, an ANPR vehicle and operator,

a communications vehicle including Force Linked Intelligence System

(FLINTS) and a site manager. They also use traffic wardens to issue FPNs

where this has been identified by the intercept officer as appropriate.

In terms of maintaining morale, checks are for a short, set period – this

proved both popular and productive. Further intercept officers are regularly

rotated between various tasks on site. As operations are carried out across

the 21 OCUs, they only have to provide arrest teams on average once per

month, and these are invariably different staff on each occasion.

Case study 2: ANPR deployment by Lincolnshire Police

Lincolnshire Police has two intercept teams dedicated to ANPR, with a

sergeant and six officers on each team. All of the ANPR officers originate

from the Roads Policing Unit, all are pursuit trained and the majority are

also motorcyclists.

The teams are, on the whole, self-sufficient. They are deployed in accordance

with the National Intelligence Model to areas selected from the BCU Tasking

and Co-ordinating Group (T&CG) meetings. Where possible they make use

of Divisional Prisoner Handling Units whenever arrests are made, but often

the arresting officer has to interview and process the prisoners themselves.

This is obviously detrimental to performance.

4.2.2 Supporting ANPR teams

In addition, most ANPR intercept teams (17 of the 23) received some form

of additional support from their force at some point during the pilot. In total,

12,000 hours were recorded as support. This is on average 42 hours of support

per week (or 32% of the dedicated team’s intercept time). This support included:

• controllers – to provide ANPR teams with checks on people and vehicles

stopped (PNC/local intelligence etc), and to deploy team members to

ANPR activations

• ANPR technicians – to offer IT support and expertise to operational teams

and to develop future IT solutions

• database managers – to develop, collate, analyse and manage intelligence

sources and to create, audit and manage information databases. To act as a

liaison point with other departments, forces and agencies. To share intelligence,

information and target profiles (eg people, vehicles and locations)

• analysts – to collate and analyse crime patterns and intelligence for ANPR

use and to inform the operational deployment of ANPR teams

• media/publicity personnel – to manage media interest. This was particularly

relevant given the high profile nature of ANPR intercept teams.

Finding 17. Forces recognised the need to support appropriately the

operations of ANPR intercept teams, in particular to provide good central

support and intelligence.

4.2.3 Senior officer commitment

As well as resources, many ANPR project managers recognised that the level

of support received from Chief Officers was vital:

“The experience of the Police Standards Unit has emphasised

the importance of ACPO involvement in the strategic direction

of ANPR at force level. It is very easy for forces not to support

ANPR in the face of the competing demands of reactive policing

and so fail to deliver the potential increases in performance that

ANPR has shown that it can produce. Active leadership by an

ACPO officer, driving delivery through a performance culture and

the adoption of a project structure, involving the major in-force

stakeholders, continues to be one of the most efficient and

effective ways of maximising the performance potential of ANPR.”

Examples of instances where there was direct intervention at Chief Officer

level include:

• Northamptonshire, where Chief Officer support led to development of and

support from partnerships with the local council (the CCTV set-up)

• West Midlands where support came in the form of BCU aid at ANPR stops

• North Wales where the support of the Chief Constable was seen as key in

driving the team.

Finding 18. Senior officer commitment to the programme was seen as

critically important.

4.3 Location deployment

4.3.1 Means of deployment

The most common deployment method was the mobile ANPR vehicles

(both vans and in-car systems) with intercept support provided by a combination

of motorcycles and cars. While motorcycles provide a valuable high-speed

response in congested urban areas, they cannot work without car support

(for example, to transport prisoners).

Furthermore, health and safety assessments have shown that motorcycles are

more vulnerable and ACPO guidelines prohibit them from engaging in pursuits.

Finally, they are less popular with intercept officers in poor weather conditions.

Alan Ford, PCU

4.3.2 Location of deployment

Another key aspect to successful ANPR operations was selecting the most

suitable locations, reflecting a range of competing factors, specifically:

• the maximisation of operational performance – for ANPR to be most effective,

it should be deployed in areas where the chance of encountering criminality

is most likely. However, a number of locations are used to ensure that criminals

do not become aware of where the ANPR units are likely to be deployed

• ensuring that officers operate in a safe manner and vehicles are stopped

within a controlled environment. The safety of the intercept team, the vehicle

occupants and the general public is paramount and it is vital that this be

taken into account prior to selecting a suitable stopping area.

In terms of good practice in selecting an appropriate location for a mobile

ANPR and intercept team, the following catalogues the key stages:

• a pattern analysis identifies hot spots of crime and when these are occurring,

for example housebreaking on the south side of the city in the mid-morning

• profiling is then carried out by relevant OCUs/BCUs to identify areas

where offenders are likely to be travelling from and, from this, the likely

routes they take

• the traffic intelligence officer liaises with local traffic units and beat officers

to identify potential ANPR intercept locations on likely routes. A database of

previously used locations is referred to, together with the results from previous

ANPR deployments. This allows productive locations to be easily identified.

In addition, new sites are visited and risk assessed for suitability. If needed,

traffic flows are measured in advance to assist with site intelligence. The ANPR

team have the final say on suitability of a site unless directed from level 2

tasking, as described in Section 2.2.1 above.

• prior to deployment, the local OCU/BCU is contacted to ensure that all

relevant intelligence databases are available and used and that all key

people know when the deployment is to take place (for example, to ensure

that there are no potential conflicts of operations and to prepare the prisoner

handling capability).

Finding 19. In discussions with ANPR managers, it was widely recognised

that a key success factor was the preparatory intelligence to ensure that

the ANPR team is sent to the most appropriate location at the right times

and, most importantly, have access to local intelligence databases.

This section of the evaluation presents analysis in relation to the 180,543

vehicle stops undertaken by the intercept teams during Laser 2 and shows:

• As a mechanism to read the registration marks of moving vehicles, the

ANPR systems used by the police proved extremely effective. They were

able to read approximately 28 million VRMs. Combined with intelligence

databases, over 1.1 million (3.9%) were identified as vehicles of interest to

the police [Section 5.1].

• The intercept teams stopped 9.2% of these vehicles of interest (101,775)

[Section 5.1].

• The intercept teams also stopped a further 78,768 vehicles as they passed

as a result of officer observations [Section 5.2].

• The most common reason (45% of stops) for stopping a vehicle on the

basis of an observation was that the vehicles or occupants looked

suspicious, followed by not displaying a VED licence (20%) and not

wearing a seatbelt (17%). This shows that in addition to addressing

criminality, intercept teams are also contributing to road safety and the

reduction of vehicle excise duty evasion [Section 5.2].

• As a proportion of all vehicle stops (44% over the duration of Laser 2),

observation stops were significantly higher than Laser 1 (where only 22%

came from observations) and increased during Laser 2. Possible reasons

for this include: [Section 5.3]

Findings: Vehicle stops

– decreased confidence in using the DVLA databases as the primary

means of stop

– the introduction of new fixed penalty notices which require observation-

based stops

– the setting of targets including for fixed penalties issued per officer

per week

• Proportionally, Asian drivers were more likely to get stopped from an

officer observation than an ANPR stop [Section 5.3].

5.1 ANPR, reads, hits and stops

During Laser 2, ANPR cameras read approximately 28 million VRMs of

which over 1.1 million (3.9%) resulted in a ‘hit’, that is a match with an

intelligence database.

While the evaluation did not look at the accuracy of these VRM reads,

feedback from the forces was that the ANPR systems were extremely accurate

– anecdotal evidence that fewer than 1 in 25 reads were incorrect. In practice,

ANPR controllers were able to confirm reads before officers intercepted hits –

this virtually eradicated stopping vehicles where the ANPR reader had misread

the VRM.

Finding 20. ANPR was seen to be an extremely effective means of

reading VRMs and, when combined with an ANPR controller confirming

VRMs before an intercept was requested, very few vehicles were

incorrectly stopped as a result of an ANPR misread.

Finding 21. During Laser 2, ANPR cameras read approximately 28 million

VRMs of which over 1.1 million (3.9%) resulted in a hit. In total, the ANPR

intercept teams stopped 101,775 vehicles (9.2%) of these vehicles as a

result of ANPR hits.

5.1.1 ANPR reads, hits and stops

Figure 5.1 shows the reads, hits and stops for each Laser 2 force.

Finding 22. Overall 3.9% of vehicles passing ANPR cameras were flagged

against the intelligence systems as of potential interest to the police.

Of these, the intercept teams stopped 9.2% – this meant that intercept

teams stopped approximately 1 in 275 of vehicles passing ANPR cameras

due to an ANPR hit. In Laser 1, the equivalent figure was 1 in 200 of

vehicles passing ANPR cameras due to an ANPR hit.

Figure 5.1: ANPR reads, hits and stops by Laser 2 force

Avon & Somerset 712,827 14,241 2.0% 4,488 31.5%

Cambridgeshire 1,258,102 79,027 6.3% 5,133 6.5%

Cheshire 1,663,304 56,851 3.4% 2,977 5.2%

City of London 137,757 5,536 4.0% 660 11.9%

Cleveland 476,284 19,133 4.0% 1,379 7.2%

Greater Manchester 949,743 33,108 3.5% 12,324 37.2%

Hampshire 1,007,023 23,431 2.3% 3,071 13.1%

Hertfordshire 894,107 15,265 1.7% 2,312 15.1%

Kent 3,715,374 92,484 2.5% 6,045 6.5%

Lancashire 1,358,883 30,208 2.2% 3,426 11.3%

Leicestershire 1,639,527 56,608 3.5% 7,011 12.4%

Lincolnshire 368,224 12,266 3.3% 7,527 61.4%

Merseyside 281,420 12,179 4.3% 4,422 36.3%

Metropolitan 281,420 12,179 4.0% 11,317 24.0%

North Wales 1,075,297 48,441 4.5% 4,277 8.8%

North Yorkshire 947,044 33,156 3.5% 2,140 6.5%

Northamptonshire 5,037,081 317,200 6.3% 4,633 1.5%

Northumbria 778,804 30,704 3.9% 3,203 10.4%

Nottinghamshire 488,029 24,004 4.9% 1,060 4.4%

Staffordshire 1,621,286 75,163 4.6% 2,662 3.5%

Warwickshire 1,237,070 27,790 2.2% 2,907 10.5%

West Midlands 684,086 18,304 2.7% 3,638 19.9%

West Yorkshire 745,771 39,586 5.3% 5,168 13.1%

Total / Average 28,262,367 1,111,752 3.9% 101,775 9.2%

% of reads

generating Vehicles % of hits

Force VRM reads VRM hits hits stopped stopped

Finding 23. The results from the forces showed a wide variation between

the proportion of ANPR hits that were stopped, ranging from 1.5%

(Northamptonshire) to 61.4% (Lincolnshire) – this reflects deployment

tactics. For example Northamptonshire Police cover a wide area of road

network through CCTV, while Lincolnshire only used mobile ANPR units

and in-car systems.

VRM reads, hits and stops may include the same vehicle on a number of

times, for example where a vehicle passes ANPR readers on many occasions

(especially in areas where there are widespread ANPR-enabled CCTV systems).

While forces collect data from the ANPR systems (for example the video feed)

which could be used to analyse the number of times individual VRMs were

read over the 13 month period (and indeed where they were read), this data is

not collated centrally within a back office facility. No analysis could therefore be

undertaken of the number of unique reads and hits from ANPR cameras.

Finding 24. If it is assumed that all vehicles have an equal chance of

passing an ANPR camera, then approximately 1 in 25 vehicles on the

road are of potential interest to the police. However, the lack of collation

of all VRM read information across forces means that this estimate

cannot be validated.

Overall the percentage of VRM reads that lead to a hit was lower than that

achieved during Laser 1 – 3.9% in Laser 2 as opposed to 4.6% in Laser 1.

This can be attributed to some forces’ lack of confidence in DVLA’s databases

(which generated the majority of the hits), resulting in the databases not being

used as the primary trigger for ANPR intercepts.

Similarly the percentage of vehicle hits stopped by the ANPR intercept teams

was lower than that achieved during Laser 1 – 9.2% in Laser 2 as opposed to

12.7%. As will be seen below, the number of stops per hour was also less.

However feedback from forces suggested that ANPR teams spent little time

waiting for hits (dead time) and most of their time investigating vehicle hits.

This suggests that in Laser 2 officers were able to stop proportionally fewer of

those vehicles of interest because they were spending more time per vehicle.

5.1.2 ANPR deployment method

The ANPR stops came from a variety of triggers. Overall, 59.7% of 101,775

ANPR stops were generated by mobile ANPR units, 34.0% from in-car

systems and 6.3% from CCTV systems. Figure 5.2 shows the ANPR means of

deployment for those hits that resulted in a stop for each of the Laser 2 forces.

Figure 5.2 ANPR stops by means of deployment

Finding 25. There was no relationship between the method of deploying

ANPR cameras (CCTV, mobile unit and in-car system) and the total

number of ANPR stops achieved. The volume of stops was dependent on

other factors, in particular staffing.

City of London

Nottinghamshire

Cleveland

North Yorkshire

Hertfordshire

Staffordshire

Warwickshire

Cheshire

Hampshire

Northumbria

Lancashire

West Midlands

North Wales

Merseyside

Avon and Somerset

Northants

Cambridgeshire

West Yorkshire

Kent

Leicestershire

Lincolnshire

Metropolitan

Greater Manchester

Mobile ANPR unit

In-car ANPR system

CCTV / fixed site ANPR system

0 10,0008,0006,0004,0002,000 12,000

ANPR stops

5.1.3 Triggering database

During the pilot, a total of 101,775 vehicles were stopped as a result of ANPR

hits, ie matches against intelligence databases. The source of the hits is shown

in Figure 5.3 together with the 78,768 vehicle stops that were not triggered by

an ANPR hit, but by an observation made by the intercept officer.

Figure 5.3: Vehicles stops generated by database

It should be noted that on a few occasions (about 9.7% of hits), a vehicle was

stopped following hits from more than one database, for example a vehicle

appeared on both PNC and DVLA’s VED database – a similar level to that

recorded during Laser 1.

Overall the two DVLA databases (no current Vehicle Excise Duty and No current

keeper details) accounted for 70% of ANPR hits that led to stops. This is a

lower proportion than that recorded during Laser 1 (75%), reflecting continued

concerns within ANPR intercept teams about the DVLA database quality.

5.2 Observation-generated

Finding 26. As in Laser 1, intercept teams did not rely entirely on ANPR

technologies for identifying vehicles to stop – the intercept teams also

stopped vehicles as they passed as a result of officer observations.

This led to an additional 78,768 vehicle stops that did not originate from

ANPR hits, ie 44% of all stops made by the intercept teams.

No information was collated on the triggering database for those 1,009,977 vehicles that

generated a hit but were not stopped.

Note that a vehicle can give rise to multiple hits, eg on PNC and DVLA No VED

80,000

70,000

60,000

50,000

40,000

30,000

20,000

10,000

56,053

22,370

14,547

18,668

78,768

DVLA:

No VED

DVLA:

No keeper

PNC marker Other (Observations)

This figure is significantly higher than Laser 1, where only 8,577 of the 39,188

(22%) of the vehicle stops were the result of officer observation. It was not

evident that there was any single reason for the increased proportion of

observation stops, however possible reasons include:

• decreased confidence in using the DVLA databases as the primary means

of stop – a number of forces used the DVLA databases in combination with

other data sources before deciding to stop vehicles. Hence, while there were

broadly similar levels of vehicle hits to Laser 2, the actual number of these

hits that were stopped was significantly lower (9.2% of hits stops in Laser 2

as opposed to 12.7% of hits stopped in Laser 1)

• the introduction of new fixed penalty notices requires observation-based

stops, for example driving while using a mobile telephone can only be

observed by an officer

• many forces set targets for arrests and fixed penalties issued per officer per

week. While this helped to ensure that performance was delivered, it meant

that greater emphasis was placed on keeping officers busy at all times rather

than waiting for ‘good quality’ ANPR hits.

Finding 27. During Laser 2, the use of observation stops as a method

of engaging criminality increased significantly at the expense of ANPR-

generated stops. This reflects on the quality of underlying intelligence

databases, which is discussed below.

Figure 5.4 shows the number of observation stops per force and the reason for

the stop which includes multiple reasons, for example where the driver is using

a mobile telephone and not wearing a seatbelt. Key points to note are:

• the largest single reason (44.8% of stops) for stopping a vehicle on the basis

of an observation was ‘other’ – primarily, vehicles or occupants that looked

suspicious but were not known to the police. The equivalent figure for Laser 1

was 49.7%

• the next largest category related to failing to display a valid VED (20.4% of

observation stops). While DVLA’s no VED database was one that all forces

used as an ANPR trigger, DVLA’s database excluded those vehicles with tax

that had expired in the last two months. Further, DVLA’s database also

excluded those vehicles that were taxed but were not displaying their tax disc

(an offence). Intercept officers were thus able to stop these vehicles on the

basis of observation. These observations were less common than they were

in Laser 1 (29.4%), despite the cost recovery element of the pilot that allowed

forces to hypothecate revenue from the offence of failing to display a VED

• failing to wear seatbelt observations saw the most dramatic increases compared

to Laser 1 – they rose from 6.0% of observations in Laser 1 to 17.0% in Laser 2

Figure 5.4: Reason for observation stops by force (percentages)

5.3 All Vehicle stops (ANPR and observations)

5.3.1 Vehicle stops by force

During the Laser 2, a total of 180,543 vehicles were stopped, 56% as a result

of ANPR triggers and 44% as a result of observation. Figure 5.5 shows the

total volume of stops by force.

Avon & Somerset 1.9% 5.1% 17.2% 9.0% 9.3% 2.6% 54.9% 2,867

Cambridgeshire 2.8% 19.2% 14.1% 3.2% 5.2% 1.7% 53.8% 4,153

Cheshire 1.1% 4.2% 19.7% 5.5% 4.3% 0.6% 64.6% 2,134

City of London 5.2% 28.5% 10.6% 8.3% 4.5% 0.0% 2.9% 902

Cleveland 7.2% 19.5% 19.6% 7.9% 7.1% 10.3% 28.4% 2,068

Greater Manchester 2.6% 0.8% 6.0% 4.1% .7% 0.2% 58.8% 3,463

Hampshire 1.9% 10.5% 32.4% 7.0% 6.6% 0.9% 40.5% 3,369

Hertfordshire 4.3% 28.5% 37.4% 4.6% 4.5% 2.1% 18.7% 3,328

Kent 2.8% 9.7% 8.6% 5.0% 3.6% 1.4% 68.8% 4,098

Lancashire 3.1% 27.6% 14.8% 7.7% 5.2% 3.2% 38.5% 8,100

Leicestershire 3.9% 9.4% 30.7% 3.8% 9.7% 1.3% 41.3% 2,837

Lincolnshire 0.9% 37.9% 4.6% 12.6% 2.5% 0.9% 40.6% 4,007

Merseyside 1.5% 4.8% 27.3% 2.7% 5.2% 1.1% 57.3% 2,335

Metropolitan 2.3% 14.7% 23.4% 8.7% 7.9% 1.6% 41.4% 6,717

North Wales 1.0% 17.5% 13.6% 5.5% 2.7% 0.8% 58.8% 6,728

North Yorkshire 2.7% 28.5% 10.8% 11.4% 6.6% 1.3% 38.6% 2,358

Northamptonshire 0.7% 2.3% 21.8% 1.9% 23.7% 14.0% 35.5% 2,365

Northumbria 2.2% 2.0% 3.9% 2.4% 9.2% 0.8% 79.6% 819

Nottinghamshire 2.9% 19.6% 4.9% 10.2% 20.7% 2.6% 39.2% 1,617

Staffordshire 10.3% 19.1% 20.9% 2.6% 25.5% 3.3% 18.2% 3,110

Warwickshire 1.7% 5.8% 20.3% 8.9% 5.8% 2.5% 55.1% 1,821

West Midlands 1.0% 21.7% 41.9% 1.2% 0.5% 0.0% 33.6% 5,109

West Yorkshire 2.1% 7.8% 37.5% 3.4% 4.9% 0.3% 44.0% 3,468

Total / Average 2.7% 17.0% 20.4% 6.0% 7.0% 2.1% 44.8% 78,768

Vehicle Known

Mobile No excise Vehicle Driving person / Total

Force phone seatbelt duty defect manner vehicle Other stops

The Metropolitan Police Service stopped the most vehicles while the City of

London stopped the fewest. Greater Manchester Police stopped the largest

number of vehicles following an ANPR trigger (78%), while as a proportion of

their total stops, City of London (74%) and Lancashire (70%) stopped the most

vehicles as a result of an observation.

Figure 5.5: Total volume of ANPR and observation stops

5.3.2 Vehicle stops over time

The balance of ANPR and observation-generated stops changed during the

pilot. Figure 5.6 shows that observations became a more prominent reason for

stopping in the later stage of the project, rising from 35% of stops in week 1 to

48% 12 months later. It is worth noting that the change is due to a decrease in

the number of ANPR stops rather than to any significant increase in

observation stops.

City of London

Nottinghamshire

Cleveland

Northumbria

North Yorkshire

Warwickshire

Cheshire

Hertfordshire

Staffordshire

Hampshire

Merseyside

Northamptonshire

Avon and Somerset

West Yorkshire

West Midlands

Cambridgeshire

Leicestershire

Kent

North Wales

Lancashire

Lincolnshire

GMP

Metropolitan

ANPR-generated stops

Observation-generated stops

02 1614121086418

Stops ('000)

Late February to the end of March saw an increase in the number of ANPR

stops. This corresponded to a period where the DVLA identified quality issues

with their databases that led to their suspension for a month. In practice this

meant that intercept teams were using out of date information. While this meant

that more vehicles were being stopped, more of these were false positive hits,

ie they were being incorrectly stopped. The quality of the databases is discussed

in more detail in section 6 of this report.

Finding 28. The greater use of observation-generated stops during

Laser 2 was a gradual development – during the first twenty weeks ANPR

methods contributed between 60-70% of vehicle stops. In the last twenty

weeks, this had fallen to between 50-60% of vehicle stops. The change in

the balance of stops is primarily down to a decrease in the number of

ANPR generated stops. This finding is counter to the drive towards

intelligence-led policing.

Figure 5.6: Percentage of stops coming from ANPR and observation by week

5.3.3 Vehicle stops by ethnicity of driver

Officers were required to fill in a form recording the ethnicity of the driver of each

vehicle stopped. Because this was done by means of questioning, drivers were

entitled not to state their ethnicity, (‘not stated’). However it was evident that in

some cases the ‘not stated’ category equated to ‘unknown’. Figure 5.7 shows

the ethnicity of the vehicle driver for the 180,543 vehicle stops during Laser 2.

Number of stops per week

Observations

ANPR

500

1,000

1,500

2,000

2,500

3,000

1 3 5 7 9 1113151719212325272931333537394143454749515355

Week

Figure 5.7: ANPR stops by ethnicity of driver – as reported by forces

In terms of benchmarking data, there is no information available on the

ethnicity of drivers on the road. However given that ANPR does not

discriminate on the basis of ethnicity, it could be argued that ethnicity

information from ANPR stops provides a surrogate baseline. On this basis,

Figure 5.6 shows that drivers self-classified as white and Asian were more

likely to be stopped as a result of observations than by ANPR.

Finding 29. The ethnicity data shows a significant difference between

observations and ANPR stops for persons self-reported as White and Asian.

5.3.4 Vehicle stops by time spent intercepting

The volume of stops (both ANPR and observation) is primarily a function of

the amount of time that teams spend intercepting. To take account of this,

Figure 5.8 shows the number of stops per intercept hour during Laser 2.

Figure 5.8: Vehicle stops per intercept hour deployed by week

Week

Stops per hour

1.4

1.2

1.0

0.8

0.6

0.4

0.2

555351494745434139373533312927252321191715131197531

Observation stops 77.4% 8.5% 5.9% 1.1% 0.8% 6.3% 100.0%

ANPR stops 71.5% 6.3% 6.9% 1.0% 0.9% 13.5% 100.0%

Total stops 133,776 13,033 11,661 1,842 1,500 18,731 180,543

Unknown /

White Asian Black Other Mixed Missing Total

Overall the number of stops per intercept hour was stable at just under one

stop throughout the 56 weeks (180,543 stops in 197,554 intercept hours

equates to 0.91 stops per intercept hour). This may seem counter-intuitive –

with more experience it might be expected that officers would become more

effective at stopping vehicles and the average number of vehicles stopped per

hour would increase. However from the beginning of the pilot, officers recognised

that it was not the quantity of vehicle stops that was key, rather it was the

quality of questioning and searching (where appropriate) that was crucial.

Discussion with intercept officers suggested that the majority of teams did not

have to wait long for a ‘hit’. The key to an effective operation was to identify

vehicle hits that were most likely to lead to arrests. Once vehicles were stopped,

the majority of an officer’s time was spent investigating the hit. This is

consistent with the fact that less than 10% of hits were actually stopped.

Finding 30. On average, just under one vehicle was stopped per officer

hour intercepting – this level of performance was maintained throughout

the pilot. In overall terms, 180,543 vehicles were stopped during 368,446

staff hours (including administration, prisoner handling and civilian time)

– this equates to one vehicle stopped for every two hours staff input.

Feedback from the field suggested that ANPR officers in intercept duty

spent little time waiting for hits (dead time) and most of their time

investigating vehicle hits – this is supported by the fact that less than

10% of vehicles that registered an ANPR hit were actually stopped.

Figure 5.9 lists the vehicle stops per intercept hour deployed by force and

shows clear differences in the stops per hour by force. For example Lincolnshire

averaged 1.78 vehicle stops per hour, while the City of London 0.4 stops

per hour.

Finding 31. Analysis of vehicle stops per hour by force does not appear

to be a strong indicator of performance – this reflects the different local

conditions, operational objectives and stages of development at which

ANPR intercept teams were operating.

Figure 5.9: Vehicle stops per intercept hour deployed by force

5.3.5 Vehicle stops by time and day

During the pilot, intercept officers recorded information on when and where

they stopped each vehicle. Analysis of this information indicates that the most

common day for stopping vehicles was Wednesday, with fewest vehicles being

stopped on Saturdays and Sundays. Figure 5.10 shows the proportion of stops

that were ANPR and observation-generated – overall there was no significant

difference in the mix for stop by day of week (ie observations/ANPR-generated).

Avon & Somerset 7,354 7,663 0.96

Cambridgeshire 9,286 10,595 0.88

Cheshire 5,111 8,025 0.64

City of London 2,562 6,483 0.40

Cleveland 3,446 5,361 0.64

Greater Manchester 15,787 11,766 1.34

Hampshire 6,439 8,611 0.75

Hertfordshire 5,640 4,912 1.15

Kent 10,143 7,095 1.43

Lancashire 11,525 15,674 0.74

Leicestershire 9,848 9,078 1.08

Lincolnshire 11,533 6,497 1.78

Merseyside 6,756 5,011 1.35

Metropolitan 18,034 16,757 1.08

North Wales 11,005 8,006 1.37

North Yorkshire 4,498 6,412 0.70

Northamptonshire 6,998 12,220 0.57

Northumbria 4,021 7,181 0.56

Nottinghamshire 2,676 4,562 0.59

Staffordshire 5,771 8,427 0.68

Warwickshire 4,727 7,411 0.64

West Midlands 8,747 11,360 0.77

West Yorkshire 8,636 8,448 1.02

Total / Average 180,543 197,554 0.91

Force Vehicle stops Intercept hours Stops per hour

Figure 5.10: ANPR and observation-generated stops by day of week

Analysis of stops by time of day, shows that nearly 40% of vehicles stops

took place between 10:00-12:00 and 13:00-15:00. This reflects the fact that

forces worked primarily during business hours. Some forces did experiment

with working later in the evening or at night. The effectiveness of operating at

different times of day will be discussed more fully in a later section of the report.

Figure 5.11 shows the number of stops that were ANPR and observation-

generated – overall there was no significant difference in the basis for stop by

time. Stops were most likely during daylight hours (60% of stops at 6pm were

ANPR-generated) and least likely between 11pm and 4am (when only 33% of

stops were ANPR-generated).

Figure 5.11: ANPR and observation-generated stops by time of day

Sunday FridayThursdayWednesdayTuesdayMonday Saturday

40,000

35,000

30,000

25,000

20,000

15,000

10,000

5,000

Vehicle stops

Day of the week

Observation-generated stop

ANPR-generated stop

data

40,000

35,000

30,000

25,000

20,000

15,000

10,000

5,000

Vehicle stops

Time of day

Observation-generated stop

ANPR-generated stop

2018161412108642 2422

5.3.6 Vehicle stops by estimated year of registration

The VRM of each vehicle stopped by officers was recorded. From this it was

possible to estimate when a vehicle was first registered, accepting that in a

small number of cases a VRM may be transferred from an older to newer

vehicle. Cars registered in 2000 were stopped most frequently – these stops

originated from hits against DVLA’s No current keeper database. Figure 5.12

shows the number of vehicle ANPR and observation-generated stops by the

(estimated) year of vehicle registration.

Figure 5.12: ANPR and observation-generated stops by year of vehicle registration

Finding 32. Overall, newer vehicles were more likely to be stopped

through an ANPR match, while older cars were more likely to be stopped

as a result of officer observation.

5.3.7 Vehicle stops by location

As part of the pilot, forces were asked to record the postcode district (eg BS5 for

Bristol) in which they deployed their ANPR intercept teams. This was not always

captured (only 80% of the stops included a postcode and two forces recorded

no valid postcode district information), however it does provide some insight

into the locations where the different forces were deployed. Overall, 15% of all

vehicle stops happened in 21 postcode districts, ie within a relatively small area.

Figure 5.13 lists:

• the number of different postcode districts recorded by forces at which vehicle

stops took place. While forces may have visited more than one location in a

particular postcode district, it is a useful surrogate as to the approximate

concentration of ANPR operations

16,000

14,000

12,000

10,000

8,000

6,000

4,000

2,000

Vehicle stops

Year of registration

Observation-generated stop

ANPR-generated stop

No Data

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

• the single postcode district leading to the most number of vehicle stops

• the number of vehicle stops within that postcode district

• the percentage of all force stops in that postcode district.

Figure 5.13: ANPR and observation-generated stops by year of vehicle registration

While City of London deployed ANPR intercept teams within only two postcode

districts, their force covers a total of only four postcode districts. At the other

extreme, the Metropolitan Police Service had very dispersed ANPR operations,

deployed in 160 different postcode districts.

Avon & Somerset 26 BS5 372 5.1%

Cambridgeshire 47 PE1 1,548 16.7%

Cheshire 30 WA8 1,472 28.8%

City of London 2 EC2 36 1.4%

Cleveland 27 TS5 353 10.2%

Greater Manchester 74 M14 1,521 9.6%

Hampshire 53 PO2 393 6.1%

Hertfordshire 45 SG1 1,841 32.6%

Kent 3 CT4 5,549 54.7%

Lancashire 42 FY3 1,373 11.9%

Leicestershire - - - -

Lincolnshire 29 PE25 1,794 1 5.6%

Merseyside 47 L36 912 13.5%

Metropolitan 160 SE1 584 3.2%

North Wales 45 LL18 3,132 28.5%

North Yorkshire 34 DL10 546 12.1%

Northamptonshire 9 NN1 1,389 19.8%

Northumbria 39 NE4 339 8.4%

Nottinghamshire - - - -

Staffordshire 31 ST1 1,383 24.0%

Warwickshire 5 CV31 969 2 0.5%

West Midlands 71 B20 633 7.2%

West Yorkshire 45 LS1 475 5.5%

Number of

different

postcodes

Postcode leading

to most stops

Stops at this

postcode

Stops at most

visited site as %

of all stops

This section of the evaluation sets out the actions that arose from the

intercepts. In total, including stops from both ANPR hits and observation,

there are records of the actions for 180,543 stops.

1. There were 13,499 people arrested, including: [Section 6.3]

• 2,263 arrests for theft and burglary

• 3,324 arrests for driving offences (for example driving whilst disqualified)·

• 1,107 arrests for drugs offences

• 1,386 arrests for auto crime (theft from and of vehicles)

• 55% of these already had a criminal record

2. Across Laser 2 forces, the average level of performance was around

91 arrests per FTE per annum (compare to a national ‘average’ of

around 10): [Section 6.3]

• as a targeted approach, ANPR teams are over nine times more effective

than conventional policing

• there were substantial differences in the number of arrests made per

FTE between forces. For example four police forces (Merseyside,

Nottinghamshire, North Wales and West Midlands) had over 140

arrests per FTE

Findings, action taken,

property recovered and

arrests made

• these differences between performance cannot be accounted for by

force type – better performing areas were characterised by strong

leadership, good intelligence and experienced officers

3. A large amount of goods, drugs and weapons were recovered,

including: [Section 6.2]

• 1,152 stolen vehicles were recovered (valued at over £7.5 million)

• drugs worth over £380,000 were seized from 740 vehicles

• stolen goods worth over £640,000 were recovered from 430 vehicles

266 offensive weapons and 13 firearms were seized

4. There was a strong correlation between vehicle documentation

offences and volume crime: [Section 6.3]

• 3,549 (26%) of arrests originated from vehicle stops for no VED or no

current keeper details

• section 8.4.3 shows that older vehicles are more likely to be guilty of

vehicle document offences. In this section we show that, per 100

vehicles stopped, more arrests are made from older vehicles

• section 5 showed that the majority of stops were conducted in normal

working hours whereas, in this section, we show that the conversion

rate increased in the evening. By changing deployment patterns,

conversion rates could improve further.

6.1 Possible actions taken at a stop site

For all the vehicle stops, the intercept officer kept a record of the actions that

were undertaken. These were:

• vehicle/person search

• recovery of property

• arrest

• reported for summons

• issuing a fixed penalty

• issuing a note requiring follow-up action – these include issuing a PG9,

HO/RT1, CLE2/6, CLE2/7, CLE2/8, V62 or VDRS

• intelligence log

• verbal advice.

Results from these actions are discussed in turn below:

6.2 Vehicle/person search

The breakdown of searches of vehicles, people, and the recovery of property

for each force can be seen in figure 6.1 overleaf.

Finding 33. Of the 180,543 vehicle stops, officers searched 4,402 (2.4%)

vehicles and 6,331 drivers or passengers at the roadside. As a result of

these searches, officers recovered 1,152 stolen vehicles (valued at over

£7.5 million), drugs on 740 occasions (with a street value of over £380,000),

stolen goods on 430 occasions (valued at over £640,000), 13 firearms and

266 offensive weapons.

There were significant differences in the value of recovered property and the

volume of items recovered across forces, reflecting the different number of

stops made by forces. Figure 6.2 overleaf shows the number of searches and

the recovery of property for each force per 100 vehicles stopped.

Finding 34. On average Laser 2 intercept officers:

• searched one out of every 41 vehicles stopped

• recovered one stolen vehicle for every 157 vehicles stopped

• recovered stolen goods from one stolen vehicle for every 420 vehicles

stopped

• found drugs in one in every 243 vehicles stopped.

Finding 35. Based on the staffing levels identified by forces, each

intercept officer would expect to recover the following over the course

of a year:

• seven stolen vehicles, with a total value of approximately £46,000

• stolen goods on three occasions, with a total value of approximately

£4,500

• drugs to be seized on four to five occasions, with a total value of

approximately £2,400

• one to two offensive weapons/firearms

• other property on two occasions.

Figure 6.1: Searches of vehicles, persons and items recovered by force

Avon & Somerset 106 208 45 6 1 32 11 9 £288,080 £8,105 £1,290

Cambridgeshire 270 368 28 10 - 20 14 11 £140,020 £46,570 £192,293

Cheshire 199 309 48 38 - 34 9 8 £380,700 £20,500 £10,928

City of London 27 28 7 3 - 3 - 10 £79,000 £50 £1,588

Cleveland 146 166 96 3 - 5 4 - £117,145 £3,675 £4,020

Greater Manchester 114 137 37 13 - 18 5 12 £245,900 £545 £1,341

Hampshire 102 158 25 10 - 14 4 11 £133,495 £480 £2,544

Hertfordshire 49 86 29 12 - 8 4 9 £188,400 £7,172 £2,143

Kent 117 168 35 25 - 10 20 24 £228,700 £530 £5,982

Lancashire 238 297 36 17 - 33 3 10 £276,950 £2,060 £85,495

Leicestershire 153 227 58 41 1 36 11 14 £402,059 £3,255 £62,764

Lincolnshire 253 126 15 16 - 9 5 24 £76,000 £3,275 £40,510

Merseyside 146 140 56 19 1 14 7 11 £371,300 £3,150 £22,926

Metropolitan 1,064 1,773 129 48 4 90 34 12 £1,159,400 £19,915 £87,486

North Wales 404 660 37 55 1 170 18 33 £243,150 £26,974 £3,997

North Yorkshire 97 166 26 10 - 14 - 4 £265,775 £33,628 £19,900

Northamptonshire 128 240 158 53 1 34 20 30 £888,550 £6,670 £39,863

Northumbria 101 91 7 5 - 15 4 10 £6,550 £54,972 £605

Nottinghamshire 250 310 88 15 1 37 20 5 £533,500 £24,670 £34,753

Staffordshire 169 227 20 8 - 19 1 44 £163,820 £3,710 £3,637

Warwickshire 144 259 26 12 1 5 2 11 £250,550 £1,145 £5,478

West Midlands 7 31 86 7 2 103 69 5 £775,025 £114,880 £14,360

West Yorkshire 118 156 60 4 - 17 1 19 £363,250 £1,645 £900

Total 4,402 6,331 1,152 430 13 740 266 326 £7,577,319 £387,576 £644,803

Searches Items recovered Value of vehicles/goods recovered

Stolen Stolen Offensive

Vehicles Persons Vehicle Goods Firearms Drugs Weapon Other Vehicles Drugs Other

Figure 6.2: Number of searches and goods recovered or seized by force per 100

vehicle stops

These findings represent a slight drop from those recorded in Laser 1. As will

be highlighted below, this reduction could be partly attributed to a decrease in

accuracy of the DVLA databases leading to more time being spent dealing with

incorrect hits.

Unsurprisingly, PNC was an extremely effective means for identifying stolen

vehicles (six times more effective than other stop reasons), and identifying

where stolen goods or drugs might be recovered (twice as effective than other

stop reasons).

Avon & Somerset 1.44 2.83 0.61 0.08 0.01 0.44 0.15 0.12 £6,402 £253 £215

Cambridgeshire 2.91 3.96 0.30 0.11 - 0.22 0.15 0.12 £5,001 £2,329 £19,229

Cheshire 3.89 6.05 0.94 0.74 - 0.67 0.18 0.16 £7,931 £603 £288

City of London 1.05 1.09 0.27 0.12 - 0.12 - 0.39 £11,286 £17 £529

Cleveland 4.24 4.82 2.79 0.09 0.15 0.12 - £1,220 £735 £1,340

Greater Manchester 0.72 0.87 0.23 0.08 - 0.11 0.03 0.08 £6,646 £30 £103

Hampshire 1.58 2.45 0.39 0.16 - 0.22 0.06 0.17 £5,340 £34 £254

Hertfordshire 0.87 1.52 0.51 0.21 - 0.14 0.07 0.16 £6,497 £896 £179

Kent 1.15 1.66 0.35 0.25 - 0.10 0.20 0.24 £6,534 £53 £239

Lancashire 2.07 2.58 0.31 0.15 - 0.29 0.03 0.09 £7,693 £62 £5,029

Leicestershire 1.55 2.31 0.59 0.42 0.01 0.37 0.11 0.14 £6,932 £90 £1,531

Lincolnshire 2.19 1.09 0.13 0.14 - 0.08 0.04 0.21 £5,067 £364 £2,532

Merseyside 2.16 2.07 0.83 0.28 0.01 0.21 0.10 0.16 £6,630 £225 £1,207

Metropolitan 5.90 9.83 0.72 0.27 0.02 0.50 0.19 0.07 £8,988 £221 £1,823

North Wales 3.67 6.00 0.34 0.50 0.01 1.54 0.16 0.30 £6,572 £159 £73

North Yorkshire 2.16 3.69 0.58 0.22 - 0.31 - 0.09 £10,222 £2,402 £1,990

Northamptonshire 1.83 3.43 2.26 0.76 0.01 0.49 0.29 0.43 £5,624 £196 £752

Northumbria 2.51 2.26 0.17 0.12 - 0.37 0.10 0.25 £936 £3,665 £121

Nottinghamshire 9.34 11.58 3.29 0.56 0.04 1.38 0.75 0.19 £6,063 £667 £2,317

Staffordshire 2.93 3.93 0.35 0.14 - 0.33 0.02 0.76 £8,191 £195 £455

Warwickshire 3.05 5.48 0.55 0.25 0.02 0.11 0.04 0.23 £9,637 £229 £456

West Midlands 0.08 0.35 0.98 0.08 0.02 1.18 0.79 0.06 £9,012 £1,115 £2,051

West Yorkshire 1.37 1.81 0.69 0.05 - 0.20 0.01 0.22 £6,054 £97 £225

Total 2.44 3.50 0.64 0.24 0.01 0.41 0.15 0.18 £6,578 £524 £1,500

Searches Items recovered Value of vehicles/goods recovered

Stolen Stolen Offensive

Vehicles Persons Vehicle Goods Firearms Drugs Weapon Other Vehicles Drugs Other

Figure 6.3 shows that the profile of driver ethnicity of the vehicles that were

searched closely matches the profile of ethnicities for all stops for the majority

of ethnic backgrounds. The major exception to this is the Black ethnic group.

While they drive only 6.5% of the vehicles stopped, 14.9% of the cars

searched were being driven by people in that ethnic category.

Figure 6.3: Profile of driver ethnicity of searched vehicles

6.3 Arrests

6.3.1 Arrests by type

The key priority for the 23 forces involved in Laser 2 was to engage with

criminals and deny them use of the roads. In practice this meant stopping and

arresting criminals. During the 13 months of Laser 2, there were 13,499

arrests by the intercept teams. Figure 6.4 shows the reason for arrest.

Note that if a person was arrested more than once then only the most serious

arrest was recorded, as opposed to recording each arrest made. Information

on arrests that were subsequently de-arrested was not collected.

Figure 6.4: Reason for arrest

Stated ethnic background % vehicles searched % vehicles stopped

White 72.4% 74.1%

Asian 6.4% 7.2%

Black 14.9% 6.5%

Other 1.0% 1.0%

Mixed 1.6% 0.8%

Not Stated/ Unknown 3.6% 10.4%

Section 25 arrests include offences that could normally be dealt with by means of a fixed penalty

or a report for summons, however the offender had a history of failing to pay or appear at Court.

Auto crime is theft from or of a vehicle.

11%

0.5%

25%

17%

15%

10%

11%

Robbery

Theft / burglary

Driving

Drugs

S25

Auto crime

Warrant

Other

Figure 6.4 shows that a small proportion (25%) of arrests of vehicle drivers were

for driving offences, with the majority of arrests for serious criminal offences,

including 17% for theft or burglary. Relative to Laser 1, this breakdown represents

a slight change in the profile of arrests:

• Theft/Burglary account for 17% of arrests (previously 21%) [2,263 arrests]

• The number of arrests for driving offences has increased from 20% to 25%

[3,324 arrests]

• Arrests for drugs offences have fallen from 11% to 8% [1,107 arrests]

• Section 25 arrests have seen a significant increase, rising from 7% to 11%

[1,486 arrests]

• Auto crime arrests accounted for 10% of arrests while previously they were

12% [1,386 arrests]

• Warrants have increased from 11% to 13% [1,812 arrests]

• Arrests for other reasons have fallen from 17% to 15% [2,043 arrests]

• Robbery has remained the smallest category for arrests staying below 1%

[78 arrests].

Finding 36. ANPR-enabled intercept officers arrested someone on average

once every thirteen vehicle stops. Only 25% of arrests related to driving

offences, ie the vast majority of arrests were for non-driving matters. It is

also worth noting that in 7,456 of the 13,499 arrests (55%) the people

arrested had previous police records.

Figure 6.5 overleaf lists the total number of arrests, arrest types, stops and

arrests per 100 vehicle stops by force.

Figure 6.5 shows a wide variation in the number of arrests made by forces –

three forces (Metropolitan Police Service, Northamptonshire and West Midlands)

achieved over 1,100 arrests during the 13 month period while one force (City

of London) achieved less than 100.

In terms of arrest types by force, again there was wide variation, for example:

• relatively few arrests were for robbery (0.6%). While Avon & Somerset and

Cleveland reported a much higher proportion of arrests for robbery, their total

number of arrests was small and these results are therefore not significant

• arrests for theft or burglary showed wide variation from the average (16.8%)

– in three forces less than 10% of roadside arrests were theft or burglary

(Avon and Somerset, North Wales and Staffordshire), while for nine forces

more than 20% of arrests were for theft or burglary

• Cleveland made nearly double (46.5%) the average number of intercept

arrests (24.6%) for driving offences

• North Wales made nearly treble (23.5%) the average number of intercept

arrests (8.2%) for drugs offences

• Northumbria made nearly double (21.2%) the average number of intercept

arrests (11.0%) for Section 25 offences, while very few (1.6%) of

Northamptonshire arrests were for Section 25 offences

• Only 1.8% of Cleveland’s arrests were for auto crime, compared to the Laser 2

average of 10.3% and Hertfordshire where 19.5% of arrests were for auto crime

• 8.1% of Herefordshire’s arrests were for outstanding warrants, compared to

the Laser 2 average of 13.4% and Northumbria where only 3.4% of arrests

were for outstanding warrants.

This spread of arrest type reflects different force operational priorities, staffing

experience and quality of local databases.

Figure 6.5: Total arrests, stops and arrests per 100 vehicle stops and arrest types by force

Avon & Somerset 343 7,354 4.66 4.47 2.0% 9.1% 37.1% 6.0% 6.3% 12.0% 20.9% 6.7%

Cambridgeshire 516 9,286 5.56 4.87 0.6% 17.5% 13.1% 8.9% 12.6% 5.5% 16.4% 25.4%

Cheshire 517 5,111 10.12 6.44 0.0% 14.2% 19.2% 8.6% 15.6% 16.2% 13.9% 12.2%

City of London 76 2,562 2.97 1.17 0.0% 21.1% 13.2% 5.3% 6.6% 5.3% 14.5% 34.2%

Cleveland 393 3,446 11.39 7.32 1.8% 17.8% 46.5% 3.6% 4.8% 1.8% 14.3% 9.4%

Greater Manchester 914 15,787 5.79 7.77 0.0% 10.2% 31.1% 5.9% 21.2% 5.4% 11.5% 14.8%

Hampshire 387 6,439 6.01 4.49 0.0% 10.1% 27.5% 7.8% 15.5% 11.9% 12.5% 14.7%

Hertfordshire 280 5,640 4.97 5.70 0.0% 19.8% 15.5% 3.9% 3.6% 19.5% 28.1% 9.6%

Kent 351 10,143 3.46 4.95 0.0% 20.9% 24.7% 3.7% 5.4% 12.5% 8.2% 24.6%

Lancashire 776 11,525 6.73 4.95 0.6% 10.6% 36.7% 8.6% 11.8% 12.0% 11.3% 8.4%

Leicestershire 599 9,848 6.08 6.59 0.3% 14.9% 35.2% 4.8% 8.4% 10.5% 11.4% 14.5%

Lincolnshire 515 11,533 4.47 7.93 0.2% 22.7% 17.8% 5.3% 17.1% 4.5% 10.2% 22.2%

Merseyside 665 6,756 9.84 13.27 0.0% 17.2% 18.1% 5.9% 18.1% 11.0% 17.6% 12.1%

Metropolitan 1,406 18,034 7.80 8.39 0.6% 21.8% 12.1% 11.0% 9.3% 8.1% 10.5% 26.6%

North Wales 862 11,005 7.83 10.77 0.2% 8.6% 33.6% 23.4% 12.8% 7.1% 6.5% 7.8%

North Yorkshire 258 4,498 5.74 4.02 0.0% 23.3% 23.6% 9.3% 11.8% 12.4% 5.0% 14.5%

Northamptonshire 1,152 6,998 16.46 9.43 1.3% 19.2% 25.6% 5.6% 1.6% 16.6% 15.5% 14.5%

Northumbria 334 4,021 8.29 4.64 0.9% 20.5% 17.2% 7.2% 30.3% 6.3% 3.4% 14.1%

Nottinghamshire 601 2,676 22.46 13.17 0.5% 20.2% 23.4% 7.9% 13.2% 9.2% 10.9% 14.6%

Staffordshire 477 5,771 8.27 5.66 0.0% 8.6% 39.0% 3.4% 13.5% 5.7% 18.2% 11.6%

Warwickshire 254 4,727 5.37 3.43 0.0% 21.3% 35.0% 2.8% 8.7% 6.7% 18.3% 7.3%

West Midlands 1,386 8,747 15.84 12.20 1.2% 23.2% 16.0% 10.6% 6.0% 14.1% 15.8% 13.1%

West Yorkshire 436 8,636 5.05 5.17 1.1% 11.3% 22.8% 5.0% 4.9% 13.9% 24.2% 16.7%

Total/Average 13,499 180,543 7.48 6.83 0.6% 16.8% 24.6% 8.2% 11.0% 10.3% 13.4% 15.1%

Arrests per Percentage of arrests

100 100 hours Theft/ Auto Other

Arrests Stops stops intercepting Robbery Burglary Driving Drugs S25 Crime Warrent reason

6.3.2 Arrests per 100 vehicle stops and per 100 hours intercepting

To understand the volume of arrests for each force in the proper context it

is important to look at them relative to the amount of effort that each force

has expended in ANPR activities (officer hours), as well as the productivity

of the stops that are made using different deployments (number of arrests

per 100 stops).

Figure 6.5 shows that on average Laser 2 forces made 7.5 arrests per 100

vehicle stops – slightly below the 7.8 arrests per 100 vehicle stops achieved

during Laser 1. Three forces (West Midlands, Northamptonshire and

Nottinghamshire) achieved more than twice the Laser 2 average of arrests

per 100 vehicles stopped, while two forces (Kent and City of London)

achieved less than half the Laser 2 average. Figure 6.5 also shows that on

average Laser 2 forces made 6.8 arrests per 100 hours of officer time spent

intercepting – again below the 8.3 arrests per 100 hours of officer time spent

intercepting during Laser 1.

In terms of how arrest levels varied during the course of the pilot, Figure 6.6

shows the number of arrests per 100 hours of officer time spent intercepting

and the arrests per 100 vehicles stopped. Overall, both these measures have

remained largely stable during the period of the pilot, showing slight increases

as forces became more experienced.

Figure 6.6: Arrests by week per 100 hours staff input and per 100 vehicles stopped

Week

1 5 9 1317212529333741454953

Arrests per 100 stops

Arrests per 100 hours deployed

Linear (Arrests per 100 hours deployed)

Linear (Arrests per 100 stops)

Figure note: the solid lines show the actual data, while the dotted lines show the trend.

Finding 37. ANPR-enabled intercept officers made 7.5 arrests per

100 vehicle stops and 6.8 arrests per 100 hours of officer time spent

intercepting. During Laser 2 there was a slight overall increase in

performance.

In terms of arrests per FTE, Laser 2 achieved 91 per FTE, compared to the

105 achieved during Laser 1. Figure 6.7 shows that during Laser 2, those

forces that had been part of Laser 1 performed significantly differently to those

that were ‘new’ ANPR forces to Laser 2. In particular:

• new Laser 2 forces started with a much lower arrest per FTE base

(approximately 52) and that this increased to a peak of well over 100

during March 2004. Their performance since then has slipped to around 80.

Throughout Laser 2 these forces have achieved on average an arrest rate

of 78 per FTE

• Laser 1 forces started with a much higher baseline arrest rate (over 100 as

per Laser 1), however this started declining at the beginning of 2004, but has

subsequently picked up. Throughout Laser 2 these forces have achieved on

average an arrest rate of 106 per FTE.

Figure 6.7: Arrests per FTE for Laser 1 and new Laser 2 forces during Laser 2

Finding 38. During Laser 2, those forces that were part of Laser 1 have,

on average, managed to replicate the high levels of performance achieved

during Laser 1 over a 13-month period. Forces new to Project Laser have

taken nearly eight months to achieve the performance benchmark set in

Laser 1 and, as yet, have been unable to sustain this for a prolonged

period. Within the forces new to Laser 2, there are however, a number

of generally underperforming forces (as classified by PSU). This lower

level of performance during Laser 2 is therefore not completely surprising.

Week

4 per. Mov. Avg. (Other 14 forces)

4 per. Mov. Avg. (9 Laser 1 forces)

100

120

140

160

180

1 4 7 10131619222528313437404346495255

Arrests per FTE

Finding 39. Comparison of arrest rates per FTE during Laser 2 against

Laser 1 shows that the introduction of cost recovery has not adversely

affected this key performance metric.

Looking at the performance of the original 9 forces alone, the number of

arrests per FTE had stayed stable (from 105 in Laser 1 to 106 in Laser 2).

This suggests that the introduction of hypothecation has not distorted

policing priorities.

Figure 6.5 showed that there are significant differences in performance

between forces in terms of arrests per 100 stops and arrests per 100 hours

deployed. Specifically are three groupings of forces as follows:

• the top performers – a group of five forces (Merseyside, North Wales,

Cleveland, West Midlands and Northamptonshire) which achieved significantly

higher levels of performance in terms of arrests per 100 stops and arrests per

100 hours deployed. Interestingly, this group includes both rural and urban

forces, those that were part of Laser 1 and those that were not, forces with

which the PSU is engaged for performance issues and those that are not,

and a spread of means of deployment (CCTV, in-car system and mobile unit).

This would suggest that high performance has more to do with ‘softer’ issues

(the experience of the team, local databases, deployment tactics, leadership,

senior management support) than general force characteristics

• outliers – specifically Nottinghamshire and City of London which recorded

the fewest number of stops:

– given its geography, size and the presence of the ring of steel, the City

of London is a unique force. In particular, given the level of security and

monitoring of roads within the Square Mile, many criminals will actively

choose to avoid the area and hence ANPR has effectively denied criminals

the use of the road

– statistically, the performance of Nottinghamshire is considered an outlier.

It is therefore debatable that their performance could be scaled up

correspondingly

– the pack – which includes the remaining 16 forces which have not achieved

the top level of performance. Within this group, there are those that have

performed significantly better than others.

These groupings are shown in Figure 6.8 below.

Figure 6.8: Arrests by week per 100 hours staff input by arrests per 100 vehicles stopped

Finding 40. Three broad categories of intercept team performance have

been identified, namely those that achieved significantly better performance

(five forces), those that achieved around average performance (sixteen

forces) and those where there were too few stops to indicate sustainable

and scalable performance. These groupings are not based on force or

ANPR characteristics; rather they appear to reflect differences in

management and deployment.

Finding 41. On the basis of the staffing inputs identified by forces, each

intercept officer full time equivalent would expect to make 91 arrests per

year. The equivalent figure for Laser 1 was 105 arrests per year.

Merseyside

51015 20 25

Hampshire

GMP

Cleveland

North Wales

Leicestershire

Cheshire

Staffo rdshi re

North Yorkshire

Warwickshire

Kent

Hert fordshire

Cambridgeshire

Lancashire

Arrests per 100 stops

Arrests per 100 hours of Intercept

Lincolnshire

Metropolitan

Northamptonshire

West Midlands

Nottinghamshire

West Yorkshire

Avon and Somerset

City of London

Cheshire

Northumbria

Figure 6.9: Arrest rate per FTE by force for officer intercept hours

6.3.3 Database productivity

Figure 6.10 shows that of the 7,319 arrests from ANPR triggers, 28% came

from PNC, 38% from DVLA no VED, 11% from No current keeper database

and 24% from local databases. These percentages vary by arrest type, with

robbery arrests being triggered primarily by PNC hits. Similarly, driving

offences are more likely to come about as a result of DVLA no VED hits.

Avon & Somerset 343 12,950 51

Cambridgeshire 516 16,694 59

Cheshire 517 12,492 79

City of London 76 8,629 17

Cleveland 393 6,394 118

Greater Manchester 914 18,321 96

Hampshire 387 12,184 61

Hertfordshire 280 6,300 85

Kent 351 9,153 74

Lancashire 776 20,792 72

Leicestershire 599 14,096 82

Lincolnshire 515 10,818 91

Merseyside 665 7,168 178

Metropolitan 1,406 21,757 124

North Wales 862 11,728 141

North Yorkshire 258 8,854 56

Northamptonshire 1,152 16,864 131

Northumbria 334 9,849 65

Nottinghamshire 601 6,448 179

Staffordshire 477 12,125 76

Warwickshire 254 11,834 41

West Midlands 1,387 17,106 156

West Yorkshire 436 12,718 66

Total/Average 13,499 285,271 91

Intercept Arrest

Force Arrest Hours

per FTE

This is the total time intercept officers were in the field, including non-intercept time. This is

consistent with the measure used in Laser 1.

Figure 6.10: Arrests by source database for ANPR stops

For arrests originating from observations, the profile of arrests is somewhat

different. Figure 6.11 shows that while the most prominent arrest type relates

to driving offences, the next most common arrests are for other reasons and

for warrants – this suggests that there is much to be gained by using the

‘policeman’s nose’. Given that the ‘other observations’ category includes

primarily intercept officers being suspicious of a particular vehicle or driver,

it is interesting to note that this category delivers the most arrests.

Figure 6.11: Arrests by source database for observation stops

Finding 42. During Laser 2, ANPR hits accounted for 56% of vehicle stops

and generated 54% of arrests. The profile of these arrests, however, was

different to that of observation generated stops, in particular ANPR lead

to more arrests for theft and burglary.

PNC 21 544 317 119 126 530 209 205 2,071

No VED 5 569 738 164 341 147 364 435 2,763

No keeper 5 120 176 51 119 47 92 176 786

Local 2 235 481 177 210 98 241 255 1,699

Total 33 1,468 1,712 511 796 822 906 1,071 7,319

% 0.5% 20.1% 23.4% 7.0% 10.9% 11.2% 12.4% 14.6% 100.0%

Theft/ Auto Other

Robbery Burglary Driving Drugs S25 crime Warrant reason Total

Mobile phone offence 0 8 15 4 5 1 11 4 48

No Seatbelt 0 26 56 25 36 19 39 33 234

VED 5 129 296 80 139 48 188 137 1,022

Vehicle Defect 0 12 53 14 27 19 21 28 174

Driving manner 0 68 231 33 49 93 69 82 625

Known person/vehicle 11 92 157 31 24 59 132 78 584

Other observation 29 460 804 409 410 325 446 610 3,493

Total 45 795 1,612 596 690 564 906 972 6,180

% 0.7% 12.9% 26.1% 9.6% 11.2% 9.1% 14.7% 15.7% 100.0%

Theft/ Auto Other

Robbery Burglary Driving Drugs S25 crime Warrant reason Total

6.3.4 Arrest type by age of vehicle

In section 4.3.6 above, it was shown that the year when a vehicle was first

registered was a strong factor in determining if was stopped and why it was

stopped. Figures 6.12 and 6.13 show that when a vehicle was first registered

was also strong indicator both of the number of arrests per 100 stops as well

as the type of arrest.

Figure 6.12: Arrests by type by estimated year of registration

Figure 6.13: Percentage of arrests by type by estimated year of registratio

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

Year of first registration

Arrests per 100 stops

Robbery

Theft / Burglary

Driving

Drugs

S25

Auto crime

Warrant

Other Reason

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

Year of first registration

Percentage of Arrests

Robbery

Theft / Burglary

Driving

Drugs

S25

Auto crime

Warrant

Other Reason

Stopping vehicles registered in 1988 was more than 3 times as likely to lead to

an arrest than stopping those vehicles registered in 2000. It is also worth noting

that the type of arrest also varies by year of car. Predictably, auto crime arrests

are more likely in newer vehicles (as these cars are more valuable), while arrests

for driving offences are most likely to arise from stops of older vehicles.

6.3.5 Arrests by time of Day

Both the time and location of deployment can play an important role in making

ANPR productive. Figure 6.14 shows that while the majority of arrests take

place between 11:00 and 16:00, the more productive times for intercept were

between 20:00 and midnight. (Note: the number of stops between midnight

and 8am was minimal).

Figure 6.14: Arrests by time of day

6.3.6 Arrest by location

As has already been mentioned, the selection of an appropriate location has

an important role to play in successful ANPR operations, both in terms of volume

and type of arrest. Figure 6.15 shows the ten most productive postcodes in

terms of arrests throughout the life of the pilot.

Arrests

Arrests per 100 stops

Total Arrests

Arrests per 100 stops

500

1000

1500

2000

2500

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Time of day

3000

Figure 6.15: The ten most productive postcodes during Laser 2

Areas where CCTV systems were in operation (such as Northamptonshire, Kent,

and Staffordshire) appeared to be very productive. This is likely to be a reflection

of the fact that more intercept hours are spent at these locations. When intercept

hours are taken into account, location is not a determinant of success.

6.3.7 Arrest case studies

While the above arrest analysis shows the quantities of arrests, it does give

any indication as to the quality of arrest. By way of example a number of

case studies provided by the ANPR-enabled intercept teams are included

below. These are from a variety of originating databases. Further examples,

as reported on force websites, are included as Appendix E to this document.

• In December 2003 during an ANPR operation, a CCTV camera registered

a PNC hit on a 4×4 vehicle. The vehicle was suspected of using false

number plates having previously gone through a speed camera in Surrey.

Following a search, several offensive weapons and class A drugs were

recovered. Furthermore, the chassis number on the vehicle could not be

found on PNC. The driver was arrested and a search of his home led to the

recovery of more drugs, a pump action shotgun and another stolen 4×4 vehicle

Postcode Force Total arrests Arrests per 100

stops at postcode

NN1 Northamptonshire 349 25.1

LL18 North Wales 258 8.2

DA1 Kent 223 4.9

WA8 Merseyside 188 12.7

CH5 North Wales 151 9.2

BB1 Lancashire 139 7.1

PE1 Cambridgeshire 134 8.7

ST4 Staffordshire 132 10.4

CT4 Kent 125 2.3

B20 West Midlands 124 19.5

• A Renault Laguna triggered an ANPR hit against a local intelligence database,

the information being that the occupants were suspected of being involved

in the supply of controlled drugs. As the intercept officer drew alongside the

vehicle, he saw the driver rapidly drinking water from a bottle. The vehicle

was stopped and the two male occupants searched with a negative result.

However, the officers detained the men for a strip search. While being conveyed

to the custody suite for this purpose, one of the males began vomiting, which

resulted in 15 wraps of heroin being brought up. Both men were arrested for

being involved in the supply of controlled drugs and in addition £815 in cash

was recovered

• Information, including part of a vehicle registration number, was received from

a neighbouring force relating to a recent Post Office robbery. The ANPR team,

which was operating close to their force border, searched recent vehicle reads

on their ANPR system and identified a possible match. This was circulated

and the vehicle stopped by two ANPR motorcyclists within 30 minutes of the

initial intelligence circulation. White overalls, masks and a baseball bat were

found in the vehicle and the two occupants were arrested and subsequently

charged with robbery and theft of a motor vehicle

• Following a DVLA no VED hit, a vehicle was intercepted and the driver’s details

taken. A check through civilian warrants revealed that there were a number

of outstanding fine default warrants. The man was arrested and taken to the

custody station. When he was told how much was outstanding he said

something to the effect of “the stupid b*****d”. It later transpired that he and

a friend had agreed to provide each other’s details if stopped by the police.

What he failed to realise was that his friend was wanted on warrant. When his

correct details were obtained, he was found to have been a disqualified driver

• A Mini Metro passed through an ANPR intercept site and showed as a hit on

a local database. Intelligence suggested that this vehicle had been spotted the

previous week in suspicious circumstances and that it was likely to be used

by a gang of local shoplifters. The vehicle was stopped and checks made on

the occupants. A search of vehicle led to the recovery of a large quantity of

brand new clothing items, all with Marks and Spencer tickets. The occupants

were arrested and subsequent enquires revealed that the items had been

stolen that morning from a local branch of Marks and Spencer

• In June 2004, a Golf TDi drove through a check site and activated a PNC

warning that the vehicle may be using false number plates. Following a short

pursuit, officers stopped the vehicle and examination revealed the vehicle to

have been stolen from the Manchester area. The driver was arrested for

stealing the vehicle and intimated that he was on route to East Midlands airport

to collect his co-offender. Officers arrived at the airport in time to arrest the

second offender as he came through customs. Both men are prolific offenders

• A Rover car was identified by ANPR (PNC) as being stolen from a burglary.

A pursuit ensued during which a carrier bag containing a loaded handgun

was thrown from the vehicle. Shortly afterwards a second stolen vehicle was

identified containing 5 persons, believed to be connected to the first. Due to

the ongoing pursuit this vehicle was lost. The Rover car was eventually stopped

and the driver and front passenger arrested for taking the car without consent

and for possession of a firearm. While awaiting recovery a mobile phone rang

in the car. An officer answered and arranged to meet the owner. 5 people

turned up and were arrested for involvement in taking the car without consent.

During processing one of the prisoners was found to have a Honda key for

which he could not account. A quick trip to the car pound established that the

key fitted a recently recovered stolen Honda Civic. It is believed that the

persons arrested were en-route to exact revenge for a stabbing that had

occurred earlier that day

• A vehicle passing the ANPR checkpoint activated the DVLA no VEL database.

The vehicle was stopped and the officer noticed the driver trying to secrete a

package down her trousers. All four occupants were detained for a drugs search

and were subsequently arrested in possession of approximately £200 worth

of heroin and cannabis and £400 in cash. During a search of the premises of

the arrested parties, a man who had been wanted for approximately two

months for supplying crack cocaine was found and arrested.

Finding 43. The ANPR-enabled intercept teams made a number of very

significant arrests. Arrests were not just for vehicle documentation crime

but for more serious crimes. It is not possible to quantify the quality of

these arrests, however the case studies give some indication as to the

value of ANPR in addressing serious crime.

6.4 Other actions

In addition to arrests, ANPR-enabled intercept officers were able to report

individuals for summons, issue a fixed penalty, issue a note requiring follow-up

action, give some verbal advice and/or prepare an intelligence log.

Finding 44. Of the 180,543 vehicle stops, in 117,492 cases (65%) the intercept

officers took some form of action. This is higher than the equivalent figure

for Laser 1 (61%). Analysis showed that the proportion of stops resulting

in some form of action improved slightly during the last 2 months of Laser 2.

Figure 6.16 shows the total number of actions taken during the pilot. It should

be noted that officers were able to take multiple actions – for example a vehicle

stop could lead to an non-endorsable fixed penalty, the driver being issued with

a request to provide their vehicle documentation at a police station (HO/RT1)

and an intelligence log being created.

Figure 6.16: Actions taken by intercept officers

Finding 45. Overall, intercept officers took 192,491 actions with respect

to 117,492 vehicle stops where an action was taken, ie approximately

1.6 actions per vehicle where an action was taken. Analysis showed the

number of other actions taken per 100 vehicles stopped increased

marginally during Laser 2.

The most commonly taken action (excluding arrests and fixed penalty notices)

was issuing a HO/RT1 (44,767 issued). The process is usually necessary when

a driver is unable to provide documentation such as a driving licence, MOT or

insurance at the roadside. The normal process is for those drivers to be issued

with an HO/RT1 and asked to produce their details at a local police station.

On average, an HO/RT1 was issued to 26% of all vehicles stops, though this

varied by force – in the West Midlands, 62.4% of vehicles stopped were issued

with a HO/RT1 while in the Metropolitan Police Service, this figure was only 9%.

The HO/RT1 process can also lead to conditional offer tickets being issued.

This came into effect in October 2003 but not all forces were able to issue these.

Where available, there were 336 conditional offers for no insurance and 316

for no MOT issued with 33 and 65 paid respectively.

Non-endorsable Fixed Penalty notices were issued at 42,867 stops (24%).

Furthermore, Endorsable fixed penalty notices were issued at 9,898 stops (5%).

It is worth remembering that for many stops, more than 1 ticket was issued and

that while officers primarily issued the fixed penalties that could be hypothecated

for ANPR, there were other tickets that could be issued (such as speeding)

which did not come under the umbrella of ANPR offences. A more detailed view

of the fixed penalty notice issuing information is covered in the next section

along with the cost recovery elements of Laser 2.

44,767

42,867

26,179

21,801

21,192

13,497

9,898

8,323

3,967

0 5 10 15 20 25 30 35 40 45 50

HO/RT1

Endorsable FP

INTEL log generated

Verbal advice given

CLE 2/6 2/7

Arrests

Non-endorsable FP

Reported for summons

VDRS / PG9

Actions Taken (’000)

In total 26,179 intelligence logs were created during the pilot. Again this varied

significantly by force. Given that these intelligence logs may be used by officers

not part of the intercept and potentially over a number of years, it is extremely

difficult to quantify the value of these logs.

Figure 6.17 overleaf lists the key actions taken per 100 vehicle stops by force.

6.4.1 Vehicle seizure

A number of forces extended their ANPR operations to include the seizure of

vehicles. Following legal advice, West Yorkshire police embarked on a scheme

of recovering uninsured vehicles. These are seen to be widely involved in serious

crime and posed a great threat to the public. If a driver was found to be uninsured

during a stop, he/she was prohibited from continuing their journey and

informed that the vehicle needs to be removed from the road within 30

minutes. If the driver could not arrange this, West Yorkshire police recovered

the vehicle at the driver’s expense. Between September 2003 and April 2004

they recovered more than 700 vehicles and experienced a drop in crime in

the area. The view of officers on this operation was that criminals were having

their vehicles taken from them and were therefore less effective as a result.

This suggests a further way in which ANPR intercept teams can potentially

impact on criminality. The findings of the Greenaway report, published in August

2004, support this view and the DfT have already embarked on a campaign that

supports police to tackle the problem of uninsured driving.

Under powers from DVLA, West Midlands Police have also begun seizing

untaxed vehicles. From the spring of 2004, more than 500 vehicles have

been recovered.

Figure 6.17: Actions taken per 100 vehicle stops by force

Avon & Somerset 4.7 35.3 2.8 15.0 3.1 18.6 4.8 2.1 11.2 29.4 26.2 74% 7,354

Cambridgeshire 5.6 27.9 0.3 8.4 0.5 26.1 3.5 1.4 6.0 5.4 43.7 57% 9,286

Cheshire 10.1 13.1 0.3 21.8 0.2 24.0 5.0 1.4 27.1 16.6 28.1 72% 5,111

City of London 3.0 10.9 0.7 4.8 2.0 60.1 10.8 1.4 10.3 4.8 10.7 90% 2,562

Cleveland 11.4 50.9 0.1 5.4 4.6 44.5 11.6 6.3 29.2 1.7 7.4 93% 3,446

Greater Manchester 5.8 19.8 1.8 4.2 0.6 10.1 1.7 6.0 6.8 8.8 55.2 45% 15,787

Hampshire 6.0 20.0 0.6 2.5 3.1 30.7 7.3 3.1 14.8 10.9 33.1 69% 6,439

Hertfordshire 5.0 29.2 1.5 1.1 1.4 35.4 6.8 1.2 10.5 12.5 33.5 67% 5,640

Kent 3.5 9.0 4.3 4.0 2.6 16.6 1.2 3.1 8.1 26.1 41.1 59% 10,143

Lancashire 6.7 31.3 4.9 6.5 7.0 42.6 7.3 1.9 6.1 13.4 12.6 88% 11,525

Leicestershire 6.1 25.6 0.1 9.3 0.9 15.3 4.9 7.7 23.3 3.2 45.5 55% 9,848

Lincolnshire 4.5 13.9 0.4 6.7 2.3 20.5 1.2 1.4 7.9 5.1 58.7 44% 11,533

Merseyside 9.8 33.4 2.6 20.5 0.9 24.4 5.0 7.7 10.1 11.0 39.8 61% 6,756

Metropolitan 7.8 9.7 2.1 13.4 1.5 6.7 7.5 8.5 10.4 9.5 44.8 56% 18,034

North Wales 7.8 24.9 0.4 20.9 2.7 38.5 7.8 3.0 5.5 4.9 36.5 65% 11,005

North Yorkshire 5.7 29.3 1.8 7.6 3.9 26.0 2.2 2.5 12.7 20.9 31.6 69% 4,498

Northamptonshire 16.5 25.3 3.9 4.0 0.8 19.9 8.1 1.4 27.2 31.0 19.7 81% 6,998

Northumbria 8.3 26.7 0.3 0.0 1.4 11.9 3.0 1.3 0.2 11.5 50.2 51% 4,021

Nottinghamshire 22.5 14.3 1.9 0.9 2.1 38.8 42.8 5.2 56.5 4.3 5.5 96% 2,676

Staffordshire 8.3 27.7 4.9 15.7 1.2 30.3 5.0 18.6 28.1 10.7 21.4 79% 5,771

Warwickshire 5.4 28.5 0.1 21.7 3.7 20.8 3.8 9.6 9.8 9.0 37.5 63% 4,727

West Midlands 15.8 62.5 12.3 0.0 4.4 36.0 3.5 4.6 55.6 5.1 10.4 96% 8,747

West Yorkshire 5.1 28.5 3.5 12.3 1.0 19.6 3.8 4.0 8.1 23.5 43.3 59% 8,636

Laser 2 Total/Average 7.5 24.8 2.5 9.3 2.2 23.7 5.5 4.6 14.5 12.1 36.0 65% 180,543

Laser 1 Total/Average 7.8 42.3 14.7 14.2 2.6 2.3 0.9 5.2 26.4 8.2 39.4 61% 39,188

Follow up action required by driver Fixed penalties other action

CLE 2/6 CLE 2/8/ VDRS/ Reported for Intel Verbal No Action Occasion Vehicle

H0/RT1 2/7 V62 PG9 NEFPN EFPN summons logged advice given taken action taken stoppe

Arrests

Findings: Database issues

This section of the evaluation sets out the results, that is to say the actions

that arose from the intercepts. In total, including stops from both ANPR hits

and observation, we have records of the actions taken from over 180,000

vehicle stops.

1. Good intelligence is at the heart of modern policing: [Section 7.1]

• Laser 1 identified a number of weaknesses in national databases,

principally with regard to the vehicle licensing. A number of initiatives

were put in place but quality remains an issue

• there are a number of concurrent pressures to improve the quality of

national and local intelligence, but much more remains to be done

• PNC and local databases were found to be around 80% accurate

compared to around 40% for DVLA

• Accuracy of DVLA databases declined over the study period

2. Teams will become more efficient as the quality of intelligence

improves: [Section 7.2]

• improvements to the database accuracy (particularly DVLA) will lead to

more efficient targeting of resources. Nevertheless, as we have seen,

around half of all ANPR arrests were as a result of hits from No current

keeper or no excise databases

3. Improving conversion rates: [Section 7.2]·

• the greatest conversion rate (hits to arrests) for ANPR database

was, not surprisingly, PNC – 19.6% of all hits resulted in arrests·

• the greatest conversion rate for observation was the category

‘known vehicles or criminals’ – 36.3% of all stops resulted in arrests.

7.1 Context

7.1.1 Laser 1 evaluation

The use of ANPR-enabled intercept teams is a prime example of an intelligence-

led policing tool – police using existing intelligence sources to direct their

activities. As identified above in section 6, targeting of police resources has

produced excellent results. However, the most critical factor that contributed

to the effectiveness of ANPR teams was the underlying intelligence on which

the stops were based.

The Laser 1 evaluation highlighted inadequacies in the accuracy

of the

various intelligence databases, in particular DVLA’s No VED and No current

keeper, and outlined a number of possible reasons for these inaccuracies.

The report identified that DVLA were undertaking a number of measures to

improve data accuracy, for example the introduction of bar-coded V11 forms,

barcode readers in Post Offices and the introduction of continuous registration.

However it went on to identify that there was a lack of in-depth understanding

as to the cause of these inaccuracies and concluded that this represented a

weakness that should be addressed.

7.1.2 Intelligence sharing

The Bichard inquiry report highlighted general weaknesses in the use and

sharing of intelligence by police. In the context of this report, it is clear that as an

intelligence-led policing tool the effectiveness of ANPR in engaging level 2 and

3 criminality will be limited by the availability of good quality and timely intelligence

across force boundaries. In the light of the Bichard findings, police must make

greater effort to effectively use and share intelligence across force boundaries.

To address these issues, the Government is taking forward a number of

actions including:

• the introduction of a National Police Intelligence Computer system (entitled

‘IMPACT’). This will ensure that all forces use the same system to manage

and share intelligence

Database accuracy was recorded by officers stopping vehicles as a result of ANPR hits.

Therefore, the accuracy of a database quoted is a reflection of the correctness of the information

for those vehicles stopped and checked as recorded by officers.

Engaging Criminality – denying criminals use of the roads, PA Consulting Group (October 2003)

100

• as an interim measure, introducing an easily searchable index of all those

persons on whom any police force holds information. This will begin in

Autumn 2004 and will be complete by Spring 2005

• a statutory code of practice on police information handling, introduced by

the end of this year to enable all 43 forces to deal with intelligence information

in the same way. This will link closely to NIM.

These recommendations should ensure that the overall quality of intelligence

collated and maintained by the police is improved and that this intelligence is

shared more effectively. While the actions relate specifically to persons, there

is latitude to include vehicles associated with these persons and therefore further

exploit ANPR. In this context, the establishment of a national data warehouse

of vehicle intelligence, including ANPR reads and hits as a further source of

intelligence, is a critical step forward and must form part of an overall national

intelligence management solution.

7.2 Data sources

7.2.1 Background

During the pilot, the ANPR readers were used with a variety of data, including:

• PNC Vehicles Index, which was provided daily or available on-line to some

intercept teams

• DVLA’s databases of No current VED and No current keeper.

Both databases were provided to the forces by DVLA on a monthly basis.

However because of the time delay between VED discs being purchased

and the DVLA systems being updated, the no VED database only included

those vehicles that had been without VED for two months or more.

In practice this meant that some vehicles without VED would be missed

• local or other ad hoc databases. These varied from force to force and

included:

– Customs and Excise databases, for example tobacco bootleggers

– outstanding speed camera tickets

– regional stolen vehicle databases, for example ELVIS which covers

Merseyside

– PIKE, a national database of LGV and commercial vehicles of interest

– Vehicle and Operator Services Agency (VOSA) databases.

101

For this pilot, the source of each database hit was recorded by the intercept

officers, namely PNC, No current VED, No current keeper or other force database.

Due to space limitations on the data collection pro forma used by the intercept

officers, it was not possible to record the source data for ‘other force databases’.

7.2.2 Vehicle stops by source

Figure 7.1 shows the accuracy of the ANPR data (as recorded by the officer

who made the stop) for the 101,775 vehicles stopped as result of database

hits. For example, 14,547 vehicles were stopped due to a PNC flag; of these,

on 79% of occasions the information that led to the stop was deemed by the

officer making the stop to be accurate. It is important to note that Figure 7.2

shows the accuracy of the ANPR data for those vehicles stopped, not the

overall accuracy of the databases (as this would have required all vehicles to

be stopped and checked).

Finding 46. During Laser 2, 111,637 checks against the intelligence

databases were made during the 101,775 ANPR stops. Overall, the

intelligence databases were shown to be accurate on 52% of occasions,

with local force databases and PNC most likely to be accurate (83% and

79% respectively), while DVLA databases least likely to be accurate

(combined accuracy of 40%).

Finding 47. Since Laser 1, the accuracy of all ANPR databases has fallen

– accuracy of local databases fell from 93% to 83%, PNC fell from 83% to

79%, and, in spite of a number of developments in the DVLA databases,

No current keeper from fell 53% to 41% and no VED fell from 51% to 40%.

102

Figure 7.1: Accuracy of information for vehicles stopped by force

Finding 48. There was significant variation in database accuracy across

forces, in particular for the national DVLA databases. This is surprising

given that there is, in principle, a consistency of approach in data

collection. In some cases, this could be explained by differing work

practices. For example, some forces chose to verify visually if a vehicle

had a valid tax disc before deciding to stop on the basis of a No VED hit.

(In most cases a visual inspection can check for failing to display a valid

VED disc, while the ANPR read checks if a vehicle has been taxed.)

However, this would not explain the difference in accuracy between

forces in the No current keeper database.

Avon & Somerset 1,081 37.0% 1,999 40.5% 1,478 79.6% 234 88.0%

Cambridgeshire 2,682 28.8% 858 30.5% 668 64.5% 1,299 76.5%

Cheshire 1,502 58.9% 258 69.8% 381 79.8% 1,211 83.7%

City of London 153 80.3% 178 81.7% 177 86.7% 185 95.1%

Cleveland 815 66.7% 39 56.4% 101 84.1% 452 90.9%

Greater Manchester 6,769 28.5% 2,905 28.6% 972 57.0% 2,626 56.0%

Hampshire 1,757 46.6% 682 48.6% 560 69.4% 380 79.4%

Hertfordshire 1,045 41.4% 615 51.8% 372 85.5% 386 83.4%

Kent 3,838 37.3% 1,702 35.8% 793 80.3% 472 90.0%

Lancashire 1,595 67.0% 833 71.1% 334 83.5% 988 90.7%

Leicestershire 4,050 44.2% 1,028 58.7% 896 88.0% 2,529 92.4%

Lincolnshire 4,651 15.8% 2,864 12.2% 477 53.9% 711 84.5%

Merseyside 3,311 40.1% 314 50.2% 411 75.2% 710 85.6%

Metropolitan 4,598 46.7% 4,781 46.3% 2,105 81.4% 640 83.4%

North Wales 2,928 44.1% 321 58.3% 355 80.7% 909 90.4%

North Yorkshire 1,036 29.9% 671 30.4% 239 53.6% 391 72.6%

Northants 2,694 42.3% 543 48.9% 1,014 77.7% 867 91.9%

Northumbria 2,026 19.1% 678 29.6% 134 65.7% 893 91.4%

Nottinghamshire 229 87.7% 237 96.2% 331 91.5% 431 95.6%

Staffordshire 1,603 51.1% 27 63.0% 301 80.1% 875 88.3%

Warwickshire 1,700 68.8% 557 80.1% 481 94.6% 454 89.4%

West Midlands 1,897 98.5% 1 100.0% 1,455 99.9% 450 99.6%

West Yorkshire 4,097 21.4% 284 35.7% 518 64.0% 579 86.9%

Total 56,053 40.1% 22,370 40.6% 14,547 78.8% 18,668 83.3%

Hits Accuracy Hits Accuracy Hits Accuracy Hits Accuracy

DVLA DVLA Local Force

No VED No current keeper PNC flag Database

103

The Laser 1 report highlighted the poor accuracy of the DVLA databases as a

factor in restricting the effectiveness of intercept teams. As part of their on-going

modernisation programme, DVLA introduced bar codes onto tax discs during

2003 – this would allow for faster, more accurate updating of records.

Furthermore, pending CJX accreditation for the DVLA will enable the electronic

transmission of this database to forces on a daily basis. This could deliver a

significant boost to the accuracy of the DVLA VED database when it is used

in ANPR operations.

As identified in section 2.5.2, in February 2004 DVLA began enforcing vehicles

that did not have continuous registration. Both developments were identified

as key to improving vehicle keeper details and information on vehicles that

were not taxed. Figure 7.2 shows the accuracy of the two DVLA databases

(as recorded from vehicle stops) during the Laser 2 pilot.

Figure 7.2: Accuracy of DVLA databases for vehicles stops (moving average over 3 weeks)

Finding 49. The accuracy of the no VED and No current keeper databases

varied throughout the pilot, with an overall drop in quality between June

2003 and February 2004. While there has been a slight improvement in

data quality since April 2004, given the wide week-on-week variations,

it is too early to say whether this improvement will be sustained.

Finding 50. Laser 2 confirmed the inadequacies of existing intelligence

databases raised in Laser 1, both in terms of overall poor quality data

and significant variations in data quality between areas. The DVLA

databases in particular were shown to be poor. Because of this, many

forces made a visual inspection of vehicles (sometimes using video images

to help) for a VED tax disc before a vehicle was stopped. Where a valid

tax disc was clearly visible, the vehicle was not stopped

Accuracy of Database (for vehicle stops)

25%

30%

35%

40%

45%

50%

55%

1 3 5 7 9 1113151719212325272931333537394143454749515355

Week

3 per. Mov. Avg. (DVLA: No VED)

3 per. Mov. Avg. (DVLA: No Current Keeper)

104

Figure 7.3 shows that in spite of the poorer data quality, the two DVLA

databases led to the largest number of valid stops.

Figure 7.3: Source database for valid vehicles stops

Finding 51. Approximately 70% of valid vehicle stops originated from the

two DVLA databases. In this respect, the DVLA data was a key element to

the operation of ANPR-enabled intercept teams.

ANPR provided 56% of the stops and approximately 54% of all arrests. Figure

7.4 shows stop reason and arrests rate per 100 vehicle stops for observations

and ANPR stops.

Figure 7.4: Accuracy of database for vehicle stops

50%

20%

17%

13%

DVLA: VED

DVLA: No current keeper

Local databases

PNC flag

19.6

12.3

8.4

12.3

1.8

2.3

3.8

6.5

10.1

11.6

36.3

1.8

1.0

0.9

0 5 10 15 20 25 30 35

PNC

DVLA No VED

DVLA NCK

Local

No Seatbelt(s)

Mobile Phone Offence

Vehicle Defect

Vehicle Excise Duty

Other Observation

Driving Manner

Known Person/Vehicle

Arrests per 100 stops

Correct hit

Incorrect hit

105

Figure 7.4 shows that observation by an officer of a known person or vehicle

was most likely to lead to an arrest – over 36 arrests per 100 vehicle stops.

A correct PNC hit was the next most productive hit, leading to 19.6 arrests

per 100 vehicle stops. Somewhat surprisingly, the no current VED database

yielded 12.3 arrests per 100 vehicles stops when there was a correct hit – as

many as the local force intelligence database. Accuracy of the database was

clearly important since incorrect hits generated only 10% of the arrests of a

correct hit.

Finding 52. Correct ANPR hits were one of the most effective means of

generating arrests per 100 stops. Correct ANPR hits yielded approximately

ten times more arrests than incorrect hits – this stresses the importance

of data quality in driving performance.

The quality of the DVLA databases declined between the 6 month pilot that

began in September 2003 and the current ANPR evaluation period. Had the

no VED and No current keeper databases been as clean in Laser 2 as they

were in Laser 1, the number of arrests generated by these databases would

have been greater. Using 12.3 arrests per 100 stops identified by a correct

VED hit, there would be an improvement in the database quality to 51% (the

level it was in Laser 1) with an additional 685 arrests. Similarly, if the No

current keeper database was the same level of accuracy as per Laser 1,

this would contribute a further 214 arrests.

Finding 53. If the DVLA databases had maintained their accuracy achieved

in Laser 1 during Laser 2, arrests per 100 stops during Laser 2 would have

been 8.0, higher both than the actual in Laser 2 (7.5) and that achieved in

Laser 1(7.8).

Following circulation of initial findings on data accuracy, the DVLA has

re-iterated its commitment to providing the police with the most accurate and

up to date information available. Phase three of the Barcoding All Relicensing

Transactions (BART) Project began a live pilot on the 26 July 2004 and was

scheduled to be rolled out to all issuing Post Office branches by the end of

August 2004. This phase of the project sees the addition of the V10 (vehicle

licence renewal application) to the BART system and the provision of an

on-line enquiry link between Post Office branches and DVLA. This will greatly

reduce the number of incorrect hits over the next few months. Furthermore,

the DVLA acknowledges that links between the agency and the police via a

CJX (Criminal Justice Extranet) would greatly improve data transfer and they

are actively working to speed up such a connection.

106

Findings: Cost recovery

This section of the evaluation provides a summary of the costs and benefits

of the ANPR operation across the 23 forces.

The costs of ANPR enforcement

• the 23 forces are recovering some of the enforcement costs from the

penalties paid for by offending motorists. Strict guidelines govern these

arrangements and these are being adhered to [Section 8.2]

• the rapid introduction of the ANPR pilot at the same time as the introduction

of new fixed penalties meant that many forces were not able to issue fixed

penalties until well into the pilot [Section 8.3]

• some of the larger fixed penalties have had much lower payment rates

(due in most part to less severe penalties being available in court)

[Section 8.4]

• in general, a higher proportion of fixed penalties were issued to older

vehicles [Section 8.4]

• around £1 million of funds were recovered to be recycled into further

enforcement (against expenditure of around £12 million). [Section 8.4]

The benefits

• in addition to addressing criminality, the use of ANPR-enabled intercept

teams also contributes to wider objectives, specifically road safety and

excise collection [Section 8.4].

107

8.1 Context

8.1.1 Background

Although the results from Laser 1 were extremely encouraging, there were only

limited resources available to fund the national roll-out and operation of ANPR.

If the benefits of ANPR were to be maximised within existing budgets, an

innovative funding approach was required. Following a submission to HM

Treasury, conditional approval was given to test a cost recovery programme

for dedicated ANPR-enabled intercept teams from 1 June 2003. This allowed:

• police to target vehicle documentation offences, and crime in general, using

ANPR-enabled dedicated intercept teams, maximising the use of and building

upon existing intelligence

• the activity to be funded through receipts from fixed penalties issued for

vehicle offences detected by the ANPR-enabled dedicated intercept teams.

If certain conditions can be satisfied, HM Treasury can grant permission for

Government Departments to recover the costs of enforcement and detection.

A case was made and was accepted. This was on the basis of substantial

evidence that ANPR contributed to the Home Office policy objectives of

tackling criminality.

This therefore allowed the Home Office (the sponsoring Department) a period

of two years to pilot the cost recovery programme, assess the benefits and come

to a policy decision as to whether the pilot should be rolled out nationally – a

decision that would require primary legislation.

8.1.2 The Laser 2 concept

In broad terms, costs are recovered for Laser 2 from Fixed Penalty Notices

issued at the roadside by ANPR-enabled intercept teams. In order to separate

the fixed penalty receipts from those generated by existing activities by officers,

it was necessary to introduce a distinguishing mark/feature on these fixed

penalty notices issued. The simplest method was for the ANPR teams to use

fixed penalty pads with a special identifier or in a different colour from the rest

of the force. In this way, the central ticket offices could separate tickets arising

from ANPR teams from other activities.

For each stop, the officer recorded on a roadside collection sheet key information

relating to the stop and its outcomes (see appendix B). This information served

two purposes: performance evaluation and for auditing fine monies. Information

from these sheets was then collated, entered onto a database by the analyst

and provided to the PSU on a monthly basis, while the original sheets were

kept for audit.

108

Force central ticket offices then placed an identifier on each fixed penalty

that the ANPR team generated when they enter the data onto their system.

When the monies had been paid (via the magistrates court), this allowed

the relevant fine revenues to be identified and ring fenced for cost recovery.

The magistrates’ courts pass all fine revenue to the DCA on a monthly basis.

The element to be cost recovered was then forwarded onto the Home Office,

together with a record stating how the revenue is divided across forces.

Appropriate revenue costs were then paid by the Home Office to the forces

on a quarterly basis in arrears.

In practice, the FPNs that could be cost recovered fell into the following basic

categories:

• endorsable ticket (6 points) and £200 fine: no insurance

• endorsable ticket (3 points) and £60 fine: driving other than in accordance

with the licence

• non-endorsable ticket, but £60 fine for no VED

• non-endorsable ticket, but £60 fine for no MOT

• non-endorsable ticket, but £30 fine: eg no seatbelt, using mobile phone,

obscured VRM.

8.2 Conditions of cost recovery

8.2.1 HM Treasury requirements

In the 1998 Public Expenditure Survey, HMT identified certain conditions

where fines and penalties could be cost recovered, specifically where:

• performance against policy objectives is likely to be improved

• arrangements are in place to ensure that the activity will not lead to the

abuse of fines and penalty collection as a method of revenue rating, and

that operational priorities remain undistorted

• revenues will always be sufficient to meet future costs, with any excess

revenues over costs being surrendered

• costs of enforcement be readily identified and apportioned without undue

bureaucracy, and with interdepartmental and inter-agency agreement,

where necessary

• savings can be achieved through the change and there are adequate

efficiency regimes in place to control costs, including regular efficiency reviews.

109

8.2.2 How Laser 2 sought to meet these requirements

In order to ensure that Laser 2 forces met these conditions, the national ANPR

Steering Group (which includes representatives from the Home Office, DfT,

ACPO, DCA, and HMT) carried out the following:

• put in place an ANPR cost recovery handbook, which set out a number of

rules and guidelines to ensure that performance against policy objectives

was measured and operational priorities were not distorted. Specifically the

rules covered:

– the objectives of Laser 2

– the arrangements for cost recovery

– what fixed penalties were covered by the cost recovery scheme (as per

Appendix D)

– what expenditure was covered by the scheme

– financial controls and governance arrangements, including efficiency

and effectiveness

– project monitoring

– asked for written submissions (‘operational cases’) from forces as to how

they planned to operate ANPR-enabled intercept teams within these rules

and guidelines

– requested that each of the forces’ ACPO officer sign a letter explicitly

agreeing to the rules and guidelines as set out in the handbook as part

of their operational case submission.

On this basis, the PSU accepted forces onto the ANPR pilot.

8.2.3 Monitoring arrangements

Where delivery or performance fell below acceptable standards by the National

ANPR Project Board, the PSU intervened as follows:

• on the first occasion where there was evidence of under-performance,

a representative from Police Standards Unit would visit the force to

undertake a high-level review of operations to identify possible reasons

for under-performance and agree steps to be taken to remedy these

• if there was ongoing under-performance, a representative from PSU would

again visit the force and undertake a more detailed review of operations.

Following this, PSU would write to the Chief Officer’s representation:

110

– setting out the findings of the review

– detailing actions that must be taken to address these

– setting out a timetable for delivery of these actions

– pointing out the consequences of failing to meet

• if there was no noticeable improvement, following this, the national ANPR

Project Board reserved the right to suspend a force from the pilot. Prior to

doing this, the PSU undertake to give at least one calendar month’s notice.

In practice this did not prove to be necessary since, as we will see later,

following the interventions, normally performance improved.

In terms of what constitutes an acceptable performance, no specific targets

were set. However indicative performance measures were:

• number of days per week intercept team(s) operated – the operational inputs

• vehicle stops per hour of intercept officer – the utilisation

• arrests per FTE officer – the effectiveness of the operation

• FPNs issued per FTE officer – cost recovery.

Finding 54. The national ANPR Programme Board set in place robust

controls and processes for Laser 2 to help ensure that the HMT

conditions for cost recovery were achieved.

8.3 Factors affecting the introduction of cost recovery

In relation to the cost recovery aspect, the following points are worth noting:

• prior to the start of Laser 2, the Home Office relaxed the guidance to allow

(ANPR) officers to issue up to three FPNs at a time, though this was limited

to one endorsable and two non-endorsable tickets

• on 1 June 2003, the Home Secretary introduced four new fixed penalties,

namely:

– failing to supply details necessary to identify an offending driver, contrary

to s172, Road Traffic Act 1998 (RTA), with a penalty of £120

– having no insurance, contrary to section 143 RTA, with a penalty of £200

– having no MOT certificate, contrary to section 47 RTA with a penalty of £60

– not displaying a vehicle excise licence, contrary to section 33 Vehicle

Excise and Registration Act (VERA) 1994 was increased to £60.

111

While this coincided with the start of Laser 2, in practice many forces had

not printed the necessary fixed penalty pads for these tickets to be issued.

Hence many forces were unable to issue these tickets for a number of months

• due to the rapid start to Laser 2 (forces were notified of their inclusion in

March 2003 for a June 2003 start), during the initial stages many forces

found themselves unable to cost recover from FPNs. Part of this was down

to a lack of preparation (tickets books were in some cases unavailable until

August 2003) while in other cases it became difficult to differentiate between

the tickets that were issued by ANPR teams and those issued by other

officers in the force. However, all these issues were eventually addressed

with new ticket books becoming available for the new offences

• following negotiations with HMT, from November 2003 forces were allowed

to issue Conditional Offers for no insurance or MOT where a HO/RT1 had

been issued at the roadside but the driver had failed to produce the relevant

documents within seven days. To support this process, forces required an

upgrade to their ticketing system. At the end of the pilot period, 12 of the 23

forces had not had carried out this upgrade, while a further two only received

the upgrade during June 2004

• on 1 December 2003, the DfT introduced a new FPN offence of using a mobile

phone while driving, with offenders subject to a £30 fine. However, ACPO

decided that, for the first two months, police forces in England and Wales

would issue verbal warnings to drivers instead of issuing a FPN. This was

included within the cost recovery process.

8.4 Fixed penalty notices issued and paid

8.4.1 FPNs issued during Laser 2

Since June 2003, Laser 2 intercept teams issued 54,035 FPNs that were cost

recoverable to 52,765 drivers, that is to say that there were only a relatively

few occasions where more than one FPN was issued to a driver. Figure 8.1

shows the breakdown of tickets by ticket type.

112

Figure 8.1: FPNs issued by ANPR intercept teams

Overall, the most frequently issued ticket type was a non-endorsable £60 FPN

for failing to display VED/no VED (22,825 tickets) and non-endorsable £30 FPNs

for a variety of minor offences (20,290). While non-compliance with insurance

and no MOT is seen to be as frequent as VED evasion (see section 2.1 above),

relatively few FPNs were issued for no insurance (6,299) and no MOT (1,496).

This probably reflects the fact that insurance and MOT documentation tends

not to be carried by drivers and would be followed up by means of an HO/RT1.

Finding 55. On the basis of the staffing levels reported by forces during

Laser 2, in one year a Full Time Equivalent intercept officer would expect

to issue:

• 36 endorsable tickets for no insurance

• 16 other endorsable tickets

• 9 non-endorsable tickets for no MOT

• 132 non-endorsable tickets for no VED and

• 117 non-endorsable tickets (£30)

This amounts to a total of 310 FPNs per annum. Given that the number

of fixed penalties available to officers increased during Laser 2, it may

be expected that the actual number is slightly higher than this.

Overall, the number of fixed penalties issued increased from week 1 (367

tickets issued) to a peak in week 25 (1,470 tickets issued) and fell away to

around 1,000 per week between weeks 48 to 55. Figure 8.2 shows the number

of tickets during Laser 2.

38%

41%

12%

Non-endorsable, £60 (No VED)

Non-endorsable, £30

Non-endorsable, £60 (No MOT)

Endorsable £200

Endorsable £60

113

Figure 8.2: FPNs issued by ANPR intercept teams by week

The most significant change in FPN tickets issued was for no VED. The increase

from week 1 (when 77 tickets issued) to the peak in weeks 33 and 34 (when 736

and 733 tickets were issued respectively) probably reflects the use of the recently

introduced offence of failure to display VED and the growing confidence of ANPR

teams to issue tickets for this.

There was significant week-on-week variation in the issue of lower value

(£30) non-endorsable tickets, with a slight upward trend after Christmas 2003

– coinciding with police enforcement of mobile phone offences. The issue of

other tickets (non-endorsable £60 tickets for no MOT and endorsable £60 and

£200 tickets for no insurance) showed a more stable profile, gradually increasing

from a low base to stable level around week 13.

Finding 56. For the two largest volume tickets issued at the roadside

(No VED and non-endorsable minor offences, which together accounted

for nearly 80% of tickets issued) there was significant variation in the

volume of tickets issued during the pilot. In particular the dramatic decline

of FPNs issued for no VED to approximately a third of the peak achieved

at the beginning of 2004 would suggest a fundamental change – this

coincides with both the temporary suspension of the DVLA database

(see section 2.5 above) and the introduction of continuous registration

(February 2004 – Week 34).

100

200

300

400

500

600

700

800

7 10131619222528313437404346495255

Weeks

FPNs issued

Non-endorsable, £60 (No VED)

Non-endorsable, £30

Non-endorsable, £60 (No MOT)

Endorsable £200

Endorsable £60

114

8.4.2 FPNs issued by type of stop generated

In terms of how FPNs were generated, 66% of FPNs issued came from officer

observation vehicle stops, while only 34% came from ANPR-generated stops.

This finding is unsurprising – £30 non-endorsable and £60 endorsable tickets

could not come from ANPR hits alone and currently the data underlying the

ANPR systems did not provide vehicle insurance or MOT details.

It was still possible for an ANPR hit to lead to these FPNs being issued – for

example a vehicle passing the ANPR may be flagged as having no VED, but

when checked an officer may find that the vehicle has recently been taxed.

The driver, however, may also have been observed driving without a seatbelt

on, and therefore received a £30 non-endorsable FPN. Figure 8.3 shows that

the majority of ANPR tickets (63%) were for no VED, while 52% of observation-

generated stops were for £30 non-endorsable offences.

Figure 8.3: FPNs issued by intercept source

Finding 57. Only a third of FPNs issued were generated by ANPR

intelligence stops as opposed to officer observation. This reflected the

data used by forces – primarily a PNC extract and DVLA data. ANPR

intercept teams did not have the motor insurance or MOT databases to

trigger vehicle stops. These databases are expected to be available in

the medium term and will provide another useful intelligence source.

8.4.3 FPNs issued by estimated year of vehicle registration

Figure 8.4 shows the number and type of FPNs issued per 100 vehicle stops

by the (estimated) year of vehicle registration. It should be noted that there

were very few vehicle stops for vehicles registered in 2004 and results for

2004 are therefore not considered to be significant.

Non-endorsable, £60 (No VED)

Non-endorsable, £30

Non-endorsable, £60 (No MOT)

Endorsable £200

Endorsable £60

17%

63%

31%

10%

52%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

ANPR Observation

115

Overall, drivers of older vehicles were more likely to be issued a FPN than

drivers of new vehicles, with vehicles registered in 1990 most likely to receive

a FPN from an ANPR intercept team. In terms of ticket types, there was also

an uneven distribution. No FPNs were issued for failure to have an MOT for

drivers of vehicles newer than 2001 – this is logical as an MOT is only a

requirement of vehicles older than 3 years. FPNs were more likely to be

issued to older vehicle than newer vehicle drivers for no insurance and no

VED, while newer vehicle drivers were more likely to receive low value (£30)

non-endorsable FPNs (both using a mobile telephone and non seat belt).

Finding 58. Analysis of FPNs issued per 100 vehicle stops by the (estimated)

year of vehicle registration confirms existing offender profile information,

much of which is already known by traffic officers – older vehicles are

more likely to be untaxed and uninsured than newer vehicles. The analysis

does, however, confirm and provide useful quantification of this relationship

and confirms the strong link with document and volume crime.

Figure 8.4: FPNs issued per 100 stops by year of vehicle registration

8.4.4 FPNs issued by force and their value

The total value of fines associated with the 54,035 FPN tickets issued by

ANPR intercept teams was £3,515,320, ie an average of £65 per ticket issued.

Figure 8.5 shows the breakdown of tickets issued and their value.

Non-endorsable, £60 (No VED)

Non-endorsable, £30

Endorsable £200

Endorsable £60

Non-endorsable, £60 (No MOT)

No Data

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

Year

of car

registration

FPN tickets issued per 100 stops

116

Figure 8.5: FPNs issued by force and their total value

On average, 12% of tickets issued were for no insurance, 6% for other

endorsable offences, 3% of tickets for no MOT, 42% for no VED and 38%

of tickets were £30 non endorsable tickets. As can be seen from Figure 8.5,

there were significant differences in the issuing force profile of tickets issued.

For example:

• 37% of the tickets issued by Nottinghamshire were for no insurance, while

only 16% of their tickets were for no VED

• 70% of Cheshire’s FPNs were for no VED, with only 10% of their tickets

were £30 non-endorsable tickets

• 34% of the Metropolitan Police Service’s tickets were for no insurance,

19% for other endorsable offences, and 16% of their tickets were for no VED.

Finding 59. On average, forces issued one FPN for every 6 officer hours

on ANPR duties. There was, however, significant force by force variation

in the number and profile of tickets issued, reflecting different operating

conditions, priorities and working practices. For example Northumbria

issued one FPN for every 24 officer hours on ANPR duties, while North

Wales issued one FPN for every 2½ officer hours on ANPR duties.

0 1,000 2,000 3,000 4,000 5,000

Lancashire

North Wales

West Midlands

Cambridgeshire

Metropolitan

Lincolnshire

Hertfordshire

Hampshire

Nottinghamshire

Staffordshire

Northants

Cleveland

Merseyside

Leicestershire

City of London

West Yorkshire

Kent

Greater Manchester

Avon & Somerset

Cheshire

North Yorkshire

£331,510

£311,760

£153,650

£179,350

£196,320

£138,580

£153,100

£125,850

£86,030

£134,890

£60,900

Warwickshire £80,000

Northumbria

FPN tickets issued

Non-endorsable, £60 (No VED)

Non-endorsable, £30

Non-endorsable, £60 (No MOT)

Endorsable £200

Endorsable £60

£261,790

£96,500

£162,930

£179,350

£196,320

£36,670

£87,270

£105,140

£126,080

£107,410

£214,810

117

8.4.5 Fine monies recovered

Laser 2 forces were only able to recover costs from those tickets issued by

ANPR intercept teams that had been paid. These monies were not paid direct

to the police rather they were collected by the magistrates’ courts who then

forwarded them on to the DCA. In the analysis below of tickets paid, it is

important to understand that:

• not all tickets are paid – where this happens, the fine is registered with the

magistrates’ court and pursued through the courts. In these circumstances,

fines monies cannot be used for cost recovery

• even where fines are paid, there is a delay between the fines being issued

and the money being made available for cost recovery, specifically:

• there is a permitted payment period of (typically) 28 days between the fine

being issued and the fine being lodged with the magistrates’ courts

• the magistrates’ courts record the payment being made and identify ANPR

payments to DCA per calendar month

• each quarter the magistrates’ courts forward all fine revenue (including fines

associated with ANPR) to the DCA. DCA then pass on the relevant monies

to the Home Office, who then assign this money on the basis of the fines

paid in each Laser area.

Thus in the analysis below, only ticket revenue up to the end of March 2004

(which had been reconciled and received by the Home Office) is presented.

Figure 8.6 shows the number and cash value of tickets paid (without the fine

being registered) by force area as recorded by the DCA. Thus of the 42,592

tickets issues between 1 June 2003 and 31 March 2004 (approximate value

£2,858,400), 20,870 tickets (42%) had been paid to magistrates’ courts without

the fine being registered and the money forwarded to the Home Office –

amounting to £925,580, ie 32% of ticket value.

Finding 60. In general, 82% of the FPNs issued by police are paid without

the fine being registered,

although this figure includes parking tickets.

The 42% payment rate achieved for FPNs issued by ANPR intercept

officers was thus low.

Unfortunately, due to the lack of connectivity between the force data systems,

it was not possible to track individual tickets and their payments. Therefore no

analysis on ticket payment rates by, for example, ethnicity, age of vehicle, and

triggering database is possible. Seventeen forces’ Central Ticket Offices (CTOs)

were, however, able to provide data on total tickets paid for the 56 week period

– this data was similar to that provided by DCA, though it was not reconciled.

Motoring Offences and Breath Test Statistics, Home Office (2004)

118

Using this CTO data for weeks 5-56 (ie one year), 130 officer FTEs issued:

• 4,150 tickets for no insurance, with a 14% payment rate

(approximately £119,600)

• 1,956 other endorsable tickets, with a 36% payment rate

(approximately £41,800)

• 1,235 tickets for no MOT, with a 35% payment rate (approximately £25,600)

• 19,563 tickets for no VED, with a 32% payment rate (approximately £378,500)

• 13,269 non-endorsable £30 tickets, with a 69% payment rate

(approximately £274,000).

Figure 8.6: FPNs issued by force (1 June 2003 and 31 March 2004)

Avon & Somerset 1,223 475 39% £23,020 £48.46

Cambridgeshire 2,176 715 33% £37,660 £52.67

Cheshire 1,285 553 43% £34,890 £63.09

Cleveland 1,344 593 44% £29,690 £50.07

City of London 1,628 435 27% £24,320 £55.91

Greater Manchester 1,073 213 20% £16,300 £76.53

Hampshire 1,915 887 46% £55,890 £63.01

Hertfordshire 1,769 1,020 58% £43,460 £42.61

Kent 1,665 904 54% £35,560 £39.34

Lancashire 4,754 2,004 42% £83,480 £41.66

Leicestershire 1,792 794 44% £45,010 £56.69

Lincolnshire 1,780 951 53% £33,670 £35.40

Merseyside 1,780 490 28% £39,340 £80.29

Metropolitan 1,896 1,008 53% £50,010 £49.61

North Wales

4,386 2,375 54% £119,950 £50.51

North Yorkshire 1,038 619 60% £24,260 £39.19

Northamptonshire 1,731 353 20% £29,300 £83.00

Northumbria 467 237 51% £12,530 £52.87

Nottinghamshire 2,068 621 30% £33,790 £54.41

Staffordshire 1,499 779 52% £39,060 £50.14

Warwickshire 920 319 35% £21,680 £67.96

West Midlands 2,661 981 37% £49,430 £50.39

West Yorkshire 1,741 739 42% £43,280 £58.57

Total 42,592 18,065 42% £925,580 £51.24

FPNs FPNs FPN payment FPN monies Average FPN

issued paid rate paid revenue

The DCA return for North Wales (5,180 tickets paid) has been amended to the figure recorded by

North Wales Central Ticket Office (2,375) as the DCA figure is not reliable. This would mean that

the average FPN revenue paid for North Wales (£23.16) is below the minimum FPN value (£30).

119

Finding 61. Payment rate appears to be a direct function of the fine level

associated with the ticket. No insurance (which incurs a £200 fine) had a

very low payment rate (14% of tickets issued) while non-endorsable £30

fines have a much higher payment rate (69%). It is interesting to note that

the ticket payment for no VED was lower (32%) than general endorsable

tickets (36%) – this could be because no VED will necessitate additional

expenditure (taxing the vehicle) and therefore by association is a larger fine.

Finding 62. CTO figures show that there was significant force variation

in payment levels. Some CTOs, for example Greater Manchester and

Hampshire, were able to achieve much higher FPN payment rates than

others, for example Northamptonshire and Avon and Somerset.

Finding 63. The differences in the payment rate of tickets, particularly

for no insurance, are in part due to the level of fine imposed at courts.

A number of forces reported an average fine level of less than £200 and

less than 6 points, with a longer period to pay. A consistent approach to

the fines imposed at court could have a significant impact on the

payment rate of ANPR tickets.

Finding 64. On this basis, a FTE ANPR intercept officer would on average

issue 310 tickets per annum (approximate value £19,900) of which £6,400

(32%) would be paid directly to magistrates’ courts. Given that staff

costs of a police constable are typically £35,500, including employer

contributions, it is clear that ANPR cost recovery does not cover basic

employment costs.

120

Figure 8.7: FPNs payment rates by force from CTO data (weeks 5-56)

Finding 65. A key point in comparing Figures 8.6 and 8.7 is that payment

rates vary between CTO and DCA for some forces. For example, DCA

record payment rate of FPNs issued by Greater Manchester Police as 20%,

while Greater Manchester’s CTO have an average payment rate of 45%.

This difference is most probably due to a clerical error or delay in filling

in the payments received form at the magistrates’ courts (ie the magistrates’

courts return to DCA) and needs to be addressed accordingly.

Avon & Somerset 8% 39% 22% 21% 51%

Cambridgeshire 14% 41% 31% – 62%

Cheshire 14% 42% 36% 33% 73%

City of London – – – – –

Cleveland 12% 45% 27% 31% 72%

Greater Manchester 31% 54% 43% 29% 83%

Hampshire 19% 41% 40% 34% 72%

Hertfordshire 9% 52% 16% 40% 79%

Kent – – – – –

Lancashire 7% 43% 31% 29% 70%

Leicestershire 17% 22% 62% 44% 41%

Lincolnshire – – – – –

Merseyside 26% 38% 34% 38% 69%

Metropolitan – – – – –

North Wales 14% – 42% 22% 75%

North Yorkshire 18% 33% 32% 40% 70%

Northamptonshire 7% – 22% 21% 57%

Northumbria – – – – –

Nottinghamshire – – – – –

Staffordshire 26% – 34% 38% 79%

Warwickshire 27% – 43% 44% 72%

West Midlands 15% – 29% 19% 54%

West Yorkshire 14% 19% 30% 24% 68%

Average 14% 36% 35% 32% 69%

Endorsable Endorsable Non-Endorsable Non-Endorsable Non-Endorsable

£200 £60 £60 (No MOT) £60 (No VED) £30

121

8.4.6 Costs incurred and fine monies recovered

As part of Laser 2, forces were able to recover their allowable costs from the

fine revenues received. Prior to Laser 2 commencing, a number of issues were

flagged up to forces by the Home Office:

• first, most forces already had an under-utilised ANPR potential, and

therefore it was not envisaged that Laser 2 would require substantial

investment in ANPR readers

• second, unlike the safety camera scheme, ANPR operations would not be

accompanied by large scale communications and publicity

• thirdly, the potential scale of cost recovery was recognised to be limited and

• finally, as Laser 2 was a pilot, there was no long-term agreement to fund

activity. Investment in equipment for the medium term was inappropriate.

On this basis, forces were asked to focus on recovering revenue costs

associated with Laser 2, rather than other revenue costs (such as ANPR

and vehicle maintenance and lease costs).

122

Figure 8.8: Staff costs by force and monies recovered by force (June 2003-March 2004)

The monies realised by forces relative to the costs incurred by the pilot were

small. However forces recognised the value of these monies, for example:

“Had it not been for the money we recovered through hypothecation

our force would not have been in the position to invest in key

support areas for ANPR such as improving our databases

investing in automating some of our processes.”

Avon & Somerset* £530,000 £23,020 £506,980

Cambridgeshire £343,216 £37,660 £305,556

Cheshire £227,332 £34,890 £192,442

City of London £258,360 £29,690 £228,670

Cleveland* £290,000 £24,320 £265,680

Greater Manchester* £900,000 £16,300 £883,700

Hampshire* £560,000 £55,890 £504,110

Hertfordshire* £240,000 £43,460 £196,540

Kent* £760,000 £35,560 £724,440

Lancashire £563,313 £83,480 £479,833

Leicestershire* £540,000 £45,010 £494,990

Lincolnshire* £680,000 £33,670 £646,330

Merseyside £291,987 £39,340 £252,647

Metropolitan £1,154,680 £50,010 £1,104,670

North Wales £792,756 £119,950 £672,806

North Yorkshire £191,450 £24,260 £167,190

Northamptonshire £867,413 £29,300 £838,113

Northumbria* £330,000 £12,530 £317,470

Nottinghamshire £335,750 £33,790 £301,960

Staffordshire £588,781 £39,060 £549,721

Warwickshire £237,660 £21,680 £215,980

West Midlands £832,590 £49,430 £783,160

West Yorkshire* £420,000 £43,280 £376,720

Total £11,935,288 £925,580 £11,009,708

Force Staff Costs Monies recovered

Deficit

Following agreement with HMT, 2% of the fines monies recovered was retained by APCO for

ANPR development

Tracie O’Gara, ANPR Project Manager, Lancashire Police

* indicates forces where staff costs for the first financial year of ANPR were estimated from business case projections.

Non estimated figures are actual costs provided by forces

123

Finding 13. Overall the cost recovery process realised an additional

£926,000 in total to the 23 Laser 2 forces over a nine-month period.

While these monies did not cover the costs of the enforcement

(approximately £12 million for the same period), these monies were

seen to be worthwhile, for example in helping to improve the intelligence

capability of the ANPR teams and providing the administrative support

for the teams.

8.4.7 Wider benefits to society

In addition to addressing criminality (in part paid for by offenders), cost recovery

of fixed penalty notices and the use of ANPR-enabled intercept teams also

contributes to wider objectives, specifically road safety (eg through enforcing

seat belt wearing and not using mobile telephones while driving) and excise

collection (eg ensuring that all vehicles on the road are appropriately taxed).

124

Findings: ANPR arrest

outcomes

This section of the evaluation provides an analysis of the arrest outcomes

(that is to say, what proportion of the arrests led to successful convictions).

In particular it highlights: [Section 9.2]

• an average ANPR FTE will contribute around 31 offences per annum

towards to the Offences Brought to Justice Target – this is over three time

more than general policing

• if ANPR intercept teams were rolled out one per BCU, this would

contribute 26,400 additional OBTJs per annum towards the target –

a contribution of around 15% to the Government’s target

• since Laser involves redeploying existing resources more effectively,

this represents little incremental cost and hence good value for money.

125

9.1 Context

Narrowing the Justice Gap (NJG) – the difference between the number of crimes

that are recorded and the number that result in the perpetrator being brought

to justice – is a key measure of the effectiveness of the criminal justice system

(CJS) and an important way in which public confidence in the CJS can be

improved. The police play a critical role in NJG through the detection of crime

and, in partnership with the Crown Prosecution Service, the successful

prosecution of offenders.

The CJS has been set a target to increase the number of offences brought

to justice from 1.02 million offences in 2000-2001 to 1.2 million by 2005-06.

Work on NJG is being led by the Justice Gap Taskforce comprised of senior

members from each of the CJS agencies.

9.2 Tracking the outcome of ANPR arrests

ANPR was identified by the Justice Gap Taskforce as having the potential to

make an important contribution to the delivery of the NJG target. In order to

assess the scale of that contribution, forces were asked to collect outcome data

for ANPR arrests. The Justice Gap Action Team also asked the evaluation team

to model the potential impact of rolling out Laser 2 nationally on the NJG target.

The findings of this work, together with an assessment of the impact of ANPR

on NJG objectives relative to other initiatives, are discussed in this section.

At this point it is also worth noting the counting rules used to measure the

number of Offences Brought To Justice (OBTJs). OBTJs consist only of

recorded or notifiable offences. These cover all ‘Indictable Only’ (the most

serious offences which must be prosecuted through the Crown Court) and

‘Either Way’ (serious offences which must be prosecuted through the Crown

Court) offences and a small number of ‘Summary’ offences. A full list of these

is included in Appendix G. It is also important to note that recorded offences

exclude a number of the misdemeanours/crimes dealt with by ANPR, such as

driving while disqualified and driving under the influence of alcohol/drugs.

Laser 2 forces were therefore asked to use the custody reference numbers

collected as part of the roadside pro forma to identify the outcome of the arrests

made. Not all forces were able to undertake this tracking process – different

IT systems adopted by police forces, courts and other agencies made case

tracking a considerable challenge for the majority of Laser forces. In total,

11 forces provided outcome information in the required format for this NJG

study. These forces, listed below, represented a cross-section of force size

and geography and could therefore be considered a reasonably representative

sample for modelling purposes.

Narrowing the Justice Gap Framework, Home Office (October 2002)

126

Cleveland Kent North Yorkshire Greater Manchester

Hampshire Lancashire West Midlands North Wales

Hertfordshire Lincolnshire Cheshire

To give the maximum opportunity for all cases to have progressed through

the system and reached some form of disposal (that is where an offence is

considered dealt with and where it exits the prosecution system), the modelling

exercise considered only those arrests made during the first three months of

Laser 2 (ie between June and August 2003). No comparable data was routinely

collected as part of Laser 1 and therefore arrests made prior to the start of

Laser 2 were not considered.

The data collected for the tracking of ANPR outcomes was based on the premise

that arrests could be disposed of at a number of points, either by police or by

the courts. While it was not possible to track the outcomes of arrests that

were handed over to other agencies (eg immigration and other police forces),

nor arrests for outstanding warrants, those forces providing data were able to

identify the outcome of cautions as well as the results of court proceedings for

recorded offences.

The findings for the ANPR arrest tracking aspect for recorded offences were

as follows:

• The forces taking part in the study tracked 840 arrests. This represented

89% of all arrests that were captured by ANPR intercept deployments in

the 11 areas. The remaining 11% could not be tracked through the system,

either due to discrepancy between the Custody Reference Number recorded

on the pro forma with the force’s tracking database or where the Custody

Reference Number was missing. These 840 arrests were delivered by 46 FTE.

• The 840 arrests that could be tracked related to 1,425 offences. Of these,

1,094 (77%) were dealt with by the arresting police force (ie not transferred

to other forces or agencies) and related to ‘new’ offences. The balance of

the 331 arrests were warrant arrests (116) or the offence was dealt with by

a force/agency other than by the arresting force (215). Because of the difficulty

in tracking across agency, the modelling exercise was based on only these

1,094 offences tracked.

• Of the 1,094 offences tracked, charges were refused on 137 occasions, ie the

police took no further action at that time and the specific charges dropped.

A further 56 (5%) were dealt with by caution. Of these, 54 were defined as

recorded offences according to Home Office guidelines (Appendix G) and

therefore contributed to OBTJ.

127

• Thus, of the 1,094 offences that were tracked, charges were preferred

(someone was charged) for 901 offences (ie 1,094-137-56). This equates

to 82% of offences resulting in charge. Of these 901 charges, 272 were

for recorded offences (30%). These 272 recorded offences resulted in the

following pleas:

– 43% (118 offences) led to an early guilty plea, that is where the defendant

admitted guilt at the first opportunity

– 5% (13 offences) led to a late guilty plea, that is where the defendant

previously pleaded not guilty but changed this to a guilty plea prior to trial

– 15% (42 offences) led to a not guilty plea, that is where the defendant

pleaded not guilty and trial proceedings began

– 36% (99 offences) for which no plea data was available.

Therefore 131 (118+13) recorded offences (48%), where charges were made,

were successfully disposed without trial proceedings beginning (though this

could be higher due to the 99 offences where no plea data was available).

• Of the 42 offences where there was a not guilty plea:

– 12 offences (29%) resulted in successful convictions

– 27 offences (64%) were acquitted

– for 3 offences (7%) the result was either on-going or no information

was provided.

• Of the 99 offences where there was no plea data available:

– 40 offences (40%) resulted in successful convictions

– 3 offences (3%) were acquitted

– for 56 offences (56%) the result was either on-going or no information

was provided.

• In terms of outcomes:

– there were successful outcomes for 237 recorded offences (54 cautions,

188 early guilty pleas, 13 late guilty pleas and 52 convictions, though not all

of these have gone to court for trial), ie 73% of recorded offences had

successful outcomes

– for a further 59 recorded offences (3 not guilty pleas and 56 no plea

information available), there either the case was still on-going and there

was no case disposal data, ie for 18% of recorded offences there was

no case disposal data

– in only 30 recorded offences (27 not guilty pleas resulting in no conviction

and 3 cases where there was no plea information available), there was no

conviction, ie 9% of recorded offences resulted in no conviction.

128

The disposal of offences is summarised in Figure 9.1 below.

Figure 9.1: How offences from a sample of 945 ANPR arrests were brought to justice

From the above analysis, the key findings are:

• in terms of new OBTJ contribution:

– The review period related to 13 weeks (one quarter year) of data collection

and involved 46 FTE officers. These officers made 840 arrests that

contributed 237 OBTJs. This equates to roughly to 21 OBTJs per ANPR

FTE (237 x 4 quarters / 46 FTE officers)

– In addition, 59 of the 272 recorded offences (22%) for which charges

were pressed were still pending an outcome or no outcome had been

recorded. Assuming that the profile of outcomes remains consistent,

there would be approximately 30 more offences brought to justice.

Over a year this would add a further 2-3 OBTJs per FTE per annum (24)

– Further, for 11% of the arrests made there was no case tracking information

(including outcome). If half of these arrests follow a similar profile (the other

50% assumed to be arrests that lead to nothing and hence unreported)

then that would mean that ANPR intercept teams would deliver 25 OBTJs

per FTE.

(137)

(901)

Refused

charge

Charges

preferred

Caution

received

(56)

Recorded

offence

(272)

Non-recorded

offence

(629)

(137) (901)

Arrests tracked

through system

Recorded

offence

(54)

Non-recorded

offence

(2)

Number of

offences

(1,425)

Other

outcome

(331)

Early guilty

(118)

Late guilty

(13)

Not guilty

(42)

Not known

(99)

Convicted

(12)

Not convicted

(27)

Not convicted

(3)

Unknown/ongoing

(56)

Unknown/ongoing

(3)

Convicted

(40)

Not TrackedNot Tracked

Number of

offences tracked

(1,094)

Arrests made by

ANPR teams

(945)(945)

(840)(840)

129

Given that the use of ANPR intercept teams is a relatively new approach

and that the data tracking exercise is related to the early stage of Laser 2,

it would not be unreasonable to expect an increase in performance over

time. For example, for those forces providing case tracking information,

the arrest rate in the first quarter of Laser 2 was 83 arrests per FTE. For the

final quarter this had risen to 117 arrests per FTE (+41%), while for all forces

the equivalent increase was +23%. It would not be unreasonable, therefore,

to argue that when operating in a steady state, ANPR intercept teams would

deliver 31 OBTJs per FTE.

Finding 67. As a conservative estimate, an average ANPR FTE will be

able to contribute 31 offences towards the OBTJ target. This compares

favourably against the 9 OBTJs that general policing duties would deliver.

• ANPR will also have an impact on disposal of existing recorded cases;

specifically during Laser 2, there were 1,813 arrests relating to outstanding

warrants. It is not possible to estimate what proportion of these warrants

related to recorded offences, however it is likely that a significant proportion

relate to offences where charges had been preferred. While some warrants

may relate to failure to produce documentation at the police station

(HO/RT/1 offences), it is probable that this is indicative of other offences.

Further, close working with courts in using the outstanding warrant database

by ANPR intercept teams will ensure that warrants are executed more quickly

and will help to maintain the integrity of the warrant process.

• The modelling exercise suggested that a further 215 (of the 1,425) offences

detected by ANPR intercept teams were passed over to other forces/agencies.

These will make a contribution to OBTJ, however there is no data on which

to base an estimate.

• The contribution of ANPR could improve further if offences such as driving

while disqualified were included in NJG figures. In the sample of 840 arrests,

there were 159 offences of driving while disqualified (118 of which have

already led to convictions) and 4 drink-driving offences all of which have

already been successfully convicted. If the above offences were to become

part of the NJG group, 15 additional offences would be brought to justice

per officer per annum.

130

If it is assumed that each ANPR intercept officer generates 31 OBTJs per year:

• When Project Laser is rolled out nationally, it is envisaged that approximately

1,200 uniformed FTE officers will be operating as part of ANPR intercept

teams (on average one intercept team per BCU). These officers will be

transferred from existing policing duties, rather than be additional staff.

• Prior to operating as ANPR intercept officers, these officers would on

average have delivered 9 OBTJs each per annum.

Finding 68. On the basis of the analysis presented, the national roll-out

of ANPR will deliver 37,200 OBTJs per annum (1,200 x 31). Because the

roll-out of the ANPR intercept teams would be resourced by existing

officers, not all of these OBTJs will be new offences, ie the officers will

have made arrests for recorded offences in their normal duties. On this

basis, the roll-out of the ANPR intercept teams will deliver an additional

26,400 OBTJ per annum, ie approximately 15% of the target of additional

offences brought to justice.

Finding 69. Project Laser is one of 13 initiatives assessed by JGAT as

contributing to the NJG target. On the basis of the above findings, the

national roll-out of Laser 2 makes a major contribution to the achievement

of the OBJT target, ie it is one of the more significant projects in terms

of OBTJ (approximately 15% of the additional offences required to meet

the OBTJ target). It is also worth noting that the roll-out of Project Laser

nationally would see existing resources (officers) deployed more effectively,

rather than the use of additional resources (and hence increased cost).

131

132

Conclusions and

recommendations

This section of the evaluation provides overall conclusions and

recommendations that can be drawn from the evaluation of Laser 2.

10.1 Conclusions

On the basis of Laser 2 evaluation findings, the following conclusions have

been made:

10.1.1 Context

C1. The use of ANPR-enabled intercept teams to engage criminality on the

road is clearly aligned with a number of key Government objectives,

including the recently published strategic plan for criminal justice, the

police service’s National Intelligence Model and ACPO’s Road Policing

strategy. The use of ANPR-enabled intercept teams also contributes to

wider objectives, specifically road safety (eg through enforcing seat belt

wearing and not using mobile telephones while driving) and excise collection

(eg ensuring that all vehicles on the road are appropriately taxed).

133

C2. The concept of ANPR-enabled intercept teams also addresses the public’s

desire to see more ‘officers on the street’ and more action taken against

illegal drivers. Given the link between vehicle documentation offences

(which can be relatively easily identified from national databases) and

wider criminality, we conclude the targeting of these offences through

the use is of ANPR-enabled intercept teams can make significant

contributions to policy objectives.

C3. While ACPO have an ANPR strategy, at present there is no overarching

Government strategy for ANPR – this would cover issues around what

investment is made in ANPR equipment, what standards apply to this

equipment cross-Government, where and when the equipment is deployed,

the use of databases with ANPR systems, data sharing from the cameras

and communications strategy. The lack of a coordinated strategy across

Government means that the full benefits of ANPR are not being realised,

especially between the Home Office and the Department for Transport.

10.1.2 Technology

C4. Both Laser 1 and Laser 2 evaluations have shown ANPR to be an effective

policing tool for reading large volumes of VRMs that can be deployed in

a number of ways (in-car systems, mobile units or via CCTV systems) to

suit operational requirements. When set up properly, the technology has

proven reliable and accurate – typically more than 95% of VRMs are

correctly read. Given the volume of traffic on the roads, without ANPR,

police could not enforce vehicle documentation offences effectively

without this capability.

C5. To improve the effectiveness of teams, forces often visually verified VRM

reads of hits before a vehicle was stopped or stopped vehicles on the

basis of officer judgement. This was an important aspect to the success

of ANPR-enabled intercept teams. Not only did this recognise the benefits

of experienced officers it also allowed for a visual check of vehicles (for a

VED tax disc) before they were stopped, and prevented an over-reliance

on technology. This type of good practice should be encouraged.

10.1.3 Deployment

C6. The expansion of Laser 1 to Laser 2 has shown that the results achieved

within a small-scale pilot can be achieved across a much wider cross-

section of forces and these results are sustainable over time.

134

C7. There was no evidence from Laser 2 as to the optimum number of

officers per team or ANPR-enabled intercept teams/officers per force

area. However given the small proportion of ANPR hits that were stopped

(less than 10%) and the fact that intelligence flags are likely to increase

(with the introduction of the MOT and the no insurance databases),

the current staffing of intercept teams could be increased considerably

across forces without the introduction of significant dead time.

C8. In terms of national roll-out, there was no information from Laser 2 to suggest

that the staffing front-line intercept activities with 2,000 officers would

produce significantly different results from those achieved during Laser 2,

providing an appropriate performance management regime is in place.

C9. ANPR has been shown to be three times more effective at bringing

offences to justice compared to conventional policing.

C10. While the focus of the Laser 2 intercept teams was engaging criminality,

they will also have had an impact on road safety matters (for example

through stopping over 20,000 vehicles where the driver was using a

mobile telephone or not wearing a seat belt) and tax evasion (through

stopping 22,000 vehicle drivers whose vehicles were not taxed).

10.1.4 ANPR team management

C11. Areas that performed the best had a combination of strong leadership

within the police, supported by teams of highly motivated officers (with

complimentary skills). The quality of local intelligence is key to the success

and relies as much on good back office support and analysis as it does

on the front-line intercept teams.

10.1.5 Intelligence and data quality

C12. Both Laser 1 and Laser 2 evaluations highlighted existing inadequacies

in the accuracy of various intelligence databases, in particular DVLA’s

‘no VED’ and ‘no current keeper’. Data used with ANPR must be as

accurate and up-to-date as possible for a number of reasons. First, poor

quality data leads to inefficient targeting of police resources. Second,

inefficient targeting means that law-abiding members of the public are

being unnecessarily stopped.

135

C13. While DVLA has undertaken a number of measures to improve data

accuracy, for example the introduction of bar-coded V11 forms and barcode

readers in Post Offices and the introduction of continuous registration,

these have not resulted in an improvement in data accuracy (as reported

for ANPR operations). As yet, there are no CJX facilities for the electronic

transfer of updated information to forces on a daily basis and this is a

further limitation on database accuracy. Further, there is a lack of rigorous

understanding as to the precise causes of data inaccuracies. As data is

key to ANPR, we conclude that this represents a weakness that should

be addressed.

10.1.6 Resources and cost recovery

C14. The controls and processes set in place by the ANPR Steering Group

have worked well – while forces were required to collect additional

information and were able to issue new fixed penalties, there was no

evidence to suggest that operational priorities were distorted – forces

achieved similar arrest rates and performance levels to Laser 1.

C15. Given the focus on recovering monies from FPNs, the Laser 2 evaluation

highlighted the low payment levels associated with some fines. In particular

the introduction of a £200 fine and 6 penalty points for no insurance was

supposed to reduce the burden on police and courts. However with a

14% payment rate this is not the case. The failure of the FPNs to reduce

the bureaucratic burden represents a weakness that should be addressed.

C16. Overall the cost recovery process realised an additional £1 million in total

to the 23 Laser 2 forces over a nine-month period. While these monies

did not cover the costs of the enforcement (approximately £12 million for

the same period), these monies were seen to be worthwhile, for example

in helping to improve the intelligence capability of the ANPR teams and

providing the administrative support for the teams. On this basis, we

conclude that the cost recovery aspect contributed to the overall success

of Laser 2.

C17. Forces did not receive any additional funding for Laser 2 other than the

monies available via cost recovery.

136

In terms of operation, the use of ANPR intercept teams represents an

innovative approach:

• Targeting vehicle documentation enforcement to engage with and disrupt

criminals

• Delivered through an intelligence-led piece of technology (an ANPR reader)

• Benefiting from officers’ experience (eg observations of vehicle drivers)

• Supported by existing policing processes (eg prisoner handling).

On this basis we conclude that ANPR-enabled intercept teams have shown

to be an extremely effective means of engaging criminals. Laser 2 has built

upon the significant success of Laser 1 by proving the concept across a wider

range of forces, over a longer time period and with a greater level of resource.

Using a range of police intelligence and experience, Laser 2 intercept teams

were able to disrupt criminal activity in an efficient and effective manner, bringing

more than three times the number of offences to justice that comparable

resources deployed through conventional policing would achieve.

While the cost recovery element realised less than 10% of the expenditure

incurred, these monies were key for example in helping to improve the

intelligence capability of the ANPR teams and providing the administrative

support for the teams. Therefore, we conclude that the cost recovery aspect

contributed to the overall success of Laser 2. The pilot identified a number

of areas where operations could be improved (in particular data). Once these

have been addressed and given the development of a good practice manual,

it would be expected that ANPR would be an even more effective policing tool

than was shown in the pilot.

The Laser 2 evaluation identified a number of both positive and negative

lessons learnt from the pilot, including:

• Effective project management – Project Laser has introduced additional

workloads on police, involved large number of parties and required delivery

across traditional organisational boundaries. Potentially the project was

risky. However, given that it has developed the ANPR intercept concept and

helped 23 forces to deliver and prove the concept this can be taken as an

indication as to the successful management of the project. Contributing

factors to this include:

– sponsorship from senior levels, in particular an ACPO “champion”

– clear terms of reference and objectives that have provided a focus for the

project. While associated issues have been raised the project (data quality)

has not sought to address all matters

137

– effective risk management, including recognition that taking risks and

developing new ways of working is a key aspect to pilots

– central support, including from both PSU and ACPO

• Demonstrating effectiveness – a key aspect to the project has been that

it has sought at all stages to demonstrate effectiveness and contribution to

Government and police targets. This has required significant and often

onerous data collection by forces, however this has lead to a significant

data and knowledge:

– to support performance management within forces

– that provides the basis for independent evaluation and demonstration

that Project Laser is effective

– supports the intelligence-led project.

• Data validation – in addition to the data collection in the field, there was a

process of data entry and subsequent validation. Given that much of the data

collected relates to each other (for example deployment data relates to stops),

then the data validation process could have been built in to the data entry

process. This would have reduced both the burden of data entry/validation

and improved the quality of data. The Home Office have recognised this and

are in the process of developing a bespoke data entry tool for forces.

• Change management – Laser 2, has by its very nature necessitated

considerable change for those taking part, including a new way of working,

additional fixed penalties, new requirements on intelligence. Many of these

changes happened within a very short timescale and in some circumstances,

in particular where there are interdependencies with other projects and the

very complexities of trying to manage change across 23 forces.

• The importance of legislation – key aspects of Project Laser (in particular

the on-going roll-out and improving effectiveness) are dependent on setting

in place enabling legislation. Given the pressures on the legislative timetable,

it has been crucial to have in place the necessary political support to carry

forward the amendments.

• Laser 2 is not a road policing project – while Laser 2 is being led from the

road policing portfolio within ACPO it is not a roads policing project. A key

issue for the project has been to engage those outside the road policing

community to convince them of the value of Laser. This has included active

engagement with ACPO crime representatives, communications to forces

and promotion of the concept by PSU. While this has been in part successful,

the knowledge of the capability of ANPR across forces appears to be patchy.

138

10.2 Recommendations

On the basis of the evaluation findings and conclusions, the following

recommendations are made:

10.2.1 Roll-out of Project Laser

R1. Project Laser has proved that ANPR intercept teams, if used appropriately,

can be an extremely effective police tool in engaging and dealing with

criminality in all its forms. There is a strong case that Laser is rolled out

nationally and this roll-out proceeds as rapidly as possible to ensure that

the benefits to police and society are achieved.

Rapid roll-out, however, creates a short-term funding problem.

The introduction of cost recovery to recycle the money requires primary

legislation. From the evidence gathered to date in the first 13 months

of Laser 2, this is never going to be sufficient to fund the rapid roll-out

of Laser, including centralised support and the possible development of

a centralised intelligence function. This leaves two options: funded

centrally or through local reprioritisation. We recommend a combination

of the two provides a way forward in the short-term with some funding

being ring-fenced centrally, perhaps for capital and infrastructure costs,

and this being matched by police forces to staff the units. A business

case for this investment is required.

Cost recovery could then be used as a means of supplementing local force

expenditure, in particular in the improvement of intelligence and its handling.

Within the context of HMT cost recovery rules, this use of monies seems

appropriate for two reasons. First it avoids any question of double funding.

Second it will lead to an improvement in ANPR team effectiveness and

achievement of policy objectives.

R2. Whichever funding route is found, it is important that Laser is extended

nationally as part of a co-ordinated programme, managed centrally. This will

allow for the spread of best practice, the development and use of appropriate

standards (eg data), co-ordination in response to issues (eg external

communications and data quality issues). Whilst it is desirable to have the

ability to recover some of the costs, this should not constrain roll-out of the

programme. One of the key successes of the programme to date has been

the central coordination of the programme by the Police Standards Unit.

The participation of 23 police forces has added to the complexity of the

data collection and analysis and considerable effort has been invested in

standardising the data collection and analysis procedures. If the programme

is to be rolled out nationally with cost recovery as an element, it seems

sensible that that the central coordination is retained within the Home Office.

139

10.2.2 Review of data used for ANPR

R3. The introduction of MOT and no insurance databases, planned for later

in the year, is an important development and should increase the

productivity of the ANPR intercept teams. These should be fully evaluated

in terms of their strengths and weakness for use with ANPR teams before

their introduction.

R4. The accuracy of the DVLA database in particular needs to be investigated.

There are also substantial variations in the quality and accuracy of local

intelligence databases that require investigation. There should be more

effective use of intelligence at a national and at a local level.

10.2.3 Deployment management

R5. Currently, most ANPR teams are tasked and deployed from a central

location. This can mean in some areas that considerable time is spent

travelling to and from ANPR intercept sites. Clearly, this is not best use

of police time and we suggest that consideration is given to co-locating

ANPR intercept teams with BCUs and roads policing units, as appropriate.

Support systems will need to be put in place to ensure best practice and

intelligence is shared and performance monitored as a whole.

10.2.4 Review of level of fines and payment rates

R6. There is an apparent disconnect between the levels of fixed penalties for

the more serious offences and the penalties that are awarded if the case

is taken to court – anecdotal evidence suggest that penalties in some

cases are less in court, both in monetary value and the number of points

awarded. This could potentially damage the effectiveness of the fixed

penalty scheme and needs to be urgently reviewed by ACPO and the DCA.

10.2.5 National vehicle intelligence data warehouse

R7. There is an urgent need to move from a heavy reliance on locally produced

and held vehicle intelligence to the provision of a national vehicle intelligence

data warehouse, which would hold all relevant vehicle intelligence and be

accessed in real time by ANPR readers. This data warehouse should also

hold ANPR reads and hits, which are themselves a vital source of vehicle

intelligence and should be accompanied by the development of data mining

tools of a more sophisticated nature. This vehicle intelligence database

must be part of, or compatible with, the National Intelligence Management

system proposed under Bichard.

140

10.2.6 Development of a national ANPR strategy

R8. Many Government departments and agencies (principally within the remit

of the Home Office and the Department for Transport) have invested heavily

in ANPR technology. There has been little coordination of this activity.

In some cases, this has lead to duplication of effort and wasted resources.

Now is a good time to take stock, to plan for future investment and make

sure that there is best use from existing infrastructure. We recommend

that the Home Office and Department for Transport, working with other

Government departments and key stakeholders, develop a detailed strategy

and implementation plan for ANPR for the next few years. This would

address a number of issues including:

– the setting of standards and protocols for information sharing across

Government

– investments in ANPR infrastructure (both cameras, communications

and back-office) to maximise value for money for Government

– a protocol for dealing with ANPR databases such errors in the data only

need to be address by one body, that multiple agencies do not end up

pursuing the same motorists

– a communication strategy for Government on ANPR cameras, what they

are used for and how to deal with press/public enquiries

– consideration of how to link to future programmes such as lorry road

user charging, road user charging, electronic vehicle identification,

the national ID card programme and biometrics would link in

– how the lessons learnt from the ANPR programme could be exploited

in other similar programmes.

141

142

Appendix

143

Appendix A: Acronyms

ABI Association of British Insurers

ACPO Association of Chief Police Officers

ACPOS Association of Chief Police Officers in Scotland

ANPR Automatic Number Plate Recognition

BCU Basic Command Unit

CCTV Closed Circuit Television

CJS Criminal Justice System

CJX Criminal Justice Extranet

CTO Central Ticket Offices

DfT Department for Transport

DCA Department for Constitutional Affairs

DVLA Driver and Vehicle Licensing Agency

EVI Electronic Vehicle Identification

FIS Force Intelligence System

FLINTS Force Linked Intelligence System

FTE Full Time Equivalents

FPN Fixed Penalty Notice

HMIC Her Majesty’s Inspectorate of Constabulary

HMT Her Majesty’s Treasury

HO/RT/1 Home Office Road Transport form 1 (document producer)

MOT Ministry of Transport

NCIS National Criminal Intelligence Service

NIM National Intelligence Model

OBTJ Offences Brought To Justice

OCU Operational Command Units

NJG Narrowing the Justice Gap

PA PA Consulting Group

PITO Police Information Technology Organisation

PNC Police National Computer

PSU Home Office Police Standards Unit

RTA Road Traffic Act

SORN Statutory Off Road Notification

VED Vehicle Excise Duty

VERA Vehicle Excise and Registration Act

VOSA Vehicle and Operator Services Agency

VRM Vehicle Registration Mark

144

Appendix B: Data collection pro forma

Figure B.1: Deployment Pro Forma

145

ADDITIONAL NOTES

ETHNICITY CODES

W - White

W1 - White - British

W2 - White - Irish

W3 - White - Welsh

W4 - White - English

W5 - White - Scottish

W9 - Any other White background

M - Mixed

M1 - White and Black Caribbean

M2 - White and Black African

M3 - White and Asian

M9 - Any other Mixed Background

A - Asian / Asian - British

A1 - Asian - Indian

A2 - Asian - Pakistani

A3 - Asian - Bangladeshi

A9 - Any other Asian background

B - Black / Black - British

B1 - Black - Caribbean

B2 - Black African

B9 - Any other Black background

O - Other

O1 - Chinese

O9 - Any other

NS - Not Stated

NUMBER OF ARRESTS MADE

Robbery

S25

Theft/Burglary

Auto Crime

Driving

Warrant

Drugs

Other

1) _______________________

CUSTODY REFERENCE NUMBER(s)

3) _______________________

5) _______________________

2) _______________________

4) _______________________

PLEASE MAKE ANY ADDITIONAL NOTES OVERLEAF

Mobile

Phone

VED

Driving

Manner

Other Observation

Seat Belt

Vehicle Defective

Known Person/

Vehicle

DATA COLLECTION SHEET v4.0

Officer __________

Date __________

Time __________

VRM ____________

TICK IF…

Number of CROs Arrested ________

CRO(s) in Vehicle How many? ________

Vehicle Searched

Person Searched How many? ________

ACTION(s) TAKEN

NO ACTION TAKEN

HO/RT1

CLE 2/6(7)

CLE 2/8 / V62

VDRS / PG9

NEFPN (pto)

EFPN (pto)

Reported for Summons

INTEL Log Generated

Verbal Advice Given

PROPERTY RECOVERED

Stolen Car £ ________

Drugs £ _______

Stolen Goods £ ________

Offensive Weapon(s)

Firearms

Other

Number of Endorsable FPNs issued: No Insurance ______

FIXED PENALTY NOTICES ISSUED (FPNs)

Number of Non-Endorsable FPNs Issued:

ETHNICITY OF DRIVER & PERSONS ARRESTED (see over for codes)

Ethnicity of Arrest 2 ________

Ethnicity of Arrest 4 ________

Ethnicity of Arrest 3 ________

Ethnicity of Arrest 5 ________

Yes

ANPR HIT

OBSERVATION

PNC

DVLA: Vehicle Excise

DVLA: No Current Keeper

Local: 1 2 3 4 Other

Database Correct

Yes / No

Figure B.2: Roadside Stop pro forma

146

Police Force:

Date:

Time

Trigger:

Police Force:

Date:

Time

Trigger:

Police Force:

Date:

Time

Trigger:

Police Force:

Date:

Time

Trigger:

Please enter case study details here.

ANPR CASE STUDY 3

Avon & Somerset Police

Please enter case study details here.

ANPR CASE STUDY 1

Avon & Somerset Police

ANPR CASE STUDY 4

Avon & Somerset Police

ANPR CASE STUDY 2

Avon & Somerset Police

Please detail cases where an ANPR hit resulted in a significant outcome. Please include why the vehicle was stopped, how ANPR

influenced the stop, what happened at the roadside, what the reason for the arrest was and what the outcome was. Please do not

enter any confidential details as some of these case studies may later be published.

Please enter case study details here.

Figure B.3: Case Study pro forma

147

Appendix C: Data completeness by field

While every effort was made to ensure that data quality remained high and all

data items were returned complete, there were instances where inconsistencies

in the data meant that they could not be considered in the analysis or where

returns were incomplete. These included stops where there were internal

inconsistencies within the information provided. Examples of this included

occasions where the reasons for stops were unrecorded and arrests were

recorded and either not categorised or no other information was captured

about the stop. The table below breaks down the number of such excluded

stops for every force where the data were considered unreliable

Figure C.1: Unreliable stops by force

Avon and Somerset 9 0.12%

Cambridgeshire 41 0.44%

Cheshire 21 0.41%

City of London 22 0.86%

Cleveland 67 1.94%

Greater Manchester 46 0.29%

Hampshire 46 0.71%

Kent 12 0.12%

Lancashire 73 0.63%

Leicestershire 16 0.16%

Lincolnshire 21 0.18%

Merseyside 26 0.38%

Metropolitan 280 1.55%

North Wales 64 0.58%

North Yorkshire 21 0.47%

Northants 21 0.30%

Northumbria 27 0.67%

Nottinghamshire 22 0.82%

Staffordshire 21 0.36%

Warwickshire 50 1.06%

West Midlands 84 0.96%

West Yorkshire 134 1.55%

Total 1,134 0.63%

Force Name Stops record omitted % of evaluated stops

Hertfordshire

0.18%

148

Figure C.2 shows the completeness of returns for key data fields.

Figure C.2: Completeness of returns for key data fields

Force Name 100%

Week Number/Date 100%

Time of Stop 99.9%

VRM 99.6%

Postcode 80.5%

Ethnicity of driver 89.6%

Field Completeness

149

Appendix D: Fixed penalty notices included under

cost recovery

• Contrary to section 47, Road Traffic Act 1988. No MOT certificate

• Contrary to section 143, Road Traffic Act 1988. No insurance

• Contrary to section 172, Road Traffic Act 1988. Failure to supply details

• Contrary to section 87(1), Road Traffic Act 1988. Drive otherwise than in

accordance with licence

• Contrary to section 42, Road Traffic Act 1988. Driver not in proper control of

vehicle and S104 Road Vehicles Regulations (Con & Use) Regulations 1986

(Previously this offence has been used to deal with mobile phone offences

the new offences relating to the use of mobile phones when driving have

received HM Treasury approval for inclusion within the scheme as a sub group

to the above offence

• Contrary to section 33, Vehicle Excise and Registration Act 1994 Failing to

exhibit excise licence

• Contrary to section 42, Vehicle Excise and Registration Act 1994

Keeping/driving without registration mark

• Contrary to section 43, Vehicle Excise and Registration Act 1994 Registration

mark obscured

• Contrary to section 59, Vehicle Excise and Registration Act 1994 Registration

mark not fixed

• Contrary to R17 Road Vehicles (Regulations and Licence) Regulations

1971 & schedule 2, Registration mark not conforming to regulations

• Contrary to section 163(3) Road Traffic Act 1988 Failing to stop for police

constable

• Contrary to R11(1) RVLR 1989 S42 Road Traffic Act 1988 Showing red light

to front

• Contrary to R25 RVLR 1989 No headlights/front fog lights not lit in poor visibility

• Contrary to R54 RV (Con & Use) Regulations 1986. S42 Road Traffic Act

1988 No silencer/defective exhaust

• Contrary to R54 RV (Con & Use) Regulations 1986. S42 Road Traffic Act

1988 Failing to maintain silencer

• Contrary to R57 RV (Con & Use) Regulations 1986. S42 Road Traffic Act

1988 Noise limits & exhaust systems on motor cycles

150

• Contrary to section 14 Road Traffic Act 1988 S5(1)(a) and (b) MV (Wearing of

Seat Belts) Regulations 1993 Failing to wear seat belt (adults) driver/passenger

• Contrary to section 15(2), Road Traffic Act 1988 Child in front passenger

seat – no seat belt and Sec 5 (1)(a)(b) MV (Wearing of seat belts by

children in front seats)

• Contrary to section 15(4), Road Traffic Act 1988 Child in rear passenger

seat – no seat belt and Sec 8 (1)(a)(b)(c) MV (Wearing of seat belts

Regulations 1993).

151

Appendix E: ANPR case studies

This section details some of the ANPR successes as reported by forces.

E.1 Avon & Somerset

Three arrested after ANPR hit (Brislington)

At 7.15pm Monday July 12 2004 the ANPR system alerted police to a Vauxhall

Tigra which had been stolen from a home in Hanham on July 10. This vehicle

failed to stop for police and a pursuit, involving two ANPR vehicles, took place.

The Tigra failed to stop at a Give Way sign at the junction of Hampstead Road

and Talbot Road in the Brislington area of Bristol and collided with an Audi 80 car.

The driver of the Tigra ran off and was lost following a foot chase. The front

seat passenger, a 26-year-old man, was arrested and found to be an escapee

from prison. The two occupants of the Audi received minor cuts and grazes

as a result of the collision. Police then learned that the driver of the Audi,

a 28-year-old man, was a disqualified driver who was also wanted by Wiltshire

police in connection with a robbery and by Bridgwater police in connection

with a theft.

The passenger of the Audi, a 22-year-old man, was found to be wanted for

escaping from Dorchester Prison and also in connection with a theft in Bridgwater.

Superintendent Lawrie Lewis, head of the force’s Road Policing Unit, said:

“We are determined to deny criminals the use of our roads. The ANPR system

is camera technology designed to target those who flout the law, and protect

those who respect it. This latest incident is an excellent example of how effective

the ANPR system can be. As a result of one identification, we have arrested

three people. This sends out a strong message to any criminals planning to

use roads in the Avon and Somerset force area; we are watching.”

E.2 Cambridgeshire Constabulary

ANPR and trading standards haul in counterfeit goods

• An ANPR stop was made recently by Cambridgeshire officers as a result

of intelligence received from Trading Standards. Officers found counterfeit

goods and over £500 from the sale of these goods at a nearby car boot

sale. Officers were able to arrest the offender and together with Trading

Standards were able to supply the evidence needed to present at trial.

152

Peterborough Police pull in serial offender thanks to ANPR

• Officers in Peterborough had pursued an individual who was driving a stolen

sports car on more than one occasion. His driving was thought to be extremely

dangerous and officers had been unable to stop the vehicle and make an

arrest. Using intelligence, two ANPR officers selected a location in order to

find and stop the vehicle. A short while later the officers identified the individual

walking away from the, now parked, stolen car. He was arrested and later

admitted nine charges, including three for dangerous driving, taking without

owners consent, aggravated burglary, driving while disqualified and driving

without insurance.

ANPR and council initiative in Peterborough a success

• During an initiative in early 2003 ANPR officers teamed up with Peterborough

City Council’s CCTV Unit to conduct a two-week operation in Peterborough.

Four CCTV cameras were linked into the ANPR system, with an officer working

in the CCTV control room to monitor and help direct resources to stop those

cars identified. During the campaign:

– 78,125 registration numbers were read

– 352 vehicles were stopped

– 13 people were arrested for various offences

– 41 people were reported for driving offences

– 153 vehicles were reported to the DVLA for no tax

– 260 people were required to produce their documents

– 24 fixed penalty notices were issued for minor traffic offences

– 22 vehicles were found to have defects which had to be corrected

– 5 searches were made

– £3,000 worth of property was recovered (a stolen car)

Record breaking find for ANPR officers

• ANPR officers on the Fletton Parkway, Peterborough, in October 2003,

found a stolen vehicle less than 40 minutes after it was stolen from an

address in another county. The vehicle – a Land Rover Defender – that

showed up as being stolen was reported stolen from Kislingbury,

Northamptonshire at 3pm. After a short follow to Boongate, Peterborough

the man driving the stolen vehicle gave himself up to officers at 3.38pm.

153

E.3 Hertfordshire

Worldwide credit card scam

Professional criminals at the centre of a worldwide credit card scam were caught

in Stevenage thanks to hi-tech equipment recently introduced in Hertfordshire.

“This is just one of the huge success stories from the sophisticated Automatic

Number Plate Recognition (ANPR) and I have no doubt that we will have many

more examples like this in Hertfordshire very soon,” said Detective Inspector

Greg Cooper, who is leading the use of ANPR in the Eastern Area.

Driving crime off the county’s roads, detecting criminals and specifically

targeting crime hotspots are the main advantages of the ANPR, which works

by scanning literally every vehicle registration that passes in front of it and

checking them against information stored in several databases, including

the Police National Computer. This identifies vehicles of interest to the police,

such as stolen cars or those involved in crimes. When a suspicious vehicle

is recognised it can be the focus of targeted interception and enquiries.

“The equipment showed that a vehicle was being used by credit card fraudsters

and was stopped by officers,” said Greg. “The car contained two men who

had ten credit cards in their possession, some of which were unsigned.

Further enquiries revealed that deceptions had already been committed locally

but it soon became clear that we were dealing with professional criminals who

were at the centre of worldwide credit card fraud approaching half a million

pounds. They were linked to a sophisticated network for obtaining credit card

numbers and manufacturing cards,” said Greg. Two men are currently awaiting

trial at Crown Court.

“Just think, would that car have been stopped if it were not for ANPR?” said

Greg. “Officers working as dedicated ANPR intercept teams can arrest ten times

as many offenders as other officers, according to recent national statistics.”

To date several ANPR operations have been run in all areas under Operation

Grip (Eastern) and Operations Raceway and Reabsorb (Central), with most

activity in the Eastern area over the past few months.

Greg said: “These successful operations clearly show the potential this

technology has to drive down crime and detect travelling criminals, making

our streets safer. If we deny criminals the use of the road then we will be

better able to enforce the law, prevent crime and detect offenders.”

154

The sophisticated cameras can be fitted to marked and unmarked cars and

the Constabulary also has a dedicated ANPR vehicle. “The technology brings

many crime fighting benefits and has a knock-on effect in increasing levels of

intelligence, a positive effect on road safety and brings reassurance to the

community with high visibility police operations,” said Greg.

The force has recently been successful in an application to expand ANPR

and join the national Project Laser. This project, which forces had to bid to join,

intends to expand the use of ANPR in a wide variety of policing environments,

including linking to CCTV control rooms. The new project is being piloted in

the Eastern Area, and specifically Stevenage. As part of the second phase

of the scheme, a dedicated ANPR unit of seven officers has been set up.

Greg said: “ANPR is now a significant weapon in our proactive armoury in

Hertfordshire and is set to revolutionise policing.”

Paul Abraham, Manager for Project Laser added: “The Constabulary intends

to exploit the use of ANPR technology to reduce crime and disorder by

detecting offenders. To support the technology, we will increase staff and

resources dedicated to ANPR interception teams and use the technology

through all delivery systems.”

Three sentenced for internationally organised counterfeit credit

card racket

Three people were sentenced at St Albans Crown Court yesterday (June 7)

for their involvement in a huge international credit card scam investigated by

Hertfordshire Constabulary. They were arrested in Gunnels Wood Road,

Stevenage on February 26, 2003 after the vehicle in which they were travelling

was flagged up by Automatic Number Plate Recognition (ANPR) as being of

interest to police during a pilot operation making use of that new technology.

Ngaih Lim (male), aged 46, from Hall Place, London W2 was sentenced to

four years’ imprisonment having pleaded guilty to conspiracy to defraud at

Luton Crown Court at a previous hearing. Kwong Wong (male), aged 34,

from Highbury Park, North London was sentenced to two years’ imprisonment

having pleaded guilty to conspiracy to defraud at Luton Crown Court at a

previous hearing. Chen Hsiang-Ching (female), aged 28, from Hall Place,

London W2 was convicted at Luton Crown Court on May 13 of Money

Laundering. She received a nine-month sentence.

155

Sergeant George Smith from the Constabulary’s Cheque and Credit Card Fraud

section said: “These three are thought to have gained a substantial amount of

money from counterfeit cards; and a large amount of high value goods, such

as Louis Vuitton, Chanel and Prada products, was recovered from their homes

during this investigation. A computer was also recovered which contained

details of 488 credit cards held by genuine cardholders around the world.”

During the court hearing, Lim and Wong admitted conspiring to defraud

national and international clearing banks and Lim admitted being part of an

organised crime syndicate. Wong admitted that he had used counterfeit cards

and passports while Chen was convicted of assisting with the laundering of

proceeds over a five-year period.

Sergeant Smith added: “This was a very complex investigation which involved

liaison with many other forces and agencies. It is pleasing to see it reach a

successful conclusion at court.”

Sergeant Dougie Fishwick who co-ordinated the pilot ANPR operation in

February 2003 said: “This outcome shows what a vital role this new technology

can play in detaining offenders who are using our road networks. It is now

routinely used across the county to provide officers with fast-time information

linked to the Police National Computer and DVLA records. Through this

technology police officers are alerted to vehicles used by criminals and

appropriate action is taken. The message is therefore clear that offenders

driving through Hertfordshire now stand a far greater chance of being detected.”

156

E.4 Merseyside

604 arrests & £370,000 of goods seized

Merseyside Police is celebrating the success of its dedicated ANPR team which

has made 604 arrests and seized goods worth more than £370,000 in the last

11 months.

The hi-tech computerised Automatic Number Plate Recognition (ANPR)

system has proved to be a powerful tool in the fight against serious crime,

enabling officers to put the intelligence they gather daily at the fingertips of

officers out and about on the roads, at the front-line of policing.

Since the beginning of the pilot in June last year, 604 people have been arrested

in Merseyside for a variety of offences including robbery, burglary and car crime

after being stopped following a ‘hit’ on the ANPR system. Additionally, over

£370,000 worth of goods, including stolen high-performance cars and a stolen

heavy goods vehicle, have been recovered.

ANPR systems instantly scan number plates and have the capability to check

whether vehicles may be involved in illegal activity. Officers can examine

intelligence at the touch of a button on a laptop computer at the roadside,

and can move at short notice to target the areas where criminals are known

to be. The ANPR system can match number plates against information stored

in databases, such as the Police National computer, DVLA databases and local

intelligence databases, to check if the vehicles are of interest to the police such

as stolen cars, or those involved in crime. And the system has the capability to

check up to 3,000 number plates per hour.

Merseyside Police’s ANPR team, codenamed ‘Operation Laser’, are out and

about in Merseyside every day, working alongside Neighbourhood colleagues

to track down those involved in crime.

The ANPR initiative is part-funded by retaining fines paid by criminals who

drive their vehicles untaxed and uninsured. While this only covers a small part

of the cost of the Force’s ANPR operation, police say this hits criminals in the

pocket and sees them paying for the police activity which catches them out.

Chief Constable Norman Bettison is keen to point out that this activity has

nothing to do with speed checks. He said: “I’ve been out on Operation Laser

when the officers put a road check in place. Motorists could be forgiven for

thinking ‘haven’t the police got better things to do’. I’d like them to know that

we’re catching serious criminals, not speeders. Criminals are our key priority

here in Merseyside and always will be while I’m here.

157

“Our message is very clear – if your vehicle is taxed and has an MOT, and

you’re entitled to be driving that vehicle and you’re not involved in crime,

then you have nothing to fear. It’s the criminals we’re after. The ones who

drive in illegal vehicles with no care or thought about other road users and

those involved in car or drug crime, which can have a devastating impact on

our neighbourhoods and blight our lives.

“Our message to our ‘other’ audience is equally clear – if you’re involved in

criminal activity, beware, because we have the technology – and the means –

to stop you. ANPR allows the police to stay one step ahead of the criminals.

We will continue to expand the number of vehicles fitted with ANPR and there

will be no hiding place for criminals on Merseyside.”

E.5 Bedfordshire

Curb crawlers in Luton are Operation Scorpion’s latest target as Bedfordshire

Police works to reduce the number of prostitutes operating in the town.

Anyone caught on ANPR curb crawling will receive a letter, sent to their home

address, saying they have been seen and their details registered. If they are

caught a second time, they will be arrested and charged and their names and

addresses will be released publicly in court. Chief Inspector Jim Saunders,

from Luton Police Station, says while it is an age-old problem in the town it is

still one he receives many complaints about.

“This operation will work very much as a deterrent and our message is ‘don’t

come to Luton because you will be caught’. Our focus is aimed primarily on

the curb crawlers rather than the prostitutes themselves. We will still be arresting

the prostitutes, but, as many of them have a drugs habit, our priority will be

to get them into treatment programmes. Prostitutes who appear at court are

normally fined, this just forces them back out on to the street to earn the money

pay their fine. We want to break the cycle and we believe we have a better

chance of doing this if we can get them off the drugs.”

Considerable funding is being ploughed into bringing ANPR technology into

Luton, including £30,000 which is to pay for Police Patrol Vehicles to be fitted

with ANPR equipment, plus a further £100,000 which will link the technology

to existing CCTV cameras in the town.

Funding for the CCTV / ANPR linked scheme is being provided by the

Neighbourhood Regeneration Fund and other partnership funding. Chief

Inspector Saunders added; “Using sophisticated technology such as this give

us a fantastic capability. We believe that this technology will be particularly

effective in tackling the problem of curb crawlers once and for all.”

158

Luton Police have already arrested 38 prostitutes for loitering and soliciting;

twenty letters have been sent to curb crawlers and six men have been

prosecuted in the past year.

E.6 Lancashire

Project Laser targets travelling criminals

Police across the county are hailing the success of a crackdown targeting

travelling criminals. Since the beginning of the month, equipment has been in

use to catch motorists who are breaking the law and there have already been

a number of significant arrests for serious offences.

The system works by scanning passing motor vehicle registrations and checking

them against information stored in a variety of databases including the Police

National Computer. This can identify vehicles of interest to the police, such as

stolen cars or those involved in crime. When a suspicious vehicle is recognised,

it can be the focus of targeted interception and further enquiries.

In Blackpool, a stolen vehicle was identified being driven on Progress Way.

The vehicle had been taken during a burglary and was also found to contain

stolen property. A man and a woman from the Manchester area were arrested

at the scene. He was later charged with handling stolen goods, resisting arrest,

obstruction, and a variety of document offences while she was charged with

allowing herself to be carried in a vehicle without the owner’s consent, going

equipped, resisting arrest, and obstruction.

A disqualified driver was stopped on Flensbury Way, Leyland and arrested for

disqualified driving. The Chorley man was charged and remanded in custody

while a woman, arrested at the same time for perverting the course of justice,

received an instant caution.

When a vehicle connected to crimes in the Cheshire area was stopped in

Clayton Brook, one of the occupants was arrested for theft.

Said Chief Inspector Tracey O’Gara: “Project Laser has already been responsible

for a number of noteworthy arrests. We are catching criminals who are using

our roads to commit crime in dangerous and illegal vehicles.

“These arrests are as a direct result of ANPR – showing what a powerful tool

this piece of equipment is in the fight against crime. We have also arrested

several people for possession of Class A drugs and criminals who have been

on-the-run from other forces.

159

“The aim of Project Laser is simple – we want to deny criminals the use of

our roads. It is known that motoring offences such as driving without tax and

insurance are often associated with other crimes such as burglary and drug

dealing. Experience has shown that when vehicles are stopped for a motoring

offence this has often led to an arrest for more serious crimes.”

A vehicle is only stopped where intelligence suggests that some form of road

traffic offence has been committed or when there is a known police interest in

that vehicle. Law-abiding citizens have nothing to fear from ANPR.

E.7 North Yorkshire

A joint operation between Road Policing officers and the Automatic Number

Plate Recognition Unit struck a series of severe blows to travelling criminals

on the borders of West Yorkshire and North Yorkshire.

For four nights in February a unit went out every night with RPG officers

along North Yorkshire’s borders with Leeds and Bradford, specifically hunting

cross-border law-breakers.

Among the successes were 18 arrests, 21 individuals reported for summons,

12 penalty tickets were issued mostly for having no tax, one stolen car was

recovered worth £9,000, £1,400 of stolen goods was recovered and cannabis

and heroin was also seized.

One of those arrested, who was stopped driving a car, stolen in a burglary which

took place in West Yorkshire a few days earlier, was wanted in connection with

a burglary and a robbery in the Selby area . He was also wanted for questioning

by South Yorkshire Police. He had been released from prison in November last

year, after serving three and a half years for dwelling house burglaries. However

this arrest meant his prison licence was revoked and he is now back in custody.

Three others were stopped after ANPR found that their vehicle did not have an

excise licence and found to be going equipped for theft after a ten-ton hydraulic

jack – frequently used to remove cash boxes from night security safes and

worth about £800 – was found in their vehicle. One man was in possession

of cannabis another was wanted for two assaults by West Yorkshire Police.

Two were released on police bail and one was taken for questioning by West

Yorkshire police officers.

And an off-duty West Yorkshire officer reported to Harrogate enquiry office that

he had seen a vehicle which was circulated as being involved in an incident of

making off without payment in his own force area. Within 45 minutes ANPR

had picked up the vehicle and the occupants were arrested by NYP intercept

officers and handed over to West Yorkshire Police.

160

“Some criminals in surrounding counties have a naïve belief that they can

stroll into North Yorkshire and help themselves,” said Detective Inspector Ian

Wills, who heads the ANPR operation. “Some of them have just had a very

rude awakening!”

DI Wills is pleased with the arrests, and also with the intelligence which has

now gone on to North Yorkshire’s ANPR database – and is being shared with

colleagues in neighbouring forces. He said: “We now know even more about

who travels where, when and with whom, and what routes they like. It all adds

to the intelligence picture that enables us to drive back the travelling criminals.”

And he was particularly happy with the efficiency of the ANPR system. He said:

“Before ANPR, night-time checks would necessarily involve stopping quite a

number of entirely innocent drivers. Now we can much better target the people

we want to know about and so minimise public inconvenience.”

E.8 Surrey

An operation to target Waverley and Guildford’s most prolific car criminals

as part of Operation Gallant II, has resulted in 48 arrests in just three weeks.

Operation Gallant II was launched in West Surrey last month in an attempt to

further reduce thefts from and thefts of cars, and linked in with the countywide

vehicle crime campaign as part of Operation Safer Surrey.

Analysis was carried out to identify the most prolific car criminals believed to be

responsible for a large proportion of the division’s vehicle crime. These offenders

were then targeted using a wide range of measures, including the use of the

ANPR system, which alone resulted in many of the arrests.

ANPR is an innovative system, which scans vehicle registrations and checks

them against information stored in various databases, including the Police

National Computer, to identify vehicles of interest to the police, such as stolen

cars and those involved in crimes. When a suspicious vehicle is recognised,

it can be intercepted by officers further down the road. Road traffic offences

are often linked to other criminal conduct and stops for such offences, often

lead to the arrest of criminals unlawfully at large.

161

Other measures used included increased use of stop and search powers and

more frequent checks of bail conditions, including non-association orders and

curfews. Maximum use was made of traffic offences to disrupt the activities of

persistent offenders, including impounding unsafe vehicles and prosecutions

for driving with no insurance. Analysis to link offenders to crimes, by the

method in which it is committed to generate arrests and search warrants was

also used. Criminal informants were asked to provide more information about

persistent offenders, and patrols were increased in hot spot areas, such as

beauty spot car parks.

Of the 48 arrests, 11 people were arrested for theft of a motor vehicle, four were

arrested for attempted theft of a motor vehicle and seven were arrested for

theft from a motor vehicle. Further arrests (two) were also made for thefts of tax

discs. Arrests were also made for aggravated vehicle taking (two), driving while

disqualified (three), driving while over the legal alcohol limit (eight) and being

unfit to drive (two). Individuals were also arrested for failing to stop, breach of

bail and assault. Three arrests were made for criminal damage and three arrests

were made for historic offences after those responsible were linked through

DNA tests.

Of those arrested, 13 were charged, 24 were released on police bail, six were

released with no further action, four were remanded in custody and one

received a caution.

The operation to target prolific car criminals was based on last year’s successful

Operation Bugle, which resulted in 116 arrests in three months. 17 offenders

who were actively targeted amassed 69 arrests between them, with seven

remanded in custody. One offender who was remanded in custody asked for

54 further car crime offences to be taken into consideration. Operation Bugle

also resulted in the recovery of four stolen cars, valued at a total of £30,000.

West Surrey Superintendent Kevin Deanus said: “The success of this operation

proves that targeting prolific offenders, and focusing our resources effectively,

does pay off. We will continue to make use of these tactics, particularly ANPR,

to target known criminals and disrupt their criminal activity. We are committed

to tackling both auto crime, a priority for Surrey Police, and persistent offenders

and this operation has demonstrated these commitments. We will not tolerate

car crime being committed in West Surrey.”

162

E.9 West Midlands

Vehicles seized for having no tax across the West Midlands will be brought to

Coventry to be crushed. Police have seized a number of vehicles since gaining

new powers to tackle unlicensed cars and will start crushing on Monday.

On the first day of gaining the new powers in May 2004, 15 vehicles were

taken off the streets for having no tax. Three owners paid up on the spot but

the remaining 12 cars were crushed. In the first four weeks of gaining the powers

officers have seized 304 vehicles and arrested 118 people. West Midlands

Police were the first force in the country to take on the new powers in May.

There are currently believed to be 86,000 untaxed vehicles on West Midlands

roads. Until now, there have been no powers available to police to seize vehicles

being used without insurance or tax. Through the new powers vehicles will be

targeted through the Automatic Number Plate Recognition (ANPR) system.

Anyone without tax will have their vehicle seized until they pay a release fee

and buy tax. If the vehicles are not claimed in 14 days they will be crushed.

Those worth more than £2,000 will be sold. Police say a large proportion of

unlicensed vehicles are used by criminals, and the majority end up being

abandoned or torched.

163

Appendix F: National ANPR project board membership

• Home Office Police Standards Unit – Chair

• ACPO representatives including:

– Head of Road Policing

– Chair ACPO ANPR Steering Group

– ACPO Crime Business Area

– National ANPR Co-ordinator

• Police Information Technology Organisation

• Department for Transport – Licensing Roadworthiness & Insurance

• Department for Constitutional Affairs

• Crown Prosecution Service

• Driver and Vehicle Licensing Agency

• HM Treasury

• ACPOS / Scottish Police Information Strategy (Observer)

164

Appendix G: Recorded offence guidance

Recorded offences are offences that must be notified by police forces to the

Home Office and then published as part of the National Crime Statistics.

These are the offences that are counted for the purposes of the Narrowing

the Justice Gap target, 1.2 million offences to be brought to justice in 2005-06.

Recorded Offences are:

• All ‘either way’ and ‘indictable only’ offences and a small number of

summary offences of:

– racially aggravated harassment

– causing intentional harassment

– fear of violence

– causing harassment, alarm or distress

– indecent exposure with intent to insult any female

– assault on a constable

– obstruction of a constable doing their duty

– common assault and battery

– assaulting a person assisting a constable

– assaulting a prison officer

– resisting / obstructing a prison officer

– assaulting a court security officer

– resisting or obstructing a court security officer

– supply of articles for administering controlled drugs

– unauthorized access to computer material

– unauthorized taking of a motor vehicle

– unauthorized taking of a conveyance other than a motor vehicle or cycle

– aggravated vehicle taking

– taking a cycle without consent

– interference with motor vehicles

– tampering with motor vehicles.

• All other summary offences are not recorded offences including:

– loitering or soliciting for the purposes of prostitution

– driving with excess alcohol / failing to provide a breath specimen

– driving while disqualified.

165

If a summary offence is not a recorded offence then it is does not count

towards the Narrowing the Justice Gap target.

Recorded offences should not be confused with ‘Recordable offences’.

Recordable offences are those that have to be recorded on the Police National

Computer by law, resulting in a person having a ‘criminal record’. The category

of Recordable offences includes indictable, either way and most summary

offences and is a much larger group of offences than the category of

Recorded offences.

166

Appendix H: References

New Research on Uninsured Drivers, Association of British Insurers

(March 2004)

Response of The Association of British Insurers on behalf of Motor Conference

and the MIB to The Greenaway Review of Compulsory Motor Insurance and

Uninsured Driving, ABI (February 2004)

Vehicle Excise Duty Evasion, DFT (2002)

Road traffic: by type of vehicle: 1992-2002, DfT (2004)

Narrowing the Justice Gap Framework, Home Office (October 2002)

Illegal Parking in Disabled Bays: A Means of Offender Targeting, Sylvia

Chenery, Chris Henshaw and Ken Pease (1999, Home Office RDS)

The Criminal History of Serious Traffic Offenders, Gerry Rose

(2000, Home Office RDS)

Motoring Offences and Breath Test Statistics, Home Office (2004)

Roles and responsibilities review Highways Agency/ACPO, PA Consulting

Group (2003)

Engaging criminality – denying criminals use of the roads, PA Consulting

Group (November 2003)

Diary of a Police Officer, PA Consulting Group (2001)

Uninsured Driving in the United Kingdom, Professor David Greenaway

(July 2004)

This document has been prepared by PA. The contents of

this document do not constitute any form of commitment or

recommendation on the part of PA and speak as at the date

of their preparation.

No part of this documentation may be reproduced, stored in

a retrieval system, or transmitted in any form or by any means,

electronic, mechanical, photocopying or otherwise without the

written permission of PA Consulting Group.

LON ~ 10444

Published by TSO (The Stationery Office) and available from:

Online

www

.tso.co.uk/bookshop

Mail,Telephone, Fax & E-mail

TSO

PO Box 29, Norwich NR3 1GN

Telephone orders/General enquiries 0870 600 5522

Order through the Parliamentary Hotline Lo-call 0845 7 023474

Fax orders 0870 600 5533

Email book.orders@tso.co.uk

Textphone 0870 240 3701

TSO Shops

123 Kingsway, London WC2B 6PQ

020 7242 6393 Fax 020 7242 6394

68-69 Bull Street, Birmingham B4 6AD

0121 236 9696 Fax 0121 236 9699

9-21 Princess Street, Manchester M60 8AS

0161 834 7201 Fax 0161 833 0634

16 Arthur Street, Belfast BT1 4GD

028 9023 8451 Fax 028 9023 5401

18-19 High Street, Cardiff CF10 1PT

029 2039 5548 Fax 029 2038 4347

71 Lothian Road, Edinburgh EH3 9AZ

0870 606 5566 Fax 0870 606 5588

TSO Accredited Agents

(See Yellow Pages)

and through good booksellers

ISBN 0 11 341299 1

First published 2004

Printed in the United Kingdom by The Stationery Office

N173583 C5 19585 10/04 990866

Corporate headquarters

123 Buckingham Palace Road

London SW1W 9SR

United Kingdom

Tel: +44 20 7730 9000

Fax: +44 20 7333 5050

E-mail: [email protected]

PA Consulting Group is a leading management, systems and technology consulting firm,

operating worldwide in more than 35 countries.

Principal national offices

www.paconsulting.com

Argentina

Buenos Aires

Cerrito 866 – Piso 9

(C1010AAR) Buenos Aires

Tel: +54 11 4813 9898

Fax: +54 11 4811 9855

Australia

Sydney

146 Arthur Street

North Sydney

NSW 2060

Tel: +61 2 9964 2222

Fax: +61 2 9964 2233

Czech Republic

Prague

Panorama Centrum

Skretova 12

120 00 Prague 2

Tel: +420 246 086 666

Fax: +420 246 086 667

Denmark

Copenhagen

Tuborg Boulevard 5

DK-2900 Hellerup

Tel: +45 39 25 50 00

Fax: +45 39 25 51 00

Finland

Helsinki

Business Park Quartetto

House Cello

3rd Floor, Linnoitustie 2 B

FIN-02600 Espoo

Tel: +358 9 530 570

Fax: +358 9 530 574 40

France

Paris

54 avenue Hoche

75008 Paris

Tel: +33 1 56 60 51 45

Fax: +33 1 56 60 56 00

Germany

Frankfurt

Fürstenbergerstrasse 3-9

60322 Frankfurt am Main

Tel: +49 69 71 70 20

Fax: +49 69 71 70 22 63

Indonesia

Jakarta

Graha Iskandarsyah, 7th Floor

Jl. Iskandarsyah Raya No 66C

Jakarta 12160

Tel: +62 21 720 9948

Fax: +62 21 720 9949

Ireland

Dublin

Embassy House

Herbert Park Lane

Ballsbridge, Dublin 4

Tel: +353 1 668 4346

Fax: +353 1 668 1771

Japan

Tokyo

Marunouchi Yaesu Building

4F, 2-6-2, Marunouchi

Chiyoda-ku, Tokyo 100-0005

Tel: +81 3 5208 1500

Fax: +81 3 5208 1508

Malaysia

Kuala Lumpur

5th Floor

Bangunan Getah Asli (Menara)

148 Jalan Ampang

50450 Kuala Lumpur

Tel: +60 3 2161 2322

Fax: +60 3 2161 8231

Netherlands

Utrecht

Coltbaan 33

3439 NG Nieuwegein

Tel: +31 30 282 96 00

Fax: +31 30 282 96 08

New Zealand

Wellington

PO Box 1659

Lambton Quay

Wellington

Tel: +64 4 499 9053

Fax: +64 4 473 1630

Norway

Lysaker

PO Box 50

Lysaker Torg 25

N-1324 Lysaker

Tel: +47 67 58 67 58

Fax: +47 67 59 09 83

People’s Republic of China

Beijing

Unit 80D

8th floor Union Plaza

No 20 Chaoyang Men Wai Avenue

Beijing 100020

Tel: +86 10 6588 8559

Fax: +86 10 6588 8557

Hong Kong

20/F St George’s Building

No 2 Ice House Street

Central

Tel: +852 2913 5700

Fax: +852 2810 1432

Russian Federation

Moscow

Bolshoi Strochenovsky Pereulok 22/25

Stroenie 1

Moscow 115054

Tel: +7 095 775 2401

Fax: +7 095 775 2406

Sweden

Stockholm

Kungsgatan 8, 4tr

SE-111 43 Stockholm

Tel: +46 8 454 19 00

Fax: +46 8 454 19 01

United Kingdom

London

123 Buckingham Palace Road

London SW1W 9SR

Tel: +44 20 7730 9000

Fax: +44 20 7333 5050

Cambridge

Cambridge Technology Centre

Melbourn

Herts SG8 6DP

Tel: +44 1763 261222

Fax: +44 1763 260023

United States

Washington, DC

1750 Pennsylvania Avenue, NW

Suite 1000

Washington, DC 20006

Tel: +1 202 442 2000

Fax: +1 202 442 2001

Boulder

390 Interlocken Crescent

Suite 410

Broomfield, CO 80021

Tel: +1 720 566 9920

Fax: +1 720 566 9680

Cambridge, MA

One Memorial Drive

Cambridge, MA 02142

Tel: +1 617 225 2700

Fax: +1 617 225 2631

Chicago

311 South Wacker Drive

Suite 6330

Chicago, IL 60606

Tel: +1 312 566 9752

Fax: +1 312 566 9753

Houston

5847 San Felipe

Suite 1700

Houston, TX 77057

Tel: +1 713 821 1492

Fax: +1 713 821 1489

Los Angeles

520 South Grand Avenue

Suite 500

Los Angeles, CA 90071

Tel: +1 213 689 1515

Fax: +1 213 689 1129

Madison

2711 Allen Boulevard

Suite 200

Middleton, WI 53562

Tel: +1 608 827 7820

Fax: +1 608 827 7815

New York

The Chrysler Building

405 Lexington Avenue

New York, NY 10174

Tel: +1 212 973 5900

Fax: +1 212 973 5959

Princeton

315A Enterprise Drive

Plainsboro, NJ 08536

Tel: +1 609 936 8300

Fax: +1 609 936 8811