Roger Clarke's Web-Site


© Xamax Consultancy Pty Ltd,  1995-2016

'Covert Implementation of ANPR in Australia'

The Covert Implementation of Mass Vehicle Surveillance in Australia

Version of 19 March 2009

For presentation at the Fourth Workshop on the Social Implications of National Security: Covert Policing, 7 April 2009, ANU, Canberra

Roger Clarke **

© Xamax Consultancy Pty Ltd, 2009

Available under an AEShareNet Free
for Education licence or a Creative Commons 'Some
Rights Reserved' licence.

This document is at

The accompanying slide-set is at


Automated Number Plate Recognition (ANPR) applies optical character recognition to photographs of vehicles, in order to extract the vehicles' registration data. This paper outlines two alternative architectures for ANPR, referred to as the 'mass surveillance' and 'blacklist-in-camera' approaches. They reflect vastly different approaches to the balance between surveillance and civil liberties.

Australian policing agencies have been variously piloting and deploying ANPR, but without public oversight or control. A national agency, Crimtrac, is proposing to develop a vast database, which would store billions of entries showing the whereabouts of vehicles about which no suspicion of wrongdoing exists. Its purpose is expressly to facilitate mass surveillance of the Australian population. This represents national security extremism, and is a gross breach of trust by law enforcement agencies in Australia.


1. Introduction

Automated Number Plate Recognition (ANPR) uses digital cameras and software that provides Optical Character Recognition (OCR) capability to automatically extract the registration data of large numbers of vehicles. The technology is related to, but differs in important ways from, that which has long been used for 'speed cameras' and 'red light cameras'. Background is provided in HTS (2008) and Clarke (2008).

ANPR can be used in a variety of settings, including the entrances and exits of parking-stations, and pricing control-points on toll-roads. It can be applied to parked cars, in order to detect unauthorised vehicles and locate stolen ones.

For traffic management and traffic law enforcement purposes, cameras can be positioned adjacent to roadways, to monitor passing traffic. For minor infringements, it can be used in the same manner as 'speed cameras', as an administrative tool.

For road safety purposes, however (e.g. to assist in catching unregistered vehicles, unlicensed drivers and stolen cars), it is essential that it be used in conjunction with a capability to intercept vehicles of interest. This is a 'real-time location' technology, as that term is used in Clarke (1999a, 1999b). The purposes ANPR is put to may extend beyond traffic matters to policing generally, and to 'real-time vehicle tracking' and 'retrospective analysis'.

In the U.K., police use of ANPR has reached epidemic proportions, has been implemented in an uncontrolled manner, and relies on seriously error-prone underlying data (e.g. Lettice 2005). This has very enormous implications for privacy, and for democratic freedoms more generally. It is crucial that Australian implementations not make the same gross mistakes as the U.K.

This paper focusses on its use by Australian law enforcement agencies, and in particular on the extent to which data is collected and then stored for subsequent analysis. Its purpose is to summarise the position in Australia in relation to ANPR deployments by policing and other agencies, and the subsequent uses of the data that are envisaged by policing and many other law enforcement agencies. The paper reflects research undertaken by the author and other Board members of the Australian Privacy Foundation, in particular APF (2008a) and APF (2008b).

The paper commences by describing the conventional architecture that has been developed and is applied in a number of countries, most intensively in the United Kingdom. The descriptive term 'Mass Surveillance ANPR' is used. An alternative architecture is described, referred to by the term 'Blacklist-in-Camera' technology. This supports targeted use of ANPR for law enforcement, without generating a vast mass surveillance database.

To date, there has been only limited deployment of ANPR in Australia. Current use by policing and other agencies is outlined, shown to have been implemented without public overview or consultation, and shown to use Mass Surveillance technology. A national agency is shown to be seeking to facilitate these implementations, and to warehouse the captured data in what would be the nation's first-ever genuine mass surveillance operation.

2. Mass Surveillance ANPR

A brief description of conventional ANPR architecture is as follows:

"The ANPR camera-unit can be designed to transmit every instance of vehicle registration-data that it is able to extract from passing vehicles. The receiving device might be a display, for example in a nearby police patrol vehicle. In practice, however, the receiving device is generally a computer with substantial data-storage. The extracted registration-data may be used for user-pays charging and/or law enforcement, ..., but is also stored, together with the date, the time and some indication of location and perhaps direction of view or of movement" (Clarke 2008).

Exhibit 1 provides a diagrammatic overview. A camera captures images that include vehicle number plates. Combined with the camera is one or more processors that extract the number-plate as text, possibly store the image and text, and transmit the image, text and meta-data relating to the image (e.g. date, time and location) to an operational policing hub. The hub selects those that it wishes to intercept and issues alerts to a police vehicle downstream from the wanted vehicle. The hub stores all of the data and/or passes it to a central location for storage and subsequent use.

Exhibit 1: Architecture of Mass Surveillance ANPR

The effectiveness of ANPR for traffic law enforcement and road safety purposes depends on a large number of factors, in particular the quality of image capture and data-extraction, the speed of alert-generation, the quality of the data on which the alert-generation is based, and the current availability of a suitable resource downstream from the point of sighting.

The primary focus of this paper is the capture and retention of images and data relating to all sightings, not just the small minority that give rise to alerts. The discussion in (Clarke 2008) continued:

"Over time, and with the proliferation of image-capture devices, the effect of this process is the accumulation of a massive database of vehicle movements. Nothing remotely resembling it has ever existed in the past, even in the old USSR (where internal passports were used to restrict freedom of movement) and East Germany (where monitoring of the population reached its then greatest extremes).

"The justification for such mass surveillance is that there is intelligence value in ANPR data. It might be feasible to locate designated vehicles, to track them in real-time, and to submit vehicles of interest to retrospective tracking. Further, proponents postulate that a wide array of (loose) inferences may be able to be drawn about vehicles being associated with one another in some manner (such as travelling in proximity, or being co-located on multiple occasions).

"Firstly, it is far from clear that any such intelligence benefits are real, and secondly, it appears that national security agencies expect their propositions to be accepted by politicians and the public without supporting evidence, and without question. Even the most cursory consideration of the claims leads to a completely contrary conclusion: vehicle registration data is unreliable, false positives will be frequent, forgery is easy, and both 'organised crime' and terrorists can readily organise themselves so as to circumvent, nullify and even subvert such monitoring".

The deployment of ANPR in the U.K. has expressly been of Mass Surveillance ANPR. See, for example, Lewis (2008). The current project represents the fulfilment of a joint police forces strategy to implement a 24x7 vehicle movement database, in order to "fully exploit Vehicle Intelligence" (ACPO 2005).

Exhibit 2 presents the problems with Mass Surveillance ANPR, as summarised by the Australian Privacy Foundation.

Exhibit 2: Extracts from the APF Policy Statement (APF 2008b)

5. As commonly practised, and as supported by currently available technologies, [Mass Surveillance] ANPR represents a gross privacy intrusion, and in some jurisdictions breaches privacy law, in the following ways:

6. As commonly practised, and as supported by currently available technologies, ANPR is a mass surveillance technique and breaches the human right of liberty of movement (UDHR 13.1, ICCPR 12.1). More specifically, with conventional [Mass Surveillance] ANPR:

7. The practice of ANPR can readily become arbitrary interference by law enforcement officers, in such ways as the following:

8. The effects of the practice of ANPR on the public reputation of law enforcement agencies and individuals can be positive, in that they will be seen to be active, and to be effective; but run a great risk of being seriously negative, in that they will be seen to be intrusive into the activities of law-abiding citizens, and a key part of a 'police state' apparatus that gathers vast quantities of information about people's movements

3. Blacklist-in-Camera ANPR

The APF's Policy Statement also drew attention to the existence of an alternative way of structuring ANPR infrastructure, which satisfies all of the law enforcement requirements in relation to interception of vehicles, but avoids many of the worst features of Mass Surveillance ANPR.

In verbal evidence, this author described the architecture as follows: "the camera has a list of number plates that it is looking for, which may be drawn from multiple sources, of course, depending on what the objectives are. Clearly, motor vehicle registrations that are suspect, that are wanted by police, that have been reported stolen and are not yet reported as unstolen or found, there could be multiples of these lists. But that black list [is] inside the camera, such that the only data that escapes from the camera, [that] is reported to wherever it is reported to, is a hit. ... That clearly would ... avoid all of these problems that arise with a mass surveillance technique gathering large quantities of data about many people's movements" (QT 2008a, pp. 11).

A slightly more technical description is as follows: "The camera-unit can be designed as a high-security device that only discloses data that satisfies tightly-defined and tightly-controlled criteria. ... Tightly-coupled processing within the camera-unit can compare the registration data extracted from images against one or more controlled blacklists that have been downloaded to it. These can contain the registration numbers of vehicles that law enforcement agencies want to intercept for specific reasons. The only data disclosed by the device would be high-probability 'hits' against those blacklists" (Clarke 2008).

Further key requirements of the 'Blacklist in Camera' design include:

Exhibit 3 presents the architecture in diagrammatic form.

Exhibit 3: Architecture of Blacklist-in-Camera ANPR

Blacklist-in-Camera architecture is not just a theoretical possibility. Applications of it have been in use for some time, at least in various parts of Canada. In a variant in Ontario, for example, "Each licence plate number is transmitted to an onboard computer and immediately compared to the database of stolen vehicles, which is stored on the hard drive of the computer" (OIPC 2003, p. 2). Using this approach the intelligent component of the apparatus, and the database, are in the police-car rather than the camera; but there is still no operational need for the image or the data to be sent to a central point unless it relates to a hit against the blacklist-in-device.

4. Covert Implementation of ANPR in Australia

A number of traffic management and law enforcement agencies in Australia have variously conducted pilots and deployed ANPR capabilities. The first sub-section below provides an overview of these activities. The second sub-section discusses the emergent coordinative mechanism whereby the law enforcement community intends to establish mass surveillance of motor vehicles along similar lines to the rapidly-emergent U.K. "surveillance society" (Ford 2004, HoL 2009).

4.1 Activities of Individual Agencies

This section identifies the extent to which ANPR has been implemented in Australia, and to which the deployments have been the subject of prior evaluation, consultation and authorisation by Parliaments.

(1) Deployments

Predecessor technologies in the forms of 'red-light cameras' and 'speeding cameras' have long been in use in Australia. Capture of an image of a vehicle and its identifier is triggered by an event (variously, detection of a vehicle in a location at a particular time, or moving at a speed, that indicates that a crime may have been committed, or by synchronisation with light-changes from green to amber or amber to red). These give rise to a range of issues (e.g. of accuracy, of data security, and in some cases of automation unmediated by human review). But the issue of mass surveillance does not arise, because no image is captured unless a trigger exists, and a relatively very small proportion of transport movements find their way into a database.

ANPR has technical differences from its predecessors. In particular, it necessarily involves digital rather than wet-chemistry photography, and automatic extraction of the registration data in real-time rather than manual and/or deferred extraction.

There have been occasional media releases or press coverage of pilots or initial implementations of ANPR. The primary consolidated sources known to this author are that made available in June 2008 at Appendix 2 on p. 17 of Crimtrac (2008c), mirrored here, and QT (2008b, pp. 5-6). These contain inconsistencies; but it appears that, in most States and Territories, one or more agencies has deployed or at least piloted ANPR, with an apparent total of between 300 and 400 cameras acquired - although not all of them are currently operational.

The only longstanding and well-established application appears to be that used by the NSW Road Transport Authority for trucks, called Safe-T-Cam, which operated using older technologies from 1989, but has since migrated from wet-chemistry photography to ANPR. It involves 24 fixed-location cameras, and is to some extent integrated with a South Australian scheme operating a further 11. There is one known instance of the application of the system to cars, in 1998, although the author is not aware of any legal authority for that use (Clarke 2000).

On the basis of the limited available information, it appears that all instances of ANPR in Australia apply Mass Surveillance ANPR architecture. None appear to use Blacklist-in-Camera technology.

(2) Legal Authority and Public Justification for Deployments

The first and to date only instance of detailed consideration by a Parliament, and of any form of public consultation, was an Inquiry by the Queensland Parliamentary Travelsafe Committee in 2007-08. The 32 Submissions to that Inquiry included two by Privacy Commissioners and two by civil liberties and privacy advocacy organisations - APF (2008a) and QCCL (2008).

The Victorian Privacy Commissioner's submission concluded that "The whole concept of an individual's right to anonymity is sacrificed: it is no longer possible to drive on a public road anonymously, even if one is doing nothing wrong" (OVPC 2008, p. 4).

The Federal Privacy Commissioner concluded that (OFPC 2008):

The Queensland Parliamentary Travelsafe Committee reported in September 1998 (QT 2008b). It concluded that:

In short, the sole public assessment of ANPR conducted in Australia to date concluded that no justification has yet been demonstrated based on either road safety or even operational efficiency. Hence:

There are further features of the current situation that give rise to very serious concerns about human rights and parliamentary laxness in exercising oversight over the actions of government agencies. To the best of the author's knowledge:

4.2 Coordinative Activities

A wide range of law enforcement agencies exist, in all nine jurisdictions. A key means of achieving coordination among them is Crimtrac, which is a federal government agency formed in 2000 as an instrument of 'collaborative federalism'. Its function is to provide information and information technology services to law enforcement agencies of all jurisdictions, and to facilitate communications among them. Among other services, it operates national fingerprint, DNA and child sex offender databases. Its clientele extends well beyond the policing agencies and the other uniformed services, to many other government agencies in all jurisdictions, and to a range of non-government organisations.

Between early 2006 and late 2008, Crimtrac conducted a $2 million project on ANPR: "The Commonwealth government has funded a scoping study to examine a national approach to ANPR through Proceeds of Crime money.  CrimTrac will prepare a report, outlining options and feasibility for a national ANPR capability which will be delivered to the Minister for Home Affairs and the Ministerial Council for Police and Emergency Management - Police by late 2008" (Crimtrac 2008a). As part of the project, consultants were hired to prepare a Privacy issues Analysis, and to conduct a Privacy Impact Assessment (PIA).

A Consultation Paper was provided to a small PIA Consultation Group (Crimtrac 2008c). That Paper stated the assumption that "At this stage it appears likely that the system will collect and store for at least a period all sightings of all vehicle passengers" (p. 6). Further: "there appears to be quite strong agreement [among law enforcement agencies] for the concept of a national APNR system that would make available the combined ANPR data from a range of sources, facilitate information sharing between State/Territory and Australian government agencies and allow searching and analysis of the national set of ANPR data over time", and "All parties considered the system must have the ability to capture data for all passing vehicles through ANPR equipment, rather than just matches against a hotlist" (p. 7).

A 'National Automated Vehicle Recognition System' (NAVR) was envisaged. Key underlying assumptions in the design were (p. 8):

Crimtrac and its PIA consultant were well-aware of the 'Blacklist-in-Camera' alternative architecture when the Consultation Paper was prepared. Among other things, the relevant Crimtrac staff were present when evidence was presented in both written and verbal form to the Queensland Parliamentary Travelsafe Committee (QCCL 2008 and QT 2008a, pp. 10, 11).

In evidence to the Queensland Parliamentary Travelsafe Committee on 14 March 2008, the Crimtrac CEO said that "We have not yet determined exactly the extent to which we would capture all data. It may well be that we only capture hot list data" (QT 2008a, p. 17). Yet only 3 months later, Crimtrac chose to exclude from its Consultation Paper any reference to such a possibility. In short, by the time the PIA was conducted, the 'scoping study' was already fully committed to the facilitation of Mass Surveillance ANPR.

It is unclear whether any written submissions were made during the course of the Crimtrac PIA. None are apparent on the web-sites of the federal, Victorian and N.S.W. Privacy Commissioners. Contrary to expectations, the PIA Report was not published promptly on conclusion of the project, and had not been published at the time of writing in March 2009. The Scoping Study Report as a whole was also due to be completed by late 2008, but neither the Report nor anything about it had been seen publicly at the time of writing.

Crimtrac's CEO, Ben McDevitt, and the Chair of its Steering Committee, Ken Moroney, intentionally created the impression that the agency had an open mind on data capture and retention, and would conduct its study openly, including a full PIA. The agency has abjectly failed to fulfil that commitment. The delays further enhance the opportunity for law enforcement agencies throughout the country to implement ANPR in its mass surveillance form. This creates the likelihood of a fait accompli argument being launched as a defence against public concerns - and, indeed, it is readily inferred that the delays are at least in part intended to achieve that end.

5. Conclusions

A wide array of serious issues arise from ANPR of any kind, including the accuracy of extraction of plate-numbers under varying operational conditions, the accuracy and timeliness of blacklists, vehicle interception procedures, police powers following such interceptions, access to the images and resulting data, the ease and incidence of falsification and duplication of plate-numbers, and the prospect of the onus of proof being inverted and people being required to prosecute their innocence. There is also a concern about the potential for the images to be used to identify people as well as vehicles (e.g. Dearne 2008).

Substantial as those concerns are, they pale into insignificance in comparison with the dramatic change in climate, from a relatively free nation to a 'surveillance society', that is inherent in the implementation of Mass Surveillance ANPR.

During the period 2002-08, the Howard Government dramatically worsened the civil rights / law enforcement powers balance, and significantly reduced the already inadequate controls over the activities of national security agencies (APL 2008). The reputation of the Australian Federal Police is in tatters following its gross over-reaction in the Mahomed Haneef affair, and its inability to acknowledge that errors had been made and to fix the problem (CI 2008). The Rudd Government has provided indications it will do something to restore the balance, by considering the rescission of unjustified measures, and the adjustment of others. After 16 months, however, it has not actually done anything, and the Independent Reviewer of Terrorism Laws Bill 2008 [No. 2] still languishes on the parliamentary table.

ANPR is a litmus test of the Rudd Government's capacity to withstand the backroom pressure put on it by the law enforcement community. The Australian public wants law enforcement agencies to have appropriate technology and appropriate powers; but not to the extent that freedoms and democracy are undermined.

The lack of effective public oversight, the substantial opaqueness of the process, and the long delays in publishing information, lead to the inevitable conclusion that national security and law enforcement agencies in Australia are using covert means to implement mass surveillance of motor vehicles. As expressed by Michael Cope, on behalf of QCCL (QT 2008a, p. 10): "It is a serious concern to us that the creation of this vast database represents a serious threat to privacy and individual liberty. It is really straight out of the Big Brother handbook".


ACPO (2005) 'Denying Criminals the Use of the Roads' Association of Chief Police Officers, March 2005, at

APF (2005a) 'Numberplate Recognition Technology' Letter to the N.S.W. Police Commissioner, Australian Privacy Foundation, January 2005, at

APF (2005b) 'Numberplate Recognition Technology', Letter to the Australian Privacy Commissioner, Australian Privacy Foundation, January 2005, at

APF (2008a) 'Automatic Number Plate Recognition Technology (ANPR)', Submission to the Queensland Parliamentary Travelsafe Committee, Australian Privacy Foundation, January 2008, at

APF (2008b) 'Automated Number Plate Recognition (ANPR)', Policy Statement, Australian Privacy Foundation, March 2008, at

APL (2008) 'Terrorism Law', Australian Parliamentary Library, 2008, at

CI (2008) 'Report of the Clarke Inquiry into the Case of Dr Mohamed Haneef' Commonwealth of Australia, 21 November 2008, at$file/Volume+1+FINAL.pdf

Clarke R. (1999a) 'Person-Location and Person-Tracking: Technologies, Risks and Policy Implications' Proc. 21st Int'l Conf. on Privacy and Personal Data Protection, pp.131-150, Hong Kong, 13-15 September 1999. Revised version in Information Technology & People 14, 2 (Summer 2001) 206-231, at

Clarke R. (1999b) 'Relevant Characteristics of Person-Location and Person-Tracking Technologies' A separately-published Appendix to Clarke (1999a), Xamax Consultancy Pty Ltd, Canberra, October 1999, at

Clarke R. (2000) 'How to Ensure That Privacy Concerns Don't Undermine e-Transport Investments' Proc. AIC e-Transport Conf., Melbourne, 27-28 July 2000, at

Clarke R. (2008) 'You Are Where You've Been Location Technologies' Deep Privacy Impact' Invited Keynote, Seminar on 'Location Privacy', University of N.S.W., 23 July 2008, at

Clarke R. & Wigan M. (2008) 'You Are Where You've Been: Location Technologies' Deep Privacy Impact' Proc. Third Workshop on Social Implications of National Security, Canberra, 23-24 July 2008. Republished in Michael K. & Michael M.G. (2008) 'Australia and the New Technologies: Evidence Based Policy in Public Administration' Research Network Secure Australia, July 2008, pp. 100-114, at

Crimtrac (2008a) 'ANPR - Automated Number Plate Recognition' Crimtrac, 2008, at

Crimtrac (2008b) 'Crimtrac Board Considers Number Plate Technology', OnTrac magazine 1, 2 (2008), pp. 10-11, at

Crimtrac (2008c) 'Privacy Impact Assessment Consultation paper: Automatic Number Plate Recognition - CrimTrac Scoping Study' Crimtrac, June 2008, at

Dearne K. (2008) 'Privacy concerns on speed cameras' The Australian IT Section, 23 September 2008, at,24897,24387179-15306,00.html

Ford R. (2004) 'Beware rise of Big Brother state, warns data watchdog' The Times, London, 16 August 2004, at

HoL (2009) 'Surveillance: Citizens and the State' House of Lords Select Committee on the Constitution, February 2009, at

HTS (2008) 'License Plate Recognition - A Tutorial' Hi- Tech Solutions, Israel, November 2008, at

Lettice J. (2005) 'No hiding place? UK number plate cameras go national' The Register, 24 March 2005, at

Lewis P. (2008) 'Fears over privacy as police expand surveillance project' The Guardian, 15 September 2008, at

OFPC (2008) 'Inquiry into Automatic Number Plate Recognition Technology: Submission to the Queensland Parliamentary Travelsafe Committee: Issues Paper No.12 ' Office of the Federal Privacy Commissioner, February 2008, at

OIPC (2003) 'Privacy Investigation: The Toronto Police Service's use of Mobile Licence Plate Recognition Technology to find stolen vehicles' Office of the Information and Privacy Commissioner, Toronto, Ontario, 29 April 2003, at, original PDF version mirrored at

OVPC (2008) 'Submission for the Travelsafe Committee Queensland Parliament Inquiry into Automatic Number Plate Recognition Technology (ANPR)', Office of the Victorian Privacy Commissioner, January 2008, at

QCCL (2008) 'Submission to the Chair TravelSafe Committee, Parliament House ' Queensland Council for Civil Liberties, 23 January 2008, at

QT (2008a) 'Inquiry into Automatic Number Plate Recognition Technology: Transcript of Proceedings - 14 March 2008' Queensland Parliamentary Travelsafe Committee, Hansard, at

QT (2008b) 'Report On The Inquiry Into Automatic Number Plate Recognition Technology' Report No. 51, Queensland Parliamentary Travelsafe Committee, September 2008, at

Wikipedia entries:

Declarations: In late 2007, this author undertook a consultancy assignment for Crimtrac, through his consultancy company, to prepare a Privacy Issues Paper on ANPR. In March 2008, he was primary author of the APF's Policy Statement on ANPR, and appeared on its behalf before the Queensland Parliamentary Travelsafe Committee. In April 2008, he responded to Crimtrac's invitation to tender for the performance of the PIA, in the process disclosing his absence overseas during part of June and July. The tender was not successful. Because of the potential for conflict of interest, he did not participate in the APF's interactions with Crimtrac during the conduct of that PIA.


This paper draws heavily on prior work by the Australian Privacy Foundation. The contributions of (in alphabetical order) Usman Iqbal, Katina Michael, Dan Svantesson, David Vaile and Marcus Wigan are gratefully acknowledged.

Author Affiliations

Roger Clarke is Principal of Xamax Consultancy Pty Ltd, Canberra, and is a Board member and currently Chair of the Australian Privacy Foundation. He is also a Visiting Professor in the Cyberspace Law & Policy Centre at the University of N.S.W., a Visiting Professor in the E-Commerce Programme at the University of Hong Kong, and a Visiting Professor in the Department of Computer Science at the Australian National University.

xamaxsmall.gif missing
The content and infrastructure for these community service pages are provided by Roger Clarke through his consultancy company, Xamax.

From the site's beginnings in August 1994 until February 2009, the infrastructure was provided by the Australian National University. During that time, the site accumulated close to 30 million hits. It passed 50 million in early 2015.

Sponsored by Bunhybee Grasslands, the extended Clarke Family, Knights of the Spatchcock and their drummer
Xamax Consultancy Pty Ltd
ACN: 002 360 456
78 Sidaway St, Chapman ACT 2611 AUSTRALIA
Tel: +61 2 6288 6916

Created: 27 February 2009 - Last Amended: 19 March 2009 by Roger Clarke - Site Last Verified: 15 February 2009
This document is at
Mail to Webmaster   -    © Xamax Consultancy Pty Ltd, 1995-2013   -    Privacy Policy