Using automated number plate recognition as a crime management strategy
August 2004, News
Automated number plate recognition (ANPR) has been around for a number of years but this technology is increasingly maturing into a highly useful tool for policing and security. I had the privilege to sit in on a presentation by John Dean, the Association of Chief Police Officers’ (ACPO) National ANPR Co-ordinator at IFSEC recently.
The presentation, appropriately labelled `Denying Criminals Use of the Road' provided some eye-opening facts on how ANPR has been implemented as a policing strategy in the UK. I was struck by the potential of such systems for the South African environment as a general method of identifying offenders linked to certain vehicles, but also as a method of combating specific crimes such as highjacking or cash-in-transit robberies through recognition of the number plates of stolen cars.
Briefly, ANPR was described by Dean as a technology to allow vehicles observed by video camera to have vehicle registration marks read (through pattern recognition of visual images) and which then compares these 'reads' to the information in a number of databases. Dean indicates that while police have been using the technologies for years, improvements in the technology and reductions in costs have led to the increasing use of ANPR to address volume crime. He described in particular the increasing frequency of use of ANPR in proactive roadside operations in conjunction with intercept teams somewhat down the road from the camera. The improvements in the technology have resulted in extremely quick processing times and provide the basis for the intercept teams to respond almost immediately when the offending vehicle has passed by the camera.
Increase in arrest rate
Dean outlines how ANPR can substantially increase the arrest rate for officers, indicating that there is potential to improve arrest figures by 900% in some cases. The types of crimes for which people have been apprehended go way beyond issues such as non payment of licence fees or traffic fines. Vehicles stopped and inspected resulted in the arrests of people for vehicle crimes (12%), theft and burglary (21%), outstanding warrants of arrest (11%), drugs (11%) and driving offences (20%). A major trial project with 23 of the regional police forces between June 2003 and March 2004 scanned 22,8 million vehicle registration marks of which 900 000 or 4% were of immediate interest to the police. Available response resources stopped 136 857 of these vehicles and recovered 874 vehicles valued at £5,6 million. They also retrieved eight firearms and 159 offensive weapons from these vehicles in a situation where although illegal firearms are becoming more widespread in the UK, they are still far from the norm.
The 'Ring of Steel' that controls vehicle access and usage into the centre of London is based on ANPR. In his presentation, Dean pointed out that detection of criminals in this area was low compared to other police regions involved in the rollout of ANPR. However, this was attributed to criminals avoiding the use of suspect vehicles in this area because of the effectiveness of the ANPR - something that the police are deliberately trying to do with the 'Denying Criminals Use of the Road' strategy.
Successful ANPR implementation
ANPR needs to be implemented as a comprehensive strategy incorporating a number of Government, policing, traffic and local authority partners. Besides the technology, the backup support is also critical to success. Database accuracy and integrity need to be extremely high and up to date and new data needs to be captured into the systems almost immediately. For example, if stolen vehicles are entered as they are received into the system, the chances of quick recovery are increased substantially. Such quick reaction and detection could also have a potential impact on crimes such as cash-in-transit robberies where use is often made of high powered stolen cars. This indicates the potential need for dedicated resources to deal with the 'hits' obtained from the system.
Dean also notes the need for sensible locations for deployment. Having a number of cameras at strategic points in the road network around Johannesburg, for example, would need some planning. Also, arranging police response teams who can respond to the supplied information, either on a roving vehicle basis or static positioning near the camera site, would be an issue in terms of which force supplies such manpower and the cost of this. Extensive development and use has also been made of mobile camera resources with attendant response teams in the UK and this can be considered as an option. Of interest also, was that about 20% of apprehensions were based on personal observations of vehicles by police in the response teams at the side of the road - showing that just the presence itself has an advantage.
The use of ANPR on an ad hoc basis here and there has a limited benefit. The ability to implement it area wide, however, has huge potential. It would appear in the South African context that the major challenge is probably not just to have the ANPR technology in place, but to have the backup and database support systems in place, coordinated and up to date. We could certainly learn a number of points on this from the UK police. Benefits of such a system could play a major part in the prevention and detection of various forms of crime on South African streets.
Acknowledgement: John Dean (ACPO National ANPR Coordinator). Denying Criminals Use of the Road. Presented at the IFSEC BSIA Conference, May 2004.
Dr Craig Donald is an industrial psychologist and specialist in human factors in security and CCTV. He is the co-developer of the Surveillance and Monitoring Assessment Exercise (SAMAE) for the selection and placement of CCTV operators and presenter of the CCTV Surveillance Skills training course.
For more information contact Craig Donald, Leaderware, 011 787 7811, firstname.lastname@example.org, www.leaderware.com