8.7 Integrating Security Technologies
Security technologies are often used alongside other crime prevention measures or integrated with other technologies to enhance their effectiveness. For example, the security solution that has been put in place to address the risk posed by terrorism in London is a convergence of target hardening and CCTV surveillance practices (Fussey, 2007). The anti-terrorist strategies put in place to counter the effects of potential major terrorist threats, such as vehicle-based bombs, include traditional situational crime prevention measures (e.g. introducing barriers and bollards, reducing the number of entrances to buildings, restricting parking) which work alongside surveillance technologies.
The three security technologies discussed above are often linked to computer databases that help process the data collected. Lyon (2002: 246) made the following comment when examining security technologies and the opportunities they provide for intensive surveillance:
It is their dependency on computer-based information infrastructures that gives them their peculiar power. Without the assistance of complex and sophisticated data processing power, these new technologies would remain relatively weak as surveillance tools
Digital CCTV systems have been linked to various computer programmes to help operators identify 'unusual events' such as a person behaving suspiciously, problematic individuals or specific vehicles (Graham, 2000: 47). The purpose of the technologies is to help law enforcement agencies 'in their decision-making coordination, control, analysis and visualisation (Ceyham, 2008: 109). Biometrics systems depend heavily on connective technologies because without the system's ability to compare the captured data to individuals in a searchable database, the process would not have the capacity to efficiently identify individuals and threats.
Facial recognition technology is a security measure that integrates CCTV and biometric systems. The process involves CCTV cameras surveying individuals and attached to the cameras are "Face Grabbers" which are devices that extract face images from the continuous video stream (Gates, 2002). The images are converted into templates that are compared to templates stored in a database of risky individuals such as terrorists or criminal suspects. When a match is found the system will alert a human operator who takes appropriate action.
Introna and Wood (2004) suggest that the processing of people through facial recognition systems has a number of political implications that are not fully understood. The systems depend on algorithms that define when individuals are matched to samples taken by the cameras. These algorithms have a profound impact on how the systems work and the systems make use of computer programmes 'to provide more than the raw data observed' (Introna and Wood, 2004: 181) through the CCTV cameras. Research has found that the algorithms display identification biases and recognition rates are higher for males and older people. Green et al (2003: 9) came to the conclusion that 'some people are easier to identify than others' and this included the finding that other races were easier to identify than white including Asians and Afro-Americans. The technology is not intrusive and data can be collected and processed without individuals' knowledge and the algorithms at the centre of the process are very difficult to scrutinise due to their complex nature.
The effectiveness of facial recognition technologies relies heavily on the quality of the image captured and in crowded or outdoor environments capturing a high quality image can be problematic. A number of other factors can impact on the identification process and these include the length of time between when the image in the database was captured and when the image was taken by the camera (ibid: 189). If systems have to compare images to large numbers of records in a database this can have a detrimental effect on the system's ability to identify people. Lyon (2003: 671) argued that facial recognition technology has only limited uses and will probably be ineffective against terrorism as capturing high quality images of terrorists to store in data is difficult and they can wear disguises to avoid recognition.
A facial recognition system was linked to 300 cameras in the Newham district of London. Introna and Wood (2003) reported that the police admitted the system had not led to any positive identifications, despite being linked to a small database of offenders. The integration of security technologies appear to offer ways of processing data collected about individuals and alerting agencies to risky individuals. The technological fix that many of the systems offer has many weaknesses and the technology has to prove itself in real world environments before it will become a widespread technique of surveillance. Many terrorist threats are not from individuals who are known to law enforcement agencies meaning their data will not be stored on databases which allows them to avoid automated recognition.