Using Artificial Intelligence and Data to Combat Knife Crime

Given that knives can be effectively hidden under clothing, it might be difficult to spot carriers. Even more so in a crowd. Stop and Search is an invaluable tool to combat knife crime and, according to the Metropolitan Police Service, has resulted in more than 3,500 arrests for weapons possession. However, the procedure has long held a reputation for contributing towards ‘biased’ policing and can also lead to false allegations of bias against police officers themselves.

Both of these scenarios undermine trust between local police forces and communities, which can ultimately make it more challenging for police officers to reduce and/or prevent knife crime.

Further integration of new technologies into policing and security could help to spot knife carriers and reduce the need for physical searches, which may well contribute to tackling both real and perceived biases.

Thermographic (infrared) cameras, which detect the emission of body heat, are a non-invasive solution to discovering hidden weapons, including knives. The cameras are able to detect weapons hidden under clothing as objects over a certain size or made of certain materials, such as metal, block some transmission of body heat.

Infrared technology can be implemented within small portable equipment and could be used as an alternative or ‘first step’ for Stop and Search procedures on the street as it’s less invasive and more discreet.

Weapons screening with thermographic cameras at entrances to terminals and buildings could help to mitigate safety risks to both personnel and the public and could ultimately save lives. Additionally, mandatory screening for all entrants removes any biases from security protocols and could help to build public trust in security and policing.

Facial recognition

Facial recognition is a useful tool for preventing crime. It helps police forces identify people that may pose a risk to public safety and speeds up the arrest of those involved in knife crime to ensure they can no longer pose a threat to the public. Police trials of the technology have used targeted facial recognition, which does not undermine the privacy of the general public.

Targeted facial recognition scans the faces of passers-by and looks for a match within a pre-programmed database of existing persons of interest (also called a ‘Watchlist’). In the case of knife crime, only people previously convicted of a knife offence or currently suspected of involvement in serious and violent crime would be included on a police watchlist. In previous facial recognition trials, the data of any person found not to be a match was deleted.

The technology alerts local police forces or security personnel when it recognises a match to its database. Early warning of potential risk can lead to swift interventions and the prevention of serious and violent crime. As facial recognition uses biometric measurements to determine identity, it can be far more accurate than CCTV images in helping police officers to quickly identify the perpetrators of violent crime.

Software platforms fusing facial recognition, thermographic imaging and other technologies go even further than this. They can recognise visible weapons such as knives and immediately alert the police, which could lead to prompt intervention in a knife attack on the street.

Making use of data

There’s immense potential for greater use of Artificial Intelligence (AI) in policing to help reduce and prevent knife crime.

A little over a year ago, research conducted by Cambridge criminologists while working with a Metropolitan Police Service detective revealed how police data could be used to predict which areas of the country may be vulnerable to knife crime. The study identified a trend in which areas with a high amount of knife assaults in the previous year were more likely to see a knife homicide in the following year than those with no knife assaults.

While ‘predictive policing’ – ie the use of computer algorithms and statistical data to aid allocation of resources - has been researched and trialled by police forces in England over the past few years, its potential still remains largely untapped. Using AI to discover patterns in knife crime statistics could help police forces allocate resources where they’re most needed, make more informed decisions on specialist training and, ultimately, prevent crime.

Additionally, there’s the potential for AI and computer algorithms to help police speed up the removal of violent content online and tackle threats of violence imparted over social media. Run by the All-Party Parliamentary Group on Knife Crime in conjunction with Barnardo’s, a study involving young people revealed they’re able to view violent content online too easily and that such content was fuelling a perception that carrying knives ‘was the norm’.

Metropolitan Police Service Commissioner Cressida Dick has criticised social media platforms, claiming that they’ve played a role in gang violence and knife crime.

There are many contributing factors to the rise in knife crime (including social and economic issues) which, arguably, will not be addressed by technology, but existing new technologies can play an important role in immediately tackling some factors in the knife crime epidemic and help the police service to protect and save lives.

Adrian Timberlake is Chief Technical Officer at Seven Technologies Group