AI could protect historic sites from vandalism

Professor Robin Bryant from Canterbury Christ Church University, collaborating with Mark Harrison, Head of Heritage Crime at Historic England, is exploring an AI-driven approach to tackle this issue. The project aims to use supervised machine learning techniques to identify graffiti offenders by recognising their unique tags. This technology could also trace the movements of these vandals by matching graffiti across various locations and even analyse paint types to pinpoint the origins of the spray cans used.

Professor Bryant's research, funded by Historic England, analysed criminal activities over four years in Kent and Medway. It revealed that 8% of all recorded crime and anti-social behaviour occurs in or near protected heritage sites. He estimates that "approximately one in five listed buildings and one in four places of worship in Kent and Medway experience some form of crime each year."

AI technology offers the potential to connect multiple offences to the same perpetrators, facilitating more effective investigations. The research also highlighted that about one in ten monuments are impacted by crime, with over half of the registered parks or gardens and nearly four in five conservation areas experiencing crime annually.

Professor Bryant explained: "Our best estimates suggest that currently approximately one in five listed buildings and one in four places of worship in Kent and Medway experience some form of crime each year.

“AI offers the possibility to link offences together to the same offenders. In classic policing terms, this is normally a very good start in terms of an investigation.”

The study suggests that machine learning could effectively link criminal damage incidents against historic buildings, aiding in offender identification and damage prevention. This method could also identify 'copycat' offences, where similar tags are used by different offenders.

The research further revealed that most crimes at these sites involve theft of historical materials or criminal damage, including graffiti. It was noted that listed buildings and places of worship most at risk are typically located on the outskirts of urban areas rather than deep in the countryside, with the transport network playing a significant role in the targeting of these sites.

Professor Bryant emphasised his commitment to working with Historic England to fulfil their strategic goals of enhancing civic pride, prosperity, safety, and well-being in economically and socially disadvantaged areas.

Mark Harrison commented on the potential of this research, stating: “This research presents an exciting opportunity to develop new and innovative approaches to prevent and investigate graffiti within the historic environment.”