To date, the use of AI services in policing has largely been in the areas of facial recognition and resource use in more effective ways, but a recent study led by the University of Northumbria highlights how to help prevent unsolved crimes, particularly by providing insights into weapons used in the crime. "Machine learning uses a range of algorithms to create complex data relationships," say the authors. "By careful fine-tuning, they can be used to estimate the salient features of ammunition used in a specific shooting event from the associated gunshot residue (GSR) deposited on surrounding surfaces or objects, including spent cases, injuries and, potentially, shooter's arms." The team believes that their approach represents a significant improvement from existing methods of GSR analysis, and the new approach represents unprecedented accuracy. This is the approach the group believes will bring new insight into some of the past high and unsolved crimes, such as the Bloody Sunday murders of 1972.