Optimising police resource utilisation for enhanced predictive policing

Police forces play a key role in ensuring the public’s safety by responding to emergencies, performing neighbourhood patrols, monitoring criminal activity, making arrests, investigating crimes and testifying in court. With today’s society in a state of austerity significant funding cuts are being implemented on public services including the police. It is vital that the police are still able to protect the public with the limited resources available and this means an increase in operational efficiency is required. This research investigates how the police response resources can be used to their highest efficiency, to ensure the public’s welfare, by investigating allocation of resources to incidents and directing patrol routes.

Two major aspects which affect the efficiency of response to incidents are the location of response resources when an incident occurs and which resource is selected to respond to an incident. These two problems are the focus of this research, which has yielded a computer aided dispatch algorithm and a computer aided patrol positioning algorithm. The computer aided dispatch algorithm automatically selects the most appropriate response resource to attend an incident. This involves multiple criteria decision making as there are many factors to consider when choosing the most appropriate resource including availability, area coverage and predicted response time. The method of dispatch developed results in decreased response times and better demand coverage as well as reductions in distance travelled.

The computer aided patrol positioning algorithm generates the ideal response resource patrolling locations by considering emergency demand coverage and targeting areas of high crime, hotspots, with patrols. Hotspots are found by analysing historical incident data and then using Kernel density estimation. The hotspots to patrol are then selected using a variation of the double standard location problem. This police response resource patrol positioning algorithm leads to decreased response times, increased demand coverage and problem area targeting.

This study has taken place in collaboration with Leicestershire police. Hence it has benefited from finding real world problems within the police force and allowed an algorithm to be developed to cater for these problems and fit in with methods already in place. Communicating with other police forces has confirmed this is a common problem, so even though details of the processes may not a line completely with all police forces, the tool developed is general and can be applied to all police forces.