Modern police force is using advanced analytical tools to bring down crime rates and predict criminal activities. The adoption of such techniques can improve their efficiency.
Speaking at the Artificial Intelligence Conclave organised by the Times Group last week, Ramit Sharma, Inspector General Police Allahabad Range, spoke about how the Indian Police is using AI for purposes such as sentiment analysis and resource allocation. Sharma has been at the forefront of technical application in the IPS.
One of the prominent application of AI in policing is the Police Emergency Management System (PEMS). PEMS currently consists of 3,200 cars in 75 districts of Uttar Pradesh. In addition to this, there are also a plan of including 1,600 motorcycles in the second phase. The PEMS periodically generates an analytical intelligence report. The report can use advanced analytics techniques such as Four Frame Analysis, Hotspot Analysis, Trend Analysis of criminal events. These services use machine learning at the backend.
Sharma explained that such reports are now used to predict criminal activities. Another important use of AI in modern policing has been social media analytics. The police monitors multiple social media channels such as Twitter, YouTube and open Facebook groups. Uttar Pradesh is only the second state in India to implement such advanced techniques. Sharma mentioned that network graph, word clouds, sentiment and geographical analysis are some of the tools used to understand complex issues in the communities.
Since the social networks are a set of nodes, Sharma was hopeful that AI could further facilitate to analyse activities and understand sentiments in a more efficient manner. In addition to PEMS, the police is also actively using video analytics in Intelligent Traffic Management System, where careful analysis of video footage is used to book offenders and better understand road conditions.
Sharma said at the event, “Allahabad Police is also working on an analytics solution to better manage crowds at the Kumbh Mela”.
The Kumbh Mela is the largest gathering of people in the world where over two crore people attend the event. Sharma mentioned several modeling techniques to model crowds such as cellular automata models and agent-based models. He also mentioned the partnership between Motilal Nehru Institute of Technology (MNIT) and the Allahabad Police. Together they are working on developing models to use AI to better manage crowds in the Kumbh Mela.
Try deep learning using MATLAB