Making a case for futuristic predictive policing in India

Bibhu Prasad Routray

New Indian Express, 9 September 2012

Jharkhand police is adopting predictive policing — an ambitious futuristic crime control method which will allow the police to foretell not just the nature of crime, but also indicate the place where such a felony is to take place. As understood by the layman, post-adoption of the method, the police would be able to reach the crime scene before the crime is committed. Quite naturally, the state police appears to be excited about the project and have indicated that the pilot project could take off in the next eight months and the plan for the entire state could be a reality by 2014.

For the staunch believers in and practitioners of conventional policing, the project, which allows computers to predict crime, is a wasteful exercise. It is being argued that the computers would be ineffective instruments of peeping into the unpredictable ‘human’ nature of the criminals. It has been made out that there is no alternative to the strengthening of the police station — the basic structure of policing — and empowering the beat constable. While some of these arguments are valid, to write off the role of technology in future crime prevention may not be a good idea.

The futuristic policing system being implemented in Jharkhand in consultation with the Ranchi campus of the Indian Institute of Management is based on sophisticated algorithms and behavioural science- which complies crime-related data of past years. The software then analyses the data and produces results, predicting future crime in terms of locations and the nature of crime. The method does sound incredible, is yet not without successes.

Several of the police departments of the developed world are gradually taking help of software for predictive policing, with impressive returns. For example, use of the software ‘PredPol’ has resulted in a 25 per cent drop in burglaries in areas under the Los Angeles Police Department. A similar software developed by the IBM for the South Carolina police has also led to a decrease in crime rates. The Scotland Yard in Britain has popularised ‘Facewatch’ — a smart phone application that provides the viewers the pictures of suspects who may be in their areas. Coming across such a suspect, people can alert the police department by simply tapping on the suspect’s image.

Critiques of the futuristic policing models argue that the analytic capacities of the software infuses no additional expertise. Police officials use past data and their own analysis to identify crime patterns.

While that is true, it is also a fact that the ability to identify crime patterns is intrinsically linked to capacity of the individual officers, which is not a standard quality associated with all police officers. Benefits from the software, thus, could be immense, especially for states which have dearth of officials and ground-level personnel, vacancies at times running as high as 25 per cent. The software would provide instantaneous analyses, free from the vagaries of delays and follies associated with the human planning. The system would, thus, release more personnel to hit the road tackling crime rather than being busy in crime trend analyses. In an age of diminishing budgets, police departments would be able to optimise force deployment.

The biggest challenge would, however, come not from the believers in conventional policing, but from inside the police establishment and the legal system. Can the system work without a highly responsive and alert police force, which wholeheartedly believes in the new operating environment? Would the courts prosecute suspects arrested on the basis of a software prediction? Would the system prevent crimes in new areas? The fact that answer to all these questions is in the negative does not take away anything from the need for policing to go futuristic and tech-savvy.

The predictive policing software does not replace the conventional policing, but acts as an enabler. It facilitates the seamless integration of the merits of technology with the skills of the police personnel. It does not aim at solving crimes, but preventing them. More number of crimes being prevented would produce less criminals in the court and aid the over-burdened judiciary.

Of all the states, Jharkhand is probably the best positioned to experiment with predictive policing. Carved out in November 2000, it is a relatively new state, thereby burdened with comparatively less data to be fed into the system. Its a theatre where organised crime has mixed dangerously with extremism and conventional criminal activities. It is also the state which has performed exceptionally well in terms of police modernisation, especially in the past couple of years. Jharkhand’s success would possibly spur the other states to follow suit.

http://newindianexpress.com/magazine/voices/article601920.ece