This innovation is aimed at introducing smart policing measures in highway patrolling and it is unique in the sense that it uses data analytics and technology tools to decipher the existing situation. Let us illustrate that using a simple example. The team used its spatial analytics tools to map and analyze public locations such as restaurants, hospitals, ATMs, residential areas and other similar areas against the details of past accidents on the NH4. This helped us make projections, which were important in order to identify the most common accident hotspots. Not just this, we used the data insights to create a patrolling schedule to be followed by the constables. We also traced the path that a patrol vehicle should take in order to be at a hotspot when it sees the maximum number of accidents. All of this is indicative of a major break from the established status quo.

Using the available data, we developed a new format elicit vital information regarding accidents from the five police stations overseeing the highway. This form comprised of the following metrics:

1. Sections of the Indian Penal Code and the Motor Vehicles Act, 1988 that were invoked

2. Number of fatalities and number of people who sustained serious injuries

3. Location of the accident (with the landmark)

4. Time of the accident

5. Parties (pedestrian and type of vehicles) involved

Of course, due to the lack of digital housekeeping of data, the first phase of the project focused on digital data conversion. This exercise comprised of trawling handwritten logs of road accident related complaints, injuries and fatalities, and extracting information about the location and timing of accidents, the type of vehicles involved and the penal sections invoked.

Let us be very honest and say that this was the most difficult phase of the project because while there was a lot of data, it was in different formats, disaggregated and unstructured. Next, the data analysis phase was initiated with the deduction of hot-spots – places with repeated occurrences of fatal and serious accidents, based on the data that was available but never utilized and processed to arrive at meaningful insights.

In addition, surveying of the NH-4 stretch passing through the district was done to know of the infrastructural defects, which would be crucial to the correlation of trends in accidents observed. This included checking for the adequacy of signposts. Photographs were clicked and landmarks common to the hotspots (like petrol pumps, restaurants, and junctions) were paid special attention. The team members also had access to GIS data and were able to capture relevant details through the same. Based onthe geo-analysis conducted through GIS data captured to determine hotspots, a patrolling schedule was devised for the constables and the patrol vehicles bearing in mind the temporal vulnerabilities of the hotspots. The timebased analysis of the accidents yielded that majority (78%) of the accidents occurred between 10:30 AM – 8:30 PM, with 45% of them between 2:30 - 5:30 PM.

These details were used to construct an efficient patrolling schedule with a revived focus on better vigilance and human resource support during the hotspot time slots. Vehicular analysis revealed that 56% of the accidents involved a 2W-4W pair, suggesting either over-speeding of the 4W or violation of lane rules by the 2W. The over-speeding insight of 4W vehicles was supported by the preponderance of skidding cases recorded by the same vehicle type. Regarding pedestrians, the data analysis confirmed the worst fears that 96% of cases involving pedestrians led to a fatality. Therefore, based on the analysis of the obtained data, there were some incisive insights generated that could be transformed into actionable inputs for the police to work upon. A dossier was prepared for each hot-spot. It contained information about the timeperiod when maximum accidents occur, the most prevalent vehicular combination involved during collisions and the efficacy of the sections invoked in deterring misdemeanors in the future.

For this project, we collaborated with the Deputy Superintendent of Police Harssh Poddar, who was the nodal authority for this innovation. Harssh ensured coordination and support from all the different stakeholders involved in this project i.e. National Highway of India officials, who were responsible for upkeep and maintenance of the highway stretch, the Maharashtra Highway Road Development Corporation which looks at roadside amenities, and the Maharashtra Highway State Police apart from the police officers. For instance, we photographed the infrastructural defects visible during our tours and a report containing the drawbacks was submitted to the deputy engineer of Maharashtra Highway Road Development Corporation, which has already started taking action on the specific points. Therefore, we were able to get the support from all the concerned stakeholders. Without this collaboration, the innovation would have suffered because the mitigation strategy focuses on the coordinated effort from all the concerned stakeholders.

The police officials helped us understand the data, fill up the data forms we had prepared as well as explain the local context and situation for a qualitative analysis. Because of the coordination between different stakeholders, as described in the previous response, there has been an increased onus on information sharing and proper planning. Even among the police stations, there is better coordination as borne by the fact that today there is a patrol squad comprising of policemen from all the 5 police stations. Also, let us provide a very specific example to illustrate the coordination. In our dossier prepared for each hot-spot, we had specified how under the Gokul Shirgaon (GS) police station, at milestone 601/00 there was a need for a rumble strip near the hotel to prevent overspeeding vehicles. Today, the Maharashtra Highway Road Development Corporationhas constructed the rumble strip. Moreover, given that 42.6% of all accidents in GS led to a fatality, there was an increased focus on providing ambulance support by NHAI.

A highway patrol squad comprising of eight constables (from the five police stations) and helmed by an Assistant Sub-Inspector of Police was constituted. Interviews and orientation sessions for the constables were conducted. The timings for manual supervision by the constables have been deduced by accounting for susceptibility of a hotspot at any given time. Utmost consideration has been given to ensure coordination between the three agencies; the highway squad, the NHAI patrolmen, and the State HSP personnel while surveying the highway.

The exercise has borne fruits, for the number of accidents and fatalities have both reduced by double digit percentage points in the first month itself. In particular, as a result of targeted patrolling, highway accidents have reduced by 40% in 4 months. Constant feedback is being obtained from the squad’s supervisor and further refinements are being pursued. Using the feedback, the team is regularly working on refining the mitigation strategy. For instance, upon interacting with the highway patrol squad to get their inputs and feedback and also understand their experience pertaining to the application of the Motor Vehicles Act, 1988, we discovered that the provisions penalizing overspeeding and drunken driving couldn’t be applied due to the paucity of odometers (speed measuring devices) and breathalyzers. Based on this finding, the Police was advised to procure the necessary equipment for mitigating road accidents and fatalities. The remedial measures are underway. In order to tend to the injured party in the aftermath of an accident the constables responsible for manning the highways are also undergoing an extensive training on administering first-aid to the aggrieved in tie-up with the local hospitals.

The pace of reforms in the policing system is very slow. What compounds the problem is the apprehension of the lower level officer towards usage of data and technology. In Kolhapur, officials at the lower level (who handle most of the data collection) were unwilling to cooperate with us. We employ a bottoms-up approach in all our programs, however, official engagement with them was not working. Hence we engaged with them in an informal setting to ascertain the reasons behind their unwillingness. We realized that the officers suffered from inadequate information about both data collection and analysis. They had been handed this work but not explained the role of data and technology in administrative systems. We decided to create awareness amongst them. Consequently, we conducted sessions focused on explaining the role of data and its advantages. There was a significant improvement in their reaction as they slowly came to appreciate their role in the project.

Data analytics, technology, policy analysis and problem-solving are the core skill set of the team. While we utilized all four tools for the project, two crucial components were also required. The first one was perseverance of the team when confronted with unexpected tasks. For example, we had to convert the hard copies of data into a digital format. The conversion to digital data required long hours of manual work, without which no analysis could have been made. A supporting hand was lent to us through Harssh. He ensured that he uses his position for the best outcome of the project. Through his efforts, the sustainability of the project was guaranteed. As we submitted our recommendations, he ensured proper implementation of patrolling, administrative and other infrastructural solutions. As a result of targeted patrolling, highway accidents reduced by 40% in 4 months. Hence the beauty of the project lay in the commitment of both parties.

The innovation is extremely simple and scalable. All police stations can start with the digital data conversion process, which also happens to be the biggest challenge. This innovation can be rolled out at all the 726 hotspots identified by the MORTH and moreover, data and technology tools leveraged to identify patterns and trends in order to come up with customized patrolling schedules and other mitigation measures. Only then can the country really have a SMART Police force and more importantly, reduce the number of accidents and associated loss of lives.

We were excited as well as apprehensive about the concerned project as this was the first time we were working with the security forces. Our approach while working on other projects relies on acquiring a combination of both macro and micro pictures. However, a macro picture is not suitable for accident analysis as the causality lies mostly in the micro picture. As we worked with officers on day to day basis, we understood the compulsions which they have to operate under. A critical learning picked up, was the absolute specificity with which, each task has to be carried out. As resources are scarce and the mandate of security is huge, man hours cannot be wasted on unsuccessful/ imprecise task completion. For example, while analyzing the data it was incredibly important to note down the exact time of accidents as traffic movements are dependent on a host of local factors. Concentrating on the time of the accidents helped us to assess the causality of most of them.