Road-Safety and Smart City
February 1, 2018 In SMART CITYThe issue of road safety in India has been an often-neglected concern, when it comes to the rapid development of urban areas and the implementation of traffic policies, as well as awareness programs. Motorization has been growing at an exponential rate in the country and has led to the rapid expansion of road networks in recent times. This, in turn, has impacted road safety levels quite adversely. According to studies conducted by the Ministry of Road Transport and Highway (MoRTH), the total number of road crashes in the country has grown by 2.5% from 4,89,400 to 5,01,423 in last year and the total number of fatalities has gone up by 4.6%. with 1,50,785 deathsin road accidents. With the 100 smart cities initiative being underway, digital technologies are now set to enable services, that earlier were just pipe dreams. Machine to Machine (M2M) communications, IoT and ICT are looking to make city life safer, healthier and more comfortable, hence, here’s taking a look at some of the ways that technology can help tackle the issue of road safety in India: How to make Indian roads safer? Data Analytics has a crucial role to play when it comes to attaining the goal of eliminating fatalities/severe injuries, due to road accidents. On a global perspective, the Swedish project Vision Zero hasfor almost 2 decades, been working on solutions that can make traffic casualties a thing of the past. 25 U.S cities are a part of the initiative and to create a road-safety model that can fit into the Indian Smart-City vision, here’s taking a look at the applicable findings:
- Any road-safety initiative needs to set a timeframe with clear and measurable strategies, acknowledging the fact that traffic-related deaths are preventable. The plans need to look at a multidisciplinary approach, that can bring together a varied set of stakeholders for the redressal of the core issue.
- Target interventions can be made on the basis of leveraging data analytics. Analyzing historical crash data can grant us a better understanding regarding the areas, where the frequency of occurrence of collisions is higher, the conditions co-related to the collisions and the road users that are most vulnerable to such occurrences.
- Using machine learning feeds from traffic cameras can be analyzed for occurrences of near collision events. As opposed to collision events, the cost of learning in which is high (with respect to damage as well as human casualties), video analytics offer a zero-cost learning opportunity, which can help in substantial understanding via leveraging of these incidents.
- o Video analytics systems generally employ a tracker technology, that can detect and follow the actual trajectory of moving objects, which can be eventually classified into categorical subsets like pedestrians, bicycles, cars etc. via deep neural networks.
- o These systems can over time, produce count reports that are able to classify vehicles, on the basis of their turning movement, direction of approach, speed, acceleration, mode of transportation etc.
- o Basing their data on a collection of near-collision conditions, the system can produce risk scores, which can enable a city to flag high-risk locations and create intervention methods for the mitigation of dangerous conditions.
- Citizen participation and crowdsourcing data is a successful method which has seen great results in cities in the U.S. Breaking down the barrier of inaccurate reportage/inequitable representation/general hesitancy for reportage; creation of apps that can help citizens report collision/near-collision incidents, can further help in the collection of collision data which would be extremely helpful for analysis and model creation.