Adoption of AI and predictive analytics to make India's roads safer

An artificial intelligence-based driver safety app could go a long way in fixing the problem

On average, 400 people die in road accidents in India every day. Between 2018 and 2020 a staggering 434,244 lives were snuffed out in road accidents, or almost 150,000 people every year. The statistics are grim — but they may be about to change, thanks to the use of artificial intelligence (AI) to make India’s roads safer.

A World Bank report pointed out last year that India has just 1 per cent of the world’s vehicles, but accounts for 11 per cent of global road-accident deaths — the highest in the world. In the last decade, 1.3 million people died and another 5 million were injured on Indian roads. 

Along with lives lost, the large number of road-accident deaths has a significant socio-economic impact, too. The World Bank report estimates that road crashes cost Rs 5.96 trillion, or 3.14 per cent of India’s gross domestic product (GDP).

To combat this problem, Nitin Gadkari, minister for road transport and highways, has set in motion a plan to reduce road accidents by 50 per cent by 2024, through the adoption of AI and predictive analytics to make the country’s roads safer.

Akhilesh Srivastava, project head, Road Safety 2.0, a project initiated by the World Economic Forum (WEF), believes that India needs a different approach, since all the conventional practices for ensuring road safety adopted in the last five decades have failed to produce results.

In December 2021, the WEF along with partners like the HumSafer Driver Safety Foundation, 3M and Keltron (a Kerala state government enterprise in collaboration with Niti Aayog), and the Union road transport ministry initiated a pilot under the Road Safety 2.0 programme. Dropping the punishment-based system, this new approach seeks to incentivise truck drivers to drive safely.

Human error

Road Safety 2.0 aims to introduce technology into the “4Es” of road safety — education, engineering, enforcement and emergency care, each of which has been ignored for decades.

The group that runs the programme found that 80 per cent of accidents happen owing to human errors such as speeding, or when the driver has been behind the wheel for more than 8 hours without a rest, or is drunk, says Srivastava.  

“We started a pilot project where we use artificial intelligence to assist people to overcome their human limitations. We started by using a mobile app from HumSafer. The driver has to download the app and place the handset on a stand at eye-level and charge it. Using AI, the app monitors the driver’s eyes and sends an alert the moment he becomes drowsy, so that he can take a break,” explains Srivastava.

Game-changer app 

The pilot was done on 6000-7000 drivers and not a single accident was reported by users of this app. The app not only tracks the alertness of drivers, but their behaviour, too. “This is a game-changer because, along with tracking, it also scores the driver’s driving habits. Is he manoeuvring too much, what is the speed limit, is the driver taking breaks? These points then get incentivised,” he adds.

Starting with a few thousand truck drivers, the project has gone on to include over 10,000 of them. HumSafer Foundation now plans to take this number to 100,000 drivers by the end of the year, and in the next four years, reach at least 10 per cent of the total number of truck drivers in the country, which is currently around 8 million.

Jehaan Kotwal, co-founder, HumSafer Foundation, says that the model works because it is incentive-based. “Our systems work on a punishment mechanism, which means that if you break a law you get punished. But what if I am a good driver, what do I get? Also, making drivers download an app has to have a motivational factor, otherwise why will they download an app?” he asks. 

Once the app is downloaded and starts tracking a driver, he gets Rs 1.50 for every 10km of safe and good driving. These rewards can be encashed at the end of the month, and the amount can go as high as Rs 1500 a month.

Going ahead, the company is planning to include features such as family tracking, sharing of correct and important information on government schemes, advances in sleep detection features, and finally, crash detection.

“We are working on a feature that allows us to detect an accident that has happened, by auto-detecting the G-force. Then we can immediately get medical assistance. We are working with IIM-Ahmedabad and IIT-Delhi in trying to humanise our approach to using this app,” Kotwal adds.

Keltron is a partner in setting up cameras and systems for automatic detection of violations and instant challan generation. However, at present there is some disconnect, as the National Highways Authority of India detects the violations and generates a challan, but enforcement has to be done by the state police machinery. With Keltron, this gap can be bridged.

The other startup that WEF roped in is Highway Delit, which works with emergency caregivers in case of accidents. Their wagons can communicate with the nearest hospital.

Next steps 

Srivastava is also looking at how AI can be used to enforce good practices, such as issuing of an e-challan the moment a camera detects a driver who is not wearing a seat belt. The other area is in emergency care.

India has the highest fatality ratio, mainly because no help is forthcoming during the golden hour (the hour immediately following an accident). So, a pilot is being run to map accidents and the nearest hospitals, to save lives.

The next big focus is to create an ecosystem that becomes a sustainable model, with participation from insurance companies, tyre companies, and oil firms, among others. “Getting insurance players involved is important. With the HumSafer app, insurance companies can also benefit, as they get access to individual data on a driver’s driving habits, and can accordingly fix premiums,” says Srivastava. “Worldwide, insurance firms have been intrinsic to road safety programmes, but in India somehow they have kept away.” 

The ministry of road transport and highways issued a notification on February 25, 2022, which mandates a detailed procedure for investigating road accidents for a quick settlement of claims by the Motor Accident Claim Tribunal. It also mandated validated mobile numbers in the Certificate of Insurance.

Srivastava believes that the technology platform that HumSafer provides can be used for accident settlements, since it monitors the entire journey. Technology, he stresses, can certainly help achieve the target of halving the number of road accidents in India by 2025.

(With reporting by Dhruvaksh Saha)