This scientist is making roads safer with mobile telematics technology

Hari Balakrishnan of MIT has been working on inferring bad driving behaviour using sensor data from phones and other IoT devices in vehicles to help reduce thousands of road crashes each year

In India, over 150,000 people lose their lives in road accidents every year. Factors such as speeding, phone usage while driving, excessive hard braking or harsh acceleration are some of the factors which lead to road accidents.

Hari Balakrishnan, Fujitsu Professor of Computer Science at Massachusetts Institute of Technology, has been working towards inferring these behaviours using sensor data from phones and other internet of things (IoT) devices in vehicles. This is based on mobile telematics technology which could help reduce the thousands of road crashes that occur annually.

Mobile telematics refers to acquiring and processing data from sensors such as position sensors, accelerometers, and other such technologies on mobile devices and assets to draw inferences about mobility patterns. “For example, by intelligently processing sensor data from participating users’ smartphones, we can measure how well they are driving,” explains Balakrishnan. It is done with full user permission with feedback provided on driving quality. It is also fully automatic, that is, the data is acquired in the background without draining the user’s battery significantly. “We can tie it with incentives to encourage them to drive better,” says Balakrishnan, who recently won the Infosys Prize 2020 in Engineering and Computer Science for his broad contributions to computer networking and commercial use of mobile telematics. 

Between 2005-2010,  the CarTel project on mobile sensing that he led at MIT introduced the idea of using sensors attached to mobile assets such as vehicles and user’s phones to measure the environment in a scalable way, helping to create the field of mobile sensing. 

Born in Nagpur, Balakrishnan belonged to a family of physicists with his father working at the Reactor Research Centre in Kalpakkam. His family moved to Chennai, when he was nine years old, where his father took up a position as a Professor of Physics at IIT Madras and his mother, who is also a theoretical physicist, was then able to re-enter the workforce and became a Professor of Physics at the Institute of Mathematical Sciences in Chennai.

When in school, his dream was to play cricket professionally. “But I’m sure I wasn’t good enough! A research career was in some ways the primary path I was exposed to my parents are academics,” remembers Balakrishnan, an IIT-Madras alumnus.

It is perhaps a little more surprising that he was also able to establish a business, although that is also heavily based on research and technology. 

The products of his company, Cambridge Mobile Telematics (CMT), are currently being used in over 25 countries. “We work worldwide with insurers, rideshare companies, cellular carriers, government agencies, and automobile makers. CMT is today the world’s largest mobile telematics provider, helping to reduce road accidents,” says Balakrishnan. He says there is scope for individual drivers, rideshares as well as fleets to utilise this technology for a safer ride in India. “We are working with various partners to bring the technology to countries around the world, including India,” he says.

He feels there is a long way to go in mobile telematics and his team is geared up for that journey to change road safety and insurance. First, it will make insurance pricing more equitable by setting prices based on how someone drives, that is focusing on causative factors such as phone distraction, speeding and not demographic factors such as age. Moreover, by providing feedback about driving, people can become better drivers, making roads safer by reducing crashes. Second, using telematics for real-time crash detection for roadside assistance and then using AI-based crash reconstruction, will help people in need much quicker and more accurately. Third, telematics will expand to include video and other emerging sensors, providing even better data for these applications. “Last, if we move toward more autonomous vehicles (AVs), all insurance will become telematics-based. That is the world we are likely to be in the next decade. At that point, ensuring an AV, which will have to share roads with traditional vehicles and pedestrians, will be based on evaluating the quality of AV sensors, the AI algorithms used in the AVs, and the quality of software,” says Balakrishnan.