About
Prof. Madhur Behl conducts research at the confluence of artificial intelligence and predictive control for autonomous systems.
He is working on improving the safety and agility of fully autonomous vehicles operating under uncertain traffic situations.
He is the co-founder of the F1/10 International Autonomous Racing Competitions — which is like the FIRST Lego league, but for self-driving cars.
PAVE: How does your work align with our mission to educate the public on AV technology and its promise?
We have created a 1/10 scale fully autonomous open-source platform called F1/10 (f1tenth) — that anyone can assemble and program, to learn about the principles of perception, planning, and control for autonomous driving. This has made autonomy easily accessible to students, researchers, labs, and enthusiasts who otherwise would not get to work on a full-scale prototype. By routinely hosting demonstrations, and workshops for students and the general public we are educating everyone about autonomy, with a focus on how self-driving cars operate and what are the safe underpinnings of the technology.
PAVE: What are current barriers you see in public acceptance of automated vehicles?
The question of safety still remains a large challenge. It may not be sufficient that the AV has driven millions of miles in testing, if it fails in a situation that most humans would never fail. So it is not only about the number of failures but also the type of ‘common sense’ failures that shape public opinion and acceptance. Therefore, transparency in communicating the safety and security elements of the vehicles will be very important. Interactions between automated vehicles and the vulnerable road users also remains a challenge. With inward facing cameras, that can capture the identity of the passengers, I also anticipate data-privacy could become a barrier to adoption.
