We’re making self-driving vehicles safer

With the ultimate driving test for autonomous vehicles

D-RISK is preparing autonomous vehicles (AVs) for the complex reality of our roads.

Our project is focused on edge cases: the one-in-a-million high-risk situations for which AVs have to be prepared.

A child dressed up as a traffic light for Halloween would be easy for a human to understand and avoid. But most modern autonomous systems will see it merely as a traffic light – hopefully not lit green.

This, and a nearly uncountable number of other high-risk scenarios, are what we mean by edge cases.

Using real-life data from traffic cameras, accident reports, and stories submitted by the public, we’re gathering the world’s largest library of edge case scenarios. Edge cases like the one shown in the video below.

These edge cases are being used to create simulations to train AVs for every scenario you can imagine – and even those you can’t. After retraining with our edge cases, AVs can recognise high-risk situations and anticipate hazards before they occur. And when regulators are ready for a driving test for self-driving cars, D-RISK will have one waiting for them.