The next generation of driverless cars will need to think about what’s on the road, not just what they see

Autonomous vehicles have made remarkable progress over the last decade. Self-driving cars and buses that once struggled to stay in lane can now navigate busy city streets, detect pedestrians and cyclists, and respond smoothly to traffic signals.

But one challenge remains stubbornly difficult. The most difficult situations on the road are not the common ones, but the rare and unpredictable events – what AI researchers call “long-tail scenarios” or “edge cases” because they appear as outliers on any event distribution curve.

Examples of this include unexpected roadworks, unusual behavior of other road users and other subtle situations where the likelihood of something happening is very low – but which would have a significant impact on the vehicle and the journey.

Addressing these problems requires more than just better sensors – it requires vehicles that can anticipate uncertainty. The most promising class of AIs yet developed for this purpose is known as…