This panel session will share an inside look at some of the most surprising and outstanding challenges in self driving, and Cruise’s ML-first approach to solving them. Leaders from Cruise AI will discuss where ML is replacing traditional methods for tracking, planning, perception and prediction and how self driving is pushing new frontiers for AI. As part of this, we will cover uncertainty modeling and the critical role this plays in enabling autonomous vehicles to make more robust and safer decisions in complex driving environments like San Francisco. We will also walk through the homegrown ML tools and the ML infrastructure we’ve built to foster fast experimentation and scale.