Safety solutions for autonomous vehicle integration in urban mobility: efficient and reliable acting in an uncertain and unreliable world.

Can I cross? Is the cyclist seeing me? Will that car actually stop? All familiar trains of thought for a pedestrian navigating an urban environment: a wrong move could have fatal consequences. Fortunately, humans are masters in predicting others' behavior and acting appropriately in ambiguous and uncertain situations. But with the advent of autonomous vehicles, like self-driving cars, autonomous ships and drones, all sharing space with humans, there are new challenges for safe navigation. These challenges not only deal with how constructing these autonomous systems, but also require an understanding how humans respond to the uncertainty and risk that these vehicles bring. In Neuroscience and Psychology, overarching probabilistic theories of how humans deal with uncertainty have only recently been proposed, most prominently the "Bayesian brain hypothesis" where the brain represents knowledge of the world in terms of probabilities, and its neural computations approximate Bayesian decision theory. New data science and AI methods allow simulating models and analyzing data within this theoretical perspective. We can now envision integrated frameworks for dealing with uncertainty based on probabilistic algorithms, and address an important array of problems. In particular, this project will focus on 1) developing frameworks for estimating and controlling uncertainty, also with limited energy and compute resources; 2) delivering performance guarantees for AI systems, allowing autonomous agents to reliably operate in a complex world; and 3) developing perception, planning and decision-making methods for safe and efficient autonomous behavior that is also predictable and acceptable for other autonomous agents and humans. Achieving these goals will enable us to successfully and safely integrate AI and robotics into society while also shedding new light on our understanding of the great mystery that is our brain.