Robots Making Contact with an Uncertain World

What happens when robots leave the lab and enter the real world? Suddenly, uncertainty is everywhere — slippery mud, bending branches, unpredictable terrain. These challenges are especially tough when it comes to contact: one moment a robot applies massive force, the next it has no grip at all.

In his talk, Aaron M. Johnson (Carnegie Mellon University / Visiting Professor at TUM) shows how robots can learn to master the unknown — from off-road driving in new environments to agile walking through vegetation. Expect cutting-edge insights into how uncertainty can be modeled, reduced, and even turned into an advantage for the future of robotics.