Robots that adapt like animals
Whether an arm on a factory floor or a robot deployed in the real world, if a robot becomes damaged we'd like it to adapt to that damage to keep doing its task and/or limp back to the repair station to be fixed. However, while animals can quickly adapt to injuries, current robots cannot find a compensatory behavior when they are damaged. I will describe our algorithm, featured on the cover of Nature, which allows robots to adapt to damage in less than two minutes. Before deployment, a robot exploits a novel algorithm to create a detailed map of the space of high-performing behaviors: This map represents the robot's ‘intuitions’ about what behaviors it can perform and their value. If the robot is damaged, it uses these intuitions to guide a learning algorithm that conducts intelligent experiments to rapidly discover a compensatory behavior that works in spite of the damage.
Jeff Clune is the Loy and Edith Harris Associate Professor in Computer Science at the University of Wyoming and a Senior Research Scientist and founding member of Uber AI Labs. He focuses on robotics, reinforcement learning, and training neural networks either via deep learning or evolutionary algorithms. He has also researched open questions in evolutionary biology using computational models of evolution, including the evolutionary origins of modularity, hierarchy, and evolvability. Prior to becoming a professor, he was a Research Scientist at Cornell University, received a PhD in computer science and an MA in philosophy from Michigan State University, and received a BA in philosophy from the University of Michigan.