Hierarchical Approaches to Robot Control
Wednesday, February 26, 2014
12:00 pm - 1:00 pm
Fitzpatrick Center Schiciano Auditorium Side B
George Konidaris - MIT
Robots are increasingly becoming a part of our daily lives, from the automated vacuum cleaners in our homes to the rovers exploring Mars. However, while recent years have seen dramatic progress in the development of affordable, general-purpose robot hardware, the capabilities of that hardware far exceed our ability to write software to adequately control. The key challenge here is one of abstraction. Generally capable behavior requires high-level reasoning and planning, but perception and actuation must ultimately be performed using noisy, high-bandwidth, low-level sensors and effectors. My talk will address the use of learned motor controllers - or skills - as a means of constructing higher-level reusable procedural knowledge suitable for use in abstract planning. The first part of my talk will present work aimed at identifying subproblems that recur, and constructing and retaining new motor skills for solving them. I will demonstrate a robot system that learns to complete a task, and then extracts components of its solution as reusable skills, which it deploys to quickly solve a second task. The second part will focus on practical methods for acquiring skill control policies, through the use human demonstration and active learning. Finally, I will present some new theoretical insights which establish a link between the skills available to a robot and the abstract representations it should use for planning.