Embodied Intelligence

3-0-9 units · Graduate

Explores the problem of designing a general-purpose intelligent robot, addressing challenges in robot design, such as long-horizon decision making, continuous action spaces, partial observability, and nonstationary and adversarial environments. Students read foundational papers on each challenge as well as state-of-the-art research results. Instruction emphasizes cutting across disciplinary boundaries. Topics include control theory, machine learning, planning, decision making, and hardware design. Students present papers, engage in discussion, and complete reading assignments, concluding with an open-ended research project. One of 6.4100, 6.4120[J] , 6.8110[J] or 6.7920[J] is recommended but not required.