How AI Helps Robots Understand the World

I am a Professor of Robotics and a Professor of Electrical Engineering and Computer Science at the University of Michigan. I am the leader of the Laboratory for Progress (Perceptive RObotics and Grounded REasoning SystemS).

My work and collaborations aim to discover methods for computational reasoning and perception that will enable robots to effectively learn from and assist people in common human environments. Essentially, we explore how to make the real world programmable by regular people through the control of autonomous robots. Critical challenges towards this goal are enabling robots to perceive our world, understand the form and function of objects, reason under uncertainty, and learn from human users. This research pertains primarily to robot learning from demonstration, semantic perception, and mobile manipulation towards enabling the usability of this technology by people in real situations.

Because science is exactly independently verifiable knowledge, open-source contributions and reproducibility are critical features of my work. My open-source approach to scholarship emphasizes a balance of theory and practice, a balance of research and teaching, and prioritization on equity and excellence.