Research

Our lab develops AI methods to advance autonomy in robots. This means allowing physical machines to solve efficiently and reliably complex, long-horizon, cognitive tasks. We are especially interested in long-lived robots, i.e., robots deployed for long, unattended periods of time, in an open world—exploring the challenges and opportunities these entail. We focus on the problems of planning and decision making and their interface with machine learning, in an attempt to bridge the gap between “model-based” and “model-free” approaches. Our contributions often involve novel theoretical frameworks, which bridge across disciplines while exploiting themes of abstraction, simplification, incremental processes, and hierarchical representations.