I am an Assistant Professor in the Computer Science department at the University of Colorado and lead the Autonomous Robotics & Perception Group, conducting research in robotics, control theory, and perception. The goal behind my work is to develop novel ways in which robots can act even more efficiently or safely than their human counterparts, especially in the context of driving vehicles, physical feats, and understanding their environment. This in turn enables robots to understand their own limitations and develop new ways to solve problems that humans otherwise would solve relatively predictably. As a part of this, I also work on algorithms and hardware that enable robots to sense and perceive their environment in unique ways. My work applies concepts of nonlinear dynamical systems to contemporary model-predictive control of robots, wielding simulation-in-the-loop control and random tree-based planning to enable small ground vehicles to navigate uncertain environments.

I’ve also been actively working in robot perception using visual-inertial SLAM, machine learning and online system identification, all with an emphasis on experimental validation. One example of this work is in the autonomous navigation of ground vehicles through obstacle-filled rooms; humans would generally avoid driving on walls or taking jumps off objects due to their apparent challenge, but in fact the robots we build can prefer these “acrobatic” routes if their hardware supports it.

I earned my PhD at Cornell University in Theoretical and Applied Mechanics, where I conducted research on nonlinear delay-differential equations and coupled oscillators with an emphasis on perturbations, differential geometry and analysis. I was also previously a National Research Council Postdoctoral Research Fellow at the Naval Research Laboratory under the direction of Ira Schwartz and in collaboration with Ani Hsieh at Drexel University. Before joining the faculty at CU-Boulder, I was a research scientist with Gabe Sibley and chiefly maintained the tools in our GitHub.