Hi, I’m Kyle!
I am particularly interested in achieving highly reactive and performant control of systems with complex dynamics, a task that fundamentally requires mutually-aware perception and control. I’m currently studying how learning-based methods, particularly variants of reinforcement learning, can be applied to these problems.
Before attending Berkeley, I was an undergraduate student at the Georgia Institute of Technology, where I worked with Professor Evangelos Theodorou at the ACDS Lab on trajectory optimization and nonlinear model-predictive control for robotics. I’ve also spent time in the autonomous vehicle industry working at the intersection of planning and prediction.