Kevin Tracy
Papers
6
Total Citations
128
H-Index
5
About
Kevin Tracy is a leading researcher in robotics and optimization, whose work is reshaping how robots perceive, plan, and control their motion in complex environments. His primary research areas span differentiable collision detection, trajectory optimization, and high-performance model-predictive control (MPC). Tracy’s most influential contribution is the development of DCOL, a fast and differentiable collision detection framework for convex primitives, which has garnered 46 citations and enables gradient-based optimization for tasks that were previously non-differentiable. He also introduced ALTRO-C, a conic MPC solver that achieves real-time control rates for complex robotic systems, and ReLU-QP, a GPU-accelerated quadratic programming solver that pushes the boundaries of high-dimensional control. His work on CALIPSO, a differentiable solver for trajectory optimization with conic and complementarity constraints, further demonstrates his ability to bridge theory and practice. Tracy’s innovations have been widely adopted in simulation, control, and learning pipelines, with his papers accumulating over 128 citations. His research is essential reading for anyone interested in the intersection of optimization, robotics, and differentiable programming.
Research Focus
Key Achievements
Top Papers
- 1Differentiable Collision Detection for a Set of Convex Primitives46 citations · 2023
- 2Planning With Attitude31 citations · 2021
- 3ALTRO-C: A Fast Solver for Conic Model-Predictive Control18 citations · 2021
- 4
- 5
- 6