Thomas Bourgeat

Massachusetts Institute of Technology, IIT@MIT

Papers

Total Citations

104

H-Index

About

Thomas Bourgeat is a leading researcher at the intersection of robotics and computer architecture, pioneering domain-specific hardware acceleration for robotic systems. His core research focuses on designing specialized computing platforms—including FPGAs, GPUs, and custom accelerators—to overcome the severe latency and computational bottlenecks in robot motion planning and control. Bourgeat’s seminal work on "robomorphic computing" introduced a design methodology for creating accelerators parameterized by a robot’s physical morphology, directly addressing the order-of-magnitude performance gap in real-time robotic tasks. His highly cited papers, including "Accelerating Robot Dynamics Gradients on a CPU, GPU, and FPGA" (39 citations) and "RoboShape" (23 citations), demonstrate how parallel computing and topology-aware patterns can scalably deploy efficient accelerators across diverse robotic platforms. By identifying and exploiting high-level computational structures prescribed by robot kinematics and dynamics, Bourgeat’s research enables faster, more efficient algorithms for state-of-the-art planning and control. His work is instrumental in making hardware acceleration practical for the growing complexity of autonomous systems, bridging the gap between robotics and custom silicon design.