Max Robotics — Global Service Robot Directory + US Certification

About

Jiayuan Gu is a researcher working at the intersection of robotics, computer vision, and 3D scene understanding, with a particular focus on bridging the gap between simulation and real-world robotic systems. His most influential contribution, "Close the Optical Sensing Domain Gap by Physics-Grounded Active Stereo Sensor Simulation," has garnered over 20 citations and introduces a fully physics-grounded pipeline for simulating active stereovision depth sensors — a critical advance for making synthetic training environments more faithful to real-world conditions. This work reflects a broader commitment to sim-to-real transfer, a challenge he also tackles in "Evaluating Real-World Robot Manipulation Policies in Simulation," which addresses scalability and reproducibility in robotic policy evaluation. His research extends into task generalization for robot learning, exemplified by "RT-Trajectory," which leverages hindsight trajectory sketches to help robots generalize across novel tasks. More recently, Gu has pushed into high-quality 3D scene reconstruction with CAST and affordance-driven diffusion policies with AffordDP, demonstrating a growing interest in enabling robots to better perceive, reconstruct, and interact with their environments. Collectively, his work represents a cohesive effort to make robotic systems more capable, generalizable, and grounded in physical reality.