Max Robotics — Global Service Robot Directory + US Certification

About

Tao Du is a computational researcher working at the intersection of differentiable simulation, soft robotics, and physics-based machine learning. His work focuses on developing differentiable physical simulators that enable gradient-based optimization for designing and controlling complex robotic systems, including soft robots, aerial vehicles, and deformable materials. Du's most impactful contributions include his development of differentiable simulation frameworks for underwater soft robots (88 citations), which elegantly couple hydrodynamic models with gradient-based control, and DiffCloth (80 citations), a differentiable cloth simulator incorporating dry frictional contact that has advanced applications in computer animation and robotic manipulation. His PlasticineLab benchmark (24 citations) has provided the research community with a standardized environment for evaluating skill learning in deformable-body physics — a significant gap previously overlooked by rigid-body-focused benchmarks. Beyond simulation, Du has pioneered co-design methodologies for both soft robots and UAVs, demonstrating how learned latent representations and graph grammars can automate the joint optimization of robot morphology and control policy. Through tools like ChainQueen, he has made differentiable soft-body simulation more accessible and extensible. With over 250 cumulative citations, Du's research is shaping how next-generation robots are designed, modeled, and controlled.