Max Robotics — Global Service Robot Directory + US Certification

About

Cheng-Yu Kuo is a robotics researcher advancing the frontier of model-based reinforcement learning (MBRL) for safe and agile robot locomotion. His primary research areas span contact-safe MBRL, energy-exchange dynamics for bipedal robots, and sample-efficient probabilistic model predictive control. Kuo’s major contributions include developing uncertainty-aware MBRL frameworks that enable robots to learn contact-safe behaviors during training—a critical step toward deploying autonomous systems in human environments. His work on spring-loaded biped robots, such as the SLIP-inspired designs, demonstrates how learning energy-exchange dynamics can achieve real-time, sample-efficient walking acquisition with high generalizability. Notably, his 2021 paper on contact-safe MBRL has garnered 18 citations, reflecting its impact on safe robot learning. Kuo also introduced a task decomposition approach for bipedal locomotion, reducing the complexity of controlling highly dynamic systems. His research bridges the gap between theoretical reinforcement learning and practical robotic control, offering solutions that are both computationally efficient and physically robust. For students and researchers, Kuo’s work exemplifies how integrating model-based methods with uncertainty quantification can unlock new capabilities in legged robotics, making autonomous locomotion safer and more adaptable.