Max Robotics — Global Service Robot Directory + US Certification

About

Tianhe Yu is a leading researcher at the intersection of robotics, machine learning, and embodied AI, whose work has fundamentally advanced how robots learn from data and generalize to new tasks. His most influential contributions center on developing scalable, real-world robotic control systems and meta-learning frameworks for rapid skill acquisition. Yu is best known as a key contributor to the Robotics Transformer (RT-1 and RT-2) series, which pioneered the integration of large-scale, vision-language-action models for end-to-end robotic control. RT-2, with over 267 citations, demonstrated how web-scale knowledge can be transferred directly to robotic manipulation, enabling emergent semantic reasoning and unprecedented generalization. His earlier work on Meta-World (282 citations) established a critical benchmark for multi-task and meta-reinforcement learning, while his foundational papers on one-shot visual imitation learning (270 citations) and domain-adaptive meta-learning (113 citations) showed how robots can acquire new skills from a single human demonstration. Yu’s research has accumulated over 2,000 citations, and his gradient surgery method for multi-task learning (108 citations) remains a widely adopted technique. Through his work at Google Robotics and as a leading voice in embodied AI, Yu has helped bridge the gap between internet-scale pretraining and real-world robot deployment.