Max Robotics — Global Service Robot Directory + US Certification

About

Yuheng Kuang is a leading researcher at the intersection of robotics, large language models, and embodied AI. His work fundamentally redefines how robots understand and act upon human language, bridging the gap between high-level instructions and real-world physical actions. Kuang is best known for his pivotal contributions to the "RT" (Robotics Transformer) series, including RT-1 (512 citations) and RT-2 (267 citations), which pioneered the integration of vision-language models into end-to-end robotic control. These models enable robots to leverage web-scale knowledge for emergent semantic reasoning and zero-shot generalization. His seminal paper "Do As I Can, Not As I Say" (516 citations) introduced the concept of grounding language in robotic affordances, addressing a critical weakness of large language models in physical domains. Kuang also led groundbreaking work in high-speed robotic learning, notably achieving amateur human-level performance in competitive table tennis—a landmark in dexterous, real-time control. His research on quadruped agility (Barkour benchmark) and the recent Gemini Robotics series further showcases his commitment to bringing AI into the physical world. With over 1,300 citations across his most influential works, Kuang’s research is shaping the future of capable, generalist robots that can learn, adapt, and perform complex tasks in dynamic environments.