Max Robotics — Global Service Robot Directory + US Certification

About

Percy Liang has established himself as a leading researcher at the intersection of robotics, natural language processing, and machine learning, with a particular focus on enabling robots to understand and act upon human language instructions. His work spans visual representation learning, vision-language-action (VLA) models, and adaptive human-robot interaction, addressing fundamental challenges in how robots learn, generalize, and communicate with humans. Among his most influential contributions is his research on language-driven representation learning for robotics (55 citations), which demonstrates how large-scale human activity video datasets combined with techniques like masked autoencoding and contrastive learning can dramatically improve robotic policy transfer. His work on OpenVLA (39 citations) advances the open-source development of vision-language-action models, enabling robots to fine-tune pretrained behaviors rather than learning from scratch. Complementing this, his adaptive language interface research (22 citations) introduces neural semantic parsing systems that learn from user decomposition, improving human-robot collaboration efficiency. Through frameworks like LILA and his studies on real-time corrective feedback ("No, to the Right," 45 citations), Liang consistently addresses the critical challenge of making robotic systems adaptable and responsive to natural human communication — research that is steadily reshaping how intelligent robots learn alongside people.