Max Robotics — Global Service Robot Directory + US Certification

About

Gilbert Feng is a robotics researcher whose work sits at the intersection of large-scale robotic learning, dexterous manipulation, and legged locomotion. His most influential contribution is as a key collaborator on the landmark **Open X-Embodiment** project (119 citations), which aggregated robotic datasets across dozens of platforms and introduced the RT-X models—a foundational step toward generalist robots that can transfer skills across different hardware. This work mirrors the paradigm shift seen in NLP and computer vision, aiming to consolidate robotic learning around shared, high-capacity models. Feng also tackles the intricate challenges of **multi-stage manipulation**, exemplified by his work on cable routing through hierarchical imitation learning (46 citations). Here, he developed methods for robots to handle deformable objects across sequential tasks, a notoriously difficult problem in robotics. Additionally, his **GenLoco** framework (17 citations) addresses the need for versatile, generalized locomotion controllers for quadrupedal robots, enabling these platforms to perform robust skills without task-specific tuning. Through these contributions, Feng is helping to push robotics from narrow, single-task systems toward more adaptable, general-purpose intelligence.