Max Robotics — Global Service Robot Directory + US Certification

About

Corey Lynch is a pioneering robotics and machine learning researcher whose work sits at the intersection of self-supervised learning, imitation learning, and natural language grounding for robotic systems. He is perhaps best known for developing **Time-Contrastive Networks (TCN)**, a groundbreaking self-supervised framework that enables robots to learn rich visual representations directly from unlabeled multi-view video — without human annotations — garnering over 600 citations across related publications and influencing a generation of robot learning research. Lynch has made substantial contributions to making robots more flexible and generalizable. His work on **Relay Policy Learning** tackled the notoriously difficult challenge of long-horizon robotic tasks by combining imitation and reinforcement learning in a hierarchical framework. He further advanced language-conditioned robot control through projects like **BC-Z**, **Language Conditioned Imitation Learning**, and **Interactive Language**, collectively enabling robots to interpret and execute natural language instructions in open-world settings. His co-authorship on **PaLM-E** (350+ citations) — Google's embodied multimodal language model — reflects his leadership in bridging large-scale AI with physical robot deployment. Across his career, Lynch's research has consistently pushed toward robots that learn efficiently from observation, generalize broadly, and communicate naturally with humans.