Max Robotics — Global Service Robot Directory + US Certification

About

Gregory Kahn is a robotics and machine learning researcher whose work spans autonomous aerial vehicles, safe reinforcement learning, and data-driven robot control. His research sits at the intersection of deep learning, model-based reinforcement learning, and real-world robotic deployment — a combination that has produced consistently influential contributions to the field. Kahn's most-cited work, "Learning Deep Control Policies for Autonomous Aerial Vehicles with MPC-Guided Policy Search" (2016, 422 citations), demonstrated how model predictive control could be harnessed to train efficient neural network policies for quadcopters. His 2017 paper on uncertainty-aware reinforcement learning (227 citations) addressed one of the field's most pressing practical challenges: keeping robots safe during the learning process itself. His early contributions to 3D mapping and Gaussian belief space planning further reveal a researcher deeply invested in probabilistic, computationally grounded approaches to robot decision-making. More recently, Kahn contributed to the landmark Open X-Embodiment project (2023–2024, 220+ combined citations), a large-scale collaborative effort to build generalist robotic learning models trained across diverse real-world datasets. His work on sim-to-real transfer and meta-reinforcement learning for payload transport underscores a sustained commitment to building robots that genuinely generalize. Across more than a decade of research, his contributions have shaped how autonomous robots learn, adapt, and operate safely in unstructured environments.