Max Robotics — Global Service Robot Directory + US Certification

About

Mohammad Norouzi is a pioneering researcher whose work bridges the frontiers of reinforcement learning and robotics. His key contributions span two transformative domains: advancing deep off-policy evaluation (OPE) for real-world decision-making, and developing motion planning strategies for reconfigurable robots operating on challenging terrains. In his highly cited 2021 paper on OPE benchmarks (25 citations), Norouzi addresses the critical challenge of leveraging offline datasets for policy evaluation—a breakthrough with profound implications for healthcare and recommender systems. His 2020 work on mastering Atari with discrete world models (23 citations) demonstrates how intelligent agents can learn from imagined outcomes, dramatically improving sample efficiency in complex environments. On the robotics side, Norouzi’s research on planning high-visibility, stable paths for reconfigurable robots (22 citations) enables autonomous exploration of uneven terrain, maximizing sensor payload height for enhanced environmental perception. His work on probabilistic stable motion planning and urban search-and-rescue robots further underscores his commitment to deploying robust, real-time solutions in safety-critical scenarios. With a total of over 100 citations across his most impactful publications, Norouzi’s dual focus on algorithmic innovation and practical robotics continues to shape how autonomous systems learn, plan, and act in the physical world.