Max Robotics — Global Service Robot Directory + US Certification

About

Neha Das is a robotics and machine learning researcher whose work sits at the intersection of robot learning, visual perception, and safe control. Her research focuses on enabling robots to learn complex manipulation skills from visual demonstrations, with a particular emphasis on inverse reinforcement learning and body schema representations that allow robots to generalize to novel environments and objects. Her most influential contribution, "Model-Based Inverse Reinforcement Learning from Visual Demonstrations" (2020, 8 citations), tackles the challenging problem of scaling model-based IRL to real robotic manipulation tasks with unknown dynamics — a significant step toward practical, data-efficient robot learning. Complementing this, her work on keypoint predictive models and extended body schemas draws inspiration from how humans develop internal representations of their bodies and tools, providing robots with analogous generalization capabilities across unfamiliar manipulation scenarios. More recently, Das has expanded into safety-critical control, developing Control Barrier Function-based frameworks for elastic joint robots to ensure safe human-robot interaction even when system dynamics are imperfectly known. Collectively, her research addresses some of the most pressing challenges in deployable robot learning — perception, generalization, and safety — making her contributions particularly relevant to the next generation of collaborative robotic systems.