Max Robotics — Global Service Robot Directory + US Certification

About

Erez Karpas is a prominent AI researcher whose work sits at the dynamic intersection of automated planning, robotics, and human-robot collaboration. His research addresses one of the central challenges in modern robotics: enabling intelligent agents to autonomously combine low-level capabilities—navigation, motion planning, and continuous control—into coherent, high-level goal-directed behavior. Karpas has made significant contributions to mixed discrete-continuous planning, developing frameworks such as ScottyActivity that integrate convex optimization with symbolic planning to allow robots to reason fluidly across both domains (cited 36 and 16 times respectively). His most cited work, "Automated Planning for Robotics" (81 citations), serves as a foundational reference bridging classical AI planning with real-world robotic systems. He has also advanced the theory of goal recognition design—analyzing and redesigning environments to ensure agents' goals become identifiable more rapidly—contributing both deterministic and stochastic formulations alongside a comprehensive survey of the field. Beyond individual robot behavior, Karpas has explored robust execution for human-robot teams and replanning under uncertainty, reflecting a commitment to practical deployment in dynamic environments. His body of work, spanning task-and-motion planning to environment redesign, has meaningfully shaped how researchers think about deployable, adaptive robotic intelligence.