Max Robotics — Global Service Robot Directory + US Certification

About

Drew McDermott is a pioneering researcher in artificial intelligence, with a career-long focus on autonomous robotics and automated planning. His work fundamentally addresses how robots can operate reliably in uncertain, dynamic environments. A central theme is the challenge of robot map learning, where he argued that error accumulation is inevitable and that systems must explicitly detect and correct errors rather than solely trying to prevent them—a key insight in his highly cited 2003 paper (136 citations). McDermott also made foundational contributions to robot planning, exploring how plans can be improved during execution and how robots can anticipate and forestall failures. His concept of "structured reactive plans" provides a framework for robots to respond to unexpected events while pursuing long-term goals. Beyond planning, he developed efficient image-based techniques for place recognition, enabling robots to recognize locations for map construction. With over 350 total citations, McDermott’s work bridges theoretical planning research with practical robotics, emphasizing adaptive, robust systems that learn from experience. His legacy lies in shifting the field from static maps and rigid plans toward dynamic, error-tolerant approaches essential for real-world autonomy.