John Winder

University of Maryland, College Park, University of Maryland, Baltimore County

Papers

Total Citations

H-Index

About

John Winder’s research sits at the intersection of robotics, artificial intelligence, and grounded language learning, with a focus on enabling robots to act intelligently in complex, human-scale environments. His most influential work, “Planning with Abstract Markov Decision Processes” (2017, 49 citations), tackles the fundamental challenge of planning under uncertainty in large state–action spaces. Winder’s key contribution here is the introduction of hierarchical abstraction, allowing robots to efficiently compute plans even when reward functions shift dynamically as goals change—a critical step toward deploying autonomous systems in real-world settings. This work has shaped how researchers approach scalable decision-making under uncertainty. More recently, Winder has ventured into grounded language acquisition, co-authoring “A Spoken Language Dataset of Descriptions for Speech-Based Grounded Language Learning” (2021, 6 citations). This multimodal dataset, combining RGB and depth data with spoken descriptions, addresses the pressing need for sample-efficient learning in robotics, bridging natural language processing, computer vision, and signal processing. By providing a resource for speech-based grounding, Winder is helping pave the way for robots that can understand and follow spoken instructions in human environments. His work is steadily building a foundation for more capable, communicative robots.