Max Robotics — Global Service Robot Directory + US Certification

About

Harel Yedidsion is a leading researcher in multi-robot systems and human-robot interaction, whose work bridges the gap between autonomous coordination and natural, teachable collaboration. His most impactful contribution, the 2020 paper "Jointly Improving Parsing and Perception for Natural Language Commands through Human-Robot Dialog" (46 citations), pioneers methods for mobile robots to use dialog to refine language understanding, fusing natural language parsing with multi-modal perceptual models of concepts like "red" or "heavy." This work establishes a foundation for robots that learn from real-time human feedback. Yedidsion also tackles the critical challenge of multi-robot navigation in confined spaces, as seen in his 2023 paper "Learning Perceptual Hallucination for Multi-Robot Navigation in Narrow Hallways" (9 citations), where he develops strategies to overcome suboptimal behaviors when robots meet in tight corridors. His research extends to optimal verbal instruction for guiding robot teams (2019, 7 citations) and the DCOP_MST framework for coordinating mobile sensing agents (2016, 5 citations). A key advocate for Human-Interactive Robot Learning (HIRL), Yedidsion envisions robots that are not just autonomous but teachable by everyday people. His work, spanning from office deployments to RoboCup@Home competitions, profoundly shapes how robots perceive, communicate, and cooperate in human environments.