Max Robotics — Global Service Robot Directory + US Certification

About

Hadi Salman is a versatile researcher whose work spans computer vision, robotics, and machine learning, with particular expertise in adversarial robustness, autonomous exploration, and robot learning. His most recognized contribution, "Unadversarial Examples: Designing Objects for Robust Vision" (2020, 25 citations), introduced a novel framework that strategically leverages object design to dramatically enhance the performance and robustness of modern vision models — a creative inversion of traditional adversarial machine learning thinking. In robotics, Salman has made meaningful strides in autonomous coverage and navigation, developing multi-agent ergodic coverage strategies with obstacle avoidance (20 citations) and stochastic trajectory optimization techniques suited for search and surveillance applications. His work on hierarchical reinforcement learning for behavior sequencing (19 citations) and LiDAR-based place recognition (19 citations) further demonstrates his ability to bridge perception and control in complex environments. Earlier in his career, Salman explored mechanical robot design, including magnetic adhesion climbing robots and single-actuator differential-drive systems. Across disciplines ranging from deep learning to physical robot engineering, his body of work reflects a rare breadth of technical curiosity and an enduring commitment to making intelligent systems more capable and reliable.