Max Robotics — Global Service Robot Directory + US Certification

About

Hany Abdulsamad is a robotics and machine learning researcher whose work sits at the intersection of reinforcement learning, optimal control, and probabilistic modeling, with a particular focus on enabling robots to learn and generalize complex motor skills. His early and most influential contribution, "Reinforcement Learning vs Human Programming in Tetherball Robot Games" (2015, 14 citations), established a rigorous framework for comparing learned versus hand-engineered robotic behaviors, highlighting the practical promise of autonomous skill acquisition. Building on this foundation, Abdulsamad has advanced contextual reinforcement learning through self-paced curriculum strategies that improve generalization across novel task scenarios, and has reframed stochastic optimal control as an inference problem, offering principled probabilistic alternatives to heuristic-driven optimization. His more recent work explores variational hierarchical mixture models for scalable inverse dynamics learning and applies active inference frameworks to address the persistent challenge of partial observability in robotic manipulation. Across his portfolio, Abdulsamad consistently bridges theoretical rigor with practical robotics applications, contributing tools that help autonomous systems learn, adapt, and reason under uncertainty — making his research particularly relevant for students working at the frontier of intelligent robot learning.