Max Robotics — Global Service Robot Directory + US Certification

About

Shiro Urushihara is a robotics researcher whose work bridges the gap between classical control theory and modern artificial intelligence, focusing on human-robot interaction, cooperative multi-agent systems, and reinforcement learning. His early contributions centered on wire rope tension control for robotic systems, where he developed dual disturbance observers to enable variable tension in human-robot applications—a concept that enhances safety and adaptability in physical human-robot collaboration. Urushihara also pioneered the use of complex-valued neural networks (CVNN) for cooperative conveyance systems, demonstrating how neural architectures can generate synchronized movement patterns among multiple mobile robots. More recently, he has advanced into multi-agent reinforcement learning, proposing a dueling Deep Q-Network (DQN) architecture that separately estimates state-value and action-value functions to improve coordination in shared environments. His work on robust load position servo systems for industrial robots, which eliminates vibration and end-effector offset, remains influential in precision manufacturing. With publications spanning 2009 to 2024, Urushihara’s research consistently integrates mechanical design, control systems, and learning algorithms, making him a notable figure in the evolution of intelligent, cooperative robotics.