James Hermus

Massachusetts Institute of Technology, IIT@MIT

Papers

Total Citations

H-Index

About

James Hermus is a leading researcher in the fields of human motor control, physical human-robot interaction, and robotic manipulation. His work bridges robotics and neuroscience, focusing on how humans exploit redundancy and dynamic primitives to achieve dexterous physical interaction—insights that are critical for designing safer, more intuitive collaborative robots. Hermus’s most cited paper, “Exploiting Redundancy to Facilitate Physical Interaction” (2021, 28 citations), challenges conventional nullspace projection methods by revealing how humans use kinematic redundancy to simplify control. His subsequent studies on dynamic primitives (2022, 8 citations; 2023, 5 citations) demonstrate that humans rely on stereotypical force patterns during constrained motion, a finding that explains both our superior interaction skills and inherent limitations. Most recently, his work on task-relevant stiffness tuning (2025, 2 citations) shows how humans rapidly modulate arm impedance to anticipate collisions—a capability robots still struggle to replicate. With a growing citation record and a focus on foundational principles of human movement, Hermus is shaping the next generation of physically interactive robots that can work safely alongside people.