Human State Estimation: Assessment and Exploitation for Humanoid Robots

Using humanoid robots and digital human models to mimic human movements provides a valuable method for understanding human motor control. Motion optimization is central to generating and simulating these movements. However, complex actions, especially locomotion, pose challenges in modeling, kinematics, and dynamics due to the necessary computation of physical quantities and derivatives. We will revisit various perspectives for motion optimization in humanoid systems and highlight applications, including human musculoskeletal analysis and the emulation of human movement through humanoid robots.

Summary of Important Discussions: Through in-depth analyses and discussions, we identify state estimation as a cornerstone of humanoid movement replication, shedding light on the complexities and potential solutions in this field. The accurate mimicry of human movement not only advances robotic capabilities but also deepens our understanding of human motor control, bearing significant implications for rehabilitation, sports science, and the development of assistive technologies.