Event
Simulation-based joint training of human-exoskeleton co-adaptation strategies
1 reports · 1 sources
Yifei Yuan and co-authors propose a staged multi-agent training (SMAT) method to jointly train a musculoskeletal human agent and bilateral hip exoskeleton control policies in physical simulation, enabling the assistance policy to learn human-machine co-adaptation. The policy was transferred to a physical hip exoskeleton without subject-specific retraining and tested in eight healthy adults. Indirect calorimetry compared three conditions—no exoskeleton, a passive device, and active assistance—and active assistance reduced net metabolic rate by 19.7% relative to the passive device (p < 0.001), with a hip mechanical power positive power ratio of 0.984. The authors state that effective assistance was maintained at walking speeds and terrains beyond the training conditions.
Report timeline
- SMAT: Jointly training human-exoskeleton co-adaptation policies in simulation and validating on real subjects
The study proposes Staged Multi-Agent Training (SMAT), a method that jointly trains a musculoskeletal human agent and a bilateral hip exoskeleton control policy in physics simulation to learn an assistance policy that accounts for human-device co-adaptation. The policy was transferred to a physical hip exoskeleton without subject-specific retraining and tested in eight healthy adults; indirect calorimetry showed that active assistance reduced net metabolic rate by 19.7% relative to the passive device (p < 0.001), the hip mechanical positive-power ratio was 0.984, and effective assistance was maintained at walking speeds and terrains beyond the training conditions.
arXiv cs.RO ·