OLIVE: Online Low-Rank Incremental Learning for Efficient Adaptive Exoskeletons
Researchers have developed OLIVE, a novel framework for online learning in wearable exoskeletons. This system efficiently adapts exoskeleton control to individual users and dynamic environments by updating only a low-rank component of the control policy. OLIVE utilizes on-body sensor feedback for personalization and dynamically adjusts its update complexity based on terrain, demonstrating significant improvements in gait smoothness, effort reduction, and motion stability. AI
IMPACT Enables more personalized and responsive exoskeleton assistance for mobility impairments.