ExoLaN improves torque and movement estimates for exoskeleton users
ExoLaN: Physics-Consistent Context-Aware Dynamics Learning for Exoskeletons
Robotics
Summary
ExoLaN is a method to better estimate how much force people apply when moving with an exoskeleton, even when they do new tasks or change how they move. The authors created a system that uses timing information and sensors in the user's shoes to understand when and how they are touching the ground. This helps predict their movement and joint forces more accurately without needing to relearn for each user. It also figures out what type of activity the user is doing without being told upfront, which could help future exoskeletons assist people more smoothly.
What this means in practice
- •For exoskeleton developers: Improve exoskeleton control algorithms to adapt assistance across different users and activities without retraining$Commercial implications: Enables development of adaptable assistive robots that fit diverse users and tasks, making commercial exoskeleton products more versatile
- •For robotics engineers: Design prediction models for human-robot interaction that account for contact events and changing contexts to enhance motion prediction accuracy
Authors
Lucas Schulze, Maximilian Schwarz, Jona Hoppe, Jan Peters, Oleg Arenz
Abstract
Task-agnostic assistive exoskeleton control based on human intention offers greater flexibility than conventional approaches that rely on predefined tasks or motion patterns. Human joint torque estimation enables task-agnostic assistance by characterizing user actions. Physics-consistent methods such as Deep Lagrangian Networks (DeLaN) have been applied to estimate the human torques in multi-user settings, but existing approaches cannot adapt to a specific user without retraining, and do not account for intermittent contacts during locomotion. We propose ExoLaN, a Context-Aware DeLaN for human-exoskeleton interaction that learns the full coupled system dynamics while adapting to changes in interaction context. ExoLaN combines temporal context with partial contact-force measurements from force-sensitive insoles to infer latent dynamics embeddings and estimate generalized contact torques. On seven unseen users performing 21 unseen tasks, ExoLaN reduces torque estimation MSE by 7% compared to a black-box baseline. Beyond inverse dynamics, ExoLaN serves as a unified model that also enables accurate forward prediction: training with a multi-step prediction loss reduces acceleration MSE by 59% and long-horizon position and velocity errors by 60% and 93%, respectively, compared with a single-step loss. Moreover, the learned latent context captures task information without explicit task labels, making it a promising signal for task-aware assistive control.