Multi-Objective Human-in-the-Loop Bayesian Optimization of a Lower-Limb Exoskeleton

2026-09-24

AI Summary

Existing exoskeleton optimization methods often neglect dynamic user preferences and struggle with multi-objective trade-offs. This work proposes MO-HILBO, a framework that achieves the first multi-objective human-in-the-loop optimization incorporating dynamic user preferences. By leveraging Bayesian optimization to infer personalized Pareto-optimal controllers, the method effectively balances metabolic cost against user comfort. Experimental validation demonstrates that this approach successfully identifies Pareto-optimal solutions aligned with user expectations, significantly enhancing the efficiency of personalized adaptation. Furthermore, the open-source mohilo toolkit is released alongside this study, providing a standardized platform for future research in exoskeleton control.

Abstract

Human-in-the-loop optimization (HILO) is a common approach for optimizing the control of assistive devices to account for the wearer's unique biomechanics and subjective preferences. However, despite research suggesting that a person may have a different prioritization of objectives depending on time-varying factors such as the environment, their mood, or energy levels, existing HILO approaches only consider a single objective or enforce a fixed weighting on a set of objectives. Neither approach is capable of representing an individual's preferences over objectives. In this work, we propose Multi-Objective Human-in-the-loop Bayesian Optimization (MO-HILBO), which builds on explicit multi-objective Bayesian optimization to efficiently infer a personalized set of Pareto-optimal controllers. We compare our approach with an existing multi-objective HILO method and experimentally demonstrate MO-HILBO on a lower-limb exoskeleton across two objectives: metabolic cost (efficiency) and ordinal human feedback (comfort). We find that MO-HILBO (1) discovers Pareto-optimal controllers, and (2) that the pairwise ordering of points on the Pareto front itself is consistent with validation trials. Lastly, we open-source mohilo, a Python package for running both HILO and MO-HILBO on wearable devices: https://dynamicmobility.github.io/mohilo/.

Problem

• Human-in-the-loop optimization
• Multi-objective optimization
• Lower-limb exoskeleton
• Personalized preferences
• Pareto-optimal controllers

Innovation

• Multi-Objective Bayesian Optimization
• Human-in-the-Loop
• Pareto-Optimal Controllers
• Lower-Limb Exoskeleton
• Ordinal Human Feedback