OLIVE: Online Low-Rank Incremental Learning for Efficient Adaptive Exoskeletons

2026-06-02

AI Summary

This work addresses the limitations of existing exoskeleton controllers, which predominantly rely on static gait policies and struggle to adapt to dynamic environments and inter-user variability. The authors propose a parameter-efficient online adaptive control framework that personalizes control policies in real time during deployment via a low-rank residual update mechanism. By integrating multimodal proprioceptive feedback—including electromyography (EMG), inertial measurement unit (IMU), and vibrotactile signals—and leveraging a context-aware gating module with a dynamic rank scheduler, the system achieves flexible and efficient policy optimization. Requiring no offline trajectory pre-training, the controller converges within approximately 1,800 steps with an end-to-end latency of only 7.4 milliseconds. Evaluated across diverse terrains, it significantly outperforms the strongest baseline, improving gait smoothness, reducing user effort, and enhancing movement stability by 13%, 22%, and 15%, respectively.

Abstract

Wearable exoskeleton systems hold promise for restoring mobility in individuals with physical impairments, yet most existing controllers rely on static gait policies that lack the ability to adapt to dynamic real-world environments or individual user characteristics. We present \olive (\underline{O}nline \underline{L}ow-rank \underline{I}ncremental Learning for Efficient Adapti\underline{ve} Exoskeletons), a parameter-efficient online adaptation framework that continuously personalizes exoskeleton control during deployment. \olive decomposes the adaptive component of the control policy into a low-rank residual form~$\dW = \At\Bt^\top$ with rank~$r!\ll!\min(d,k)$, reducing online update cost from $\mathcal{O}(dk)$ to $\mathcal{O}(r(d{+}k))$ while preserving the stability of a pretrained base controller~$\Wz$. Parameters are updated via a reward-shaped policy gradient driven purely by on-body sensor feedback (EMG, IMU, vibration), eliminating dependence on offline reference trajectories. A gating mechanism modulates the strength of personalization based on contextual state, and a dynamic rank scheduler adapts the update dimensionality to terrain complexity -- allocating minimal capacity on simple flat terrain and expanding to higher-rank updates on demanding uneven surfaces -- enabling robust performance across diverse activities: flat walking, stair navigation, slopes, and uneven terrain. Experiments on the wearable platform demonstrate that \olive achieves +13, +22, and +15 percentage-point improvements in gait smoothness, effort reduction, and motion stability over the strongest baseline, converging within $\sim$1{,}800 walking steps at 7.4,ms end-to-end latency. Our code implementation is available at https://github.com/FastLM/OLIVE.

Problem

• exoskeleton control
• adaptive learning
• personalization
• dynamic environments
• gait adaptation

Innovation

• low-rank adaptation
• online incremental learning
• exoskeleton control
• parameter-efficient learning
• sensor-driven policy gradient