A Continual Learning Framework for Adaptive Control of Modular Soft Robots

2026-07-07

AI Summary

This work addresses the challenge of controlling modular soft robots, whose nonlinear dynamics, hyper-redundant structures, and morphological variability hinder generalization of conventional controllers and necessitate retraining upon configuration changes. To overcome this, the authors propose an adaptive control framework inspired by continual learning: for a fixed morphology, distributed policies learn local dynamics of individual modules to enable precise control; when the morphology changes, the system incrementally learns the new configuration while preserving previously acquired knowledge to mitigate catastrophic forgetting. This approach represents the first application of continual learning to modular soft robot control and supports on-demand module activation to reduce computational overhead. Experiments on both simulated tendon-driven systems and a physical three-module pneumatic soft arm demonstrate superior trajectory tracking accuracy, selective module activation capability, and enhanced computational efficiency.

Abstract

Soft robots have attracted significant attention in applications such as medical intervention, rehabilitation, and robotic manipulation due to their inherent compliance, flexibility, and high degrees of freedom. Modular soft robots (MSRs), composed of multiple interconnected segments, represent an emerging class of robotic systems with highly deformable and reconfigurable structures capable of performing complex tasks. However, designing controllers for MSRs remains challenging due to their nonlinear dynamics, modeling complexity, and hyper-redundant nature. Existing approaches typically require controllers to be retrained from scratch whenever the robot morphology changes. In this work, we address these challenges through a continual learning inspired control framework capable of incrementally adapting to changes in robot morphology while preserving previously acquired knowledge. Specifically, the proposed framework enables the controller to sequentially learn new MSR configurations without forgetting previously learned ones. In addition, for MSRs with fixed configurations, the same framework can be employed in a distributed manner to learn module-specific dynamics, enabling localized control and improved precision. The proposed approach is validated through closed-loop trajectory tracking experiments in simulation using a tendon-driven soft robot, as well as on a real-world three-module pneumatic soft robotic arm. Furthermore, we demonstrate the adaptive capabilities of the framework through a reaching experiment in which the controller selectively activates only the necessary modules to reach a virtual target position, thereby reducing computational overhead.

Problem

• modular soft robots
• continual learning
• adaptive control
• morphology changes
• hyper-redundant systems

Innovation

• continual learning
• modular soft robots
• adaptive control
• morphology adaptation
• distributed learning