Octopus-inspired Distributed Control for Soft Robotic Arms: A Graph Neural Network-Based Attention Policy with Environmental Interaction

TL;DR

This paper introduces SoftGM, a novel distributed control architecture inspired by octopus tentacles, utilizing graph neural networks to enable soft robotic arms to learn contact-rich reaching tasks with environmental interaction without global obstacle maps.

Contribution

SoftGM is the first octopus-inspired distributed control system using graph attention networks for soft robotic arms that adaptively discover obstacles and coordinate contact-rich movements.

Findings

SoftGM outperforms six MARL baselines in complex obstacle scenarios.

The approach maintains robustness under noise and actuation failures.

SoftGM achieves the best performance in a wall-with-hole environment.

Abstract

This paper proposes SoftGM, an octopus-inspired distributed control architecture for segmented soft robotic arms that learn to reach targets in contact-rich environments using online obstacle discovery without relying on global obstacle geometry. SoftGM formulates each arm section as a cooperative agent and represents the arm-environment interaction as a graph. SoftGM uses a two-stage graph attention message passing scheme following a Centralised Training Decentralised Execution (CTDE) paradigm with a centralised critic and decentralised actor. We evaluate SoftGM in a Cosserat-rod simulator (PyElastica) across three tasks that increase the complexity of the environment: obstacle-free, structured obstacles, and a wall-with-hole scenario. Compared with six widely used MARL baselines (IDDPG, IPPO, ISAC, MADDPG, MAPPO, MASAC) under identical information content and training conditions, SoftGM matches strong CTDE methods in simpler settings and achieves the best performance in the wall-with-hole task. Robustness tests with observation noise, single-section actuation failure, and transient disturbances show that SoftGM preserves success while keeping control effort bounded, indicating resilient coordination driven by selective contact-relevant information routing.