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

Source: [Octopus-inspired Distributed Control for Soft Robotic Arms: A Graph Neural Network-Based Attention Policy with Environmental Interaction](<https://www.sota2.com/research/paper/69b24870d030084c22e2f7c3>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-07-18T05:00:13.368Z

Authors: Linxin Hou, Qirui Wu, Zhihang Qin, Yongxin Guo, Cecilia Laschi

Published: 2026-03-10T19:56:07.000Z

## 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 assuming full obstacle geometry at the start of each episode. 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 maintains comparable performance under non-ideal simulated conditions while keeping control effort bounded, suggesting that selective contact-relevant information routing improves resilience in the tested settings.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Soft Rod Reaching on Soft Rod Wall-with-hole Scenario](<https://www.sota2.com/research/sota/soft-rod-reaching-on-soft-rod-wall-with-hole-scenario.md>): Soft Rod Reaching; Soft Rod Wall-with-hole Scenario; Mean Episode Length: 1297.5.

- [Soft Rod Reaching on Soft Rod Structured Obstacles Scenario](<https://www.sota2.com/research/sota/soft-rod-reaching-on-soft-rod-structured-obstacles-scenario.md>): Soft Rod Reaching; Soft Rod Structured Obstacles Scenario; Mean Episode Length: 16.1.

- [Soft Rod Reaching on Soft Rod Basic Scenario](<https://www.sota2.com/research/sota/soft-rod-reaching-on-soft-rod-basic-scenario.md>): Soft Rod Reaching; Soft Rod Basic Scenario; Mean Episode Length: 17.1.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2603.10198>)

[PDF](<https://arxiv.org/pdf/2603.10198.pdf>)
