mlsanityData to 14 Oct 2026

Morphology-Aware Graph Reinforcement Learning for Tensegrity Robot Locomotion

arXiv 2510.26067 ↗First posted 30 Oct 2025cs.ROcited by 0

On the map: Deep reinforcement learning methods → RL for robot locomotion and manipulation

Abstract (arXiv)

Tensegrity robots combine rigid rods and elastic cables, offering high resilience and deployability but at the same time posing major challenges for locomotion control due to their underactuated and highly coupled dynamics. This paper introduces a morphology-aware reinforcement learning framework that integrates a graph neural network (GNN) into the Soft Actor-Critic (SAC) algorithm. By representing the robot's physical topology as a graph, the proposed GNN-based policy captures coupling among components, enabling faster and more stable learning than conventional multilayer perceptron (MLP) policies. The method is validated on a physical 3-bar tensegrity robot across three locomotion primitives, including straight-line tracking and bidirectional turning. It shows superior sample efficiency, robustness to noise and stiffness variations, and improved trajectory accuracy. Additionally, the learned policies transfer directly from simulation to hardware without fine-tuning, achieving stable real-world locomotion. These results demonstrate the advantages of incorporating structural priors into reinforcement learning for tensegrity robot control.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in RL for robot locomotion and manipulation, one of 15 directions in Deep reinforcement learning methods. Builds on 17 of its 31 references that are in the map. First posted on arXiv 30 Oct 2025.

Builds on

17 of 31 references.

References · 17

  1. 2017Proximal Policy Optimization Algorithmscited by 15,215
  2. 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,885
  3. 2015Continuous control with deep reinforcement learningcited by 3,407
  4. 2018Addressing Function Approximation Error in Actor-Critic Methodscited by 2,417
  5. 2017Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real Worldcited by 1,498
  6. 2018Learning Dexterous In-Hand Manipulationcited by 687
  7. 2019Learning agile and dynamic motor skills for legged robotscited by 504
  8. 2017Learning to Fly by Crashingcited by 41
  9. 2023Universal Morphology Control via Contextual Modulationcited by 23
  10. 2021Learning Modular Robot Control Policiescited by 22
  11. 2016Deep Reinforcement Learning for Tensegrity Robot Locomotioncited by 7
  12. 2020Sim2Sim Evaluation of a Novel Data-Efficient Differentiable Physics Engine for Tensegrity Robotscited by 6
5 more
  1. 2022Real2Sim2Real Transfer for Control of Cable-driven Robots via a Differentiable Physics Enginecited by 6
  2. 2018Adaptive Tensegrity Locomotion on Rough Terrain via Reinforcement Learningcited by 2
  3. 2024Design of a Variable Stiffness Quasi-Direct Drive Cable-Actuated Tensegrity Robotcited by 2
  4. 2024Learning Differentiable Tensegrity Dynamics using Graph Neural Networkscited by 1
  5. 2024Tensegrity Robot Proprioceptive State Estimation with Geometric Constraintscited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025GCNT: Graph-Based Transformer Policies for Morphology-Agnostic Reinforcement Learning
  2. 2025McARL:Morphology-Control-Aware Reinforcement Learning for Generalizable Quadrupedal Locomotion
  3. 2024DittoGym: Learning to Control Soft Shape-Shifting Robots
  4. 2025Beyond Topology: A Morphological Symmetry Graph Representation for Locomotion Policy Learning
  5. 2026PolyMorph: Learning Locomotion Policy for Cross-Embodiment Adaptation
  6. 2026ContractionPPO: Certified Reinforcement Learning via Differentiable Contraction Layers
  7. 2024Leveraging Symmetry in RL-based Legged Locomotion Control
  8. 2024One Policy to Run Them All: an End-to-end Learning Approach to Multi-Embodiment Locomotion