Tensegrity robots navigate tricky terrain with smarter control and learning
Model Predictive Control of Tensegrity Robots via Contact-Aware Graph Neural Dynamics Model
Robotics
Summary
Tensegrity robots, which are lightweight and flexible, have trouble moving around because their movements are affected by many contacts and obstacles. The authors improved how these robots move by teaching a computer model to understand the robot’s parts and how they touch things using a special network called a graph neural network. They combined this with a planning method that tries many possible movements to find the best one, even on bumpy or obstacle-filled terrain. Testing in simulated environments showed this new approach helps the robots move better and avoid problems compared to older methods.
tensegrity robotmodel predictive controlgraph neural networkcontact dynamicspath integral controlsimulationnavigationpartial observabilitymotion planningcollision detection
Authors
Nelson Chen, Patrick Meng, Charles Tang, Angelina Degay, Zachary Brei, Rebecca Kramer-Bottiglio, Kostas E. Bekris, Mridul Aanjaneya
Abstract
Tensegrity robots offer lightweight, compliant mobility over challenging terrain but remain difficult to model and control due to complex contact-rich dynamics and partial observability. This work presents a model predictive path integral (MPPI) controller for a three-bar tensegrity robot driven by a learned graph neural network (GNN) dynamics model. This work first extends prior GNN-based models with a differentiable contact detection module. The extension allows the dynamics model to reason over non-horizontal planar terrains, obstacles, as well as self-collisions. Then, the learned dynamics model and the MPPI controller operate in a closed data-collection loop, iteratively improving model accuracy and control performance. This work further introduces a hybrid MPPI strategy that combines MPPI with turning motion primitives to improve maneuverability. Experiments are performed in MuJoCo across five navigation tasks, which include, wall obstacles, inclines, narrow corridors, low-clearance structures, and a composite 3D obstacle course. The experiments demonstrate that the hybrid MPPI controller operating over the learned GNN dynamics model improves predictive accuracy over a flat-ground baseline model and achieves superior navigation performance compared to $A^*$-based re-planning and MPPI-only variants. Results show that the contact-aware learned dynamics combined with the sampling-based model predictive control enable robust tensegrity navigation in complex, contact-rich environments.