Research Analyzer
← Back ICRA 2026

Feasible Rolling Trajectory Generation and Control for Tensegrity Robots

Songyuan Liu, Qingkai Yang, Zichen Tao, Yun Gui, Jiaxu Shi, Hao Fang

PDF

AI summary

Key figure (auto-extracted from paper)
A novel iLQR-based control framework enables continuous, reset-free rolling for tensegrity robots while reducing actuator load by 34.4% and increasing speed by 18.3%.
Tensegrity robots continuous rolling iLQR control trajectory generation form-finding actuator load reduction

Problem

Existing tensegrity robot controllers rely on open-loop strategies and require reset operations between rolling cycles, preventing efficient, continuous locomotion in complex environments.

Approach

The method decomposes rolling into stance and rolling phases, uses form-finding to compute critical configurations, and applies an iLQR controller with load-minimizing penalties to generate and track feasible trajectories.

Key results

  • Enables continuous rolling without inter-cycle resets
  • Achieves 18.3% higher rolling speed than prior methods
  • Reduces actuator load by 34.4% through optimized trajectory generation
  • Validated through both simulation and physical prototype experiments

Why it matters

Advances autonomous mobility for compliant robots, making tensegrity systems more viable for disaster response and unstructured terrain navigation.

Abstract

Due to the multi-node and multi-contact motion characteristics of tensegrity robots, existing methods fail to generate feasible reference rolling trajectories, and controllers are also limited to open-loop approaches. To address this issue, we utilize motion decomposition to extract the motion phase that should be the primary focus. Subsequently, we propose a method combining form-finding-based critical configuration search and polynomial trajectories to generate feasible trajec- tories. Then, an iLQR controller that accounts for reducing actuator load is designed for trajectory tracking control. A key distinction from existing methods is that our approach eliminates the need for reset operations after each rolling cycle. The results of simulations and physical experiments demonstrate that the robot achieves continuous rolling, with improvements of 18.3% in speed and 34.4% in actuation load compared to existing works.

Index terms

Modeling Control and Learning for Soft Robots Motion Control Dynamics

Related papers