History-based model predicts uncertain motions of soft robotic arms
History-Conditioned Flow Matching for Probabilistic Dynamics of Tendon-Driven Continuum Robots
Robotics
Summary
It is difficult to predict how soft robotic arms controlled by tendons will move because of unknown factors like friction and material differences. The authors introduce a way to learn from past movements and commands to better estimate all possible next positions of the robot’s joints. Their method produces accurate predictions in computer simulations and tests on a real robot, capturing the typical sequences of motion even when some conditions change. This helps to understand and control tendon-driven robots despite unseen internal variations.
What this means in practice
- •For robotics engineers: Predict the range of future movements of soft tendon-driven robots under uncertain mechanical conditions to improve control and safety.
- •For robotic simulation developers: Build simulation tools that model uncertain dynamics of continuum robots using history-conditioned probabilistic methods to test new control strategies.
Authors
Hang Yang, Tingcong Liu, Junjie Xiong, Fangju Yang, Ke Wu
Abstract
Deterministic dynamics modeling of tendon-driven continuum robots remains challenging owing to uncertainties in material behavior, tendon transmission, friction, and contact. Measured joint configurations and nominal tendon commands do not fully characterize these internal mechanical factors, leaving uncertainty in the subsequent motion. We therefore develop a history-conditioned, physics-informed flow-matching framework for probabilistic dynamics prediction, using motion and actuation histories to predict the distribution of the next complete joint configuration. By recursively sampling next-step configurations under prescribed commands, the model predicts distributions of future whole-body motions. In simulation, scenario-specific models achieve five-second trajectory Energy Scores (lower is better) of 12.05 mm under internal friction variation and 9.29 mm under unobserved actuation disturbances. Relative to the conditional variational autoencoder and diffusion baselines, Flow attains lower Energy Scores and coverage closer to the nominal level in both scenarios. Ablations support history and structural conditioning in both scenarios. On the physical robot, predictions under two tendon-command profiles excluded from training capture the principal motion sequences, with five-second Energy Scores of 11.91 and 11.42 mm, lower than the compared baselines. The predicted-to-measured spread ratios are 1.65 and 1.22 (closer to 1 is better). These results support history-conditioned probabilistic dynamics prediction under incomplete mechanical observations.