Trajectory Optimization for Thermally-Actuated Soft Planar Robot Limbs

TL;DR

This paper introduces a novel trajectory generation method for thermally-actuated soft robotic limbs, combining simplified modeling, experimental calibration, and optimization to enable complex motions in hardware.

Contribution

It presents the first approach to generate trajectories for thermally-actuated soft robots using a discretized rigid model and direct collocation optimization.

Findings

Simulation aligns well with hardware tests.

Feasible trajectories can be generated from reference paths.

Open-loop control shows promising results.

Abstract

Practical use of robotic manipulators made from soft materials requires generating and executing complex motions. We present the first approach for generating trajectories of a thermally-actuated soft robotic manipulator. Based on simplified approximations of the soft arm and its antagonistic shape-memory alloy actuator coils, we justify a dynamics model of a discretized rigid manipulator with joint torques proportional to wire temperature. Then, we propose a method to calibrate this model from experimental data and demonstrate that the simulation aligns well with a hardware test. Finally, we use a direct collocation optimization with the robot's nonlinear dynamics to generate feasible state-input trajectories from a desired reference. Three experiments validate our approach for a single-segment robot in hardware: first using a hand-derived reference trajectory, then with two teach-and-repeat tests. The results show promise for both open-loop motion generation as well as for future applications with feedback.