Trajectory Generation for Underactuated Soft Robot Manipulators using Discrete Elastic Rod Dynamics

A team of researchers from multiple institutions has published a new paper, "Trajectory Generation for Underactuated Soft Robot Manipulators using Discrete Elastic Rod Dynamics," introducing a critical advancement for controlling soft robots. The core challenge they address is the trade-off between physical modeling accuracy and computational efficiency. Existing methods either use complex continuum rod models that are too slow for real-time control or simplified approximations that sacrifice fidelity. The team's novel contribution is a control-oriented reformulation of Discrete Elastic Rod (DER) dynamics, which yields a mathematically tractable, control-affine representation while preserving key first-principles force-deformation relationships.

This new formulation allows for the generation of dynamically feasible trajectories that are consistent with the robots' underactuation—meaning they have fewer control inputs than degrees of freedom. The researchers validated their framework experimentally on a pneumatic soft robotic limb. Hardware results demonstrated a consistently improved trajectory tracking performance compared to a standard constant-curvature baseline model. The improvement was particularly notable under complex actuation conditions, where simpler models fail. This work bridges a significant gap in soft robotics, providing a practical tool for planning precise, physically accurate motions for applications in delicate manipulation, medical devices, and exploration in unstructured environments.