Pseudo-rigid body networks: learning interpretable deformable object dynamics from partial observations

TL;DR

This paper introduces Pseudo-rigid Body Networks, a model that predicts deformable linear object dynamics using interpretable rigid body chains and physics-informed encoding, achieving accurate and physically meaningful predictions from partial data.

Contribution

It proposes a novel pseudo-rigid body inspired model that combines interpretability with accuracy for deformable object dynamics prediction from partial observations.

Findings

Provides physically interpretable predictions in robot experiments.

Achieves prediction accuracy comparable to black-box models.

Demonstrates effectiveness with partial observational data.

Abstract

Accurately predicting deformable linear object (DLO) dynamics is challenging, especially when the task requires a model that is both human-interpretable and computationally efficient. In this work, we draw inspiration from the pseudo-rigid body method (PRB) and model a DLO as a serial chain of rigid bodies whose internal state is unrolled through time by a dynamics network. This dynamics network is trained jointly with a physics-informed encoder that maps observed motion variables to the DLO's hidden state. To encourage the state to acquire a physically meaningful representation, we leverage the forward kinematics of the PRB model as a decoder. We demonstrate in robot experiments that the proposed DLO dynamics model provides physically interpretable predictions from partial observations while being on par with black-box models regarding prediction accuracy. The project code is available at: http://tinyurl.com/prb-networks