Efficient Transfer Learning of Robot Dynamic Models Using Morphological Similarity

2026-07-06

AI Summary

This work addresses the challenge of efficiently transferring dynamic models across soft underwater robots that exhibit similar morphology yet differ significantly in scale and hydrodynamic characteristics. To this end, the authors propose an unsupervised domain adaptation approach based on neural networks. The method leverages labeled data from a source domain (a larger robot) to train a dynamics model and employs an autoencoder to construct a shared latent space, enabling effective transfer to a target domain (a smaller robot) with limited or no labeled data. By innovatively integrating morphological priors with domain adaptation mechanisms, the framework achieves high-accuracy ego-velocity estimation without requiring target-domain labels. Experimental validation on physical robots demonstrates that the proposed approach substantially improves both the accuracy and efficiency of cross-platform dynamic modeling.

Abstract

This study proposes a neural network-based transfer learning framework for modeling the dynamics of soft, fin-actuated underwater robots. We focus on morphologically similar robots that differ in scale and hydrodynamic properties. A model trained on data from a larger robot (source domain) is adapted to a smaller one (target domain) with limited labeled data. To enable label-efficient transfer, we develop an autoencoder-based domain adaptation approach that learns a shared latent representation aligning the dynamics of both robots. Experiments on two real underwater robots show that the proposed method enables accurate state estimation of the body-frame velocities on a target platform without labeled data, highlighting its potential for efficient cross-robot dynamics transfer among morphologically similar platforms.

Problem

• transfer learning
• robot dynamic models
• morphological similarity
• domain adaptation
• soft underwater robots

Innovation

• transfer learning
• morphological similarity
• domain adaptation
• soft robotics
• underwater robots