Data-Efficient Model Learning and Prediction for Contact-Rich Manipulation Tasks

Shahbaz Abdul Khader, Hang Yin, Pietro Falco, Danica Kragic

Abstract: In this letter, we investigate learning forward dynamics models and multi-step prediction of state variables (long-term prediction) for contact-rich manipulation. The problems are formulated in the context of model-based reinforcement learning (MBRL). We focus on two aspects-discontinuous dynamics and data-efficiency-both of which are important in the identified scope and pose significant challenges to State-of-the-Art methods. We contribute to closing this gap by proposing a method that explicitly adopts a specific hybrid structure for the model while leveraging the uncertainty representation and data-efficiency of Gaussian process. Our experiments on an illustrative moving block task and a 7-DOF robot demonstrate a clear advantage when compared to popular baselines in low data regimes.

External IDs:doi:10.1109/lra.2020.2996067