Learning to Live Life on the Edge: Online Learning for Data-Efficient Tactile Contour Following

TL;DR

This paper introduces an online learning method using Gaussian Process Latent Variable Models for tactile contour following, enabling robots to learn efficiently from limited data and generalize to new stimuli.

Contribution

The paper presents a novel online learning approach with GP-LVMs for tactile sensing, improving data efficiency and robustness in robotic contour following tasks.

Findings

Successful contour following with limited data

Robust performance on novel tactile stimuli

Advantages of latent variable models in tactile learning

Abstract

Tactile sensing has been used for a variety of robotic exploration and manipulation tasks but a common constraint is a requirement for a large amount of training data. This paper addresses the issue of data-efficiency by proposing a novel method for online learning based on a Gaussian Process Latent Variable Model (GP-LVM), whereby the robot learns from tactile data whilst performing a contour following task thus enabling generalisation to a wide variety of tactile stimuli. The results show that contour following is successful with comparatively little data and is robust to novel stimuli. This work highlights that even with a simple learning architecture there are significant advantages to be gained in efficient and robust task performance by using latent variable models and online learning for tactile sensing tasks. This paves the way for a new generation of robust, fast, and data-efficient tactile systems.