Evaluating Continual Learning on a Home Robot

TL;DR

This paper demonstrates how continual learning methods can be adapted for real home robots, enabling them to learn multiple tasks sequentially with minimal data, using novel methods SANER and ABIP.

Contribution

The paper introduces SANER and ABIP, novel methods for continual skill learning on low-cost robots with few demonstrations, addressing real-world non-iid data challenges.

Findings

Successfully learned four kitchen tasks sequentially

Achieved learning with only a handful of demonstrations per task

Demonstrated feasibility of continual learning on real home robots

Abstract

Robots in home environments need to be able to learn new skills continuously as data becomes available, becoming ever more capable over time while using as little real-world data as possible. However, traditional robot learning approaches typically assume large amounts of iid data, which is inconsistent with this goal. In contrast, continual learning methods like CLEAR and SANE allow autonomous agents to learn off of a stream of non-iid samples; they, however, have not previously been demonstrated on real robotics platforms. In this work, we show how continual learning methods can be adapted for use on a real, low-cost home robot, and in particular look at the case where we have extremely small numbers of examples, in a task-id-free setting. Specifically, we propose SANER, a method for continuously learning a library of skills, and ABIP (Attention-Based Interaction Policies) as the backbone to support it. We learn four sequential kitchen tasks on a low-cost home robot, using only a handful of demonstrations per task.