Simulation-Aided Policy Tuning for Black-Box Robot Learning

TL;DR

This paper introduces a simulation-aided black-box policy search algorithm that leverages both real robot data and simulation to enable fast, data-efficient learning and adaptation in robotic tasks.

Contribution

It presents a novel dual-information source optimization algorithm that improves policy learning efficiency by combining real-world experiments with simulation data.

Findings

Reduces robot interaction time significantly

Achieves high-probability policy improvements

Demonstrates successful real robot task learning

Abstract

How can robots learn and adapt to new tasks and situations with little data? Systematic exploration and simulation are crucial tools for efficient robot learning. We present a novel black-box policy search algorithm focused on data-efficient policy improvements. The algorithm learns directly on the robot and treats simulation as an additional information source to speed up the learning process. At the core of the algorithm, a probabilistic model learns the dependence of the policy parameters and the robot learning objective not only by performing experiments on the robot, but also by leveraging data from a simulator. This substantially reduces interaction time with the robot. Using this model, we can guarantee improvements with high probability for each policy update, thereby facilitating fast, goal-oriented learning. We evaluate our algorithm on simulated fine-tuning tasks and demonstrate the data-efficiency of the proposed dual-information source optimization algorithm. In a real robot learning experiment, we show fast and successful task learning on a robot manipulator with the aid of an imperfect simulator.