GoSafe: Globally Optimal Safe Robot Learning

Proceedings - IEEE International Conference on Robotics and Automation/Proceedings, pp. 4452–4458

Abstract

When learning policies for robotic systems from data, safety is a major concern, as violation of safety constraints may cause hardware damage. SafeOpt is an efficient Bayesian optimization (BO) algorithm that can learn policies while guaranteeing safety with high probability. However, its search space is limited to an initially given safe region. We extend this method by exploring outside the initial safe area while still guaranteeing safety with high probability. This is achieved by learning a set of initial conditions from which we can recover safely using a learned backup controller in case of a potential failure. We derive conditions for guaranteed convergence to the global optimum and validate GoSafe in hardware experiments.

Authors 4

1. Dominik Baumann Aachen Institute for Data Science in Mechanical Engineering

   Max Planck Institute for Intelligent Systems · RWTH Aachen University

   Affiliation as printed

   Institute for Data Science in Mechanical Engineering, RWTH Aachen University, Germany

   Max Planck Institute for Intelligent Systems, Stuttgart, Germany

   RWTH Aachen University

2. Alonso Marco Aachen

   Max Planck Institute for Intelligent Systems · RWTH Aachen University

   Affiliation as printed

   Max Planck Institute for Intelligent Systems, Stuttgart, Germany

   RWTH Aachen University

3. Matteo Turchetta

   Max Planck Society · ETH Zurich

   Affiliation as printed

   Department of Computer Science, ETH Zürich, Switzerland

   Max-Planck-Society

4. Sebastian Trimpe Aachen Institute for Data Science in Mechanical Engineering

   ETH Zurich · Max Planck Institute for Intelligent Systems · RWTH Aachen University

   Affiliation as printed

   Institute for Data Science in Mechanical Engineering, RWTH Aachen University, Germany

   Max Planck Institute for Intelligent Systems, Stuttgart, Germany

   ETH Zürich