Robotics: Science and Systems XVI

Compositional Transfer in Hierarchical Reinforcement Learning

Markus Wulfmeier, Abbas Abdolmaleki, Roland Hafner, Jost Tobias Springenberg, Michael Neunert, Noah Siegel, Tim Hertweck, Thomas Lampe, Nicolas Heess, Martin Riedmiller


The successful application of general reinforcement learning algorithms to real-world robotics applications is often limited by their high data requirements. We introduce Regularized Hierarchical Policy Optimization (RHPO) to improve data-efficiency for domains with multiple dominant tasks and ultimately reduce required platform time. To this end, we employ compositional inductive biases on multiple levels and corresponding mechanisms for sharing off-policy transition data across low-level controllers and tasks as well as scheduling of tasks. The presented algorithm enables stable and fast learning for complex, real-world domains in the parallel multitask and sequential transfer case. We show that the investigated types of hierarchy enable positive transfer while partially mitigating negative interference and evaluate the benefits of additional incentives for efficient, compositional task solutions in single task domains. Finally, we demonstrate substantial data-efficiency and final performance gains over competitive baselines in a week-long, physical robot stacking experiment.



    AUTHOR    = {Markus Wulfmeier AND Abbas Abdolmaleki AND Roland Hafner AND Jost Tobias Springenberg AND Michael Neunert AND Noah Siegel AND Tim Hertweck AND Thomas Lampe AND Nicolas Heess AND Martin Riedmiller}, 
    TITLE     = {{Compositional Transfer in Hierarchical Reinforcement Learning}}, 
    BOOKTITLE = {Proceedings of Robotics: Science and Systems}, 
    YEAR      = {2020}, 
    ADDRESS   = {Corvalis, Oregon, USA}, 
    MONTH     = {July}, 
    DOI       = {10.15607/RSS.2020.XVI.054}