Zero-Shot Option Invention for Robot Planning Under Uncertainty
Keywords: Learning generalized abstractions, robot planning, stochastic motion planning, neuro-symbolic learning, learning and planning for robots
TL;DR: This paper presents a novel approach for learning to automatically create zero-shot abstractions for stochastic robot problems that are generalizable across different robots, environments, and problem instances.
Abstract: This paper addresses the problem of inventing and using hierarchical representations for stochastic robot-planning problems. Rather than using hand-coded state or action representations as input, it presents new methods for learning how to create a generalizable high-level action representation for long-horizon, sparse reward robot planning problems in stochastic settings with unknown dynamics. After training, this system yields a robot-class-specific but environment independent planning system that generalizes to different robots, environments, and problem instances. Given new problem instances in unseen stochastic environments, it first creates zero-shot options (without any experience on the new environment) with dense pseudo-rewards and then uses them to solve the input problem in a hierarchical planning and refinement process. Theoretical results identify sufficient conditions for completeness of the presented approach. Extensive empirical analysis shows that even in settings that go beyond these sufficient conditions, this approach convincingly outperforms baselines by $2\times$ in terms of solution time with orders of magnitude improvement in solution quality.
Submission Number: 6
Loading