Bridging the Gap: Regularized Reinforcement Learning for Improved Classical Motion Planning with Safety Modules

TL;DR

This paper introduces a reinforcement learning approach guided by classical algorithms for mobile navigation, enhancing efficiency, safety, and human norm compliance without needing expert demonstrations.

Contribution

It combines classical planning with reinforcement learning and safety fallback to improve navigation performance and safety guarantees.

Findings

Enhanced convergence rate of RL algorithms

No need for human expert demonstrations

Guaranteed safety through fallback system

Abstract

Classical navigation planners can provide safe navigation, albeit often suboptimally and with hindered human norm compliance. ML-based, contemporary autonomous navigation algorithms can imitate more natural and humancompliant navigation, but usually require large and realistic datasets and do not always provide safety guarantees. We present an approach that leverages a classical algorithm to guide reinforcement learning. This greatly improves the results and convergence rate of the underlying RL algorithm and requires no human-expert demonstrations to jump-start the process. Additionally, we incorporate a practical fallback system that can switch back to a classical planner to ensure safety. The outcome is a sample efficient ML approach for mobile navigation that builds on classical algorithms, improves them to ensure human compliance, and guarantees safety.