Enhancing Safety for Autonomous Agents in Partly Concealed Urban Traffic Environments Through Representation-Based Shielding

TL;DR

This paper introduces a new state representation for reinforcement learning in autonomous vehicles, improving safety and efficiency in complex urban traffic scenarios with partial visibility and unpredictable elements.

Contribution

It presents a novel perception-based state representation that enhances safety and energy efficiency for autonomous agents navigating complex urban environments.

Findings

Significant safety improvements over baseline models

Reduced energy consumption in navigation tasks

Maintained competitive travel speeds

Abstract

Navigating unsignalized intersections in urban environments poses a complex challenge for self-driving vehicles, where issues such as view obstructions, unpredictable pedestrian crossings, and diverse traffic participants demand a great focus on crash prevention. In this paper, we propose a novel state representation for Reinforcement Learning (RL) agents centered around the information perceivable by an autonomous agent, enabling the safe navigation of previously uncharted road maps. Our approach surpasses several baseline models by a sig nificant margin in terms of safety and energy consumption metrics. These improvements are achieved while maintaining a competitive average travel speed. Our findings pave the way for more robust and reliable autonomous navigation strategies, promising safer and more efficient urban traffic environments.