NavG: Risk-Aware Navigation in Crowded Environments Based on Reinforcement Learning with Guidance Points

2025-03-03

AI Summary

To address degraded navigation safety and efficiency caused by perception errors in dense dynamic environments, this paper proposes a risk-aware reinforcement learning navigation framework. The core innovation introduces “guidance points” as directional priors robust to perception uncertainty: guidance points are systematically generated via boundary extraction, candidate detection, and redundancy elimination, and a unified mapping mechanism integrates LiDAR data, human trajectory tracking, and guidance points to enable perception-planning co-modeling. Built upon the PPO algorithm, the framework achieves the highest task success rate and near-optimal traversal time in simulation. Real-world experiments in corridors and lobbies demonstrate robust pedestrian avoidance and high-confidence obstacle circumnavigation, significantly improving navigation robustness and safety.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Motion planning in navigation systems is highly susceptible to upstream perceptual errors, particularly in human detection and tracking. To mitigate this issue, the concept of guidance points--a novel directional cue within a reinforcement learning-based framework--is introduced. A structured method for identifying guidance points is developed, consisting of obstacle boundary extraction, potential guidance point detection, and redundancy elimination. To integrate guidance points into the navigation pipeline, a perception-to-planning mapping strategy is proposed, unifying guidance points with other perceptual inputs and enabling the RL agent to effectively leverage the complementary relationships among raw laser data, human detection and tracking, and guidance points. Qualitative and quantitative simulations demonstrate that the proposed approach achieves the highest success rate and near-optimal travel times, greatly improving both safety and efficiency. Furthermore, real-world experiments in dynamic corridors and lobbies validate the robot's ability to confidently navigate around obstacles and robustly avoid pedestrians.

Problem

• Mitigates perceptual errors in navigation systems
• Introduces guidance points for risk-aware navigation
• Improves safety and efficiency in crowded environments

Innovation

• Reinforcement learning with guidance points
• Perception-to-planning mapping strategy
• Obstacle boundary extraction and redundancy elimination