Finding the Easy Way Through -- the Probabilistic Gap Planner for Social Robot Navigation

2025-06-25

AI Summary

Existing social robot navigation methods in dense pedestrian crowds focus predominantly on myopic, reactive interactions, limiting their ability to proactively plan long-horizon trajectories through inter-pedestrian gaps. To address this, we propose a hierarchical trajectory planning framework. At the high level, an enhanced probabilistic collision risk model—incorporating a cooperation-aware prior—actively identifies and steers the robot toward traversable gaps in the crowd. At the low level, a local collaborative avoidance mechanism integrates conflict avoidance with cooperative collision avoidance (CCA), augmented by an optimized probabilistic gap planning (PGP) algorithm. In simulation, our approach significantly reduces collision rates and social tension while increasing pedestrian safety distances. Although resulting paths are marginally longer, the overall navigation performance improves substantially. The framework has been successfully deployed in real time on the Honda WaPOCHI robot.

Technology Category

• Intelligent Robots: Motion and Path Planning
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• Search and Retrieval-Augmented AI: Agentic search

Abstract

In Social Robot Navigation, autonomous agents need to resolve many sequential interactions with other agents. State-of-the art planners can efficiently resolve the next, imminent interaction cooperatively and do not focus on longer planning horizons. This makes it hard to maneuver scenarios where the agent needs to select a good strategy to find gaps or channels in the crowd. We propose to decompose trajectory planning into two separate steps: Conflict avoidance for finding good, macroscopic trajectories, and cooperative collision avoidance (CCA) for resolving the next interaction optimally. We propose the Probabilistic Gap Planner (PGP) as a conflict avoidance planner. PGP modifies an established probabilistic collision risk model to include a general assumption of cooperativity. PGP biases the short-term CCA planner to head towards gaps in the crowd. In extensive simulations with crowds of varying density, we show that using PGP in addition to state-of-the-art CCA planners improves the agents' performance: On average, agents keep more space to others, create less tension, and cause fewer collisions. This typically comes at the expense of slightly longer paths. PGP runs in real-time on WaPOCHI mobile robot by Honda R&D.

Problem

• Social robots struggle with long-term crowd navigation strategies
• Existing planners lack focus on finding gaps in dense crowds
• Need efficient real-time planning for cooperative collision avoidance

Innovation

• Decompose planning into conflict and collision avoidance
• Probabilistic Gap Planner biases towards crowd gaps
• Real-time performance on mobile robot platform