Learning to Wait: Preventing Global Congestion from Local Observations in Real-Time Crowd Navigation
Loading...
Date
2025
Journal Title
Journal ISSN
Volume Title
Publisher
The Eurographics Association
Abstract
We present a real-time crowd simulation approach based on reinforcement learning (RL), addressing congestion prevention in confined spaces. We learn a local navigation policy that uses compact, fast-to-compute per-agent observations of a small set of neighbors, including their desired directions. Alongside goal progress and inter-agent spacing, we reward agents for waiting when neighbors ahead pursue similar goals. This formulation fosters global self-organization from purely local interactions. Preliminary results show reduced congestion and consistent goal attainment for large crowds with hundreds of agents.
Description
CCS Concepts: Computing methodologies → Real-time simulation; Multi-agent reinforcement learning
@inproceedings{10.2312:stag.20251341,
booktitle = {Smart Tools and Applications in Graphics - Eurographics Italian Chapter Conference},
editor = {Comino Trinidad, Marc and Mancinelli, Claudio and Maggioli, Filippo and Romanengo, Chiara and Cabiddu, Daniela and Giorgi, Daniela},
title = {{Learning to Wait: Preventing Global Congestion from Local Observations in Real-Time Crowd Navigation}},
author = {Ruprecht, Irena and Michelic, Florian and Preiner, Reinhold},
year = {2025},
publisher = {The Eurographics Association},
ISSN = {2617-4855},
ISBN = {978-3-03868-296-7},
DOI = {10.2312/stag.20251341}
}