The coupled spread of information and potential-driven motion in evacuating agent populations
Abstract
Whether a population evacuates safely depends on two coupled processes: how a warning spreads through it, and how that information moves people. We model this coupling with an agent-based framework, in which each agent's information state--susceptible, exposed, informed or reverted--gates its motion. Only informed agents respond to a global background potential encoding hazards and safe zones, while all agents interact locally through a Lennard-Jones potential. The total potential experienced by an agent drives an Ito stochastic differential equation. Tests show that landscapes with potential minima leave far fewer agents never exposed than in a flat, uniform potential control landscape, as informed agents driven toward the minima sweep past susceptible neighbors. In our tests, conduit routes raise the fraction of agents in the safe-zone minima, but leave greater numbers of agents never exposed, when compared to an unguided evacuation. This is attributed to conduits shortening trips at the cost of informed agents having fewer encounters with susceptible neighbors who could receive their information. Independently, the effective information transmission rate estimated from these encounters falls roughly sixty-fold below its well-mixed value within the first time unit, because state transitions create spatial mechanisms of net transmission suppression that mean-field closure does not capture.
Related
- Optimal Strategies to Contain and Prevent the Spread of Diseases in Group-structured Populations
- Role differentiation as ignition of a collective information engine: Structuration in Agent Populations
- Early inflation in Gauss–Bonnet gravity coupled with a scalar field in an inverse-square potential
- A GRACE Interatomic Potential for W--Ta Systems: Application to Ta Segregation and Pinning of Shear-Coupled Grain-Boundary Migration
- Information Potential: A Variational Approach to Quantum Information Complexity
BibTeX
@misc{jasperson2026coupled,
title = {{The coupled spread of information and potential-driven motion in evacuating agent populations}},
author = {Benjamin A. Jasperson and Logan Dehay and Rahul Gulati and Siddhartha Srivastava and Krishna Garikipati},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.40081},
eprint = {2610.40081},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.40081},
}
RIS
TY - UNPB
TI - The coupled spread of information and potential-driven motion in evacuating agent populations
AU - Benjamin A. Jasperson
AU - Logan Dehay
AU - Rahul Gulati
AU - Siddhartha Srivastava
AU - Krishna Garikipati
PY - 2026
DA - 2026/10/29
PB - arXiv
DO - 10.48550/arxiv.2610.40081
UR - https://arxiv.org/abs/2610.40081
AB - Whether a population evacuates safely depends on two coupled processes: how a warning spreads through it, and how that information moves people. We model this coupling with an agent-based framework, in which each agent's information state--susceptible, exposed, informed or reverted--gates its motion. Only informed agents respond to a global background potential encoding hazards and safe zones, while all agents interact locally through a Lennard-Jones potential. The total potential experienced by an agent drives an Ito stochastic differential equation. Tests show that landscapes with potential minima leave far fewer agents never exposed than in a flat, uniform potential control landscape, as informed agents driven toward the minima sweep past susceptible neighbors. In our tests, conduit routes raise the fraction of agents in the safe-zone minima, but leave greater numbers of agents never exposed, when compared to an unguided evacuation. This is attributed to conduits shortening trips at the cost of informed agents having fewer encounters with susceptible neighbors who could receive their information. Independently, the effective information transmission rate estimated from these encounters falls roughly sixty-fold below its well-mixed value within the first time unit, because state transitions create spatial mechanisms of net transmission suppression that mean-field closure does not capture.
ER -