Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Simulation Environment for the Neuroevolution of Ant Colony Dynamics

Michael Crosscombe, Ilya Horiguchi, Norihiro Maruyama, Shigeto Dobata, Takashi Ikegami

TL;DR

A simulation environment to facilitate research into emergent collective behaviour, with a focus on replicating the dynamics of ant colonies, is introduced, focusing on evolving neural architectures that encode domain-specific behaviours in the network topology.

Abstract

We introduce a simulation environment to facilitate research into emergent collective behaviour, with a focus on replicating the dynamics of ant colonies. By leveraging real-world data, the environment simulates a target ant trail that a controllable agent must learn to replicate, using sensory data observed by the target ant. This work aims to contribute to the neuroevolution of models for collective behaviour, focusing on evolving neural architectures that encode domain-specific behaviours in the network topology. By evolving models that can be modified and studied in a controlled environment, we can uncover the necessary conditions required for collective behaviours to emerge. We hope this environment will be useful to those studying the role of interactions in emergent behaviour within collective systems.

A Simulation Environment for the Neuroevolution of Ant Colony Dynamics

TL;DR

A simulation environment to facilitate research into emergent collective behaviour, with a focus on replicating the dynamics of ant colonies, is introduced, focusing on evolving neural architectures that encode domain-specific behaviours in the network topology.

Abstract

We introduce a simulation environment to facilitate research into emergent collective behaviour, with a focus on replicating the dynamics of ant colonies. By leveraging real-world data, the environment simulates a target ant trail that a controllable agent must learn to replicate, using sensory data observed by the target ant. This work aims to contribute to the neuroevolution of models for collective behaviour, focusing on evolving neural architectures that encode domain-specific behaviours in the network topology. By evolving models that can be modified and studied in a controlled environment, we can uncover the necessary conditions required for collective behaviours to emerge. We hope this environment will be useful to those studying the role of interactions in emergent behaviour within collective systems.
Paper Structure (5 sections, 1 equation, 2 figures)

This paper contains 5 sections, 1 equation, 2 figures.

Table of Contents

  1. Introduction
  2. Simulating Ant Dynamics
  3. Encoding Collective Behaviours
  4. Conclusions
  5. Acknowledgements

Figures (2)

  • Figure 1: Left: Top-down video recordings, cropped and scaled to $1280$x$1280$ resolution. The ants are filmed from above in an evenly-lit $100$ mm diameter arena. Individual ant positions are extracted from a $4$ hour recording. Right: Our simulation environment using https://www.pygame.org/wiki/about and Gymnasium to reproduce ant colony dynamics with an agent that can interface with a policy network.
  • Figure 2: Close-up view of the controllable agent (blue) and the agent's corresponding trail $P_\alpha$(light blue) next to the target trail (red). The blue circle surrounding the agent represents the agent's vision divided into segments.