Simulation #2: Ant Mill
This experiment models individual ants using local movement rules. Each ant senses nearby trail pheromone, deposits its own trail, responds to neighbours and searches for food. Under some conditions, positive feedback can reinforce a circular procession. The model is illustrative: it does not claim that every natural ant mill forms by exactly these mechanisms.
Simulation #2: Ant Mill
Agent-based model · local rules · collective outcomes
Experiments
Try starting with dispersed ants, then use "Seed circular trail" as a comparison. A seeded circle is an experimental intervention, not spontaneous emergence. Compare circularity and loop duration across settings. In food mode, ants can follow a food-associated trail as well as other ants' deposited trail signal.
How to run an experiment
- Start with the default population of 300 ants and the default settings.
- Observe whether the trail field strengthens into a circular pattern without seeding one.
- Select Seed circular trail to compare an imposed circle with spontaneous movement.
- Increase randomness or evaporation, then use Disturb the trail to see whether the pattern persists.
- Change the environment to obstacles and barriers to observe how the agents alter their routes.
The circularity score is a heuristic based on the distribution and direction of movement. A high score is not, by itself, proof of a biologically realistic ant mill.
📄 This page was created on 18 October 2026. You can view its history on GitHub, preview the fileTip: Press Alt+Shift+G, or inspect the .