Proceedings of the Annual Meeting of the Cognitive Science Society

Ant colonies exemplify rigorous and efficient distributed cognition, providing profound inspirations for understanding collective intelligence. Specifically, inter-colony competitions and foraging strategies of ants offer insights into how biological systems search resources. Although it is well-known that ants naturally create tree-like structures in foraging paths, little is investigated regarding how multiple colonies optimize these structures simultaneously under competitions. Hence, we propose MT-ACO, a Dual-Colony Competition Model driven by a Pheromone Repulsion Mechanism. In this model, agents treat the pheromone trails of competitors as dynamic environmental constraints, simulating the biological principle of Competitive Exclusion. Simulations reveal that this simple rule leads to near-optimal Edge-Disjoint Directed Trees. This emergent topology maximizes resource coverage while minimizing conflict, providing a decentralized heuristic for the NP-hard Minimal Directed Edge-Disjoint Double Tree problem. Our findings demonstrate that avoidance functions offering a computational framework to understand how swarms resolve the tragedy of the commons through adaptive niche differentiation.