A Multi-Agent System for Modeling and Enhancing Collective Intelli-gence in Indian File Caterpillars

This paper presents a multi-agent simulation framework inspired by the collective behaviors of ant colonies and Indian file caterpillars to compare their coordination mechanisms and collective dynamics. Two bio-inspired models were implemented using the NetLogo simulation environment. The first reproduces the hierarchical linear formation observed in caterpillar processions, whereas the second emulates the stigmergic coordination mechanism of ant colonies through pheromone-based trail following. The remainder of this paper is organized as follows. Section 2 reviews the fundamentals of bio-inspired optimization algorithms, emphasizing their relationship with collective coordination and group dynamics. Subsequently, the proposed multi-agent model is described, incorporating social interactions, chemosensory communication, and environmental factors such as wind and rainfall. A series of simulations comparing optimized (structured and coordinated) and non-optimized (random and independent) movement strategies was conducted to evaluate energy efficiency, locomotion speed, and foraging success. Experimental results demonstrate that optimized collective behaviors consistently outperform non-optimized strategies across both biological systems. These findings highlight the ability of bio-inspired multi-agent models to accurately reproduce both simple and complex collective behaviors, supporting their potential application in autonomous robotics, distributed artificial intelligence, swarm intelligence, and multi-agent system management.