Interactions-based method to detect emergent behavior in ongoing simulations

Abstract

This study investigates the use of interactions as a metric to detect emergent phenomena in the ongoing agent-based simulations. The proposed algorithm aims to detect the flocking behavior in the Boids model. Boids model is simple to understand and to use and always verifies emergence. In recent years, several metrics have been proposed to investigate emergent behaviors in complex systems. Two main limitations of these studies are, how to choose a proper metric to avoid a high coast of computation, and how to detect emergence as it happens i.e., live detection. In this paper, we propose an interaction-based algorithm to detect emergent behavior in Boids model. The three main contributions of this work are: (1) Using interaction as metric, (2) Detect emergence when it happens, (3) No need to store simulation data for post-mortem analysis. A prior interactions analysis was done to classify the system interactions. The results show that the algorithm was able to detect the rise of the flocking, i.e., emergence in the Boids model. Simulation results are provided to illustrate the effectiveness of the proposed approach.