Bio-Inspired Drone Swarms for Target Neutralization Using Intelligent Munitions

This article proposes a novel approach to the detection and neutralization of military targets using bio-inspired drone swarms deployed from rockets equipped with intelligent cluster munitions. While the use of cluster munitions has faced international restrictions due to the residual risks they pose to civilian populations, the bioinspired swarm strategy offers a promising pathway for more effective and safer military reconnaissance and engagement operations in complex combat environments. Drawing on recent advances in swarm robotics, artificial intelligence, and bioinspired algorithms, this study explores the deployment of multiple drones upon the detonation of smart munitions launched by rockets, enabling them to pursue and identify ground targets. The method was implemented in a two-dimensional simulation combining the CACOC (Chaotic Ant Colony Optimization for Coverage) mobility model with a novel Weighted Target Approximation (WTA) strategy. Predators follow unpredictable trajectories while depositing virtual pheromone traces to avoid redundant coverage, while WTA coordinates capture intentions and prioritizes high-value targets. In this setting, agents are modeled as sensing units governed by distributed control rules, enabling broad area coverage and adaptive detection accuracy. Moreover, the parameters of both UAVs and adversarial agents are optimized using coevolutionary algorithms, ensuring continuous adaptation and improved performance. Statistical analyses using the Kruskal–Wallis and Dunn–Bonferroni tests were employed to validate the results. In the no-WTA scenario, all pairwise comparisons among infantry, vehicle, and tank captures were statistically significant (<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">p</i> < 0.01). Under WTA, infantry vs vehicle and infantry vs tank remained highly significant (<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">p</i> < 10<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">−15</sup>), but vehicle vs tank did not differ significantly (<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">p</i> = 1.000). This confirms that communication shifts predator focus toward higher-value targets rather than uniformly altering all capture behaviors. The CACOC-WTA based method demonstrated that the use of intelligent munitions successfully neutralized ground targets across the 50 evaluated generations.