Task Allocation and Improvement of Collective Work Efficiency in Swarm Robotics Using Autonomous Decentralized Algorithms

This study proposes a decentralized, dynamic task allocation algorithm for swarm robotics to enhance efficiency in multi-foraging tasks. Inspired by the adaptive foraging behavior of social insects, the proposed method introduces a utility-based metric called "Umami," which evaluates food sources based on their quantity and distance from the nest. Each robot dynamically transitions between three roles—Searcher, Picker, and Carrier—based on local information and threshold-based switching. Simulation experiments conducted under various environment sizes and robot densities demonstrated that the proposed integrated model achieved notably higher foraging efficiency, especially in large-scale environments. These findings suggest that the proposed approach offers a scalable and adaptable solution for efficient task execution in swarm robotic systems.