Task-focused robotic swarming: Maximising effectiveness via specialised roles

In the realm of distributed systems such as swarm robotics and other multi-agent setups, the principle of labour division stands as a fundamental cornerstone. In this study, we propose a task-specialised approach to robotic swarming. The framework integrates an anonymous hedonic game for task allocation and the Hilbert Space-Filling Algorithm for systematic area coverage. Focusing on the problem of efficient foraging, we conduct an in-depth comparative analysis between our proposed specialised approach and the conventional multi-tasking strategy. We establish two distinct swarm configurations: the multi-tasking swarm, where agents perform various sub-tasks, and the specialised swarm, comprising sub-teams focused on specific sub-tasks namely deposit searching and transportation. Throughout our investigation, encompassing diverse environmental scenarios with varying deposit distributions, we evaluate performance metrics such as deposit collection and distance travelled. The specialised approach harnesses the power of hedonic games to allocate tasks optimally within sub-teams, ensuring efficient resource transportation to the base. Our results reveal the advantages of the specialised swarm strategy. Through task specialisation, we achieve heightened scalability, reduced resource consumption, adaptability to dynamic environments, and resilience in the presence of agent unavailability.