my students and i are excited and grateful that a paper we co-authored has been accepted at the 2nd International Conference on Metaverse and AI Companions in Education and Society. The conference will be held at the National Central University in Taiwan, from the 19th to the 21st of June.
the paper is titled Modelling and Optimising evolution of self-organising task specialisation in ant-inspired AI swarms.
its abstract reads:
This research endeavours to model and optimise the evolution of self-organising task specialisation in ant-inspired AI swarms, drawing inspiration from the intricate behaviour and evolution observed in ant colonies. The primary motivation stems from the remarkable efficiency and adaptability of Pheidole megacephala (Fabricius) ant colonies in task allocation, particularly the evolution of specialised castes such as soldiers with larger heads for nest defence and smaller workers for food collection. We use an agent-based simulation model, set in a grid-based environment, to simulate ant colonies tasked with collecting and transporting food. Intriguingly, colonies face decisions on energy allocation for ant production, introducing a size mutation rate that influences the percentage of ants produced with larger sizes. Larger ants are not only more energy-intensive to produce but also serve a defensive role against predators. We aim to investigate the impact of such size mutations on the colony's overall efficiency and survival, drawing parallels to Pheidole megacephala ants, where bigger heads aid in nest defence. The simulation provides a platform to study the optimal spread of mutation rates that maximises resource gathering, and through generational analysis, this study attempts to identify adaptive strategies that balance energy allocation, defence mechanisms, and foraging efficiency in ant-inspired AI swarms.