DECENTRALIZED PATH FORMATION TECHNIQUE FOR SWARM ROBOTS USING B.A.T.M.A.N. APPROACH

This work presents a swarm robotic algorithm that enables to move an object in one target location to another location by following the shortest possible route found. The algorithm is inspired by neighboring node existence identification concept used by Better Approach To Mobile Ad-hoc Networking (B.A.T.M.A.N.) protocol and foraging behavior of ants. The algorithm consists of two distinct sets of robots. Nodes are path formation robots between food location and nest location. Worker robots move foods from food location to nest location by following the path formed by nodes. The algorithm is modelled on Net Logo agent-based modelling environment and measured the performance of the algorithm. Performance analysis shows that path completion by node robots, and food collected, and food returned by workers robots also improved its performance with large robot group sizes in worlds with and without obstacles. Additionally, it has been observed that increasing node switching probability (probability of leaving path formed from food location or nest location before joining both to form a single path) is not desirable as it decreases the path forming capability of the node robots