This paper investigates coherent collective motion, collective decision-making and behaviour transitions in networks of wireless connected mobile robots. Emergent global behaviour is achieved through the multiple interactions of locally communicating agents. Collectively, the robots hunt a moving target, immobilise that target and then hunt the next target. Using explicit communication over limited local range, the robots exchange their internal state vectors with neighbouring robots and update their internal states using deterministic transition rules. Configurations of a robot’s internal states and local sensor information are then translated into motion. The results demonstrate that the developed swarming algorithm is scalable and by exploiting the distributedness of the swarm shows an increasing success rate for increasing swarm sizes in domains of high noise. We have begun to validate our findings from simulation with real-robot experiments.