Effects of probabilistic local vision on morphogenesis in multiple robots

We used experimental computer simulation to address effects of probabilistic area in a local vision of individual robot on circle formation in multiple robots. Individual robot uses 1-bit signal communication to perceive other robots within a specific range. The robot takes reflective actions according to other robots' locations: it proceeds or turns left or right. The action selection rules of the robot allowed circle formation in the multiple robots. Conditions exist in which circle formation emerges with perfect sensing on robots' local vision. We describe experimental computer simulations to investigate sensing error effects, which signify probabilistic area between vision areas to such circle formation. We clarify that the characteristics on the emergence of circle formation are influenced by the probabilistic area in the local vision. Stability of circle formation changes for a certain amount of probabilistic area compared with the case of perfect sensing. And the probabilistic area provides the actual front-view angle that works adaptively as a specific view angle according to population-size of a circle formation. Finally, we show an example that the target population-size is given, the morphogenesis capability of that improves with the probabilistic area than without one.