String-like traveling in an autonomous running by skid-steering 2D robots using a neural network based on hue image data

Abstract

String-like traveling is a collective behavior that can be observed in living organisms, including humans. Group behavior is self-assembly movement formed by the interaction of each individual. However, the mechanism of self-assembly has not been fully elucidated. The purpose of this research is to run a running robot using a neural network based on hue image data in an enviroment without external walls, and to observe string-like traveling. String-like traveling was also observed in automous running using a neural network based on hue image data. This suggests that the generalization performance mey be improved.