Synthesis of pheromone-oriented emergent behavior of a silkworm moth

Abstract

The purpose of this research is to clarify moth emergent behavior by synthesis with currently developed tools, including living sensors and recurrent neural networks. The antennae on a silkworm moth are very sensitive compared with artificial gas sensors. These living antennae can be used as living gas sensors that can detect pheromone molecules with high sensitivity. Recurrent artificial neural networks are applied for controlling the pheromone tracing. As a result, a mobile robot with living antennae can follow a stream of pheromone like a male silkworm moth. The small numbers of neurons can generate moth-like behavior, including casting, and turning while interacting with the environment.