Artificial neural circuit integration for MEMS microrobot system

Abstract

This paper discussed about less than 5mm in size hexapod locomotive type microrobot system. The microrobot system consisted by micro-mechanical systems which were fabricated by micro fabrication technology and micro-electro systems which was constructed by integrated circuit (IC) technology. Micro-mechanical systems were equipped with small size rotary type actuators, body frame, link mechanisms, and 6 legs to realize the ant-like switching behavior. Micro-electro system was the locomotion rhythm generator of the microrobot using artificial neural circuit. Both systems were made from silicon wafer. Therefore, both systems could integrate on same silicon wafer using micro-electro-mechanical systems (MEMS) technology. Artificial neural circuit consisted by 4 cell body models and 12 inhibitory synaptic models. Cell body model was analog circuit model which could output oscillatory patterns such as the biological neuron. Cell body models were connected mutually by the inhibitory synaptic models. Thus, artificial neural circuit could generate the locomotion rhythms using synchronization phenomena of the cell body models such as biological neural networks. Locomotion rhythm generator using artificial neural circuit realized the locomotion of the robot without using any software programs or analog digital converters. As a result, MEMS microrobot performed forward and backward locomotion, and also changes direction by inputting an external single trigger pulse to the artificial neural circuit.