3D printed neuromorphic sensing systems

Author: 
Date created: 
2021-09-01
Identifier: 
etd21637
Keywords: 
Neuromorphic
3D printing
Electrochemical sensing
Tactile sensing
Synaptic transistor
Abstract: 

Thanks to the high energy efficiency, neuromorphic devices are spotlighted recently by mimicking the calculation principle of the human brain through the parallel computation and the memory function. Various bio-inspired 'in-memory computing' (IMC) devices were developed during the past decades, such as synaptic transistors for artificial synapses. By integrating with specific sensors, neuromorphic sensing systems are achievable with the bio-inspired signal perception function. A signal perception process is possible by a combination of stimuli sensing, signal conversion/transmission, and signal processing. However, most neuromorphic sensing systems were demonstrated without signal conversion/transmission functions. Therefore, those cannot fully mimic the function provides by the sensory neuron in the biological system. This thesis aims to design a neuromorphic sensing system with a complete function as biological sensory neurons. To reach such a target, 3D printed sensors, electrical oscillators, and synaptic transistors were developed as functions of artificial receptors, artificial neurons, and artificial synapses, respectively. Moreover, since the 3D printing technology has demonstrated a facile process due to fast prototyping, the proposed 3D neuromorphic sensing system was designed as a 3D integrated structure and fabricated by 3D printing technologies. A novel multi-axis robot 3D printing system was also utilized to increase the fabrication efficiency with the capability of printing on vertical and tilted surfaces seamlessly. Furthermore, the developed 3D neuromorphic system was easily adapted to the application of tactile sensing. A portable neuromorphic system was integrated with a tactile sensing system for the intelligent tactile sensing application of the humanoid robot. Finally, the bio-inspired reflex arc for the unconscious response was also demonstrated by training the neuromorphic tactile sensing system.

Document type: 
Thesis
Rights: 
This thesis may be printed or downloaded for non-commercial research and scholarly purposes. Copyright remains with the author.
File(s): 
Supervisor(s): 
Woo Soo Kim
Department: 
Applied Sciences: School of Mechatronic Systems Engineering
Thesis type: 
(Thesis) Ph.D.
Statistics: