Back to search

University of Sheffield

A Contribution Towards Intelligent Autonomous Sensors Based on Perovskite Solar Cells and Ta2O5/ZnO Thin Film Transistors

Abstract

dc:description.abstract

Many broad applications in the field of robotics, brain-machine interfaces, cognitive computing, image and speech processing and wearables require edge devices with very constrained power and hardware requirements that are challenging to realize. This is because these applications require sub-conscious awareness and require to be always “on”, especially when integrated with a sensor node that detects an event in the environment. Present day edge intelligent devices are typically based on hybrid CMOS-memristor arrays that have been so far designed for fast switching, typically in the range of nanoseconds, low energy consumption (typically in nano-Joules), high density and endurance (exceeding 1015 cycles). On the other hand, sensory-processing systems that have the same time constants and dynamics as their input signals, are best placed to learn or extract information from them. To meet this requirement, many applications are implemented using external “delay” in the memristor, in a process which enables each synapse to be modeled as a combination of a temporal delay and a spatial weight parameter. This thesis demonstrates a synaptic thin film transistor capable of inherent logic functions as well as compute-in-memory on similar time scales as biological events. Even beyond a conventional crossbar array architecture, we have relied on new concepts in reservoir computing to demonstrate a delay system reservoir with the highest learning efficiency of 95% reported to date, in comparison to equivalent two terminal memristors, using a single device for the task of image processing. The crux of our findings relied on enhancing our capability to model the unique physics of the device, in the scope of the current thesis, that is not amenable to conventional TCAD simulations. The model provides new insight into the redox characteristics of the gate current and paves way for assessment of device performance in compute-in-memory applications. The diffusion-based mechanism of the device, effectively enables time constants that have potential in applications such as gesture recognition and detection of cardiac arrythmia. The thesis also reports a new orientation of a solution processed perovskite solar cell with an efficiency of 14.9% that is easily integrable into an intelligent sensor node. We examine the influence of the growth orientation on film morphology and solar cell efficiency. Collectively, our work aids the development of more energy-efficient, powerful edge-computing sensor systems for upcoming applications of the IOT.

Degree

thesis:*
Name dc:type.qualificationname
Ph.D
Level dc:type.qualificationlevel
doctoral
Grantor dc:publisher.institution
University of Sheffield
Year dc:date.issued
2023

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Song, Xiaoyao
Advisor dc:contributor.advisor
  • De Souza, Merlyne
ORCID person:identifier.orcid

Rights

Language dc:language.dcterms_iso639_2
eng

Chain of custody

source
Harvested from
White Rose University Consortium
Base URL
etheses.whiterose.ac.uk/cgi/oai2
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
related terms
citation

Song, Xiaoyao. A Contribution Towards Intelligent Autonomous Sensors Based on Perovskite Solar Cells and Ta2O5/ZnO Thin Film Transistors. doctoral thesis, University of Sheffield, 2023.