Multi-sensor people counting for remote sanitary facility monitoring

Fonseca, Dominic 2023. Multi-sensor people counting for remote sanitary facility monitoring. MPhil Thesis, Cardiff University. Item availability restricted.

Preview	PDF - Accepted Post-Print Version Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (5MB) | Preview
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (135kB)

Abstract

This MPhil thesis has involved collaboration with iPoint Technology, a company working with the South African (SA) government and water board to find methods to help manage and estimate the amount of water used within sanitation facilities. This is achieved by logging the number of people that used these sanitation facilities over a period of time, supported through sensor-based technologies and subsequent analysis of data from these sensors. Due to the limited availability of fresh and clean water in SA due to changing climatic conditions, water conservation is a key requirement for the government. The lack of water and the potential for an outbreak of disease have been heightened due to its geography and placement. SA has been looking at different methods to improve water conservation from public sanitation facilities, transport vehicles and sanitation tanks. This research contributes to this emerging subject area, of interest to a number of other developing countries, in several ways including identifying Internet of Things (IoT) devices that would be used effectively to support this study. Physical and cyber security of the developed devices emerged as one of the key requirements within this study, alongside the benefits and types of data analysis, which could be supported. Users needed to be both safe and secure with minimal chances of theft or damage to devices and leakage of data from the devices that could compromise user privacy. This work involves a comparison of different methods for conducting data analytics, such as the use of a cloud-based platform. It also involves comparing different methods to support people counting and identifying where to store information before sending it to a cloud platform. A key research focus is on the creation of data fusion by using multiple sensors to minimize false positives. Throughout this work, there have been key contributions achieved whilst working with key limitations of the real-world location, internet signal and government regulations on the utilisation of cameras. The primary contributions of this research are as follows: Novel Distributed Technique: Developed and deployed a novel distributed technique, which can efficiently be used alongside edge computing for people counting. This test was able to achieve an accuracy of 96.2%. Case Study Implementation: Conducted multiple case studies based in a real-world setting to validate the proposed method of people counting. Framework for Future Research: Developed a robust framework, this can be taken by future researchers which can use the information and method to look more in depth at people counting without the utilisation of cameras further. The contributions of this thesis provide a foundation for people counting and sending information to a cloud platform by working in a limited means of signal strength and speed.

Item Type:	Thesis (MPhil)
Date Type:	Completion
Status:	Unpublished
Schools:	Computer Science & Informatics
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software
Funders:	Knowledge Economy Skills Scholarships (KESS) 2
Date of First Compliant Deposit:	16 October 2024
Date of Acceptance:	17 September 2024
Last Modified:	18 Oct 2024 14:29
URI:	https://orca.cardiff.ac.uk/id/eprint/172910

Actions (repository staff only)

Edit Item