Category: Uncategorized

Kudzai Promise Tenderere

Social Media Payments for Minibus Taxis

The minibus taxi industry has been resistant to formalisation. Literature has some attempts by other people and organisations to try and formalise this industry. The most common first step in all their attempts to this formalisation has been to eliminate the use of cash. Unfortunately, their methods of eliminating cash have been around the use of bank cards and bank account tied methods which excludes the poor and unbanked majority that actually use minibus taxis on a daily.

My MEng research is to develop and propose a poor people inclusive, and modern way to eliminate the use of cash, to be specific, I explore the combination of mobile money and social media chatbots.

Keegan Hull

Tunnel temperature control and thermal modelling, with experimental support with generic control and sensing devices.

Grow tunnels (or high tunnels) are used in agricultural research to extend growing seasons or to emulate optimal conditions for experiments. At Stellenbosch University, the Departments of Forestry and Agronomy of the Agriscience Department are using such tunnels for their experimental trials. However, these tunnels are not equipped to measure the internal conditions nor are they able to control temperature conditioning apparatuses such as fans and wet walls.

This project aims to transform these tunnels into smart, autonomous systems. Firstly, generic sensing and control devices called “hats” will be created to support researchers in sensing and controlling their tunnels. These hats, however, will not be limited to agriculture and will extend to any situation where sensing and control is needed. Secondly, using these hats, thermal models of the two tunnels at Agronomy and Forestry will be made using these hats. These thermal models will create a predictive model that will allow researchers to use outside weather data to better predict the potential conditions of the internals of the tunnel.

Read more about Keegan’s work oLinkedIn.

Johan Giliomee

External batteries as a solution to the energy requirements for an electric mini-bus taxi fleet in South Africa

It is 9am on a workday morning and 100 electric mini-bus taxis (eMBTs) retreat to the Stellenbosch taxi rank after rush hour. In preparation for the afternoon trips, they all need to recharge their depleted batteries. Charging at a fair rate of 22kW, these 100 taxis add 2 200kW of demand to an already crippled grid. Blowing up this number to a national scale creates unimaginable issues, which we simply cannot solve as consumers. Alternative methods of energy supply to the eMBTs are thus needed.

The reality we are facing is that optimal charging times for the grid and EVs does not align. As the interference of the operation of a MBT needs to be minimized, the use external batteries are suggested to bridge the ‘optimal charging time’ barrier. By eliminating grid-to-vehicle charging, we can control the load put on the grid, whilst still ensuring sufficient power is available to the taxis at their optimal charging time.

Two solutions are proposed: stationary – and swappable batteries. In both scenarios the use of second-life batteries is investigated.

Find out more about Johan’s work on LinkedIn.

Alex Muthua

Using infrared to improve face recognition of individuals with highly pigmented skin

Face recognition technology has become commonplace in security and access control applications. However, their performance leaves a lot to be desired when working with highly pigmented skin tones.

The aim of this work is to show how this can be enhanced by incorporating the infrared spectrum, which electronic sensors can perceive.  By augmenting existing datasets with images captured using the visible, infrared and full spectra, the performance of existing state-of-the-art face recognition algorithms can be compared for each of these three spectra.

Rita van der Walt

Assessment of the quality of learning environments in South African classrooms

South African schools stretch over a wide spectrum of income ranges, and the classrooms and school grounds across the country clearly reflect this. This research investigates the different environments of South African classrooms, and how these environments may or may not affect the productivity and quality of work of learners and teachers. The results of this research can also be used to reassess classroom building standards and regulations that classrooms need to adhere to in order to create a more habitable environment throughout the school day.

Parameters affecting human comfort and health are measured and will be assessed, including carbon dioxide, temperature, humidity, volatile organic compounds (VOCs) and ambient light. This will help gain significant insights to learning challenges that both teachers and learners face during school hours from an angle which has not necessarily been studied before. The results of this study can help achieve greater equity in South African educational environments.

Find out more about Rita on LinkedIn.

Pierre Labuschagne


The Development of a Low-Cost IoT Platform for Sustainable Development Applications in Developing Countries

The world desperately needs to improve the sustainability of the agriculture, energy, and other industries to reduce the negative impact we have on the planet. Through the rapid development of interconnected devices and associated technologies, collectively known as Internet of Things (IoT) technologies, industries can develop smart monitoring and control solutions that both improve sustainability and reduce operating costs.

Unfortunately, the development of IoT solutions is focused towards developed countries that have the capital to invest in these solutions, as well as the large-scale operations that can benefit from expensive high-tech solutions. These are too costly and often not suitable for implementing in developing countries.

My project focuses on the development of an IoT platform that is suitable for implementation in developing countries. The aim is to make the platform universal by allowing standardized addons be used to change the functionality of the final device. Such a product will need to be low-cost, easy to use/implement, and scalable so that it can be accessible and suitable for use in developing countries. A successful outcome would be increased use/development of IoT solutions in developing countries and ultimately improved sustainability amongst all industries.

Read more on Pierre’s work on LinkedIn.

Daniel van Schalkwyk

Optimization models for energy- and EWH usage costs of schools and public buildings. Forecasting the internal temperature profile and energy profile of EWH units using control and machine learning techniques.

An opportunity is identified to use machine learning and control techniques to investigate the improvement of the two-node thermal model with the aid of a climatic chamber for environmental control and data acquisition. The project will determine whether these new models can estimate electrical energy consumption more accurately and whether it can be extended to provide other forecasted information such as the temperature of the water at different regions within the EWH. Firstly, a thermo-fluid model using FEM and CFD methods will be developed, followed by the further improvement of a climatic chamber to accommodate the data acquisition of more data points. The project investigates both horizontally- and vertically mounted EWHs.