Towards the synthesis of trimethoprim using continuous flow Chemistry
- Authors: Ngwenya, Sandiso
- Date: 2020
- Subjects: Trimethoprim
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/47475 , vital:39994
- Description: The continuous flow synthesis of trimethoprim, an antibiotic, was investigated in this study from simple commercially available reagents, such as aniline and dimethylamine. The first step of the synthesis involves a highly exothermic Michael reaction, between acrylonitrile and dimethylamine; process intensification through miniaturization provides a safe and efficient method of synthesis through the use of microreactors. This is afforded by their large surface area-to-volume ratio; one of the many characteristics of continuous flow synthesis that allows for rapid heat dissipation allowing for exothermic reaction to be conducted safely at ambient or higher temperatures. This study was optimized on a Chemtrix Labtrix® Start unit using a Chemtrix reactor achieving a conversion of 99% in 12 sec residence time at 25oC. Subsequently the base catalyzed condensation of the intermediate with 3,4,5- trimethoxybenzaldehyde was investigated using Little Things Factory reactors (Channel diameter: 1.0mm). Maximum conversion (100%) was obtained at 80oC in 20 mins. Due to the poor cyclisation observed with this intermediate, according to literature findings, the 3-dimethylamino group was substituted for aniline. The optimisation for this acid catalysed substitution reaction was performed on a homemade PTFE tubing reactor (Diameter: 0.5mm) and it was found that at 120oC in 10 mins residence time a conversion of 97% was obtained. From the results obtained, an investigation into an integrated process for the condensation and substitution reaction was done using the Little Things Factory system (Channel Diameter: 1.0mm) coupled with a homemade PTFE coil reactor (Diameter: 0.5mm). This process, under optimum conditions of temperature 80oC and residence time of 30 mins, attained a conversion of 76%.
- Full Text:
- Date Issued: 2020
- Authors: Ngwenya, Sandiso
- Date: 2020
- Subjects: Trimethoprim
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/47475 , vital:39994
- Description: The continuous flow synthesis of trimethoprim, an antibiotic, was investigated in this study from simple commercially available reagents, such as aniline and dimethylamine. The first step of the synthesis involves a highly exothermic Michael reaction, between acrylonitrile and dimethylamine; process intensification through miniaturization provides a safe and efficient method of synthesis through the use of microreactors. This is afforded by their large surface area-to-volume ratio; one of the many characteristics of continuous flow synthesis that allows for rapid heat dissipation allowing for exothermic reaction to be conducted safely at ambient or higher temperatures. This study was optimized on a Chemtrix Labtrix® Start unit using a Chemtrix reactor achieving a conversion of 99% in 12 sec residence time at 25oC. Subsequently the base catalyzed condensation of the intermediate with 3,4,5- trimethoxybenzaldehyde was investigated using Little Things Factory reactors (Channel diameter: 1.0mm). Maximum conversion (100%) was obtained at 80oC in 20 mins. Due to the poor cyclisation observed with this intermediate, according to literature findings, the 3-dimethylamino group was substituted for aniline. The optimisation for this acid catalysed substitution reaction was performed on a homemade PTFE tubing reactor (Diameter: 0.5mm) and it was found that at 120oC in 10 mins residence time a conversion of 97% was obtained. From the results obtained, an investigation into an integrated process for the condensation and substitution reaction was done using the Little Things Factory system (Channel Diameter: 1.0mm) coupled with a homemade PTFE coil reactor (Diameter: 0.5mm). This process, under optimum conditions of temperature 80oC and residence time of 30 mins, attained a conversion of 76%.
- Full Text:
- Date Issued: 2020
Understanding the relationship between mangrove composition, stand dynamics and resource use in the Ntafufu estuary, Eastern Cape, as basis for sustainable resource use
- Mvakade, Monica, Rajkaran, Anusha
- Authors: Mvakade, Monica , Rajkaran, Anusha
- Date: 2020
- Subjects: Estuarine ecology -- South Africa -- Eastern Cape , Aquatic ecology , Marine sciences , Fresh water
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/49215 , vital:41611
- Description: Traditionally, coastal communities along the Wild Coast in the Eastern Cape, used mangrove forests for their livelihoods and socio-economic needs. Mangroves along the Eastern Cape coast are threatened by both natural and human factors. The wood obtained from the mangroves by rural households was used as poles for house and fence construction, firewood, and traditional medicine. This is a threat to the status of several mangrove species in this and surrounding areas. The harvesting of mangrove trees will have a direct and indirect effect on the other biotic and abiotic components of the mangrove ecosystems. The Department of Agriculture, Forestry and Fisheries (DAFF) (currently the Department of Environment, Forestry and Fisheries, or DEFF), through the National Forest Act no 84 of 1998, was directly responsible for the wellbeing of the national forests in South Africa. This also includes the mangroves as a specific national forest type. Several estuaries with mangroves occur within state forests. The harvesting and clear-felling of trees for timber and land conversion and associated waste of harvested mangrove poles and laths causes the many changes in mangrove forests. The aim of this study was to develop a framework for sustainable management of the mangroves for the Mtambalala forest management unit. This was done in relation to the requirements of the current legislation, policy guidelines and business plan of the management authority. The aim was pursued through two specific objectives. The first objective was to understand the variation in the composition in different parts of the Ntafufu Estuary, and the population dynamics of the main mangrove species, as basis for the development of sustainable resource use practices. Plots along transects were used to measure seedling to adult ratios, population structure and abiotic characteristics. The second objective was to understand the community use and requirements of mangrove wood from the Ntafufu Estuary. A participatory approach was followed to collect demographic data from each community, the requirements of mangrove poles and behaviour of harvesters. This was used to set up a framework to guide the sustainable management of mangroves based on the current status and stand dynamics of the mangrove communities. There are three mangrove species found in the Ntafufu Estuary: Bruguiera gymnorhiza (black mangrove) regenerates under the mangrove canopy and forms layers of even-aged stands. Avicennia marina (white mangrove) does not regenerate under a mangrove canopy i.e. the seedlings establish outside the canopy. A few Rhizophora mucronata (red mangrove) trees exist in the lower intertidal zone and form multiple stems on stilt roots. The mangrove associate, Hibiscus tiliaceus (wild cotton tree- mangrove associate), forms dense clusters of scrambling stems. Some of the Avicennia individuals show signs of mangrove bleeding and one individual of Bruguiera showed indications of disease. Mngazana and Ntafufu Estuaries had the highest tree densities of 2594 and 1402 trees ha-1 respectively, typical of riverine mangrove forest. The seedling to adult ratio was high in this study: 12.8 for B. gymnorrhiza and 24.3 for A. marina. Abandoned cut stems and laths were found during the study. Bark harvested for traditional medicine and cattle feeding in mangroves was also prevalent and expected to reduce the health of the forest. Mangroves in the Ntafufu Estuary are expanding, and the population structure of the three mangrove species showed a healthy balance between young and mature stems. The harvesting of Black mangrove (B. gymnorrhiza) poles for mainly roof construction is relatively limited. As part of the functions of the DAFF, the use of the mangrove resources has to be balanced with the resource availability from the mangroves, and this should result in a healthy functioning mangrove forest. Development of a participatory forest management (PFM) plan for a forest estate and its forest management unit, require an understanding of the mangrove communities and their dynamics, and the current level of resource use associated with the mangroves. The study recommended that a Mangrove Pole Harvesters Group is needed to practice silviculture through guidance of relevant governmental personnel.
- Full Text:
- Date Issued: 2020
- Authors: Mvakade, Monica , Rajkaran, Anusha
- Date: 2020
- Subjects: Estuarine ecology -- South Africa -- Eastern Cape , Aquatic ecology , Marine sciences , Fresh water
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/49215 , vital:41611
- Description: Traditionally, coastal communities along the Wild Coast in the Eastern Cape, used mangrove forests for their livelihoods and socio-economic needs. Mangroves along the Eastern Cape coast are threatened by both natural and human factors. The wood obtained from the mangroves by rural households was used as poles for house and fence construction, firewood, and traditional medicine. This is a threat to the status of several mangrove species in this and surrounding areas. The harvesting of mangrove trees will have a direct and indirect effect on the other biotic and abiotic components of the mangrove ecosystems. The Department of Agriculture, Forestry and Fisheries (DAFF) (currently the Department of Environment, Forestry and Fisheries, or DEFF), through the National Forest Act no 84 of 1998, was directly responsible for the wellbeing of the national forests in South Africa. This also includes the mangroves as a specific national forest type. Several estuaries with mangroves occur within state forests. The harvesting and clear-felling of trees for timber and land conversion and associated waste of harvested mangrove poles and laths causes the many changes in mangrove forests. The aim of this study was to develop a framework for sustainable management of the mangroves for the Mtambalala forest management unit. This was done in relation to the requirements of the current legislation, policy guidelines and business plan of the management authority. The aim was pursued through two specific objectives. The first objective was to understand the variation in the composition in different parts of the Ntafufu Estuary, and the population dynamics of the main mangrove species, as basis for the development of sustainable resource use practices. Plots along transects were used to measure seedling to adult ratios, population structure and abiotic characteristics. The second objective was to understand the community use and requirements of mangrove wood from the Ntafufu Estuary. A participatory approach was followed to collect demographic data from each community, the requirements of mangrove poles and behaviour of harvesters. This was used to set up a framework to guide the sustainable management of mangroves based on the current status and stand dynamics of the mangrove communities. There are three mangrove species found in the Ntafufu Estuary: Bruguiera gymnorhiza (black mangrove) regenerates under the mangrove canopy and forms layers of even-aged stands. Avicennia marina (white mangrove) does not regenerate under a mangrove canopy i.e. the seedlings establish outside the canopy. A few Rhizophora mucronata (red mangrove) trees exist in the lower intertidal zone and form multiple stems on stilt roots. The mangrove associate, Hibiscus tiliaceus (wild cotton tree- mangrove associate), forms dense clusters of scrambling stems. Some of the Avicennia individuals show signs of mangrove bleeding and one individual of Bruguiera showed indications of disease. Mngazana and Ntafufu Estuaries had the highest tree densities of 2594 and 1402 trees ha-1 respectively, typical of riverine mangrove forest. The seedling to adult ratio was high in this study: 12.8 for B. gymnorrhiza and 24.3 for A. marina. Abandoned cut stems and laths were found during the study. Bark harvested for traditional medicine and cattle feeding in mangroves was also prevalent and expected to reduce the health of the forest. Mangroves in the Ntafufu Estuary are expanding, and the population structure of the three mangrove species showed a healthy balance between young and mature stems. The harvesting of Black mangrove (B. gymnorrhiza) poles for mainly roof construction is relatively limited. As part of the functions of the DAFF, the use of the mangrove resources has to be balanced with the resource availability from the mangroves, and this should result in a healthy functioning mangrove forest. Development of a participatory forest management (PFM) plan for a forest estate and its forest management unit, require an understanding of the mangrove communities and their dynamics, and the current level of resource use associated with the mangroves. The study recommended that a Mangrove Pole Harvesters Group is needed to practice silviculture through guidance of relevant governmental personnel.
- Full Text:
- Date Issued: 2020
Upwelling on the Southeast Madagascan shelf: frequency, extent, and driving mechanisms
- Authors: Collins, Matthew
- Date: 2020
- Subjects: Upwelling (Oceanography) -- Madagascar , Oceanography -- Research -- Madagascar Marine biology -- Indian Ocean
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/49060 , vital:41597
- Description: The southeast coast of Madagascar, namely the region surrounding the town of Fort Dauphin, is an area that regularly experiences upwelling events. Two unique high frequency data sets, UTR data, and wind data retrieved from a nearby airport, are the primary data sets used to investigate the nature of the upwelling, i.e. the area that experiences upwelling, how often it occurs, and its mechanisms. To complement the in-situ data, several different satellite data sources are used alongside the in-situ data to study this relatively remote region, i.e. SST, chlorophyll-a concentration, current, and wind data. The results refine upon previous literature on the topic. The upwelling covers an area of 1o latitude by 1.3o longitude, and occurs on average 14 times per annum. The mechanism driving the upwelling can be attributed to both wind and current, with wind the more dominant driver along the south coast, and current more dominant along the east coast. Separate to the main objective of the thesis is the importance of in-situ data when analysing small scale events such as upwelling events. While the in-situ wind data were found to be unreliable due to its location within a valley, the UTR data were of vital importance in identifying and studying the finer details of the upwelling events. The satellite SST data, by comparison, were lacking in their ability to investigate the upwelling events. The satellite derived wind data, however, did prove useful when analysing the wind as a driver of the upwelling.
- Full Text:
- Date Issued: 2020
- Authors: Collins, Matthew
- Date: 2020
- Subjects: Upwelling (Oceanography) -- Madagascar , Oceanography -- Research -- Madagascar Marine biology -- Indian Ocean
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/49060 , vital:41597
- Description: The southeast coast of Madagascar, namely the region surrounding the town of Fort Dauphin, is an area that regularly experiences upwelling events. Two unique high frequency data sets, UTR data, and wind data retrieved from a nearby airport, are the primary data sets used to investigate the nature of the upwelling, i.e. the area that experiences upwelling, how often it occurs, and its mechanisms. To complement the in-situ data, several different satellite data sources are used alongside the in-situ data to study this relatively remote region, i.e. SST, chlorophyll-a concentration, current, and wind data. The results refine upon previous literature on the topic. The upwelling covers an area of 1o latitude by 1.3o longitude, and occurs on average 14 times per annum. The mechanism driving the upwelling can be attributed to both wind and current, with wind the more dominant driver along the south coast, and current more dominant along the east coast. Separate to the main objective of the thesis is the importance of in-situ data when analysing small scale events such as upwelling events. While the in-situ wind data were found to be unreliable due to its location within a valley, the UTR data were of vital importance in identifying and studying the finer details of the upwelling events. The satellite SST data, by comparison, were lacking in their ability to investigate the upwelling events. The satellite derived wind data, however, did prove useful when analysing the wind as a driver of the upwelling.
- Full Text:
- Date Issued: 2020
Using sound localization to gain depth perception for the visually impaired through sensory substitution
- Authors: De Klerk, James Carmichael
- Date: 2020
- Subjects: People with visual disabilities Directional hearing--Physiological aspects
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/50779 , vital:42674
- Description: The visually impaired do not have the visual ability to localize objects in three-dimensional space, rather, they rely on their other senses to gain depth perception. Sensory substitution is the concept of substituting one sense for another, normally substituting an impaired sense with a functioning sense. Visual-to-auditory sensory substitution substitutes an impaired visual sense with a functioning auditory sense. This research aimed to investigate and develop techniques for visualto-auditory sensory substitution – using sound localization as a sensory substitution for depth perception. The research started by investigating the characteristics of human audition with a particular focus on how humans localize sounds. It then looked at existing visual-to-auditory sensory substitution systems and the techniques they used. From the existing systems, a system known as MeloSee was chosen as a baseline for developing and evaluating further sensory substitution prototypes. The baseline prototype (𝑃0) was then implemented and a preliminary study performed. Based on the knowledge gained from the preliminary study, baseline implementation and the background research, a set of improvement recommendations were generated. The next iteration – Prototype 1 (𝑃1) – was then developed based on the recommendations. A comparative study between 𝑃0 and 𝑃1 was then performed. Based on the study, another set of improvement recommendations were generated. From the recommendations, a final prototype was developed – Prototype 2 (𝑃2). The last comparative study was then performed between 𝑃0 and 𝑃2, with a third set of recommendations being generated as a result. From the studies it was found that participants using 𝑃0 were able to identify when they were approaching large objects such as walls. 𝑃1 built on that, improving the ability to identify the quadrant of a nearby isolated object. 𝑃2 built on 𝑃0 and 𝑃1, achieving similar results to 𝑃1 for identifying the quadrant of nearby isolated objects, and improving on 𝑃0 and 𝑃1 with regard to depth discrimination – especially for navigation tasks where there were no obstacles. Based on the three sets of recommendations and what was learnt over the course of the research, a set of visual-to-auditory sensory substitution techniques were presented. The techniques aim to be useful for implementing visual-to-auditory sensory substitution systems, which would provide the visually impaired with the visual ability to localize objects in three-dimensional space through sound.
- Full Text:
- Date Issued: 2020
- Authors: De Klerk, James Carmichael
- Date: 2020
- Subjects: People with visual disabilities Directional hearing--Physiological aspects
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/50779 , vital:42674
- Description: The visually impaired do not have the visual ability to localize objects in three-dimensional space, rather, they rely on their other senses to gain depth perception. Sensory substitution is the concept of substituting one sense for another, normally substituting an impaired sense with a functioning sense. Visual-to-auditory sensory substitution substitutes an impaired visual sense with a functioning auditory sense. This research aimed to investigate and develop techniques for visualto-auditory sensory substitution – using sound localization as a sensory substitution for depth perception. The research started by investigating the characteristics of human audition with a particular focus on how humans localize sounds. It then looked at existing visual-to-auditory sensory substitution systems and the techniques they used. From the existing systems, a system known as MeloSee was chosen as a baseline for developing and evaluating further sensory substitution prototypes. The baseline prototype (𝑃0) was then implemented and a preliminary study performed. Based on the knowledge gained from the preliminary study, baseline implementation and the background research, a set of improvement recommendations were generated. The next iteration – Prototype 1 (𝑃1) – was then developed based on the recommendations. A comparative study between 𝑃0 and 𝑃1 was then performed. Based on the study, another set of improvement recommendations were generated. From the recommendations, a final prototype was developed – Prototype 2 (𝑃2). The last comparative study was then performed between 𝑃0 and 𝑃2, with a third set of recommendations being generated as a result. From the studies it was found that participants using 𝑃0 were able to identify when they were approaching large objects such as walls. 𝑃1 built on that, improving the ability to identify the quadrant of a nearby isolated object. 𝑃2 built on 𝑃0 and 𝑃1, achieving similar results to 𝑃1 for identifying the quadrant of nearby isolated objects, and improving on 𝑃0 and 𝑃1 with regard to depth discrimination – especially for navigation tasks where there were no obstacles. Based on the three sets of recommendations and what was learnt over the course of the research, a set of visual-to-auditory sensory substitution techniques were presented. The techniques aim to be useful for implementing visual-to-auditory sensory substitution systems, which would provide the visually impaired with the visual ability to localize objects in three-dimensional space through sound.
- Full Text:
- Date Issued: 2020
Water balance processes in indigenous and introduced spaces three production system in the Southern Cape region of South Africa
- Authors: Mapeto, Tatenda
- Date: 2020
- Subjects: Hydrology -- Research -- South Africa , Groundwater -- Research -- South Africa Trees -- Breeding Sustainable forestry
- Language: English
- Type: Thesis , Doctoral , DPhil
- Identifier: http://hdl.handle.net/10948/49325 , vital:41621
- Description: Forests and trees are important modulators of water flow and their position in the landscape impacts the spatial and temporal availability of ground water and surface water resources. As such, the global extent of natural forests is linked to the availability of water. In the circumstances where tree production systems have been introduced and are managed in landscapes previously populated with other forms of vegetation, it is necessary to understand their hydrological impact so as to ensure that the benefits derived from these tree systems do not outweigh the costs of water availability for sustaining human and environmental water requirements.
- Full Text: false
- Date Issued: 2020
- Authors: Mapeto, Tatenda
- Date: 2020
- Subjects: Hydrology -- Research -- South Africa , Groundwater -- Research -- South Africa Trees -- Breeding Sustainable forestry
- Language: English
- Type: Thesis , Doctoral , DPhil
- Identifier: http://hdl.handle.net/10948/49325 , vital:41621
- Description: Forests and trees are important modulators of water flow and their position in the landscape impacts the spatial and temporal availability of ground water and surface water resources. As such, the global extent of natural forests is linked to the availability of water. In the circumstances where tree production systems have been introduced and are managed in landscapes previously populated with other forms of vegetation, it is necessary to understand their hydrological impact so as to ensure that the benefits derived from these tree systems do not outweigh the costs of water availability for sustaining human and environmental water requirements.
- Full Text: false
- Date Issued: 2020
Wind damage impacts on Eucalyptus species performance in South Africa
- Authors: Hechter, Heinrich
- Date: 2020
- Subjects: Commercial forests -- South Africa
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/48368 , vital:40853
- Description: Commercial forestry plantations in South Africa play an important role in the economy of the country, contributing 1.2% towards the gross domestic product, as well as with job creation in rural communities. Currently plantation forests cover 1% (1.2 million hectares) of the South African land surface, of which ±75 000 hectares (±43 000 hectares to Eucalyptus species) are re-established each year. Different timber companies afforest areas with different species to satisfy specific management objectives, with the two main objectives being for either the production of saw-timber or pulpwood. Species of the Eucalyptus genera are preferred as they are fast growing and have desirable wood and pulping properties. However, their productive potential can be negatively impacted by numerous abiotic and biotic risk factors. Of the various risk factors that forest stands are exposed to, wind and wind-related tree damage has received very little attention, both nationally and internationally. Wind-related tree damage may result in wind-throw (uprooting or stem breakage), or trees that tilt (lean off vertical). Trees that are tilting can recover to an upright position, but are likely to retain some level of stem sinuosity or butt-sweep. Although strong winds increase the risk of tree damage, a number of other factors can also act to predispose trees to wind-related damage. These include choice of planting stock (genotype and type of plant stock), planting practices (including soil cultivation), site factors (wind exposure, rainfall, soil texture and soil fertility) or excessive weed competition. Most of the literature dealing with wind-related tree damage has focused on trees growing in natural forests, whereas trees in natural forests differ from forest grown plantation trees in terms of above- and below-ground morphological differences, as well as the stability factors of the trees. There is a need to not only understand the impacts of severe (catastrophic) wind events on mature trees, but also to test management strategies that prevent, or minimize damage prior to any severe wind events. Two existing eucalypt trials were used to provide information on the influence of selected re-establishment silvicultural practices on short-term pulpwood and long-term saw-timber survival, growth and uniformity when influenced by catastrophic wind events, within South Africa.
- Full Text:
- Date Issued: 2020
- Authors: Hechter, Heinrich
- Date: 2020
- Subjects: Commercial forests -- South Africa
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/48368 , vital:40853
- Description: Commercial forestry plantations in South Africa play an important role in the economy of the country, contributing 1.2% towards the gross domestic product, as well as with job creation in rural communities. Currently plantation forests cover 1% (1.2 million hectares) of the South African land surface, of which ±75 000 hectares (±43 000 hectares to Eucalyptus species) are re-established each year. Different timber companies afforest areas with different species to satisfy specific management objectives, with the two main objectives being for either the production of saw-timber or pulpwood. Species of the Eucalyptus genera are preferred as they are fast growing and have desirable wood and pulping properties. However, their productive potential can be negatively impacted by numerous abiotic and biotic risk factors. Of the various risk factors that forest stands are exposed to, wind and wind-related tree damage has received very little attention, both nationally and internationally. Wind-related tree damage may result in wind-throw (uprooting or stem breakage), or trees that tilt (lean off vertical). Trees that are tilting can recover to an upright position, but are likely to retain some level of stem sinuosity or butt-sweep. Although strong winds increase the risk of tree damage, a number of other factors can also act to predispose trees to wind-related damage. These include choice of planting stock (genotype and type of plant stock), planting practices (including soil cultivation), site factors (wind exposure, rainfall, soil texture and soil fertility) or excessive weed competition. Most of the literature dealing with wind-related tree damage has focused on trees growing in natural forests, whereas trees in natural forests differ from forest grown plantation trees in terms of above- and below-ground morphological differences, as well as the stability factors of the trees. There is a need to not only understand the impacts of severe (catastrophic) wind events on mature trees, but also to test management strategies that prevent, or minimize damage prior to any severe wind events. Two existing eucalypt trials were used to provide information on the influence of selected re-establishment silvicultural practices on short-term pulpwood and long-term saw-timber survival, growth and uniformity when influenced by catastrophic wind events, within South Africa.
- Full Text:
- Date Issued: 2020
A blended learning toolkit that accommodates multiple learning styles
- Authors: Mills, Steven Christopher
- Date: 2019
- Subjects: Blended learning , Learning strategies Learning
- Language: English
- Type: Thesis , Masters , MCom
- Identifier: http://hdl.handle.net/10948/30452 , vital:30945
- Description: The purpose of this study was to identify how blended learning can be designed and incorporated to accommodate multiple learning styles within modules in the Department of Computing Sciences. A design theory was created through an analysis of literature and exploration into the backgrounds of students and lecturers within the Department of Computing Sciences. The design theory is: Blended learning can be a useful approach to accommodate multiple learning styles. Guidelines, and by extension a toolkit, facilitate the development of blended learning and provide effective tools to enable lecturers to successfully incorporate blended learning into their modules. Design-Based Research (DBR) was followed in this study, using a mixed-methods and iterative approach to determine the accuracy of the design theory. For the first iteration, the toolkit was implemented in two modules within the Department of Computing Sciences and for the second iteration, four modules. DBR produces a theoretical contribution and a practical artefact. The most important theoretical contributions are the design theory and guidelines for incorporating blended learning that accommodates multiple learning styles. The practical artefacts are the toolkit and tools therein. The toolkit, which was accessed via a website, guides lecturers through the process of incorporating blended learning that accommodates multiple learning styles and provides them with the necessary tools to do so. The design theory was proven in the evaluation that used a questionnaire to understand the lecturers’ experiences regarding the toolkit and the design theory. Therefore, the guidelines for applying blended learning is a useful approach to address multiple learning styles.
- Full Text:
- Date Issued: 2019
- Authors: Mills, Steven Christopher
- Date: 2019
- Subjects: Blended learning , Learning strategies Learning
- Language: English
- Type: Thesis , Masters , MCom
- Identifier: http://hdl.handle.net/10948/30452 , vital:30945
- Description: The purpose of this study was to identify how blended learning can be designed and incorporated to accommodate multiple learning styles within modules in the Department of Computing Sciences. A design theory was created through an analysis of literature and exploration into the backgrounds of students and lecturers within the Department of Computing Sciences. The design theory is: Blended learning can be a useful approach to accommodate multiple learning styles. Guidelines, and by extension a toolkit, facilitate the development of blended learning and provide effective tools to enable lecturers to successfully incorporate blended learning into their modules. Design-Based Research (DBR) was followed in this study, using a mixed-methods and iterative approach to determine the accuracy of the design theory. For the first iteration, the toolkit was implemented in two modules within the Department of Computing Sciences and for the second iteration, four modules. DBR produces a theoretical contribution and a practical artefact. The most important theoretical contributions are the design theory and guidelines for incorporating blended learning that accommodates multiple learning styles. The practical artefacts are the toolkit and tools therein. The toolkit, which was accessed via a website, guides lecturers through the process of incorporating blended learning that accommodates multiple learning styles and provides them with the necessary tools to do so. The design theory was proven in the evaluation that used a questionnaire to understand the lecturers’ experiences regarding the toolkit and the design theory. Therefore, the guidelines for applying blended learning is a useful approach to address multiple learning styles.
- Full Text:
- Date Issued: 2019
A clip composition case study of communal wool production in the Eastern Cape, South Africa
- Authors: Dreyer, Susanna Maria
- Date: 2019
- Subjects: Wool industry -- South Africa -- Eastern Cape , Wool -- South Africa -- Eastern Cape Sheep industry -- South Africa -- Eastern Cape
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/39480 , vital:35265
- Description: Previous studies conducted on wool production in the communal areas of the Eastern Cape have concluded future research should focus on whether the observed differences in average wool price received between shearing sheds are due to differences in clip preparation (i.e. classing) or differences in the quality of the wool being produced (i.e. genetics).The objective of this study was to establish whether there are indeed geographical differences in the average wool price received by communal wool producers in the Eastern Cape and, if so, whether this difference in earnings is attributable to differences in the quality of wool produced and/or clip preparation. The current research explored eight years of wool production records of communal wool farmers (2009/10 to 2016/17) obtained from BKB, the largest wool broker in South Africa. Statistical analyses included descriptive statistics, analysis of variance, parametric tests (i.e. t-test) and multiple regression analysis. The conclusions drawn from this study are: that there are indeed statistically significant differences in the average wool price received between the different geographic areas of the Eastern Cape communal regions, and that the primary determining factor is wool quality, while clip preparation also plays an important but secondary role. The study identified specific geographic areas where developmental interventions are most critical, with specific emphasis on improving the genetic composition of the wool sheep flocks. Specific strategies are recommended, namely:Removing inferior-quality rams and replacing them with superior-quality rams to prevent any inferior nondescript rams from breeding and producing progeny; Replacing the inferior-quality nondescript ewes in specific areas with superior-quality ewes, and Improving the survival rate of lambs and thereby contributing towards expediting the rate of genetic improvement of the communal wool sheep flocks.
- Full Text:
- Date Issued: 2019
- Authors: Dreyer, Susanna Maria
- Date: 2019
- Subjects: Wool industry -- South Africa -- Eastern Cape , Wool -- South Africa -- Eastern Cape Sheep industry -- South Africa -- Eastern Cape
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/39480 , vital:35265
- Description: Previous studies conducted on wool production in the communal areas of the Eastern Cape have concluded future research should focus on whether the observed differences in average wool price received between shearing sheds are due to differences in clip preparation (i.e. classing) or differences in the quality of the wool being produced (i.e. genetics).The objective of this study was to establish whether there are indeed geographical differences in the average wool price received by communal wool producers in the Eastern Cape and, if so, whether this difference in earnings is attributable to differences in the quality of wool produced and/or clip preparation. The current research explored eight years of wool production records of communal wool farmers (2009/10 to 2016/17) obtained from BKB, the largest wool broker in South Africa. Statistical analyses included descriptive statistics, analysis of variance, parametric tests (i.e. t-test) and multiple regression analysis. The conclusions drawn from this study are: that there are indeed statistically significant differences in the average wool price received between the different geographic areas of the Eastern Cape communal regions, and that the primary determining factor is wool quality, while clip preparation also plays an important but secondary role. The study identified specific geographic areas where developmental interventions are most critical, with specific emphasis on improving the genetic composition of the wool sheep flocks. Specific strategies are recommended, namely:Removing inferior-quality rams and replacing them with superior-quality rams to prevent any inferior nondescript rams from breeding and producing progeny; Replacing the inferior-quality nondescript ewes in specific areas with superior-quality ewes, and Improving the survival rate of lambs and thereby contributing towards expediting the rate of genetic improvement of the communal wool sheep flocks.
- Full Text:
- Date Issued: 2019
A comparative study of artificial neural networks and physics models as simulators in evolutionary robotics
- Pretorius, Christiaan Johannes
- Authors: Pretorius, Christiaan Johannes
- Date: 2019
- Subjects: Neural networks (Computer science)
- Language: English
- Type: Thesis , Doctoral , PhD
- Identifier: http://hdl.handle.net/10948/30789 , vital:31131
- Description: The Evolutionary Robotics (ER) process is a technique that applies evolutionary optimization algorithms to the task of automatically developing, or evolving, robotic control programs. These control programs, or simply controllers, are evolved in order to allow a robot to perform a required task. During the ER process, use is often made of robotic simulators to evaluate the performance of candidate controllers that are produced in the course of the controller evolution process. Such simulators accelerate and otherwise simplify the controller evolution process, as opposed to the more arduous process of evaluating controllers in the real world without use of simulation. To date, the vast majority of simulators that have been applied in ER are physics- based models which are constructed by taking into account the underlying physics governing the operation of the robotic system in question. An alternative approach to simulator implementation in ER is the usage of Artificial Neural Networks (ANNs) as simulators in the ER process. Such simulators are referred to as Simulator Neural Networks (SNNs). Previous studies have indicated that SNNs can successfully be used as an alter- native to physics-based simulators in the ER process on various robotic platforms. At the commencement of the current study it was not, however, known how this relatively new method of simulation would compare to traditional physics-based simulation approaches in ER. The study presented in this thesis thus endeavoured to quantitatively compare SNNs and physics-based models as simulators in the ER process. In order to con- duct this comparative study, both SNNs and physics simulators were constructed for the modelling of three different robotic platforms: a differentially-steered robot, a wheeled inverted pendulum robot and a hexapod robot. Each of these two types of simulation was then used in simulation-based evolution processes to evolve con- trollers for each robotic platform. During these controller evolution processes, the SNNs and physics models were compared in terms of their accuracy in making pre- dictions of robotic behaviour, their computational efficiency in arriving at these predictions, the human effort required to construct each simulator and, most im- portantly, the real-world performance of controllers evolved by making use of each simulator. The results obtained in this study illustrated experimentally that SNNs were, in the majority of cases, able to make more accurate predictions than the physics- based models and these SNNs were arguably simpler to construct than the physics simulators. Additionally, SNNs were also shown to be a computationally efficient alternative to physics-based simulators in ER and, again in the majority of cases, these SNNs were able to produce controllers which outperformed those evolved in the physics-based simulators, when these controllers were uploaded to the real-world robots. The results of this thesis thus suggest that SNNs are a viable alternative to more commonly-used physics simulators in ER and further investigation of the potential of this simulation technique appears warranted.
- Full Text:
- Date Issued: 2019
- Authors: Pretorius, Christiaan Johannes
- Date: 2019
- Subjects: Neural networks (Computer science)
- Language: English
- Type: Thesis , Doctoral , PhD
- Identifier: http://hdl.handle.net/10948/30789 , vital:31131
- Description: The Evolutionary Robotics (ER) process is a technique that applies evolutionary optimization algorithms to the task of automatically developing, or evolving, robotic control programs. These control programs, or simply controllers, are evolved in order to allow a robot to perform a required task. During the ER process, use is often made of robotic simulators to evaluate the performance of candidate controllers that are produced in the course of the controller evolution process. Such simulators accelerate and otherwise simplify the controller evolution process, as opposed to the more arduous process of evaluating controllers in the real world without use of simulation. To date, the vast majority of simulators that have been applied in ER are physics- based models which are constructed by taking into account the underlying physics governing the operation of the robotic system in question. An alternative approach to simulator implementation in ER is the usage of Artificial Neural Networks (ANNs) as simulators in the ER process. Such simulators are referred to as Simulator Neural Networks (SNNs). Previous studies have indicated that SNNs can successfully be used as an alter- native to physics-based simulators in the ER process on various robotic platforms. At the commencement of the current study it was not, however, known how this relatively new method of simulation would compare to traditional physics-based simulation approaches in ER. The study presented in this thesis thus endeavoured to quantitatively compare SNNs and physics-based models as simulators in the ER process. In order to con- duct this comparative study, both SNNs and physics simulators were constructed for the modelling of three different robotic platforms: a differentially-steered robot, a wheeled inverted pendulum robot and a hexapod robot. Each of these two types of simulation was then used in simulation-based evolution processes to evolve con- trollers for each robotic platform. During these controller evolution processes, the SNNs and physics models were compared in terms of their accuracy in making pre- dictions of robotic behaviour, their computational efficiency in arriving at these predictions, the human effort required to construct each simulator and, most im- portantly, the real-world performance of controllers evolved by making use of each simulator. The results obtained in this study illustrated experimentally that SNNs were, in the majority of cases, able to make more accurate predictions than the physics- based models and these SNNs were arguably simpler to construct than the physics simulators. Additionally, SNNs were also shown to be a computationally efficient alternative to physics-based simulators in ER and, again in the majority of cases, these SNNs were able to produce controllers which outperformed those evolved in the physics-based simulators, when these controllers were uploaded to the real-world robots. The results of this thesis thus suggest that SNNs are a viable alternative to more commonly-used physics simulators in ER and further investigation of the potential of this simulation technique appears warranted.
- Full Text:
- Date Issued: 2019
A comparison of different strategies to control pests and diseases in Brassica spp. production in the Western Cape
- Van Niekerk, Abraham Johannes
- Authors: Van Niekerk, Abraham Johannes
- Date: 2019
- Subjects: Pesticides , Pests -- Control Agricultural pests , Plant diseases
- Language: English
- Type: Thesis , Masters , MTech
- Identifier: http://hdl.handle.net/10948/44122 , vital:37113
- Description: Brassica spp. are cultivated all over the world, commercial species include: cabbage, broccoli, kale, kohlrabi and turnip. In this study the focus was on broccoli (Brassica oleracea) production in the Western Cape province of South Africa and its economically important pests and diseases: sugar beet cyst nematode (Heterodera schachtii), diamondback moth (Plutella xylostella), white blister (Albugo candida) and clubroot (Plasmodiophora brassicae), and the different methods to control these pests and diseases. The control methods focused on in this study included a commercial chemical control programme, a biological control programme and a holistic approach. Other factors were bio-fumigation and chemical fumigation and different crop rotation practices including rotation crops versus no rotation crops. The experimental design was a strip split plot design, with different pest and disease management strategies as the main plot treatment and fumigation and rotation treatment combinations arranged in strips across the main plot treatments. The main plot design was a randomized complete block with four programmes (Control, Holistic, Chemical and Biological) replicated four times and laid out in a Randomised Complete Block Design (RCBD). The treatment design of the strip plot factors was a 2x2 factorial with two fumigations (fumigated chemically and fumigated biologically) and two rotations (crop rotated and monoculture) randomly allocated across main plot treatments. Each experimental unit consisted of 40 plants. Plants were evaluated weekly for the incidence of white blister and diamondback moth. Incidences of clubroot and white blister infection of heads of broccoli were recorded 78 days after planting. Baseline soil samples were analysed to establish the soil chemical properties. Post-trial soil samples were also analysed to investigate the effect of the different practices and programmes on the soil chemical properties. Nematodes were extracted pre-trial and post-trial, and the effect of fumigation and crop rotation on plant parasitic nematodes and the nematode population diversity investigated. In the post-trial soil chemical analysis, a significantly higher concentration of Na was recorded for the biological programme when crop rotation was included compared to the no rotation treatment. The concentration of K was also significantly higher in the no rotation compared to the rotation treatment. The nematode results showed very high numbers of bacterial feeders in all the samples. Overall nematode diversity was lacking and showed very few fungal feeders, omnivores and predators. The nematode indices for all of the samples showed that nematodes were highly enriched and unstructured. Of the plant parasitic nematodes, only Heterodera spp. were obtained in the pre-trial analysis, and incidences of these nematodes were lower at the end of the trial. Low numbers of other plant parasitic nematodes viz. Pratylenchus, Paratrichodorus and Tylenchorhynchus were reported for the post-trial analysis of the soil. Because of the low numbers of plant parasitic nematodes, it was not possible to analyse the data statistically. With regard to diamondback moth and the fungal diseases, crop rotation and fumigation did not significantly affect the incidence of white blister and diamondback moth. A significant “days after planting by control programme” interaction was reported for the incidence of white blister on foliage and the incidence of diamondback moth. All three control programmes significantly reduced the incidence of diamondback moth with the chemical programme being significantly more effective than the other two programmes. All three programmes also significantly reduced the incidence of white blister on foliage and the holistic and biological programmes significantly reduced the incidence of white blister on broccoli heads with the holistic programme being significantly more effective than the biological programme. There was no clubroot infection in the trial for any of the treatments. Results of this study showed that it is possible to manage diseases and pests of broccoli using a holistic approach. However, long term trials are needed to confirm the results obtained in this study.
- Full Text:
- Date Issued: 2019
- Authors: Van Niekerk, Abraham Johannes
- Date: 2019
- Subjects: Pesticides , Pests -- Control Agricultural pests , Plant diseases
- Language: English
- Type: Thesis , Masters , MTech
- Identifier: http://hdl.handle.net/10948/44122 , vital:37113
- Description: Brassica spp. are cultivated all over the world, commercial species include: cabbage, broccoli, kale, kohlrabi and turnip. In this study the focus was on broccoli (Brassica oleracea) production in the Western Cape province of South Africa and its economically important pests and diseases: sugar beet cyst nematode (Heterodera schachtii), diamondback moth (Plutella xylostella), white blister (Albugo candida) and clubroot (Plasmodiophora brassicae), and the different methods to control these pests and diseases. The control methods focused on in this study included a commercial chemical control programme, a biological control programme and a holistic approach. Other factors were bio-fumigation and chemical fumigation and different crop rotation practices including rotation crops versus no rotation crops. The experimental design was a strip split plot design, with different pest and disease management strategies as the main plot treatment and fumigation and rotation treatment combinations arranged in strips across the main plot treatments. The main plot design was a randomized complete block with four programmes (Control, Holistic, Chemical and Biological) replicated four times and laid out in a Randomised Complete Block Design (RCBD). The treatment design of the strip plot factors was a 2x2 factorial with two fumigations (fumigated chemically and fumigated biologically) and two rotations (crop rotated and monoculture) randomly allocated across main plot treatments. Each experimental unit consisted of 40 plants. Plants were evaluated weekly for the incidence of white blister and diamondback moth. Incidences of clubroot and white blister infection of heads of broccoli were recorded 78 days after planting. Baseline soil samples were analysed to establish the soil chemical properties. Post-trial soil samples were also analysed to investigate the effect of the different practices and programmes on the soil chemical properties. Nematodes were extracted pre-trial and post-trial, and the effect of fumigation and crop rotation on plant parasitic nematodes and the nematode population diversity investigated. In the post-trial soil chemical analysis, a significantly higher concentration of Na was recorded for the biological programme when crop rotation was included compared to the no rotation treatment. The concentration of K was also significantly higher in the no rotation compared to the rotation treatment. The nematode results showed very high numbers of bacterial feeders in all the samples. Overall nematode diversity was lacking and showed very few fungal feeders, omnivores and predators. The nematode indices for all of the samples showed that nematodes were highly enriched and unstructured. Of the plant parasitic nematodes, only Heterodera spp. were obtained in the pre-trial analysis, and incidences of these nematodes were lower at the end of the trial. Low numbers of other plant parasitic nematodes viz. Pratylenchus, Paratrichodorus and Tylenchorhynchus were reported for the post-trial analysis of the soil. Because of the low numbers of plant parasitic nematodes, it was not possible to analyse the data statistically. With regard to diamondback moth and the fungal diseases, crop rotation and fumigation did not significantly affect the incidence of white blister and diamondback moth. A significant “days after planting by control programme” interaction was reported for the incidence of white blister on foliage and the incidence of diamondback moth. All three control programmes significantly reduced the incidence of diamondback moth with the chemical programme being significantly more effective than the other two programmes. All three programmes also significantly reduced the incidence of white blister on foliage and the holistic and biological programmes significantly reduced the incidence of white blister on broccoli heads with the holistic programme being significantly more effective than the biological programme. There was no clubroot infection in the trial for any of the treatments. Results of this study showed that it is possible to manage diseases and pests of broccoli using a holistic approach. However, long term trials are needed to confirm the results obtained in this study.
- Full Text:
- Date Issued: 2019
A crowdsensing method for water resource monitoring in smart communities
- Authors: Banda, Clara Mloza
- Date: 2019
- Subjects: Data collection platforms
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/42449 , vital:36658
- Description: Crowdsensing aims to empower a large group of individuals to collect large amounts of data using their mobile devices, with the goal of sharing the collected data. Existing crowdsensing studies do not consider all the activities and methods of the crowdsensing process and the key success factors related to the process. Nor do they investigate the profile and behaviour of potential participants. The aim of this study was to design a crowdsensing method for water resource monitoring in smart communities. This study opted for an exploratory study using the Engaged Scholarship approach, which allows the study of complex real-world problems based on the different perspectives of key stakeholders. The proposed Crowdsensing Method considers the social, technical and programme design components. The study proposes a programme design for the Crowdsensing Methodwhich is crowdsensing ReferenceFrameworkthat includes Crowdsensing Processwith key success factors and guidelines that should be considered in each phase of the process. The method also uses the Theory of Planned Behaviour (TPB) to investigate citizens’intention to participate in crowdsensing for water resource monitoring and explores their attitudes, norms and perceived behavioural control on these intentions. Understanding the profiles of potential participants can assist with designing crowdsensing systems with appropriate incentive mechanisms to achieve adequate user participation and good service quality. A survey was conducted to validate the theoretical TB model in a real-world context. Regression and correlation analyses demonstrated that the attitudes, norms and perceived behavioural control can be used to predict participants’ intention to participate in crowdsensing for water resource monitoring. The survey results assisted with the development of an Incentive Mechanism as part of the Crowdsensing Method. This mechanism incorporates recruitment and incentive policies, as well as guidelines derived from the literature review and extant system analysis. The policies, called the OverSensepolicies, provide guidance for recruitment and rewarding of participants using the popular Stackelberg technique. The policies were evaluated using simulation experiments with a data set provided by the case study, the Nelson Mandela Bay Municipality. The results of the simulation experiments illustrated that the OverSenserecruitmentpolicycan reduce the computing resources required for the recruitment of participants and that the recruitment policy performs better than random or naïve recruitment policies. The proposed Crowdsensing Method was evaluated using an ecosystem of success factors for mobile-based interventions identified in the literature and the Crowdsensing Method adhered to a majority (90%) of the success factors. This study also contributes information systems design theory by proposing several sets of guidelines for crowdsensing projects and the development of crowdsensing systems. This study fulfils an identified need to study the applicability of crowdsensing for water resource monitoring and explores how a crowdsensing method can create a smart community.
- Full Text:
- Date Issued: 2019
- Authors: Banda, Clara Mloza
- Date: 2019
- Subjects: Data collection platforms
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/42449 , vital:36658
- Description: Crowdsensing aims to empower a large group of individuals to collect large amounts of data using their mobile devices, with the goal of sharing the collected data. Existing crowdsensing studies do not consider all the activities and methods of the crowdsensing process and the key success factors related to the process. Nor do they investigate the profile and behaviour of potential participants. The aim of this study was to design a crowdsensing method for water resource monitoring in smart communities. This study opted for an exploratory study using the Engaged Scholarship approach, which allows the study of complex real-world problems based on the different perspectives of key stakeholders. The proposed Crowdsensing Method considers the social, technical and programme design components. The study proposes a programme design for the Crowdsensing Methodwhich is crowdsensing ReferenceFrameworkthat includes Crowdsensing Processwith key success factors and guidelines that should be considered in each phase of the process. The method also uses the Theory of Planned Behaviour (TPB) to investigate citizens’intention to participate in crowdsensing for water resource monitoring and explores their attitudes, norms and perceived behavioural control on these intentions. Understanding the profiles of potential participants can assist with designing crowdsensing systems with appropriate incentive mechanisms to achieve adequate user participation and good service quality. A survey was conducted to validate the theoretical TB model in a real-world context. Regression and correlation analyses demonstrated that the attitudes, norms and perceived behavioural control can be used to predict participants’ intention to participate in crowdsensing for water resource monitoring. The survey results assisted with the development of an Incentive Mechanism as part of the Crowdsensing Method. This mechanism incorporates recruitment and incentive policies, as well as guidelines derived from the literature review and extant system analysis. The policies, called the OverSensepolicies, provide guidance for recruitment and rewarding of participants using the popular Stackelberg technique. The policies were evaluated using simulation experiments with a data set provided by the case study, the Nelson Mandela Bay Municipality. The results of the simulation experiments illustrated that the OverSenserecruitmentpolicycan reduce the computing resources required for the recruitment of participants and that the recruitment policy performs better than random or naïve recruitment policies. The proposed Crowdsensing Method was evaluated using an ecosystem of success factors for mobile-based interventions identified in the literature and the Crowdsensing Method adhered to a majority (90%) of the success factors. This study also contributes information systems design theory by proposing several sets of guidelines for crowdsensing projects and the development of crowdsensing systems. This study fulfils an identified need to study the applicability of crowdsensing for water resource monitoring and explores how a crowdsensing method can create a smart community.
- Full Text:
- Date Issued: 2019
A deep learning approach to classifying tyres using sidewall images
- Authors: Gifford, Dean
- Date: 2019
- Subjects: Image processing -- Digital techniques , Image processing Computer science
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/39720 , vital:35351
- Description: End of Life Tyres (ELT's) pose a potential health and environmental risk when dumped in illegal stockpiles. For recycling to be considered feasible, a profitable business opportunity needs to be created. One method of making the recycling process of tyres more profitable is by understanding the compounds found within each tyre. This study aims at classifying these tyres in order to achieve this knowledge. A literature review was done to investigate neural networks, convolutional neural networks as well as existing deep learning architectures for image classification. A deep learning approach was applied in order to classify the logos of tyres as these approaches have proved their success in both image classification and more specifically logo classification. Although tyre classification has been implemented in the past, a deep learning approach has not been applied and the logo has not been the classifying element in any other studies. The main difference of this study compared to previous research surrounding deep learning and logo classification is the properties of the tyre logo. Logos on tyres are very similar in colour as they are purely formed in rubber and very seldom have any colour to them. Additionally, the embossed logos can contain variation among same branded tyres due to small inconsistencies in the moulds. The implementation of this deep learning solution saw multiple convolutional neural networks implemented. Some of these architectures were also implemented using transferred learning. The metrics obtained as outputs from training and testing the architectures were the accuracy, precision, recall, and F1-score. These metrics were compared in conjunction with the confusion matrix produced from testing. To ensure that variance was accounted for in the experiments, the k-fold cross-validation technique was adopted. The results of this study identified that one convolutional neural network model, MobileNet, was particularly well suited for the context of classifying logos on tyre sidewalls. The MobileNet architecture had the highest performance metrics for both training from scratch (96.7% accuracy) and transferred learning (98.8% accuracy). Three other models performed particularly well when trained from scratch, these were a modification of the LeNet architecture, ResNet50 and InceptionV3. The transferred learning results were also impressive with four out of the 5 models achieving an accuracy above 94%. Interestingly, the ResNet50 architecture failed to train when transferred learning was applied. Contrasting to this, the two models VGG16 and VGG19 failed to train when trained from scratch but performed equally as well as the other models when transferred learning was implemented. This indicates that although transferred learning can improve the performance of models, it is highly dependent on the task as well as the model. Overall the results obtained proved that a deep learning approach could be applied in order to classify tyres accurately.
- Full Text:
- Date Issued: 2019
- Authors: Gifford, Dean
- Date: 2019
- Subjects: Image processing -- Digital techniques , Image processing Computer science
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/39720 , vital:35351
- Description: End of Life Tyres (ELT's) pose a potential health and environmental risk when dumped in illegal stockpiles. For recycling to be considered feasible, a profitable business opportunity needs to be created. One method of making the recycling process of tyres more profitable is by understanding the compounds found within each tyre. This study aims at classifying these tyres in order to achieve this knowledge. A literature review was done to investigate neural networks, convolutional neural networks as well as existing deep learning architectures for image classification. A deep learning approach was applied in order to classify the logos of tyres as these approaches have proved their success in both image classification and more specifically logo classification. Although tyre classification has been implemented in the past, a deep learning approach has not been applied and the logo has not been the classifying element in any other studies. The main difference of this study compared to previous research surrounding deep learning and logo classification is the properties of the tyre logo. Logos on tyres are very similar in colour as they are purely formed in rubber and very seldom have any colour to them. Additionally, the embossed logos can contain variation among same branded tyres due to small inconsistencies in the moulds. The implementation of this deep learning solution saw multiple convolutional neural networks implemented. Some of these architectures were also implemented using transferred learning. The metrics obtained as outputs from training and testing the architectures were the accuracy, precision, recall, and F1-score. These metrics were compared in conjunction with the confusion matrix produced from testing. To ensure that variance was accounted for in the experiments, the k-fold cross-validation technique was adopted. The results of this study identified that one convolutional neural network model, MobileNet, was particularly well suited for the context of classifying logos on tyre sidewalls. The MobileNet architecture had the highest performance metrics for both training from scratch (96.7% accuracy) and transferred learning (98.8% accuracy). Three other models performed particularly well when trained from scratch, these were a modification of the LeNet architecture, ResNet50 and InceptionV3. The transferred learning results were also impressive with four out of the 5 models achieving an accuracy above 94%. Interestingly, the ResNet50 architecture failed to train when transferred learning was applied. Contrasting to this, the two models VGG16 and VGG19 failed to train when trained from scratch but performed equally as well as the other models when transferred learning was implemented. This indicates that although transferred learning can improve the performance of models, it is highly dependent on the task as well as the model. Overall the results obtained proved that a deep learning approach could be applied in order to classify tyres accurately.
- Full Text:
- Date Issued: 2019
A framework to measure human behaviour whilst reading
- Salehzadeh, Seyed Amirsaleh, Greyling, Jean
- Authors: Salehzadeh, Seyed Amirsaleh , Greyling, Jean
- Date: 2019
- Subjects: Computational intelligence , Machine learning Artificial intelligence Neural networks (Computer science)
- Language: English
- Type: Thesis , Doctoral , DPhil
- Identifier: http://hdl.handle.net/10948/43578 , vital:36921
- Description: The brain is the most complex object in the known universe that gives a sense of being to humans and characterises human behaviour. Building models of brain functions is perhaps the most fascinating scientific challenge in the 21st century. Reading is a significant cognitive process in the human brain that plays a critical role in the vital process of learning and in performing some daily activities. The study of human behaviour during reading has been an area of interest for researchers in different fields of science. This thesis is based upon providing a novel framework, called ARSAT (Assisting Researchers in the Selection of Appropriate Technologies), that measures the behaviour of humans when reading text. The ARSAT framework aims at assisting researchers in the selection and application of appropriate technologies to measure the behaviour of a person who is reading text. The ARSAT framework will assist to researchers who investigate the reading process and find it difficult to select appropriate theories, metrics, data collection methods and data analytics techniques. The ARSAT framework enhances the ability of its users to select appropriate metrics indicating the effective factors on the characterisation of different aspects of human behaviour during the reading process. As will be shown in this research study, human behaviour is characterised by a complicated interplay of action, cognition and emotion. The ARSAT framework also facilitates selecting appropriate sensory technologies that can be used to monitor and collect data for the metrics. Moreover, this research study will introduce BehaveNet, a novel Deep Learning modelling approach, which can be used for training Deep Learning models of human behaviour from the sensory data collected. In this thesis, a comprehensive literature study is presented that was conducted to acquire adequate knowledge for designing the ARSAT framework. In order to identify the contributing factors that affect the reading process, an overview of some existing theories of the reading process is provided. Furthermore, a number of sensory technologies and techniques that can be applied to monitoring the changes in the metrics indicating the factors are also demonstrated. Only, the technologies that are commercially available on the market are recommended by the ARSAT framework. A variety of Machine Learning techniques were also investigated when designing the BehaveNet. The BehaveNet takes advantage of the complementarity of Convolutional Neural Networks, Long Short-Term Memory networks and Deep Neural Networks. The design of a Human Behaviour Monitoring System (HBMS), by utilising the ARSAT framework for recognising three attention-seeking activities of humans, is also presented in this research study. Reading printed text, as well as speaking out loudly and watching a programme on TV were proposed as activities that a person unintentionally may shift his/her attention from reading into distractions. Between sensory devices recommended by the ARSAT framework, the Muse headband which is an Electroencephalography (EEG) and head motion-sensing wearable device, was selected to track the forehead EEG and a person’s head movements. The EEG and 3-axes accelerometer data were recorded from eight participants when they read printed text, as well as the time they performed two other activities. An imbalanced dataset consisting over 1.2 million rows of noisy data was created and used to build a model of the activities (60% training and 20% validating data) and evaluating the model (20% of the data). The efficiency of the framework is demonstrated by comparing the performance of the models built by utilising the BehaveNet, with the models built by utilising a number of competing Deep Learning models for raw EEG and accelerometer data, that have attained state-of-the-art performance. The classification results are evaluated by some metrics including the classification accuracy, F1 score, confusion matrix, Receiver Operating Characteristic curve, and Area under Curve (AUC) score. By considering the results, the BehaveNet contributed to the body of knowledge as an approach for measuring human behaviour by using sensory devices. In comparison with the performance of the other models, the models built by utilising the BehaveNet, attained better performance when classifying data of two EEG channels (Accuracy = 95%; AUC=0.99; F1 = 0.95), data of a single EEG channel (Accuracy = 85%; AUC=0.96; F1 = 0.83), accelerometer data (Accuracy = 81%; AUC = 0.9; F1 = 0.76) and all of the data in the dataset (Accuracy = 97%; AUC = 0.99; F1 = 0.96). The dataset and the source code of this project are also published on the Internet to help the science community. The Muse headband is also shown to be an economical and standard wearable device that can be successfully used in behavioural research.
- Full Text:
- Date Issued: 2019
- Authors: Salehzadeh, Seyed Amirsaleh , Greyling, Jean
- Date: 2019
- Subjects: Computational intelligence , Machine learning Artificial intelligence Neural networks (Computer science)
- Language: English
- Type: Thesis , Doctoral , DPhil
- Identifier: http://hdl.handle.net/10948/43578 , vital:36921
- Description: The brain is the most complex object in the known universe that gives a sense of being to humans and characterises human behaviour. Building models of brain functions is perhaps the most fascinating scientific challenge in the 21st century. Reading is a significant cognitive process in the human brain that plays a critical role in the vital process of learning and in performing some daily activities. The study of human behaviour during reading has been an area of interest for researchers in different fields of science. This thesis is based upon providing a novel framework, called ARSAT (Assisting Researchers in the Selection of Appropriate Technologies), that measures the behaviour of humans when reading text. The ARSAT framework aims at assisting researchers in the selection and application of appropriate technologies to measure the behaviour of a person who is reading text. The ARSAT framework will assist to researchers who investigate the reading process and find it difficult to select appropriate theories, metrics, data collection methods and data analytics techniques. The ARSAT framework enhances the ability of its users to select appropriate metrics indicating the effective factors on the characterisation of different aspects of human behaviour during the reading process. As will be shown in this research study, human behaviour is characterised by a complicated interplay of action, cognition and emotion. The ARSAT framework also facilitates selecting appropriate sensory technologies that can be used to monitor and collect data for the metrics. Moreover, this research study will introduce BehaveNet, a novel Deep Learning modelling approach, which can be used for training Deep Learning models of human behaviour from the sensory data collected. In this thesis, a comprehensive literature study is presented that was conducted to acquire adequate knowledge for designing the ARSAT framework. In order to identify the contributing factors that affect the reading process, an overview of some existing theories of the reading process is provided. Furthermore, a number of sensory technologies and techniques that can be applied to monitoring the changes in the metrics indicating the factors are also demonstrated. Only, the technologies that are commercially available on the market are recommended by the ARSAT framework. A variety of Machine Learning techniques were also investigated when designing the BehaveNet. The BehaveNet takes advantage of the complementarity of Convolutional Neural Networks, Long Short-Term Memory networks and Deep Neural Networks. The design of a Human Behaviour Monitoring System (HBMS), by utilising the ARSAT framework for recognising three attention-seeking activities of humans, is also presented in this research study. Reading printed text, as well as speaking out loudly and watching a programme on TV were proposed as activities that a person unintentionally may shift his/her attention from reading into distractions. Between sensory devices recommended by the ARSAT framework, the Muse headband which is an Electroencephalography (EEG) and head motion-sensing wearable device, was selected to track the forehead EEG and a person’s head movements. The EEG and 3-axes accelerometer data were recorded from eight participants when they read printed text, as well as the time they performed two other activities. An imbalanced dataset consisting over 1.2 million rows of noisy data was created and used to build a model of the activities (60% training and 20% validating data) and evaluating the model (20% of the data). The efficiency of the framework is demonstrated by comparing the performance of the models built by utilising the BehaveNet, with the models built by utilising a number of competing Deep Learning models for raw EEG and accelerometer data, that have attained state-of-the-art performance. The classification results are evaluated by some metrics including the classification accuracy, F1 score, confusion matrix, Receiver Operating Characteristic curve, and Area under Curve (AUC) score. By considering the results, the BehaveNet contributed to the body of knowledge as an approach for measuring human behaviour by using sensory devices. In comparison with the performance of the other models, the models built by utilising the BehaveNet, attained better performance when classifying data of two EEG channels (Accuracy = 95%; AUC=0.99; F1 = 0.95), data of a single EEG channel (Accuracy = 85%; AUC=0.96; F1 = 0.83), accelerometer data (Accuracy = 81%; AUC = 0.9; F1 = 0.76) and all of the data in the dataset (Accuracy = 97%; AUC = 0.99; F1 = 0.96). The dataset and the source code of this project are also published on the Internet to help the science community. The Muse headband is also shown to be an economical and standard wearable device that can be successfully used in behavioural research.
- Full Text:
- Date Issued: 2019
A gatherer's paradise? Early humans and plant foraging on the cape south coast, South Africa
- Authors: Gummow, William Thomas
- Date: 2019
- Subjects: Soil amendments , Plants -- Nutrition Plant physiology
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/39687 , vital:35348
- Description: South Africa is a semi-arid country that is experiencing population growth and the associated pressure of producing sufficient crops to sustain the population. Furthermore, poor land management has also led to increased degradation of valuable arable land. To improve poor soil quality and possibly avoid or reverse degradation and ultimately desertification, it was suggested that soil amendments are a potential solution to some of these problems. Although soil amendment is a relatively well-known practice to improve soil quality, the use of rubber crumb as a soil amendment is not well studied. Rubber crumb is produced from end-of-life tyres that have been ground up to fine particles and as such can be used as a secondary resource to replace the use of virgin materials. Rubber crumb has been used in an array of industries, however, the present study investigated the possibility that rubber crumb might be used as a soil amendment to alleviate various forms of soil degradation, specifically soil compaction and crusting, and potentially improve crop establishment and growth.To determine this, a soil that is prone to crusting and compaction was amended with rubber crumb at rates of 5%, 15% and 25%. Various soil physical characteristics, such as bulk density, particle density, porosity, field capacity and penetration resistance were investigated and it was determined that as the percentage rubber crumb increases in the soil, so do the beneficial characteristics of the soil. Furthermore, the rubber crumb was also treated in various ways in order to determine how the treatments might affect the chemical properties of the soil, as well as the amount of zinc that is leached from the rubber crumb. It was found that washing the rubber with a weak acid or even water prior to adding it to the soil resulted in more desirable chemical condition compared to untreated rubber crumb. Lastly, maize (Zea mays L.) was used to determine whether amending soil with pre-treated rubber crumb in different percentages had an influence on the germination success and growth rate of maize. In terms of germination success, amending soil with rubber crumb does not seem to have a significant influence, however, growth rate decreased as the percentage rubber crumb increased, while treating the rubber with a weak acid or water prior to adding it to soil had the most positive effect on plant growth. It can therefore be concluded that rubber crumb amendment rates between 5% and 15% resulted in the most desirable soil physical properties and plant growth amongst the amendments. However the best plant growth was still seen in the control soil, to which no rubber crumb had been added.
- Full Text:
- Date Issued: 2019
- Authors: Gummow, William Thomas
- Date: 2019
- Subjects: Soil amendments , Plants -- Nutrition Plant physiology
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/39687 , vital:35348
- Description: South Africa is a semi-arid country that is experiencing population growth and the associated pressure of producing sufficient crops to sustain the population. Furthermore, poor land management has also led to increased degradation of valuable arable land. To improve poor soil quality and possibly avoid or reverse degradation and ultimately desertification, it was suggested that soil amendments are a potential solution to some of these problems. Although soil amendment is a relatively well-known practice to improve soil quality, the use of rubber crumb as a soil amendment is not well studied. Rubber crumb is produced from end-of-life tyres that have been ground up to fine particles and as such can be used as a secondary resource to replace the use of virgin materials. Rubber crumb has been used in an array of industries, however, the present study investigated the possibility that rubber crumb might be used as a soil amendment to alleviate various forms of soil degradation, specifically soil compaction and crusting, and potentially improve crop establishment and growth.To determine this, a soil that is prone to crusting and compaction was amended with rubber crumb at rates of 5%, 15% and 25%. Various soil physical characteristics, such as bulk density, particle density, porosity, field capacity and penetration resistance were investigated and it was determined that as the percentage rubber crumb increases in the soil, so do the beneficial characteristics of the soil. Furthermore, the rubber crumb was also treated in various ways in order to determine how the treatments might affect the chemical properties of the soil, as well as the amount of zinc that is leached from the rubber crumb. It was found that washing the rubber with a weak acid or even water prior to adding it to the soil resulted in more desirable chemical condition compared to untreated rubber crumb. Lastly, maize (Zea mays L.) was used to determine whether amending soil with pre-treated rubber crumb in different percentages had an influence on the germination success and growth rate of maize. In terms of germination success, amending soil with rubber crumb does not seem to have a significant influence, however, growth rate decreased as the percentage rubber crumb increased, while treating the rubber with a weak acid or water prior to adding it to soil had the most positive effect on plant growth. It can therefore be concluded that rubber crumb amendment rates between 5% and 15% resulted in the most desirable soil physical properties and plant growth amongst the amendments. However the best plant growth was still seen in the control soil, to which no rubber crumb had been added.
- Full Text:
- Date Issued: 2019
A gatherer’s paradise? early humans and plant foraging on the Cape south coast, South Africa
- Authors: Botha, Maria Susan
- Date: 2019
- Subjects: Plant remains (Archaeology) -- South Africa , Plant physiology Plant ecology Botany -- South Africa
- Language: English
- Type: Thesis , Doctoral , PhD
- Identifier: http://hdl.handle.net/10948/37139 , vital:34124
- Description: Humans were driven to refugia during the cold and dry glacial Marine Isotope Stage 6 (MIS6) (~195–125 ka); only a few places, including the Cape south coast, show archaeological evidence of continuous human occupation. It has been hypothesised that the Cape south coast provided the requisite shelter to ensure human’s earliest survival because it provided all the right ingredients. The shores offer abundant shellfish, the land a diverse array of plants, and the climate was ameliorated due to the proximity of the then exposed Palaeo-Agulhas Plain to the warm Agulhas current. The aim of this study is to determine whether the indigenous flora of this region could have provided a sufficient edible resource for early humans. Residents of the Cape south coast have genetic ancestry linking them to the Khoe-San, the original inhabitants of the area, and still have an extensive knowledge of the local edible plants. With their help, I set out to determine a) whether humans have been utilising the same plant species over time, b) what the foraging potential of the edible plants in the region is and c) how resilient these plants are to human foraging. If we know whether the plant species known and used today were the same as those that were utilised by past humans, we can then use the contemporary knowledge to make predictions about past utilisation. To answer this question, I collated two databases: archaeological (all plant species found in archaeological sites [dating 0 to 80,000 BP]) and contemporary (all plant species in the modern-day ethnographic literature [last 400 years]) that occur within the Greater Cape Floristic Region(GCFR). I found a significant number of plant species shared between the two databases, which suggests that at least some plant species have been used by humans over a long period of time. To determine the indigenous plant foraging potential of the region, I foraged for food in the Cape south coast (451 bouts) monthly over a two-year period with the help of local inhabitants. The findings show that edible plant resources are distributed patchily and focusing on specific vegetation types would greatly enhance chances of harvesting 2,000 kcal per day, which is considered the daily calorific requirements for a typical hunter-gatherer. I then sought to understand how resilient these plants [(with an emphasis on plants with an underground storage organ (USO)] would be to human foraging. To do this, I set out plots and harvested all edible foods for three consecutive years with the help of foragers. Results indicate that there was a significant reduction in edible weight only in the third year of consecutive harvesting. In conclusion, using various approaches, this study investigates the plant food potential of the Cape south coast from the perspective of early human consumers. The findings suggest that knowledge regarding useful plants dates back to at least 80,000 BP. Food resources are patchily distributed across the main vegetation types found within the Cape south coast and occur in hotspots, i.e. concentrated areas hosting high densities of edible plant foods, surrounded by areas with very low plant food densities. Foragers could have harvested their daily calorific quota more easily if they focused their harvesting efforts in specific vegetation types found in the Cape south coast. Furthermore, many USOs circumvent climatic fluctuations, herbivory or both by staggering their emergence over multiple years, which implies they have some resilience to human foraging.
- Full Text:
- Date Issued: 2019
- Authors: Botha, Maria Susan
- Date: 2019
- Subjects: Plant remains (Archaeology) -- South Africa , Plant physiology Plant ecology Botany -- South Africa
- Language: English
- Type: Thesis , Doctoral , PhD
- Identifier: http://hdl.handle.net/10948/37139 , vital:34124
- Description: Humans were driven to refugia during the cold and dry glacial Marine Isotope Stage 6 (MIS6) (~195–125 ka); only a few places, including the Cape south coast, show archaeological evidence of continuous human occupation. It has been hypothesised that the Cape south coast provided the requisite shelter to ensure human’s earliest survival because it provided all the right ingredients. The shores offer abundant shellfish, the land a diverse array of plants, and the climate was ameliorated due to the proximity of the then exposed Palaeo-Agulhas Plain to the warm Agulhas current. The aim of this study is to determine whether the indigenous flora of this region could have provided a sufficient edible resource for early humans. Residents of the Cape south coast have genetic ancestry linking them to the Khoe-San, the original inhabitants of the area, and still have an extensive knowledge of the local edible plants. With their help, I set out to determine a) whether humans have been utilising the same plant species over time, b) what the foraging potential of the edible plants in the region is and c) how resilient these plants are to human foraging. If we know whether the plant species known and used today were the same as those that were utilised by past humans, we can then use the contemporary knowledge to make predictions about past utilisation. To answer this question, I collated two databases: archaeological (all plant species found in archaeological sites [dating 0 to 80,000 BP]) and contemporary (all plant species in the modern-day ethnographic literature [last 400 years]) that occur within the Greater Cape Floristic Region(GCFR). I found a significant number of plant species shared between the two databases, which suggests that at least some plant species have been used by humans over a long period of time. To determine the indigenous plant foraging potential of the region, I foraged for food in the Cape south coast (451 bouts) monthly over a two-year period with the help of local inhabitants. The findings show that edible plant resources are distributed patchily and focusing on specific vegetation types would greatly enhance chances of harvesting 2,000 kcal per day, which is considered the daily calorific requirements for a typical hunter-gatherer. I then sought to understand how resilient these plants [(with an emphasis on plants with an underground storage organ (USO)] would be to human foraging. To do this, I set out plots and harvested all edible foods for three consecutive years with the help of foragers. Results indicate that there was a significant reduction in edible weight only in the third year of consecutive harvesting. In conclusion, using various approaches, this study investigates the plant food potential of the Cape south coast from the perspective of early human consumers. The findings suggest that knowledge regarding useful plants dates back to at least 80,000 BP. Food resources are patchily distributed across the main vegetation types found within the Cape south coast and occur in hotspots, i.e. concentrated areas hosting high densities of edible plant foods, surrounded by areas with very low plant food densities. Foragers could have harvested their daily calorific quota more easily if they focused their harvesting efforts in specific vegetation types found in the Cape south coast. Furthermore, many USOs circumvent climatic fluctuations, herbivory or both by staggering their emergence over multiple years, which implies they have some resilience to human foraging.
- Full Text:
- Date Issued: 2019
A Model for Crime Management in Smart Cities
- Authors: Westraadt, Lindsay
- Date: 2019
- Subjects: Smart cities , Computer networks -- security measures
- Language: English
- Type: Thesis , Doctoral , PHD
- Identifier: http://hdl.handle.net/10948/45635 , vital:38922
- Description: The main research problem addressed in this study is that South African cities are not effectively integrating and utilising available, and rapidly emerging smart city data sources for planning and management. To this end, it was proposed that a predictive model, that assimilates data from traditionally isolated management silos, could be developed for prediction and simulation at the system-of-systems level. As proof of concept, the study focused on only one aspect of smart cities, namely crime management. Subsequently, the main objective of this study was to develop and evaluate a predictive model for crime management in smart cities that effectively integrated data from traditionally isolated management silos. The Design Science Research process was followed to develop and evaluate a prototype model. The practical contributions of this study was the development of a prototype model for integrated decision-making in smart cities, and the associated guidelines for the implementation of the developed modelling approach within the South African IDP context. Theoretically, this work contributed towards the development of a modelling paradigm for effective integrated decision-making in smart cities. This work also contributed towards developing strategic-level predictive policing tools aimed at proactively meeting community needs, and contributed to the body of knowledge regarding complex systems modelling.
- Full Text:
- Date Issued: 2019
- Authors: Westraadt, Lindsay
- Date: 2019
- Subjects: Smart cities , Computer networks -- security measures
- Language: English
- Type: Thesis , Doctoral , PHD
- Identifier: http://hdl.handle.net/10948/45635 , vital:38922
- Description: The main research problem addressed in this study is that South African cities are not effectively integrating and utilising available, and rapidly emerging smart city data sources for planning and management. To this end, it was proposed that a predictive model, that assimilates data from traditionally isolated management silos, could be developed for prediction and simulation at the system-of-systems level. As proof of concept, the study focused on only one aspect of smart cities, namely crime management. Subsequently, the main objective of this study was to develop and evaluate a predictive model for crime management in smart cities that effectively integrated data from traditionally isolated management silos. The Design Science Research process was followed to develop and evaluate a prototype model. The practical contributions of this study was the development of a prototype model for integrated decision-making in smart cities, and the associated guidelines for the implementation of the developed modelling approach within the South African IDP context. Theoretically, this work contributed towards the development of a modelling paradigm for effective integrated decision-making in smart cities. This work also contributed towards developing strategic-level predictive policing tools aimed at proactively meeting community needs, and contributed to the body of knowledge regarding complex systems modelling.
- Full Text:
- Date Issued: 2019
A multi-factor model for range estimation in electric vehicles
- Authors: Smuts, Martin Bradley
- Date: 2019
- Subjects: Electric vehicles , Hybrid electric vehicles Energy consumption Machine learning Information technology -- Management
- Language: English
- Type: Thesis , Doctoral , DPhil
- Identifier: http://hdl.handle.net/10948/43589 , vital:36926
- Description: Electric vehicles (EVs) are well-known for their challenges related to trip planning and energy consumption estimation. Range anxiety is currently a barrier to the adoption of EVs. One of the issues influencing range anxiety is the inaccuracy of the remaining driving range (RDR) estimate in on-board displays. RDR displays are important as they can help drivers with trip planning. The RDR is a parameter that changes under environmental and behavioural conditions. Several factors (for example, weather, and traffic) can influence the energy consumption of an EV that are not considered during the RDR estimation in traditional on-board computers or third-party applications, such as navigation or mapping applications. The need for accurate RDR estimation is growing, since this can reduce the range anxiety of drivers. One way of overcoming range anxiety is to provide trip planning applications that provide accurate estimations of the RDR, based on various factors, and which adapt to the users’ driving behaviour. Existing models used for estimating the RDR are often simplified, and do not consider all the factors that can influence it. Collecting data for each factor also presents several challenges. Powerful computing resources are required to collect, transform, and analyse the disparate datasets that are required for each factor. The aim of this research was to design a Multi-factor Model for range estimation in EVs. Five main factors that influence the energy consumption of EVs were identified from literature, namely, Route and Terrain, Driving Behaviour, Weather and Environment, Vehicle Modelling, and Battery Modelling. These factors were used throughout this research to guide the data collection and analysis processes. A Multi-factor Model was proposed based on four main components that collect, process, analyse, and visualise data from available data sources to produce estimates relating to trip planning. A proof-of-concept RDR system was developed and evaluated in field experiments, to demonstrate that the Multi-factor Model addresses the main aim of this research. The experiments were performed to collect data for each of the five factors, and to analyse their impact on energy consumption. Several machine learning techniques were used, and evaluated, for accuracy in estimating the energy consumption, from which the RDR can be derived, for a specified trip. A case study was conducted with an electric mobility programme (uYilo) in Port Elizabeth, South Africa (SA). The case study was used to investigate whether the available resources at uYilo were sufficient to provide data for each of the five factors. Several challenges were noted during the data collection. These were shortages of software applications, a lack of quality data, technical interoperability and data access between the data collection instruments and systems. Data access was a problem in some cases, since proprietary systems restrict access to external developers. The theoretical contribution of this research is a list of factors that influence RDR and a classification of machine learning techniques that can be used to estimate the RDR. The practical contributions of this research include a database of EV trips, proof-of-concept RDR estimation system, and a deployed machine learning model that can be accessed by researchers and EV practitioners. Four research papers were published and presented at local and international conferences. In addition, one conference paper was published in an accredited journal: NextComp 2017 (Appendix C), Conference Paper, Pointe aux Piments (Mauritius); SATNAC 2017 (Appendix F), Conference Paper, Barcelona (Spain); GITMA 2018 (Appendix B), Conference Paper, Mexico City (Mexico); SATNAC 2018 (Appendix G), Conference Paper, George (South Africa), and IFIP World Computer Congress 2018 (Appendix E), Journal Article.
- Full Text:
- Date Issued: 2019
- Authors: Smuts, Martin Bradley
- Date: 2019
- Subjects: Electric vehicles , Hybrid electric vehicles Energy consumption Machine learning Information technology -- Management
- Language: English
- Type: Thesis , Doctoral , DPhil
- Identifier: http://hdl.handle.net/10948/43589 , vital:36926
- Description: Electric vehicles (EVs) are well-known for their challenges related to trip planning and energy consumption estimation. Range anxiety is currently a barrier to the adoption of EVs. One of the issues influencing range anxiety is the inaccuracy of the remaining driving range (RDR) estimate in on-board displays. RDR displays are important as they can help drivers with trip planning. The RDR is a parameter that changes under environmental and behavioural conditions. Several factors (for example, weather, and traffic) can influence the energy consumption of an EV that are not considered during the RDR estimation in traditional on-board computers or third-party applications, such as navigation or mapping applications. The need for accurate RDR estimation is growing, since this can reduce the range anxiety of drivers. One way of overcoming range anxiety is to provide trip planning applications that provide accurate estimations of the RDR, based on various factors, and which adapt to the users’ driving behaviour. Existing models used for estimating the RDR are often simplified, and do not consider all the factors that can influence it. Collecting data for each factor also presents several challenges. Powerful computing resources are required to collect, transform, and analyse the disparate datasets that are required for each factor. The aim of this research was to design a Multi-factor Model for range estimation in EVs. Five main factors that influence the energy consumption of EVs were identified from literature, namely, Route and Terrain, Driving Behaviour, Weather and Environment, Vehicle Modelling, and Battery Modelling. These factors were used throughout this research to guide the data collection and analysis processes. A Multi-factor Model was proposed based on four main components that collect, process, analyse, and visualise data from available data sources to produce estimates relating to trip planning. A proof-of-concept RDR system was developed and evaluated in field experiments, to demonstrate that the Multi-factor Model addresses the main aim of this research. The experiments were performed to collect data for each of the five factors, and to analyse their impact on energy consumption. Several machine learning techniques were used, and evaluated, for accuracy in estimating the energy consumption, from which the RDR can be derived, for a specified trip. A case study was conducted with an electric mobility programme (uYilo) in Port Elizabeth, South Africa (SA). The case study was used to investigate whether the available resources at uYilo were sufficient to provide data for each of the five factors. Several challenges were noted during the data collection. These were shortages of software applications, a lack of quality data, technical interoperability and data access between the data collection instruments and systems. Data access was a problem in some cases, since proprietary systems restrict access to external developers. The theoretical contribution of this research is a list of factors that influence RDR and a classification of machine learning techniques that can be used to estimate the RDR. The practical contributions of this research include a database of EV trips, proof-of-concept RDR estimation system, and a deployed machine learning model that can be accessed by researchers and EV practitioners. Four research papers were published and presented at local and international conferences. In addition, one conference paper was published in an accredited journal: NextComp 2017 (Appendix C), Conference Paper, Pointe aux Piments (Mauritius); SATNAC 2017 (Appendix F), Conference Paper, Barcelona (Spain); GITMA 2018 (Appendix B), Conference Paper, Mexico City (Mexico); SATNAC 2018 (Appendix G), Conference Paper, George (South Africa), and IFIP World Computer Congress 2018 (Appendix E), Journal Article.
- Full Text:
- Date Issued: 2019
A Statistical assessment of available solar resource across multiple sites in South Africa
- Authors: Eastwood, Kirstie
- Date: 2019
- Subjects: Mathematical statistics
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/39907 , vital:35505
- Description: Around the globe, fossil fuels remain the primary source of energy at around 78% of the world’s total energy consumption. However, the associated carbon emissions, environmental impact, the depletion of fossil fuels and price and cost volatility are factors influencing the growing popularity of research into renewable energy. Solar power is acknowledged as the fastest-growing renewable energy, but the uncertainty surrounding the long-term projections of solar irradiance available for energy conversion is a hindrance when discussing the financial risk with potential investors. This study investigates the quality of freely available solar resource data in South Africa as well as proposes techniques for potential solar farm site comparisons. Tolerance intervals derived within a Bayesian framework provide information on the future available solar resource across multiple sites. These techniques capture the inherent variability in the available solar resource which equips investors with statistical methods that lead to the better understanding of the solar resource and thus aids in better decision-making.
- Full Text:
- Date Issued: 2019
- Authors: Eastwood, Kirstie
- Date: 2019
- Subjects: Mathematical statistics
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/39907 , vital:35505
- Description: Around the globe, fossil fuels remain the primary source of energy at around 78% of the world’s total energy consumption. However, the associated carbon emissions, environmental impact, the depletion of fossil fuels and price and cost volatility are factors influencing the growing popularity of research into renewable energy. Solar power is acknowledged as the fastest-growing renewable energy, but the uncertainty surrounding the long-term projections of solar irradiance available for energy conversion is a hindrance when discussing the financial risk with potential investors. This study investigates the quality of freely available solar resource data in South Africa as well as proposes techniques for potential solar farm site comparisons. Tolerance intervals derived within a Bayesian framework provide information on the future available solar resource across multiple sites. These techniques capture the inherent variability in the available solar resource which equips investors with statistical methods that lead to the better understanding of the solar resource and thus aids in better decision-making.
- Full Text:
- Date Issued: 2019
A suitability study for the implementation of a corporate geographic information system at the King Sabata Dalindyebo Municipality
- Authors: Triegaardt, Inge
- Date: 2019
- Subjects: Geographic information systems , Information storage and retrieval systems -- Geography Information technology Sustainable development
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/${Handle} , vital:37044
- Description: A tender in the Government Gazette in 2013 for the implementation of a Corporate Geographic Information System (CGIS) at the King Sabata Dalindeybo (KSD) Local Municipality, created an opportunity for a study to be conducted for the implementation of a Corporate Geographic Information System (CGIS). A CGIS is a suitable management system for implementation at a municipality, given that the municipality is ready. Internationally the implementation of CGIS at municipalities has been done successfully. In South Africa, however, the implementation of said software has been slow and at times unsuccessful. The purpose of this research study was to evaluate the current GIS knowledge and infrastructure of the KSD Local Municipality for the municipality to be a suitable candidate for the implementation of a CGIS. The evaluation was done by the collection of data through three questionnaires, two interviews and observations. The findings of this research study are discussed according to the qualitative data collected from the KSD Local Municipality and recommendations based on the results of the data collection have been made for the municipality to be implementation ready. The qualitative instruments such as pre-training, post-training/pre-implementation and postimplementation questionnaires, observations and interviews with the IT Technician and the CGIS Specialist were processed by the researcher to obtain insight to the perceptions of the employees and the current state of GIS in the KSD Local Municipality. The results from this research study were in-line with the benefits and barriers of implementation of a CGIS in municipalities in international literature.
- Full Text:
- Date Issued: 2019
- Authors: Triegaardt, Inge
- Date: 2019
- Subjects: Geographic information systems , Information storage and retrieval systems -- Geography Information technology Sustainable development
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/${Handle} , vital:37044
- Description: A tender in the Government Gazette in 2013 for the implementation of a Corporate Geographic Information System (CGIS) at the King Sabata Dalindeybo (KSD) Local Municipality, created an opportunity for a study to be conducted for the implementation of a Corporate Geographic Information System (CGIS). A CGIS is a suitable management system for implementation at a municipality, given that the municipality is ready. Internationally the implementation of CGIS at municipalities has been done successfully. In South Africa, however, the implementation of said software has been slow and at times unsuccessful. The purpose of this research study was to evaluate the current GIS knowledge and infrastructure of the KSD Local Municipality for the municipality to be a suitable candidate for the implementation of a CGIS. The evaluation was done by the collection of data through three questionnaires, two interviews and observations. The findings of this research study are discussed according to the qualitative data collected from the KSD Local Municipality and recommendations based on the results of the data collection have been made for the municipality to be implementation ready. The qualitative instruments such as pre-training, post-training/pre-implementation and postimplementation questionnaires, observations and interviews with the IT Technician and the CGIS Specialist were processed by the researcher to obtain insight to the perceptions of the employees and the current state of GIS in the KSD Local Municipality. The results from this research study were in-line with the benefits and barriers of implementation of a CGIS in municipalities in international literature.
- Full Text:
- Date Issued: 2019
A technical survey of the South African harvesting machine population
- Authors: Mathelele, Ouma Florah
- Date: 2019
- Subjects: Harvesting -- South Africa , Energy harvesting -- South Africa
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/40999 , vital:36281
- Description: South Africa adopted the practice of conducting technical surveys from Austria three decades ago. The Austrian technical survey was used to model the South African technical survey. The survey was conducted with the aim of providing the forestry industry with a detailed analysis of the state of timber harvesting operations. The first technical survey in South Africa was conducted in 1987, and the second one in 1998. The most recent study was carried out in 2007. Since 2007, there has been no other technical survey of a similar nature conducted on harvesting and transport operations. There is no recent information quantifying the volumes harvested, identifying harvesting systems and methods used, and determining volumes harvested using these systems and methods. Furthermore, no updated information exists on the type and quantities of timber trucks transporting the volume. A fourth technical survey that quantifies and analyses these aspects in harvesting operations, therefore, was due. The fourth technical survey aimed at quantifying the number of harvesting machines and timber trucks. The survey further investigated the drivers behind the mechanisation and system preference. The four technical surveys were compared to identify trends in harvesting machines and methods. The data were collected from grower companies, harvesting and transport contractors and small-scale timber growers. Questionnaires were sent to all potential respondents from the three target groups, and interviews were later conducted with selected respondents. The harvesting machines were quantified per function i.e. felling, extraction, processing and loading. The results showed that some machines were more common compared to other machines performing the same function. Mechanised felling machines sampled had more units in the fourth survey than the previous surveys had. Of the 10.6 million m3 volume surveyed in 2017, chainsaws had 1 048 units, and their number had decreased from the 5 000 sampled in the first technical survey in 1987 that sampled about 12.8 million m3. Harvesters and feller bunchers also increased from zero in the first technical survey to 94 and 13, respectively in the fourth survey. The drivers towards the increased use of mechanised equipment were identified as safety and health, ergonomics, productivity, labour issues, operational costs, terrain, and company policies. Despite the introduction of mechanisation, however, the use of manual labour was evident in hardwood processing and loading operations. Manual debarking was more common than mechanised debarking in the fourth technical survey. When comparing all four technical surveys, the cut to length (CTL) harvesting method was dominant in hardwood; harvesting 79% of the volume sampled in the 2017 survey. The full tree method was barely applied in the first three technical surveys and in the 2007 survey, it was absent. It was only in the fourth technical survey that the method was applied, felling 27% and 9% of the sampled volumes in softwood and hardwood, respectively. The technical survey provides information to the different stakeholders in the forest industry. Such include harvesting managers and contractors, grower companies, machine distributors and researchers. The information will bring awareness to harvesting managers and contractors about the different machine options available to assist in decision making when selecting systems. Machine distributors will know the number of machines, where machines are operating and their target market. A gap in knowledge will be bridged between the different provinces regarding the different types of harvesting machines out there.
- Full Text:
- Date Issued: 2019
- Authors: Mathelele, Ouma Florah
- Date: 2019
- Subjects: Harvesting -- South Africa , Energy harvesting -- South Africa
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/40999 , vital:36281
- Description: South Africa adopted the practice of conducting technical surveys from Austria three decades ago. The Austrian technical survey was used to model the South African technical survey. The survey was conducted with the aim of providing the forestry industry with a detailed analysis of the state of timber harvesting operations. The first technical survey in South Africa was conducted in 1987, and the second one in 1998. The most recent study was carried out in 2007. Since 2007, there has been no other technical survey of a similar nature conducted on harvesting and transport operations. There is no recent information quantifying the volumes harvested, identifying harvesting systems and methods used, and determining volumes harvested using these systems and methods. Furthermore, no updated information exists on the type and quantities of timber trucks transporting the volume. A fourth technical survey that quantifies and analyses these aspects in harvesting operations, therefore, was due. The fourth technical survey aimed at quantifying the number of harvesting machines and timber trucks. The survey further investigated the drivers behind the mechanisation and system preference. The four technical surveys were compared to identify trends in harvesting machines and methods. The data were collected from grower companies, harvesting and transport contractors and small-scale timber growers. Questionnaires were sent to all potential respondents from the three target groups, and interviews were later conducted with selected respondents. The harvesting machines were quantified per function i.e. felling, extraction, processing and loading. The results showed that some machines were more common compared to other machines performing the same function. Mechanised felling machines sampled had more units in the fourth survey than the previous surveys had. Of the 10.6 million m3 volume surveyed in 2017, chainsaws had 1 048 units, and their number had decreased from the 5 000 sampled in the first technical survey in 1987 that sampled about 12.8 million m3. Harvesters and feller bunchers also increased from zero in the first technical survey to 94 and 13, respectively in the fourth survey. The drivers towards the increased use of mechanised equipment were identified as safety and health, ergonomics, productivity, labour issues, operational costs, terrain, and company policies. Despite the introduction of mechanisation, however, the use of manual labour was evident in hardwood processing and loading operations. Manual debarking was more common than mechanised debarking in the fourth technical survey. When comparing all four technical surveys, the cut to length (CTL) harvesting method was dominant in hardwood; harvesting 79% of the volume sampled in the 2017 survey. The full tree method was barely applied in the first three technical surveys and in the 2007 survey, it was absent. It was only in the fourth technical survey that the method was applied, felling 27% and 9% of the sampled volumes in softwood and hardwood, respectively. The technical survey provides information to the different stakeholders in the forest industry. Such include harvesting managers and contractors, grower companies, machine distributors and researchers. The information will bring awareness to harvesting managers and contractors about the different machine options available to assist in decision making when selecting systems. Machine distributors will know the number of machines, where machines are operating and their target market. A gap in knowledge will be bridged between the different provinces regarding the different types of harvesting machines out there.
- Full Text:
- Date Issued: 2019