OpenUCT is the open access institutional repository of the University of Cape Town (UCT). It preserves and makes UCT scholarly outputs digitally and freely available, including theses and dissertations, journal articles, book chapters, technical and research reports, as well as open educational resources.
Recent Submissions
Item
Open Access
Evaluating the success of the Western Cape Provincial Treasury MyContent Central Repository in Cape Town, South Africa
(2026) Netshisaulu, Murunwa; Higgs, Richard
The era of digital technologies necessitates that organisations leverage the capabilities of Enterprise Content Management Systems solutions for the efficient management and preservation of digital content. However, a review of the literature shows that users are reluctant to adopt ECM system solutions, which hampers the digital transformation agenda. The study aimed to supplement the WCG ECM system post-implementation evaluations at the site of study by gathering insight regarding the impediments to the adoption of the WCG ECM system. Specifically, to understand the key adoption factors that influence the adoption of the WCG ECM system at the provincial department level, the key factors that impede the adoption of the WCG ECM system, the alternative systems users are adopting to manage official electronic records, and their reasons for doing so. Guided by the Delone and McLean Information Success model and adopting the multi-method approach and purposive sampling technique, the study collected data using a questionnaire and interviews with employees at the Western Cape Provincial Treasury programme level. The study affirms that the ECM system is embedded in the organisational ecosystem. However, training, ease of use and legacy systems impede user uptake, which drives users to seek alternatives to bridge system shortfalls. As ECM systems preserve the institutional memory of the organisational ecosystem, the adoption issues compromise the integrity of information assets as digital content is segmented across multiple platforms, which goes against the one-system objective. The study recommends that addressing the ECM adoption problem not only entails strengthening training interventions and management support but also necessitates that the use of legacy and alternative systems be reviewed, and evaluation tools and organisational autonomy be a key part of the ECM system agenda. Therefore, future research can explore the influence of the authoritative agency DCAS on ECM user adoption at the WCG.
Item
Open Access
Searching for giant radio galaxies in the MIGHTEE survey
(2026) Netshiavha, Ndivhuwo Theophilus; Delhaize, Jacinta; Jarvis, Matt
Giant radio galaxies (GRGs) are a class of radio galaxies that are larger than 0.7 Mpc in physical extent. They serve as ideal tools to study the influence of jets on galaxy evolution. In the past, GRGs were thought to be rare, but MeerKAT and other radio interferometers such as the Low Frequency Array (LOFAR) are rapidly proving otherwise. Thanks to its excellent sensitivity to diffuse, low surface brightness emission, MeerKAT is unveiling GRGs that were previously undetected. The MeerKAT International GHz Tiered Extragalactic Exploration (MIGHTEE) is a MeerKAT survey that operates in the L-band at a central frequency of 1.28 GHz. The aim of this work is to leverage the excellent sensitivity of the MeerKAT telescope, along with the deep multiwavelength data available, to identify and characterise GRGs in the XMM-LSS and CDFS fields of MIGHTEE Data Release 1 (DR1). In this thesis, we present the discovery of ten previously unknown GRGs, seven from the MIGHTEE XMM-LSS and three from the MIGHTEE CDFS field. We also revisit five previously discovered GRGs in the MIGHTEE XMM-LSS field. Additionally, we obtained the Southern African Large Telescope (SALT) long slit spectroscopy of two GRGs in our sample and determined their spectroscopic redshifts, likely merging state, and excitation modes. This sample of fifteen GRGs spans a redshift range from 0.15 to 0.92, with physical extents ranging from 0.73 to 1.44 Mpc. We found that these GRGs have similar projected physical extents and continuum luminosities to others in the literature. This demonstrates that the MIGHTEE DR1 survey achieves a surface brightness sensitivity comparable to that of other prominent surveys such as the Low Frequency Array Two metre Sky Survey (LoTSS). All GRGs that we could assess their host morphologies from optical images exhibit characteristics consistent with early-type galaxies. Ten of our GRGs had sufficient multiwavelength coverage to classify their excitation modes. We classified all ten as low excitation radio galaxies (LERGS). Our results align with expectations from the literature that GRGs are predominantly LERGs hosted by early type galaxies. Our findings, together with other recent literature discoveries, suggest that the number of GRGs were previously underestimated, possibly due to the limited surface brightness sensitivity of previous radio surveys.
Item
Open Access
The role of technological innovation in overcoming barriers faced by women in the South African built environment
(2026) Ndlovu, Amanda; Le Jeune, Karen
The underrepresentation of women in the South African built environment highlights the need to understand and address the barriers that inhibit their full participation. Despite existing legislation and policies implemented to support women in the built environment, challenges in retaining female professionals persist. While existing studies have identified various barriers and success strategies for improving women's retention, significant gaps remain in addressing the underlying factors affecting their participation. This research investigated how technological innovations could help overcome these gender barriers and foster greater equality in all aspects of women's careers within the South African built environment sector. The study explored how emerging technologies could be used to enhance existing solutions and effectively address the challenges faced by women professionals. A qualitative study was undertaken, and a focus group discussion was held with six women registered with the professional councils under the Council of the Built Environment (CBE). The data collected was analysed using thematic analysis to identify key themes and insights. The study found that professional women in the South African built environment face persistent barriers, including a male-dominated culture, the existence of glass ceilings, and a lack of work-life balance and mentorship. While existing solutions such as mentorship, professional development programmes, flexible working arrangements, and inclusive recruitment practices were acknowledged, their implementation has been inadequate. Technological innovations such as the use of AI and VR were identified as key tools to enhance these solutions by offering e-mentorship opportunities, promoting fair recruitment practices, enabling remote working opportunities and increasing awareness of unconscious bias. Despite concerns about potential biases and accessibility challenges, the study highlights technology's potential to drive gender equality in the BE sector. This research aims to highlight critical areas for intervention and offer practical recommendations for policymakers and organisations to better support women in the Built Environment sector.
Item
Open Access
Assessing the association between depression and viral load suppression among adolescents and youth living with HIV (15–24 years) in Malawi: Insights from Malawi Population-based HIV Impact Assessment (MPHIA) 2020–2021 survey
(2026) Ndhlovu, Grace; de Voux, Alex
Background Despite advancements in HIV antiretroviral therapy (ART), suboptimal viral load suppression (VLS) persists among adolescents and youth living with HIV (AYLWH) in Malawi. Depressive symptoms are hypothesized to affect VLS, complicating HIV management. However, limited data exist from Malawi and the broader African region examining the relationship between depressive symptoms and VLS. This study examines the association between depressive symptoms and VLS among AYLWH, leveraging secondary data from the Malawi Population-based HIV Impact Assessment (MPHIA) 2020–2021. Methods A cross-sectional analysis of AYLWH aged 15–24 years was conducted. VLS was defined as HIV RNA <200 copies/mL. Depressive symptoms were assessed using the Patient Health Questionnaire-2 (PHQ-2), with a cut-off score of ≥3 indicating the presence of depressive symptoms. Weighted univariable and multivariable logistic regression analyses were performed to explore associations between depressive symptoms and VLS, adjusting for covariates such as gender, education level, type of residence, wealth quintile, ever using alcohol, psychosocial intervention participation, and awareness of HIV status. Results Among 201 AYLWH, 68% achieved VLS and 3% reported depressive symptoms. Among participants who had achieved VLS, 70% lived in rural areas, and 91% were aware of their HIV status. Depressive symptoms were not significantly associated with VLS in both univariable (OR: 0.8, 95% CI: 0.1–4.6, p=0.801) and multivariable models (OR: 0.3, 95% CI: 0.1–2.4, p=0.276). Significant predictors of VLS included type of residence (urban vs rural; adjusted OR: 0.2, 95% CI: 0.1–0.5, p=0.002) and awareness of HIV status (adjusted OR: 0.1, 95% CI: 0–0.1, p<0.001). Gender, education, wealth quintile, ever using alcohol, and participation in psychosocial interventions were not significantly associated with VLS. Conclusion While depressive symptoms were not significantly associated with VLS, the findings highlight the critical role of structural factors such as residence and awareness of HIV status in improving VLS. The low prevalence of depressive symptoms may reflect limitations in the assessment tools or underreporting. Integrating robust mental health interventions into HIV care services may address barriers to VLS and improve outcomes for AYLWH in Malawi. Further longitudinal studies are recommended to elucidate the interplay between mental health and HIV outcomes.
Item
Open Access
Short-term wind speed prediction to mitigate strong wind induced disruptions at the Cape Town container terminal
(2026) Ndayi, Vuyo Derrick; Verrinder, Robyn; Troch, Christophe
The Port of Cape Town is frequented by strong winds which regularly disrupt its operations, particularly at the Cape Town Container Terminal (CTCT). There, container cargo is handled by wind speed sensitive equipment, most notably overhead gantry cranes for ship to shore handling, which have a safe operating wind speed thresholds of 20 m/s and 25 m/s. During these adverse wind events, the safe operational thresholds of the container handling equipment is often exceeded leading to a suspension of terminal operations. To improve the port terminal's resilience to these adverse wind events, a short-term wind speed prediction model is proposed and developed with the aim of providing a proof of concept of an additional means for the port terminal operator to plan for, navigate through, and recover from these events. For this, the Long Short-Term Memory Network (LSTM) neural network is used in ensemble with the Bootstrap model. LSTM is used for the point predictions and the Bootstrap model is used to provide a prediction interval. Archived 10-minute resolution maximum wind speed and mean direction data spanning 2.5 years from an anemometer at the crane height level at the CTCT Admin Building are used to develop the model. For the short-term wind speed forecast, four forecast time horizons are considered: 30-minutes, 1-hour, 3-hours, and 6-hours. The same base model is used for each forecast horizon but with a slight modification to account for the forecast time range interval or lead time, thus four separate models are developed and assessed. In terms of point prediction, the models have mixed performance when assessed using a number of regression performance metrics. Overall, they perform best at the shortest forecast horizon (30-minutes), and the worst at the longest (6-hours). Considering mean absolute error alone, the 30-minute model had an error of: 1.546 m/s; 1-hour model: 1.883 m/s; 3-hour model: 2.860 m/s; and the 6-hour model: 3.632 m/s. The 3-hour and 6-hour models showed a notable time lag in their respective predictions. The addition of signal decomposition as a data pre-processing step significantly improved this time lag as well as the same error metric to, 3-hour: 1.908 m/s; and the 6-hour: 2.431 m/s. The prediction performance of the exceedance thresholds is also assessed in terms of how soon before or after the event the model made an accurate prediction of the exceedance. For the 20 m/s threshold, the 30-minute and 1-hour model's could on average predict the exceedance events within 69 and 73 minutes, respectively. The 3-hour model predicted half within 180 minutes whereas as the 6-hour model could not predict the events within the tolerable 360 minutes. For the 25 m/s threshold, no useful results were achieved. The prediction interval quality of the models was good. The 90% prediction interval coverage probability was 89.9% for the 30-minute model, 80.6% for the 1-hour model, 64.8% for the 3-hour model, and 69.5% for the 6-hour model. The models also had an average mean prediction interval width of 5.971 m/s. This translates to 3 m/s above and below the point prediction value which is good considering the critical threshold range of 5 m/s. The port terminal operational gains and application of the short-term wind speed forecast and medium to long-term forecasts are discussed in the context of the decisions associated with port terminal planning and logistics. The value and practical use of the developed model is described, and its limitations are also highlighted. This thesis ends with an appreciation of the developed solution in providing a simple means to mitigate disruptive wind events by aiding port operator decision making but also notes the solution's limitations in its current form.