African Leadership Institute

Work-Integrated Learning (WIL) is an essential part of electrical engineering education, providing students with the opportunity to take their theoretical knowledge and apply it to create professional competencies. Implementing good quality WIL requires overcoming some major hurdles in Open Distance Learning (ODL) environments, such as limited access to the physical industry, unequal mentorship opportunities and challenges in evaluating practical competencies remotely. This study outlines a proposed WIL framework using Artificial Intelligence (AI) that will be developed specifically for electrical engineering students in ODL environments. The proposed framework combines Intelligent Tutoring Systems (ITS), Learning Analytics (AL) and adaptive assessment models, allowing students to simulate actual workplace tasks aligned to defined industry competencies. Additionally, Machine Learning (ML) algorithms will analyse student interaction data to support personalised feedback, competency gap analysis and adaptive sequencing of tasks. A pilot initiative with ODL undergraduate electrical engineering students has proven that the AI-based WIL model is a superior method for developing practical competencies, encouraging greater knowledge retention, enhancing assessment reliability through robust validation techniques, etc., compared to previous WIL logbook models. The data from the pilot indicated that the AI/ML Driven WIL model has positively impacted the competencies, task accomplishment, and engagement of all student cohorts participating in the pilot project. Furthermore, many students provided qualitative feedback indicating they felt much more confident after participating in the AI-based WIL model in terms of preparedness for employment, ability to complete assignments/tasks, etc., compared to using the WIL logbook ODL. Thus, the AI-supported WIL approach provides solutions to the major weaknesses experienced in traditional methods of WIL delivery in ODL, providing scalable, data-driven supervision and transparent competency evaluations. Based on the study’s findings, there is evidence that AI-supported WIL provides an effective alternative to maintaining professional engineering accreditation while improving access to education through ODL. The structure of this study provides a clear theoretical model for incorporating AIinto competency-based engineering education in ODL environments with limited resources.

Background: Artificial intelligence (AI) is increasingly integrated into everyday communication, extending beyond functional and academic uses to influence romantic relationships, particularly among university students who are deeply immersed in digital technologies. As AI systems become more emotionally responsive and easily accessible, they play an increasingly significant role in intimate interactions, raising important questions about their psychological and relational consequences. Purpose: This study aims to quantitatively examine the relationship between AI-assisted relationship communication and boundary confusion among university students, addressing a critical gap in the existing literature on artificial intelligence and intimate relationships. Methods: This quantitative cross-sectional study involved 420 university students (210 females, 210 males) currently engaged in romantic relationships. Participants completed standardized questionnaires measuring AI-Assisted Relationship Communication (AIRCS), Boundary Confusion in AI-Assisted Relationships (BCARS), and attachment anxiety and avoidance using the Experiences in Close Relationships–Revised (ECR-R). Data were analyzed using SPSS 29.0 and AMOS 29.0, applying correlation, regression, and structural equation modeling analyses. Results: AI-assisted relationship communication was strongly associated with boundary confusion and emerged as its strongest predictor. Attachment anxiety and avoidance also significantly predicted both AI-assisted communication and boundary confusion. Conclusion: The results show that students who frequently use AI for emotional support and relationship communication are more likely to experience confusion about boundaries in their romantic relationships, especially those with insecure attachment styles. While AI can help meet emotional needs, it can also blur the line between human relationships and technology. These findings highlight the importance of teaching students how to use AI responsibly in relationships, developing clear ethical guidelines, and providing targeted support through university counseling services.

Natural compounds derived from invasive plant species are increasingly recognized as promising sources of bioactive agents for drug discovery and therapeutic applications. This work explored the chemical diversity and biological potential of Tithonia diversifolia flowers (TDF) and leaves (TDL). Sequential extraction was performed using a dichloromethane/methanol mixture, followed by phytochemical screening. Quantitative analysis revealed that TDF contained higher phenolic and flavonoid levels but lower proanthocyanins compared to TDL. Antioxidant assays (2,2-diphenyl-1-picrylhydrazyl, 2,2´-azino-bis(3-ethylbenzothiazoline)-6-sulfonic acid, and ferric reducing antioxidant power) confirmed that TDF exhibited superior radical scavenging potential. Anticancer activity assessments indicated that both extracts showed cytotoxic effects against liver, colon, breast, and lung cancer cells, with minimal toxicity to normal cells. Besides, substantial α-amylase and α-glucosidase enzyme inhibition was displayed by TDL (half-maximal inhibitory concentration values of 38.99 ± 1.10and 49.62 ± 1.21µg/mL), compared to TDF (45.26 ± 2.61and 50.53 ± 1.82µg/mL), respectively, and the acarbose standard. Antifungal evaluation demonstrated subtle potency differences, with TDF (800.0µg/mL) displaying a lower minimum inhibitory concentration than TDL (900.0µg/mL) at day 13. Collectively, these findings highlight T. diversifolia extracts as multifunctional agents with significant antioxidant, anticancer, antidiabetic, and antifungal potential, supporting their application in pharmaceutical drug development.

In this paper, we propose a novel deep Q-reinforcement learning framework employing self-attention models to address dynamic resource allocation problems in optical wireless networks. The proposed framework overcomes the limitations of traditional recurrent neural networks, which struggle with sequential data execution and capturing dependencies within sequences. By leveraging a self-attention mechanism, the model enables users to learn channel access policies in an online, fully distributed manner without requiring coordination or global network information. The key innovation of this work is the integration of a single-agent DQL with a low-complexity self-attention mechanism. Unlike multi-agent systems, which often require complex coordination, the proposed single-agent framework simplifies the learning process while maintaining high performance. Numerical results demonstrate the superiority of the proposed framework. The model achieves a 153% improvement in cumulative rewards compared to state-of-the-art GRU and Long short-term memory based models. Additionally, it reduces collisions to less than 0.5% and achieves approximately twice the channel throughput of the slotted-Aloha protocol with optimal attempt probability.

In this paper we present the development of an energy and power system modelling, simulation, and analysis (ePowerSim.jl) package in Julia programming language. ePowerSim.jl is designed to present a uniform data interface for static, quasi-static, dynamic analysis, as well as network operation optimisation. It provides a co-simulation framework for the further development and experimentation of various types of models of electric power systems components or abstract entities that have mathematical formalism or data representation. ePowerSim.jl makes extensive use of cutting edge packages such as DifferentialEquations.jl, Dataframes.jl, NamedTupleTools.jl, Helics.jl, ForwardDiff.jl, JuMP.jl, and BifurcationKit just to mention a few in the Julia ecosystem. Models of synchronous generator, synchronous condenser, excitation systems, and governors developed in the package were used to model IEEE 9 bus and IEEE 14 bus test networks and subsequently validated by a real-time digital simulator of electric power systems (RTDS). The results obtains for static and dynamic models simulation in ePowerSim.jl show a close match with a simulation of the same system in RTDS. A maximum error of 0.00001 pu and 0.0001 pu were obtained for steady states and transient state respectively. Similarly, a maximum deviation of 0.0001 pu was obtained during validation for voltage magnitude during transient state at buses in the network.

Summary Trophy hunting (TH) is a form of wildlife use in which individuals pay to hunt under regulated conditions and retain selected body parts as trophies. Tanzania permits TH in areas with different land-use designations, covering diverse habitats. However, human activities such as settlement, agriculture and livestock grazing can threaten habitats and wildlife populations, particularly as revenue from TH declines. Using satellite data from between 2013 and 2023, we quantified changes in functional area and land cover across 15 hunting blocks in northern Tanzania. Over this period, functional area in hunting blocks decreased by c. 22.4% in Game Controlled Areas (GCAs) and by c. 18.0% in Open Areas (OAs), attributed to human encroachment. Longido North GCA experienced the most substantial land-cover change, with a 76% increase in dense vegetation. An analysis comparing the 2013–2019 and 2019–2023 data revealed consistent shrinkages in grassland and mixed and dense vegetation within the Longido North GCA, Masai OA and Irkishbor OA hunting blocks. Although individual land-cover losses in the given periods remained below 50% (our threshold for major habitat loss), these cumulative changes reduce habitat suitability, posing risks to biodiversity and the sustainability of TH. This study reveals the extent of human pressures on hunting blocks and calls for integrated regional planning, active management and targeted conservation strategies.

Food insecurity and gender inequality are the major threats to achieving the United Nations Sustainable Development Goals (SDGs) 2 and 5. A cross-sectional study was conducted to assess the empowerment of women, food security, and the relationship between women empowerment and food security in the City of Ekurhuleni Metropolitan Municipality, Gauteng Province, South Africa. Quantitative data were acquired from 216 randomly sampled urban crop farmers using a survey. Food Consumption Score (FCS) was employed to measure food security, and women’s empowerment in agriculture was measured using Women’s Empowerment in Agriculture Index (WEAI). Data were analysed using descriptive statistics and an ordered logistic regression model. The results showed that the majority of the households (89.4%) had an acceptable FCS, 9.3% had a borderline FCS, and only 1.3% had a poor FCS. The results also revealed that over half of the women (57.9%) were empowered. Age, educational level, employment status, business, garden size, autonomy in production, access to and decisions on credit, control over use of income, and workload were significantly associated with food security. In conclusion, the results show that an increase in women empowerment indicators, such as access to and decisions on credit, control over use of income and workload, and food security, reflects a profound relationship, influenced by socio-demographic factors. The results propose the urgency of concerted policy implications to address the synergies between women empowerment and household food security. Addressing gender gaps in agriculture and improving women’s financial autonomy is recommended to improve household food security.