Zambia ICT Journal

Using Artificial Intelligence to Mitigate Monkey-Human Conflicts in Hospitality Spaces in Zambia

2025-12-03T10:11:07+00:00

This study aimed to assess the effectiveness of an Artificial Intelligence (AI)-based deterrent system in reducing monkey incursions at a hospitality establishment in Livingstone, Zambia. The research sought to understand not only the behavioural changes in the monkeys but also the perceptions of hospitality personnel regarding the system's impact on their work environment and guest experiences. The target population included free-ranging monkeys regularly intruding on the premises and 30 hospitality staff members employed at the establishment. The staff represented diverse roles, genders, and experience levels, and all had been employed for at least six months prior to the intervention. A mixed-methods approach was adopted. Quantitative data were gathered through systematic behavioural observations of monkey activity before and after system implementation. Qualitative data were collected through 30 in-depth semi-structured interviews with staff to explore their perceptions of the system’s effectiveness. Data were analysed using statistical techniques and thematic content analysis, respectively. Findings revealed a substantial decrease in both the frequency and severity of monkey incursions following the installation of the AI-based deterrent. Observable monkey behaviours shifted significantly from habituated and aggressive patterns to avoidance and flight responses. Interview data indicated improved staff morale, reduced workplace stress, enhanced guest satisfaction, and a more professional atmosphere. While some participants expressed concerns about potential long-term monkey adaptation, the overall sentiment remained strongly positive. The AI-based deterrent system proved effective in mitigating human-wildlife conflict within a hospitality setting. It created a safer, more controlled work environment and contributed positively to both operational efficiency and guest experiences. The study demonstrates that intelligent, context-sensitive technologies can yield meaningful behavioural changes in non-human species while supporting human-centred hospitality operations. This contribution extends prior Zambian AI/IoT field deployments for wildlife and pest monitoring by demonstrating effective, real-time deterrence in a hospitality context near Mosi-oa-Tunya.

Leveraging Biometric Data and Artificial Intelligence to Enhance Beneficiary Identification in Social Cash Transfer Programs: A Case Study of Crystalised Applications in Zambia

2025-11-23T17:16:17+00:00

This paper investigates the integration of biometric data and artificial intelligence (AI) to improve beneficiary identification within Social Cash Transfer (SCT) programs in Zambia. Focusing on the Crystalized Apps platform, the research examines how AI-driven biometric technologies, such as facial recognition, fingerprint scanning, and iris detection, significantly enhances accuracy, operational efficiency, and security of SCT disbursements. Utilizing a mixed-methods approach, the study combines interviews with program administrators and an analysis of transaction data to evaluate the effectiveness of AI-enhanced biometric systems in beneficiary verification, fraud reduction, and payment security. The anticipated results aim to demonstrate that biometric AI can mitigate identity-related fraud, optimize the transfer process, and promote transparency within the system. The study also acknowledges challenges including data privacy, infrastructure limitations, and digital literacy gaps, providing a holistic perspective. Ultimately, this research seeks to provide valuable insights on how AI-based biometric authentication can strengthen social protection mechanisms, improve financial inclusion, and foster greater accountability in public welfare programs.

Impact of Low Laboratory Assessment Weight on AI and IoT Skills in Engineering Education in Zambia: A Case Study of Electrical and Electronics Department

2025-12-12T08:42:00+00:00

Laboratory sessions in engineering education are very essential in giving students hands-on training to complement theoretical learning. Unfortunately, many Zambian universities attribute less than 10% of the overall course grades to practical work, leading to poor engagement and limited skill acquisition and transfer. This study investigates the impact of low laboratory weighting on student outcomes within Electrical and Electronics Engineering programs. Emphasis is placed on the increasing need for practical competence in implementing Artificial Intelligence (AI) and Internet of Things (IoT) systems, which are driving modern electrical engineering innovations. This paper analyses the engineering curriculum, case studies, and prototypes developed by final-year students, highlighting the consequences of minimal lab exposure leading to reduced innovation capacity and industry unpreparedness. A model is proposed that increases laboratory assessment weight to 30–50%, integrates interdisciplinary project-based learning, simulations, and aligns skill development with industry demands. The findings suggest that rebalancing theoretical components can significantly enhance students’ technical proficiency, system-level thinking, and readiness for AI- and IoT-driven environments.

Multimodal Deep Hashing Biometric Authentication Systems Based on Neural Networks Regional Applications in Digital IDs

2025-11-23T17:36:06+00:00

With a focus on applications related to digital identities, this paper provides an extensive overview of multimodal deep hashing biometric authentication systems. We lay out precise research goals and examine the most recent approaches, such as privacy-preserving strategies and deep neural architectures. Modern multimodal hashing frameworks are identified, template security and system interoperability issues are evaluated, and future research directions are recommended. We employ a systematic literature search with clear inclusion/exclusion criteria and categorize the works by technique (e.g., CNN, RNN, Transformer), application domain, and modality (e.g., face, fingerprint, iris). We discuss recent developments, including transformer-based biometric models [2][3] and privacy techniques (secure sketches, homomorphic encryption) [4][5]. Key studies are compiled in a standardized comparative table. With an emphasis on open-source platforms (like MOSIP [6][7]), privacy-by-design, and economic effects, we cover policy frameworks (GDPR, eIDAS, and African Union privacy charters) and provide helpful suggestions for implementing digital ID systems in Africa. Future studies and the implementation of safe, privacy-conscious biometrics for identity programs are intended to be guided by our findings.

A Context-Aware End-to-End Predictive Analytics Architecture for Cholera Early Warning and Medical Resource Allocation Using Machine Learning

2025-12-12T09:11:14+00:00

This paper details the architecture, development, and assessment of a comprehensive predictive analytics platform designed to forecast cholera outbreaks and optimize the allocation of medical resources within Zambia's public health system. Conventional cholera management in the region is predominantly reactive, resulting in operational delays and suboptimal resource deployment. This research confronts this issue by creating a localized, proactive decision-support tool. The system utilizes a hybrid modeling approach, combining supervised learning algorithms (Logistic Regression, Random Forest, XGBoost) with a linear programming (LP) model for resource optimization. A comparative analysis was performed using a synthetic dataset from 2017-2024 that mirrors Zambia's epidemiological trends. The XGBoost model yielded the most effective performance for an early warning system, attaining an accuracy of 84.69%, a flawless recall of 1.0, and an AUC-ROC score of 0.9361. Conversely, the Random Forest model provided perfect precision (1.0) but with a minimal recall of 0.125, underscoring significant performance trade offs. The resulting prototype, which includes an interactive Streamlit dashboard, effectively transforms predictive outputs into actionable resource allocation strategies, offering a scalable and data driven solution for epidemic preparedness in resource limited settings.

Deep Learning Applications in Maize Disease Detection: A Systematic Review of Trends, Gaps, and Future Research

2025-11-23T17:58:01+00:00

Maize, a staple crop globally, faces significant threats from various diseases that can drastically reduce yield and quality, impacting food security and economic stability, particularly in regions heavily reliant on agriculture. The search for automated and effective diagnostic tools is driven by the labor-intensive, time-consuming, and error-prone nature of traditional illness detection methods, which frequently rely on visual inspection and expert knowledge. In recent years, deep learning methodologies have emerged as a transformative force in plant disease detection, exhibiting remarkable capabilities in image recognition and classification, surpassing the limitations of conventional machine learning techniques that necessitate manual feature extraction. Deep learning models, highlight key trends, including the increasing use of convolutional neural networks, transfer learning techniques, data augmentation methods, and real- time disease detection using mobile applications. The paper also identifies several gaps in the current research, such as limited diversity in maize disease datasets, insufficient focus on early-stage detection, lack of standardized evaluation metrics, and inadequate consideration of environmental factors. These models have demonstrated proficiency in learning intricate features directly from raw image data, enabling accurate and rapid identification of diseases in maize crops. The paper further outlines potential future directions for research, including the development of more comprehensive datasets, exploration of multi-modal deep learning approaches, investigation of explainable AI techniques, integration with IoT devices, adaptation of models for different maize varieties and growing conditions, incorporation of temporal data, and development of hybrid models. To address the identified gaps, the paper suggests collaborating with agricultural experts, developing models for multiple disease detection, and investigating unsupervised and semi- supervised learning approaches.

Cyber Physical Framework for Mining Pollution Remediation Using RISC V and AI in Zambia

2025-12-12T09:41:26+00:00

Anthropogenic discharge of heavy metals, acids and suspended particulates from intensive mining operations along Zambia’s Kafue River has triggered a persistent ecological crisis. Conventional monitoring—periodic grab sampling followed by laboratory analysis—offers limited temporal resolution and cannot capture rapid pollution spikes.