Browse
Recent Submissions
Item Advanced production and security(2026-08) Koen, Hildegarde SThe rapid integration of AI copilots into productivity and professional software platforms, including Microsoft 365, integrated development environments (IDEs), Adobe applications, and engineering tools, is transforming the way researchers and knowledge workers perform their tasks. These technologies offer substantial gains in efficiency, creativity, and productivity by assisting with content generation, coding, data analysis, and decision-making processes. However, the increasing adoption of AI-enabled solutions introduces significant information security and privacy concerns, particularly when cloud-based processing is involved. Sensitive organisational data may be inadvertently exposed through user prompts, uploaded documents, or AI-generated outputs. As institutions seek to remain innovative and competitive, they must also implement robust governance, security controls, and user awareness measures to protect confidential information. Achieving an appropriate balance between technological advancement and information protection remains a critical challenge in the responsible deployment of AI copilots.Item LCBE AI assessment & opportunity framework(2026-08) CSIRThis presentation assesses the current state of artificial intelligence (AI) adoption across the six departments within the LCBE portfolio and identifies opportunities for advancing AI-enabled transformation. The findings indicate that AI use remains largely informal and fragmented, with most applications focused on report writing, document drafting, meeting summaries, data analysis, and productivity enhancement. Despite varying levels of maturity, all departments share a common ambition to transition from reactive and periodic processes toward integrated, predictive, and intelligence-driven operations. Potential applications include risk intelligence, customer engagement, operational monitoring, compliance management, legal decision support, facilities management, and business resilience analytics. The assessment highlights key barriers to adoption, including data quality challenges, system integration limitations, governance requirements, confidentiality concerns, unequal access to AI tools, and skills gaps among staff. The study concludes that successful AI implementation requires a coordinated enablement strategy encompassing people, data, technology, governance, and continuous human oversight to ensure responsible, secure, and scalable deployment across the portfolio.Item AI in ICT: What's real today, and what it takes to go further(2026-08) CSIRThis presentation examines the current state of artificial intelligence (AI) adoption within information and communication technology (ICT) environments and outlines the requirements for scaling AI-driven innovation. It presents a structured approach comprising vendor-provided AI solutions, business-built AI applications, and ICT-engineered AI systems, supported by an enabling layer focused on adoption, governance, and value realization. The discussion highlights practical AI use cases already delivering benefits, including software development acceleration, operational efficiency, document processing, intelligent chatbots, workflow automation, and productivity enhancement through AI copilots. A bursary screening case study demonstrates how AI can support document analysis while maintaining human oversight and decision-making authority. The presentation further emphasizes the critical success factors for sustainable AI implementation, including leadership alignment, data readiness, enterprise governance, responsible AI practices, workforce capability development, and effective change management. It concludes that realizing the full potential of AI in ICT requires balancing technological innovation with security, ethical considerations, and organizational readiness.Item Anemometer calibration: Accuracy, traceability and impact on wind resource assessment(2026-10) Dikgale, Moyahabo SAccurate wind resource assessment forms the foundation of every wind energy investment decision, with anemometer-based wind-speed measurements serving as the primary input for evaluating site viability and predicting future energy production. The precision and reliability of these measurements directly influence energy yield estimates, project financing, and overall bankability. Even small sensor-level inaccuracies can propagate through resource assessment models, resulting in significant deviations in projected megawatt output and financial returns over a project's operational lifespan. Consequently, reducing measurement uncertainty at the data acquisition stage is critical for enhancing confidence in performance forecasts and supporting informed investment decisions. By implementing robust measurement practices, calibration procedures, and quality assurance measures, stakeholders can minimise uncertainty, improve energy production accuracy, and safeguard long-term asset value. Ultimately, high-quality wind-speed measurement not only strengthens project feasibility assessments but also contributes to the sustainable and profitable operation of wind energy assets.Item A compact, array-ready dual-polarized stacked patch element for C-band UAV SAR(2026-05) Mafa, Ike M; Winberg, S; Ferreira, RiaanThis paper presents a compact dual-polarized stacked microstrip patch antenna designed for C-band spaceborne and UAV-mounted Synthetic Aperture Radar (SAR) systems operating in the ITU-allocated 5.25–5.57 GHz band (320 MHz). The antenna employs a novel coupling-excited feed architecture in which two orthogonal coaxial probes excite inverted U-shaped slots etched into a driven patch, enabling strong coupling and improved impedance shaping. A parasitic patch—scaled 5% smaller than the driven patch and incorporating matching rounded corners—is placed above the radiator to form a vertically coupled two-resonator system that broadens the impedance bandwidth while preserving planar symmetry suitable for array integration. Electromagnetic simulations performed in Altair FEKO demonstrate a −8 dB impedance bandwidth of 5.32–5.78 GHz (≈480 MHz), fully covering the 320 MHz SAR allocation. The antenna maintains inter-port isolation better than −17 dB across the band, with a peak isolation of −24.3 dB at 5.55 GHz; such isolation levels significantly reduce crosstalk in dual-polarized SAR channels, improving polarimetric accuracy without requiring external decoupling networks. The realized gain remains stable between 3–5.3 dBi, and cross-polarization levels stay below −18 dB in the main beam region. These results confirm that the proposed architecture offers a compact, low-profile, and wideband dualpolarized radiator well suited for next-generation space-borne and UAV SAR front ends.Item An Integrated Value-Addition in Supply Chain Network for Metal-based Additive Manufacturing(2023-10) Nzengue, ACB; Mpofu, K; Mathe, Ntombizodwa R; Oyesola, MThe increasing speed of product development and the ability to deliver complex, near-net-shaped engineering metal parts are key benefits of additive manufacturing (AM). Producing or replacing parts by leveraging AM relies heavily on supply chain (SC) functions. Inadequately managed SC could lead to lower productivity and process waste. Hence, the aim is to integrate value-adding activities into the SC network using the principles of lean manufacturing, such as Value Stream Mapping (VSM) and flowchart. A case study of a braking system manufacturer company in South Africa was used to exploit VSM and the flowchart for reducing process waste. The result of the case study revealed that the frequent use of expedited shipping increased transportation costs and lead times required excess raw material inventory, and lengthy supplier order processing. The approach demonstrated the need to adjust the business model by virtually synchronising the flow of information across the SC players in the configuration of the entire AM production and distribution activities. The study provides a framework that can guide AM organisations in improving efficiency and achieving significant cost and time savings.Item AI-Enabled R&D in ACLS: Better scientific decisions, faster translation(2026-08) Cho, Moses A; Tsekoa, Tshepo LArtificial Intelligence (AI) is becoming a transformative enabler of research and development (R&D) within the Advanced Chemicals and Life Sciences (ACLS) domain. This presentation explores how AI can accelerate the journey from scientific discovery to practical impact by enhancing decision-making, improving research efficiency, and supporting faster translation of innovation into real-world applications. Key examples include the use of computer vision in precision agriculture and food safety for automated quality assessment, yield estimation, defect detection, and waste reduction; the application of African health data to inform diagnostics, therapeutic development, and healthcare decision-making; and the use of predictive AI models in chemicals and materials science to forecast outcomes before synthesis and scale-up. The study highlights 35 identified AI applications across ACLS, demonstrating varying levels of maturity and strategic value. To maximize benefits, four strategic opportunity areas are proposed to focus investment, capability development, and organizational learning. The presentation further emphasizes the importance of tailored acceleration pathways and a shared institutional AI backbone supported by trusted external networks. Collectively, these approaches position AI as a critical catalyst for innovation, sustainability, and competitiveness in scientific research and industrial development.Item Modelling of the fluid structure interaction of the horizontal axis bio-Inspired wind turbine blade and effect on annual energy production(2026-09) Netshivhulana, A; Ngwangwa, H; Pandelani, T; Nemavhola, F; Modungwa, Dithoto MRenewable energy is needed to solve South Africa's energy problem. This paper presents an investigation on the performance of a Horizontal Axis Wind Turbine (HAWT) with bio-inspired blade design. The corrugated albatross-inspired airfoil, GOE174, was generated by modifying the baseline coordinates with amplitude functions. The 2D profile was recreated in 3D blade using Solid Edge and examined in Ansys Fluent 2025R1 CFD simulations for the angles of attack from -60° to 60°. The results showed that lift increased with increasing angle of attack and maximum performance was obtained at about AoA of 50°. Flow separation beyond this critical angle resulted in stall and a drop in lift coefficient. The airfoil was placed in the wind tunnel and held in place by a force gauge rig where lift and drag forces were measured. The amount of lift that an airfoil creates has to do with the angle of attack, speed, and other conditions. The error analysis of CFD and wind tunnel findings was quantified using variance and standard deviation for three test velocities which are 2.08 m/s, 3.47 m/s and 5.36 m/s. The models showed good agreement at low velocity with little inaccuracy o (sigma) = 0.156. Variability increased at moderate velocity 0 = 0.322 and discrepancies rose substantially at high velocity, o = 0.908, indicative of turbulence and experimental sensitivity. The comparison with smooth profiles will be done in future research, the study is limited'to revealing improved lift stability, delayed stall and increased aerodynamic efficiency. The practical implication is that such designs might be used in real world turbine blades allowing for increased energy capture and operational robustness under changeable wind conditions.Item Advanced Production & Security(2026-08) Koen, Hildegarde SThis presentation examines the strategic adoption of Artificial Intelligence (AI) within the Advanced Production and Security (APS) division of the CSIR, with a particular focus on balancing innovation, competitiveness, and the protection of sensitive information. It highlights the growing integration of AI tools into research, engineering, and operational environments, while emphasizing the significant risks associated with cloud-based AI systems, including data leakage, intellectual property loss, regulatory concerns, and the exposure of defence and sovereign information. A three-tier AI adoption model is proposed, comprising Public AI, Enterprise AI, and Sovereign/Secure AI environments, to ensure that AI usage is aligned with information sensitivity and security requirements. The presentation advocates for a “move the AI to the data” approach, whereby sensitive information remains within trusted environments. It further identifies strategic opportunities for the CSIR to become a national leader in Trusted Sovereign AI through the development of air-gapped large language models, secure knowledge management systems, and AI assurance capabilities. The document also outlines short- and medium-term AI value propositions across defence, mining, manufacturing, robotics, and cybersecurity sectors, and presents research themes and practical use cases demonstrating the transformative potential of AI in secure and mission-critical applications. Overall, the presentation underscores the importance of responsible, secure, and human-centred AI adoption to advance research, innovation, and national strategic objectives.Item Decentralized elliptic curve identity for authenticated software defined radio mesh messaging(2026-09) Sebetoa, Matlapane R; Nyareli, Teboho N; Thaba, Janes MSoftware Defined Radio mesh networks are useful for disconnected and mobile communication. However, a receiver still needs a way to verify the sender, ensure that the message was not changed, and reject old messages that are replayed. This paper combines a structured literature review with an offline experiment for a decentralized elliptic curve identity layer. The review covers security implementations for the Software Defined Radio physical layer, physical layer authentication and key generation, authentication in mobile and vehicular ad hoc networks, and trust systems based on blockchain. It identifies a gap between radio-layer protection, transmitter authentication, and cryptographic authentication at the message level for Software Defined Radio mesh systems. The proposed layer derives node identities from public keys, signs outgoing payloads, and verifies the trusted key list, signature, and replay state after the radio frame is recovered. The experiment measures payload growth, Low-Density Parity-Check block impact, payload efficiency, signing time, verification time, and security behavior for tampered, replayed, incorrect-key, and unknown-node messages. The results show that signing and verification complete in microseconds, while packet growth is the main cost for small payloads that make inefficient use of message blocks. The results indicate that decentralized elliptic curve authentication is practical for low-rate authenticated Software Defined Radio messaging when compact encoding, key identifiers, and batching are used to reduce airtime overhead.Item A review of software-defined machine learning-based congestion controls in wireless sensor networks(2026-08) Nthai, R; Kobo, Hlabishi I; Maswikaneng, S; Ndlovu, LThe rapid proliferation of the Internet of Things (IoT) has made wireless sensor networks (WSNs) a cornerstone technology for modern applications, from smart cities to industrial automation. However, the inherent constraints of WSNs, such as limited energy, bandwidth, and computational power, make them highly susceptible to network congestion. This issue leads to packet loss, increased latency, reduced throughput, and significant degradation in quality of service (QoS). Although traditional congestion control mechanisms offer some relief, they often lack the adaptability required for the complex and dynamic nature of modern WSNs. This research examines the critical role of emerging technologies in mitigating congestion in WSNs, focusing on the integration of software-defined networking (SDN) and machine learning (ML). This study provides a review of this integration, often termed ML-SDWSN, analysing how centralised, programmable control of SDN can be enhanced by predictive and adaptive capabilities of ML to create more robust congestion control frameworks. It presents a comparative analysis of traditional methods against these novel SDN/MLbased strategies, highlighting performance improvements in throughput, energy efficiency, and network optimisation. Furthermore, this study identifies key limitations and open challenges, particularly the computational overhead of implementing ML algorithms in SDWSN environments for real-time control. By synthesising the current state of the art and outlining future research directions, this review underscores the significant potential of lightweight, hybrid artificial intelligence (AI)-driven solutions to alleviate congestion, thereby enhancing network capacity, ensuring QoS, and improving the overall efficiency and longevity of these networks.Item Information ethics and social context as drivers of cybersecurity resilience in South Africa’s uneven digital landscape(2026-06) Mukondeleli, Elekanyani; Siphambili, Nokuthaba; Molema, Mmamolele SSouth Africa’s rapid digital transformation has expanded socio-economic participation while simultaneously intensifying exposure to cyber risks. These risks are unevenly distributed across society due to persistent digital inequalities, infrastructure gaps and variations in digital literacy. This paper argues that cybersecurity resilience in South Africa cannot be achieved through technical controls alone but must be grounded in information ethics and social context. Using an integrative systemic literature review, the study synthesises research on cybersecurity resilience, information ethics and socio-technical systems within South Africa and comparable environments. The findings reveal recurring socio-ethical patterns in which ethical governance, trust, digital justice, and community-level practices either enable or constrain resilience at the individual, institutional, and societal levels. The paper proposes a Socio-Ethical Cybersecurity Resilience Framework that integrates ethical governance, social behaviour, and institutional responsibility. The framework offers context-sensitive guidance for policymakers, educators and practitioners seeking to strengthen inclusive and sustainable cybersecurity resilience.Item Measurement integrity for cleaner air: Calibration and compliance in industrial emission monitoring(2026-09) Dikgale, Moyahabo SAccurate measurement of emissions from industrial stacks is critical for effective clean air management and compliance with regulatory standards. Instruments used to measure gas velocity and sample pollutants, including Pitot tubes and Cegrit samplers, provide dependable results only when supported by thorough calibration. This paper outlines calibration practices conducted using wind tunnel facilities at the Council for Scientific and Industrial Research (CSIR), which provide traceable reference conditions across diverse operational ranges. By documenting calibration protocols, correction factors, and uncertainty budgets, the study demonstrates how technical accuracy enhances the reliability of emission data. Case studies show that calibrated instruments improve dispersion assessments, facilitate compliance with South Africa’s National Air Quality Act, and furnish industries with credible evidence for cleaner operations. The results highlight that calibration extends beyond being a technical necessity; it is a critical bridge between measurement integrity, regulatory enforcement and the broader mission of sustainable air quality management.Item AI-BEI Portfolio(2026-08) Business Excellence and Integration, CSIRArtificial Intelligence's (AI) value is already being realised in the following ways: Productivity augmentation & reporting, Knowledge & information services, R&D, ethics & IP management. Its common value proposition: Speed + quality support + traceability + more time for professional judgement.Item A computational model for analysis of the dynamic response of a quarter-vehicle suspension system(2026-09) Sikhakhane, Phakamani L; Joni, Emmanuel L; Mokgotho, Lebone Y; Modungwa, Dithoto MOff road vehicle suspension performance directly affects ride behaviour, structural loading, and vehicle control under irregular terrain excitation. This paper presents a cross-platform computational framework for analyzing the vertical dynamic response of a quarter-vehicle suspension system using three modeling approaches: a reduced-order MATLAB model and multibody models in Project Chrono and MSC ADAMS. A common suspension definition and terrain input were used to compare sprung mass displacement, unsprung mass displacement, suspension travel, and sprung mass acceleration. MATLAB and Project Chrono show the closest agreement. while MSC ADAMS predicts a smoother, more damped response. The framework provides a practical basis for comparing modelling fidelity in quarter vehicle suspension analysis under off-road excitation.Item Artificial Intelligence adoption in HPC: Driving innovation, performance & development(2026-08) Nchabeleng, Phasoane R; Auret, Marius; Courtnage, Julie LHigh-performance computing (HPC) is leveraging Artificial Intelligence (AI) to: Enhance efficiency, research, innovation and decision-making across its Programmes and for content creation, reporting, scientific analysis, automation, knowledge management and technology development. HPC Programmes are adopting AI while maintaining appropriate human oversight.Item AI use & AI RDI in the Smart Society Division(2026-08) Moodley, AvashlinAI is positioned in the cluster as an operational productivity lever, an RD&I capability and a cyberinfrastructure-enabled capability across NGEI and NICIS. NGEI applies machine learning, deep learning and NLP to conduct RD&I, with emphasis on adoption for productivity, performance and thought leadership.Item Status of AI Application, Adoption and Integration within the Communications Portfolio(2026-08) CSIR Strategic Communication, CSIRCSIR Strategic Communication is drafting a communications-specific AI usage guideline to assist both practitioners and researchers with ethical and efficient use of AI.Item Evaluating effective data governance frameworks for the digital environment(2026-09) Mlambo, Tlangelani P; Zuva, TAs digital technologies rapidly advance and data generation accelerates across sectors, effective data governance has become a critical enabler of organizational performance and innovation. However, despite substantial investments in digital transformation, many organizations continue to experience fragmented data practices, weak governance structures, and inconsistent data quality, limiting their ability to leverage data as a strategic asset. While numerous data governance frameworks exist, there remains limited consolidated evidence on their effectiveness within dynamic, AI-driven digital environments. This study addresses this gap by systematically evaluating contemporary data governance frameworks and their ability to support key dimensions such as data integrity, security, interoperability, accountability, and value creation in modern digital ecosystems. Adopting the PRISMA 2020 guidelines, the study reviews peer-reviewed literature published between 2022 and 2026, sourced from major databases including ResearchGate, ACM, Elsevier, IEEE, and Scopus. After applying rigorous inclusion and exclusion criteria, 10 relevant studies were selected from an initial pool of 50 articles. The findings reveal that data governance frameworks designed for cloud-based, metadata-driven, and real-time architectures are better suited to the demands of the digital and generative AI era. In particular, the Atlan Data Governance Framework emerges as highly adaptable due to its alignment with these characteristics. More broadly, the study highlights the critical features required for scalable and resilient governance in complex digital environments. By providing an evidence-based assessment of current frameworks, this research contributes to both academic literature and industry practice, offering actionable insights for organizations seeking to strengthen data governance and maximize data value. The study further recommends the adoption of modern, flexible governance models tailored to AI-driven ecosystems. Future research should expand the analysis to other technological and organizational contexts beyond digital environments for broader findings.Item Practical deployment of a private NB-IoT network with open-source platforms(2026-08) Makhalanyane, Thapelo; Hlabishi, Kobo INarrowband Internet of Things (NB-IoT) has emerged as a Low Power Wide Area Network (LPWAN) technology for industrial and enterprise IoT applications, offering licensed spectrum operation with guaranteed Quality of Service (QoS). This work presents a comprehensive, step-by-step guide to deploying a private NB-IoT network using open-source platforms. Existing literature on open-source NBIoT implementations is reviewed, and candidate platforms, including OpenAirInterface (OAI) and srsRAN, are evaluated against commercial solutions. A suitable open-source NB-IoT prototyping platform is then selected based on its demonstrated compatibility with commercial offthe-shelf (COTS) user equipment and robust support for 3GPP Release 13/14 features. The complete system architecture is detailed, including the eNodeB implementation on USRP B210 software-defined radio hardware and its integration with a lightweight Evolved Packet Core (EPC) optimised for IoT applications. A comprehensive testing methodology is employed to progressively validate the system, from basic signal generation and synchronisation through broadcast channels and random access procedures to end-to-end data connectivity. Experimental results using commercial off-the-shelf equipment demonstrate successful deployment, achieving a measured uplink throughput of 11 kbps, a downlink throughput of 17 kbps, and latency ranging from 300 ms to 15 s, depending on signal conditions. This work provides researchers and practitioners with a fully reproducible framework for establishing cost-effective private NB-IoT networks suitable for industrial automation, smart agriculture, environmental monitoring, and enterprise IoT deployments.