Arthur (Articles)

Permanent URI for this collectionhttp://hdl.handle.net/2263/51909

Browse

Recent Submissions

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    The implications of integrating artificial intelligence into data-driven decision-making
    (South African Institute of Industrial Engineers, 2024-11-29) Sutherns, J.; Fanta, Getnet Bogale
    Integrating artificial intelligence (AI) into data-driven decision-making offers advantages like increased performance, reduced costs and improved organisational efficiency; however, there are associated risks. The study employs a PRISMA protocol to systematically review academic articles from Scopus, ScienceDirect, and Web of Science databases to determine whether the risks AI pose are worth the rewards they offer. Literature trends reveal a growing interest in AI-driven decision-making, with significant research gaps in African contexts. The study indicates that AI is highly utilized for decision-making to foster competitiveness in manufacturing, finance, healthcare, education, and transport. Identified risks include bias, discrimination, privacy issues, and cybersecurity threats. It is highlighted that businesses need to address concerns about privacy, fairness, and transparency. Policymakers must develop ethical and legal standards besides regular monitoring and auditing of AI uses to mitigate risks.
  • Thumbnail Image
    Item
    Multifactorial optimization enables the identification of a greener method to produce (+)-nootkatone
    (Elsevier, 2024-09) Makhubela, Ida M.; Zawaira, Alexander; Brady, Dean; Pienaar, Daniel P.; daniel.pienaar@up.ac.za
    The natural aroma compound (+)-nootkatone was obtained in selective conversions of up to 74 mol% from inexpensive (+)-valencene substrate by using a comparatively greener biocatalytic process developed based on modifications of the previously published Firmenich method. Buffer identity and concentration, pH, temperature and downstream work-up procedures were optimized to produce a crude product in which >90 % of (+)-valencene had been converted, with high chemoselectivity observed for (+)-nootkatone production. Interestingly, the biotransformation was carried out efficiently at temperatures as low as 21 ºC. Surprisingly, the best results were obtained when an acidic pH in the range of 3–6 was applied, as compared to the previously published procedure in which it appeared to be necessary to buffer the pH optimally and fixed throughout at 8.5. Furthermore, there was no need to maintain a pure oxygen atmosphere to achieve good (+)-nootkatone yields. Instead, air bubbled continuously at a low rate through the reaction mixture via a submerged glass capillary was sufficient to enable the desired lipoxygenase-catalyzed oxidation reactions to occur efficiently. No valencene epoxide side-products were detected in the organic product extract by a standard GCMS protocol. Only traces of the anticipated corresponding α- and β-nootkatol intermediates were routinely observed.