Book Chapters

Permanent URI for this collection

http://hdl.handle.net/10204/11484

Browse

Now showing 1 - 8 of 8

Active nanocomposite films based on low density polyethylene/organically modified layered bouble hydroxides/thyme oil to retain retail shelf life and quality of hass avocados
(2020-11) Kesavan Pillai, Sreejarani; Sivakumar, D; Ray, SS; Obianom, P; Eggers, SK; Mhlabeni, T
In this study, the ability of an active film containing volatile bioactives in post-harvest disease control and preservation of quality in avocados is explored as a non-traditional treatment method. Antimicrobial transparent flexible trilayer low density polyethylene (LDPE) films containing organically modified layered double hydroxides (OLDH) and plant bioactive-thyme oil (TO) were made using single step blown film extrusion. Antifungal effects of the packaging in comparison to commercial treatment and untreated control showed considerable reduction in anthracnose disease events in ‘Hass’ cultivar of avocados while improving the fruit quality. 2wt% OLDH loading improved the oxygen and moisture barrier properties while not affecting the transparency of the film. The results suggest that the synergistic effect of barrier and antimicrobial properties of the controlled volatile bioactive release of the nanocomposite film can be utilised as a prospective strategy to modify the headspace gas composition to combat anthracnose disease in avocados.
Development of a new asphalt mix design manual for South Africa
(2015-08) Anochie-Boateng, J; O’Connell, JS; Verhaeghe, BMJ; Myburgh, P
A need existed to update the South African design methods for asphalt mixes, particularly in the light of current developments in the country such as the revision of the South African Road Design System (SARDS), and the increasing demand for, and the use of products such as high modulus asphalt mixes with reclaimed asphalt and slags, warm mix, and cold mix as alternative mixes to conventional hot-mix asphalt. Generally, there is a worldwide shift from empirical-based asphalt mix design approach towards performance-related design approaches, due to advances in asphalt technology, increased volumes of heavy vehicles on roads, and there is also a demand for higher performance mixes, and a need to review the current criteria for asphalt layers in contract specifications. This paper presents the development of a new asphalt design manual. The paper presents key highlights such as (a) the performance grade binder selection methodology in which the binder is selected based on loading and environmental conditions, as a replacement of the traditional penetration grade binder selection method, (b) the move from aggregate grading bands (as per the current South African Committee of Transport Officials specifications) towards the use of control points to select the design aggregate grading, and (c) the three levels of asphalt mix designs proposed for the manual. The manual however, requires validation through additional laboratory testing before it can be used with confidence by industry.
Ground-based surveillance and classification radar for wildlife protection
(2023-11) Berndt, Robert J; Wabeke, Leon O; Van Rensburg, Vanessa; Potgieter, Pieter F; Kloke, Kevin H
Wildlife poaching has become an increasingly large problem over the last decade in southern Africa. There have consequently been worrying declines in rhinoceros population levels. As part of a multi-year effort to investigate and develop technology to aid in the prevention of poaching activities, a C-band phased array radar, with integrated ground target classification capability, has been produced. In a field trial that took place in the Kruger National Park Game Reserve in South Africa, it successfully detected and classified human and animal movement at ranges of up to 10 km. The misclassification rate of humans was less than 7%, and non-human movement, predominantly from animals and vehicles, was incorrectly classified as human 3.8% of the time. This shows that modern radar, without assistance from other sensors, has great potential to contribute towards the preservation of scarce wildlife resources.
Integrated framework for enhancing SDN security and reliability
(2023-11) Isong, B; Ratanang, T; Gasela, N; Abu-Mahfouz, Adnan MI
This paper addresses the issues of fault tolerance (FT) and intrusion detection (ID) in the Software-defined networking (SDN) environment. We design an integrated model that combines the FT-Manager as an FT mechanism and an ID-Manager, as an ID technique to collaboratively detect and mitigate threats in the SDN. The ID-Manager employs a machine learning (ML) technique to identify anomalous traffic accurately and effectively. Both techniques in the integrated model leverage the controllerswitches communication for real-time network statistics collection. While the full implementation of the framework is yet to be realized, experimental evaluations have been conducted to identify the most suitable ML algorithm for ID-Manager to classify network traffic using a benchmarking dataset and various performance metrics. The principal component analysis method was utilized for feature engineering optimization, and the results indicate that the Random Forest (RF) classifier outperforms other algorithms with 99.9% accuracy, precision, and recall. Based on these findings, the paper recommended RF as the ideal choice for ID design in the integrated model. We also stress the significance and potential benefits of the integrated model to sustain SDN network security and dependability.
Machine learning – Imaging applications in transport systems: A review
(2023-11) Adams, A; Abu-Mahfouz, Adnan MI; Hancke, GP
Transport systems are fundamental to supporting economic growth, and the need for smarter, safer, more efficient and climate neutral transport systems continues to grow. Maintenance and operation of transport infrastructure is expensive and has many difficulties. In recent years, the application of machine learning to solve problems has been pursued with varying success rates. Many open problems still remain at this stage. This paper provides a review of deep learning applications in transport systems. Multiple deep learning applications are discussed e.g. railway safety, self-driving cars, pedestrian crossing and traffic light detection. Reviewed papers are evaluated in terms of challenges, contribution, weakness, research gaps. Key research questions for future study are proposed: performance optimization, data set improvement and the need for research on real-time performance on edge devices.
Recommendations for sustainable post COVID-19 eLearning development in rural schools in South Africa
(2023-12) Mthethwa, Siyabonga; Nthatheni, Ronewa G; Maremi, Keneilwe J; Thulare, Tumiso
Prior to the COVID-19 pandemic, many ICT initiatives were implemented in rural schools across the country to improve teaching and learning. This paper discusses the various factors that contributed to the failure to sustain ICT initiatives. Due to the risk of infection associated with face-to-face contact during the COVID-19 Pandemic, eLearning proved to be the sole means of ensuring educational continuity. The objective of this paper is to highlight factors affecting sustainability and provide recommendations that can assist in ensuring sustainable ICT initiatives in rural schools after the COVID-19 pandemic. This is accomplished by conducting a scoping review to investigate the sustainability of ICT infrastructure to enable eLearning. The following academic databases are used for scoping literature reviews: ScienceDirect, IEEE Xplore, and search engines such as Google. A total of 41 papers out of 800 publications were deemed appropriate for the analysis. The findings show that physical, management, social, and educational conditions all have an impact on the sustainability of ICT initiatives; thus, addressing these components can ensure sustainability.
A review of intrusion detection techniques in the SDN environment
(2021-11) Sebopelo, R; Isong, B; Gasela, N; Abu-Mahfouz, Adnan MI
Despite the advantages of Software-defined networking (SDN) over the traditional networks, SDN is facing several challenges such as security threats and attacks, dominated by a distributed denial of service (DDoS) attacks that target the controller. In recent years, the SDN has witnessed several research attentions leading to proposals and the development of countermeasures such as intrusion detection systems (IDS). IDS plays a critical role in detecting and preventing malicious activities on the networks. Several detection techniques have been exploited for the effectiveness of the IDS such as pattern matching, anomaly-based and specification-based. With the nature of SDN architecture, flow-based anomaly detection has been effective and commendable. Therefore, this paper conducted a review of some of the IDS schemes in the SDN environment. It was aimed to identify the solution offers, techniques, challenges and provide research directions. The findings show that IDS in the SDN is an active research area and several techniques exist and are dominated by machine learning (ML) which exploits the network traffic flow to detect abnormal behaviours. Intrusion detection on the SDN is still at large and more ML techniques needs to be explored, considering the critically of the SDN controller.
Towards a privacy compliance assessment toolkit
(2023-06) Moabalobelo, Terrence; Ngobeni, Sipho J; Molema, Bokang C; Pantsi, Phumeza N; Dlamini, Thandokuhle M; Nelufule, Nthatheni
The South African Protection of Personal Information (POPI) Act No.4 of 2013 makes it illegal to collect, use, process or store personal information unless it is done in accordance with the prescribed legal and regulatory clauses enriched in the Act. Organizations should take stock of the personal information they collect and who they share it with before they can put controls in place to safeguard it. Failure to comply with POPI Act may potentially expose the responsible party and its associated third parties to steep legal penalties including possibly imprisonment of up to 10 years or R10 million fine which is imposed by the Information Regulator of South Africa. This paper presents the results of a system called Protection of Personal Information Act Compliance Assessment Toolkit (PCAT). The main aim of this system is to assist organisations to assess their current state of compliance to the POPI Act. The PCAT solution followed an experimental research and development process, where three existing similar technologies in the market were analysed and compared to the PCAT. The results of the PCAT shows that it simplifies the POPIA compliance requirements compared to the other three existing technologies. Future work will include development of executive summary report which will make it easier for executive to view the compliance assessment in a high-level format.

Browse

Browsing Book Chapters by browse.metadata.type "Conference Presentation"

Results Per Page

Sort Options