WIReDSpace

Welcome to WIReDSpace(Wits Institutional Repository on DSpace)

For technical support contact Lewatle Phaladi by email: lewatle.phaladi@wits.ac.za, Tel : +27 11 717 6594.

 

Communities in WIReDSpace

Select a community to browse its collections.

Recent Submissions

No Thumbnail Available
Item
Digitalization in the logistics industryas a support to business continuity amid black swan events
(2022) Fourie, Marisna
Business continuity presents itself as a challenge, especially given the global economic impact and supply chain disruptions caused by the novel Coronavirus (COVID-19), a black swan event. Digitalization refers to the implementation of digital tools that transform the current business model of an enterprise. The aim of digital tools is to provide innovative opportunities that ultimately add value to the organization. Digital tools can deliver distinct organizational advancements such as improving the information flow within an enterprise and reducing waste across the enterprise. These organizational advancements are particularly useful and even necessary in supportingbusiness continuity amid black swan events. This study aimed to investigate challenges and threats experienced by Company X, a multinational logistics company, as a direct result of the Coronavirus outbreak and how a lack of digitalization couldhinder the ability of Company X to react rapidly in response to these changes. Additionally, relevant digital technologies and its influence on the business model and culture of company X were explored. A generic qualitative research design was used to explore this phenomenon. Data collection comprisedof six semi-structured interviews, whereby the role of digitalization in the supply network of company X was explored. The findings confirmed consistency with previous research indicating that strong leadership, process standardization and data integration, a strong supportive enterprise culture with a low Resistance to change factor, employee and partner engagement, alignment in business and IT strategies, a strong emphasis on training and skills development of employees, agile transformation management and the leveraging of internal and external technological knowledge ensure successful digital transformation within an organization. The study found that a lack of resources and a high degree of complexity in underlying processes prevent logistics service providers from experiencing digital transformation. Additionally, the study found that digital tools supported business continuity during the COVID-19 Pandemic.
No Thumbnail Available
Item
The aerodynamic effect of wheel fairings applied to a generic formula one car
(2022) Davkin, Liam Michael
A reason for the lack of overtaking in Formula One racing is often attributed to the loss of downforce for a trailing car. This is due to the car operating in a highly turbulent wake aft of a car ahead of it. A loss of downforce results in a trailing car losing control margins in corners, thus failing to overtake. Early concept images for the 2022 technical regulations suggested the use of wheel coverings or fairings over the rear wheels of the cars, in an attempt to reduce the impact downstream of a car’s highly turbulent wake. In order to ascertain the aerodynamic impact of such fairings, numerical models were employed to examine the wake structure of the car. This involved numerical wake surveys of the three-dimensional velocity components in various planes downstream of a car. A car model with no wheel fairings was tested in order to develop a baseline to which comparisons could be made. Two fairing configurations were then tested, namely a partial and full fairing. Both fairings covered part of the rear wheel tread. The partial fairing differed from the full fairing by the absence of a covering over the sidewalls of the wheels. Simulations that measured the change in drag and lift coefficient of a trailing car operating in the wake further highlighted the aerodynamic impact of the fairings. 1/6 scale wind tunnel wake surveys were conducted using a traverse system and five-hole probe to measure the three-dimensional velocity components at various planar probing locations. The data from the wind tunnel wake surveys was then compared to scale numerical models in order to provide validation to the numerical model setup. The numerical model setup was further verified by means of evaluating domain size sensitivity and mesh independence. By analysing two-dimensional velocity contours and streamtraces, it was seen that the addition of both partial and full wheel fairings altered the wake structure. The addition of wheel fairings increase the size of low axial speed vortex cores with increased circulation to the sides of the wheels. The upwash flux towards the midplane of the car also increased due to the addition of fairings. Both these changes in wake structure would have adverse affects on downforce production for a trailing car by reducing the magnitude of the oncoming flow and decreasing the effective angle of attack of the front and rear wings. The trailing car simulations highlighted an increase in drag coefficient and decrease in lift coefficient due to the addition of both fairing configurations. Thus, both rear wheel fairing configurations were concluded to be ineffective in increasing the overtaking opportunities in Formula One racing. This was due to the effect of inhibiting a trailing car to follow a leading car closely because of the increase in drag and understeering in to corners due to the decrease in downforce. The traditional open-wheel configuration is thus the recommended rear wheel configuration for future car regulations
No Thumbnail Available
Item
An investigation into the selection of an optimised maintenance strategy for conveyor systems within the port of Richards Bay
(2022) Naidoo, Laventhran
This research aimed to use conveyor failure data from the PORB to: 1) Review and identify effects of the existing maintenance approach, 2) Highlight failure causes and consequences and 3) To determine the most suitable maintenance strategy for conveyors in the PORB that will reduce failures thereby reducing downtime and loss of revenue. The scope of research was narrowed to the routes which had experienced the most failures which was supported through use of the Pareto principle. The failure data was put through a logical sequence of failure analysis tools to identify cause-and-effect relationships and survey questionnaires were sent to key personnel which formed the basis for the selection of a maintenance strategy. The research has shown that a modified version of Reliability Centred Maintenance (RCM), with dominant predictive methods, will provide benefit to the business by applying the appropriate combination of maintenance strategies (RM, PM or PdM) and prioritising maintenance tasks based on the equipment and components that pose significant consequential risks.
No Thumbnail Available
Item
Demand management in health care: the case for failure demand
(2022) Hartmann, Dieter
Failure demand has been shown to have a material impact in many service industries – leading to increased waiting times and reduced system capability. The nature and impact of failure demand in health systems has however not been studied in great depth. This study proposes managing demand, and more finely, failure demand as an alternative focus for closing the gap between capacity and demand. This is contrasted against the traditional focuses on system capacity, which is raised through investment or efficiency improvements. To manage demand, the context must be understood, so a definition of the demand population for the health system is proposed, out of which a proposal is made for a mental model that describes the demand-modalities that exist in health systems. This model contains four key demand classes, namely, value- and failure - demand (using Seddon’s terminology) and expanded by adding escalation demand and false demand. Failure demand is selected for development and an algorithm is proposed that defines failure demand in a complex hierarchical organisation such as health care. A table is presented of common events that drive failure demand in health care. Leading out of this model, a health care setting is selected, in this case, a national pharmaceutical supply-chain in a developing country. The analysis was conducted by data mining order- and dispatch-documents and virtually recreating the operating history. For this, custom code was developed in Visual Basic for Applications, using a Sequential Pattern Mining approach. The Wholesale- and Distribution-networks were analysed and failure demand levels of 56 % and 29 % respectively were found in these networks. Significant service delivery improvements are foreseen if the root causes of failure demand are addressed, which in this case are mainly procurement-policy related. The study shows that failure demand in health systems represents an opportunity to narrow the capacity-demand gap by managing demand through targeted interventions.
No Thumbnail Available
Item
Machine learning approaches to estimating BMI from a small set of photographs
(2022) Pantanowitz, Adam
Body Mass Index is a common anthropometric tool for screening obesity. Instead of direct measurement, it may be useful for public health research, screening, and intervention to perform screening from common photographs. Machine learning models are used to predict body mass index from controlled photographs in a relatively small dataset consisting of 161 unique participants. Deep learning with convolutional neural networks are compared and outperform feature engineering based machine learning by approximately an order of magnitude on unseen test data, with high correlation between predicted and actual values with greater than 0.94 correlation coefficient and a mean absolute error of 1.20. Given the small set of unseen test data, data augmentation techniques are then explored: the use of automatically generating the image set from source images and combining it with the original set yielding an approximate 2 % improvement on average correlation values; and a new technique called \validation bootstrapping". This technique enables the entire model to be trained with equivalent performance from a subset of as few as seven source images with performance similar to training with a regular set with 44 or more images.