Doctoral Degrees (Industrial Engineering)

Permanent URI for this collection

https://scholar2.stb.sun.ac.za/handle/10019.1/775

Browse

Now showing 1 - 5 of 112

The development of a platform using digitalisation and networked modules for forecasting, planning and management to facilitate long-term success of SMEs in South Africa
(Stellenbosch : Stellenbosch University, 2023-12) Du Plessis, Carl Jan; Prof. Schutte, Corné; Prof. Dr. Eng Hummel, Vera; Stellenbosch University. Faculty of Engineering. Dept. of Industrial Engineering. Engineering Management (MEM).
ENGLISH ABSTRACT: The high attrition rate of Small and Medium Enterprises (SMEs) in South Africa, with a staggering 70% ceasing operations within their initial two years, is a pressing concern extensively highlighted in academic literature. This dissertation presents an innovative approach to support the growth and success of these SMEs by holistically addressing the multifaceted challenges they face, such as limited education, restricted _nancial access, and inadequate management skills. Current interventions often fall short due to their sector-speci_c focus, lack of adaptability to diverse user contexts, and scalability challenges, further exacerbated by low adoption rates among SMEs. To bridge the identi_ed gap in existing solutions, the study poses a central research question: How can a con_gurable, adaptable, and accessible platform be developed to holistically address the challenges faced by South African SMEs, thereby bolstering their prospects for long-term growth and success? The proposed platform is then developed into a prototype, which is validated in real-world use cases across the services, online retail, and subsistence agriculture sectors. The _ndings from these implementations underscore the platform's potential in facilitating long-term success. This research lays the foundation for further advancements aimed at strengthening the SME sector in South Africa, with the overarching ambition of fostering a vibrant and resilient national economy. This research introduces _ve unique contributions. Foremost is the development of a comprehensive set of networked modules, tailored specifically for South African SMEs. These modules holistically address the multifaceted challenges that SMEs encounter. Bolstering this is a novel platform design, informed by a synthesis of insights from earlier research objectives. This design serves as a roadmap for devising solutions essential to the long-term success of SMEs. The third contribution, inherent in the platform design, is the integration of strategic business management systems, machine learning, and digitalisation. This multi-pronged approach, drawing on the core tenets of industrial engineering, has culminated in a platform tailored to augment SME success in South Africa. Furthermore, the establishment of a database backend for operational planning and operations management dispels the conventional complexities SME stakeholders face, facilitating seamless business performance management. Lastly, building on these foundational elements, an automated mechanism for deriving use case-speci_c KPIs has been introduced. This mechanism leverages the intricate relationships among measures, sensors, and objectives, with machine learning serving as the catalyst for producing KPIs precisely attuned to speciFIc business objectives.
Qualification and certification of laser powder bed fusion for aerospace applications: a model-based production systems engineering approach.
(Stellenbosch : Stellenbosch University, 2023-12) Gibbons, Duncan William; van der Merwe, Andre Francois; Stellenbosch University. Faculty of Engineering. Dept. of Industrial Engineering. Engineering Management (MEM).
ENGLISH ABSTRACT: Qualification approaches to aid the certification of additive manufacturing are being widely researched by academia and the aerospace industry due to the potential benefits this technology offers once industrialised. Such benefits include the ability to produce lightweight structures, reduced material waste, the ability to produce unique and complex structures, and production is economical to produce small batches when compared with some traditional manufacturing processes that are reliant on extensive tooling. However, there are challenges hindering the wider adoption of metal additive manufacturing processes in the industry. Such challenges include production controls, data management, process characterisation, material and product traceability, and a general lack of additive manufacturing qualification and certification guidance material, particularly for sub-tiered production and manufacturing organisations. This research aims at developing a production system model that defines the production system lifecycle in terms of qualification and certification, and the standard production operations for laser powder bed fusion production. This model aims at capturing the current additive manufacturing and aerospace production best practices to reduce the steep learning curve that organisations experience when implementing and industrialising new production processes such as laser powder bed fusion. A mixed-method research approach utilising both qualitative and quantitative methods was erformed. A systems engineering methodology was applied which utilised elements of design science research and model-based tools and techniques. Interviews, surveys, observations and benchmarking, and case study research methods were used during the design of conceptual production system models and during model evaluation phases. The production system model was implemented at local industrial and academic facilities. Four test cases were carried out to gather test data and evaluate the production system operation to assess the quality of the developed model. Mechanical and material testing was performed to evaluate the material and articles produced by the developed production system. The developed production system model consists of context and conceptual, operational, logical, physical, and instantiated architectural views. The model addresses production activities from an aerospace part manufacturer and producer perspective, design activities are excluded from the scope of this research. An operational architecture was modelled that defines the production system lifecycle from installation through qualification phases to ongoing production. A production system architecture was modelled that defines the standard laser powder bed fusion production operations. The production system produced material that conforms with industry specification requirements and is comparable to its wrought counterparts. An initial production run of structural components was performed to demonstrate the production system for the full product lifecycle. The use of a model-based system engineering approach for production system design improves information traceability, structuring production facilities, mapping information and material flows, controlling processes and parameters, and implementing production processes. Such aspects are important for achieving qualification and certification in the aerospace industry. Using the model, production and process controls are defined and part quality can be controlled. The developed production system model acts as a single source of truth and a mechanism for communicating production information with stakeholders. The developed architecture and model provide value as a reference for the industry for laser powder bed fusion production. The model can be used as a benchmark for future additive manufacturing and production system development undertakings and for the design and structuring of additive manufacturing quality management systems.
Development of an indoor positioning system to create a digital shadow of production plant layouts
(Stellenbosch : Stellenbosch University, 2023-12) Hermann, Julian; von Leipzig, Konrad; Hummel, Vera; Basson, Anton Herman; Stellenbosch University. Faculty of Engineering. Dept. of Industrial Engineering. Engineering Management (MEM).
ENGLISH ABSTRACT: Plant layouts were historically considered to be very static and non-volatile structures. With the increasing influence of internal and external factors causing turbulence in enterprises, it became necessary for enterprises to be able to adapt themselves to short-term influences in order to sustain optimal operational efficiency. One possible adjustment tool for such an adaptation is the plant layout through the rearrangement of the individual facilities inside the plant. Consequently, the life cycles of plant layouts in modern industries have considerably shortened due to the impact of dynamic factors causing turbulences. However, changing the plant layout is a complex undertaking as it has a considerable effect on the underlying interconnected processes in production. As the information about the plant layout is crucial and considered as master data in production systems, it is essential that this data remains accurate and up-to-date. The objective of this dissertation is therefore to develop an indoor positioning system that allows the creation of a digital shadow of the plant layout in order to continuously represent the actual state of the physical layout in the virtual space. In order to define the requirements for such a system, potential stakeholders who could benefit from a digital shadow in the context of the plant layout were analysed. In order to generate added value for their work, the requirements were derived from their perspective. As the core of an indoor positioning system is the sensory aspect to capture the physical layout parameters, different potential technologies were compared and evaluated in terms of their suitability for this particular application. Derived from this analysis, the selected concept is based on the use of a pan-tilt-zoom (PTZ) camera in combination with fiducial markers. In order to determine specific camera parameters, a series of experiments were conducted which were necessary to develop the measurement method as well as the mathematical calculation method and coordinate transformation for the determination of poses (positions and angular orientations) of the respective facilities in the plant. In addition, an experimental validation was performed to ensure that the limit values for individual parameters determined in the requirements analysis can be met. Subsequently, the overall system was developed and implemented as a prototype. Especially the interfaces and data formats for the transfer of the collected data to typical software systems used by the stakeholders were considered. The functionality and the added value of the indoor positioning system were then validated in a case study. Thereby, the benefit of the created digital shadow of the plant layout was examined from different perspectives. The findings revealed that utilising the developed indoor positioning system for the automated capture of the plant layout and its subsequent transfer to other systems offers significant advantages in terms of accuracy and effort reduction as compared to other existing methods. This integration of the automatically updated plant layout into the production master data proved to be particularly beneficial for plant simulation, material flow analysis and route planning.
Development of a quality management framework for powder-based additive manufacturing systems
(Stellenbosch : Stellenbosch University, 2023-12) du Rand, Francois; van der Merwe, Andre Francois; van Tonder, Petrus Jacobus Malan; Stellenbosch University. Faculty of Engineering. Dept. of Industrial Engineering. Engineering Management (MEM).
ENGLISH ABSTRACT: With the rise of the fourth industrial revolution, powder-bed-based Additive Manufacturing (AM) technologies have been rising alongside conventional manufacturing technologies in regulated industries such as aerospace and medicine. In recent years, the global drive has been to guarantee the quality of parts manufactured using these AM technologies to the same level as conventional technologies. While a significant portion of research was conducted on verifying part quality as part of the postprocessing process, this can usually only be done using non-destructive testing (NDT) methods. However, these processes are often expensive and time-consuming; thus, a requirement was identified for the in-situ monitoring of the part manufacturing process. There have been several studies that have attempted to address this requirement. Still, most of these studies have only focused on detecting defects that may occur during the build process and, in some cases, the classification of defects according to the defect type. The aim of this study was focused on developing a monitoring system that can be used to monitor the quality of the powder bed surface and, in the future, provide closed-loop feedback to the machine control system about the state of the powder bed surface. For the development of such a closed-loop feedback system, it is necessary to classify defects based on their type, severity, and position on the powder bed surface. This type of closed-loop feedback system is not yet implementable due to the proprietary nature of the machine control systems and manufacturer hesitance toward un-validated autonomous feedback systems. However, it is envisioned that with the correct frameworks in place, this may soon become a reality. Based upon these requirements, the first half of this study was primarily focused on developing a framework that can be used to classify defects according to the defect's type, severity, and position on the powder bed surface. The study also focused on how the framework could possibly be used in the future to implement an autonomous closed-loop feedback system that can apply corrective actions to the defects on the powder bed surface. The second half of this study was focused on the physical development of a monitoring system that could be used to monitor the powder bed surface. This monitoring system had the capability to autonomously detect and classify the defects present on the powder bed and then further process these defects according to the developed framework. This physical implementation of the monitoring system was then used to process images that were captured of real-world build jobs. The results recorded using this monitoring system were then evaluated, and it was proven that the proposed framework could be used to successfully classify these powder bed surface defects and provide feedback to the machine operator. These results demonstrated that the proposed framework could be used to create the foundation for further developing a closed-loop feedback system for powder-bedbased AM technologie
A new vehicle routing problem for increased driver-route familiarity
(Stellenbosch : Stellenbosch University, 2023-12) King, Jacobus Coenraad Petrus; van Vuuren, J. H.; Toth, P.; Stellenbosch University. Faculty of Engineering. Dept. of Industrial Engineering. Engineering Management (MEM).
ENGLISH ABSTRACT: Practical challenges often arise when implementing solutions that stem from solving vehicle routing problem instances. Unplanned external events can result in increased vehicle travel times and subsequent degradations in supply chain operational efficiency. Moreover, drivers tend to get lost and/or often travel on roads that are not suitable for their delivery vehicles when they are unfamiliar with delivery routes, especially when these routes differ significantly from one day to the next. A possible solution, aimed at streamlining the practical implementation of planned delivery routes, is therefore to attempt to increase driver-route familiarity. A novel framework, called the familiarity vehicle routing problem (FVRP) framework, is proposed in this dissertation for improving the practical implementation of planned delivery routes by introducing increased driver-route familiarity into vehicle delivery routes. The FVRP framework consists of two phases — a strategic phase and an operational phase. During the strategic phase, a set of standard delivery routes visiting each customer along a specified number of different approaches is generated for a depot and the customers it services. These routes are called master routes and are then used as blueprints for daily planning purposes when actual delivery routes are computed during the subsequent operational phase. Delivery vehicle drivers are thus afforded the opportunity to become familiar with the master routes, which is anticipated to increase the efficiency with which they will be able to perform deliveries in the long term (if their actual delivery routes do not deviate too much from these master routes). Two novel mathematical models and accompanying approximate solution approaches are proposed for the different phases of the FVRP. The (single-objective) mathematical model for the strategic phase is concerned with generating a minimum-cost set of master routes for a given depot and the customers it services. The set of arcs that form these master routes represent road links with which delivery vehicle drivers may become increasingly familiar as they continue to travel along them during future deliveries. The set of master route arcs are provided as input to the (bi-objective) mathematical model proposed for the operational phase of the FVRP. This model is concerned with computing multiple trade-off solutions which can serve as actual delivery routes along which the objectives are to minimise transportation cost and to maximise the portion of the total distance travelled along the master route arcs. Furthermore, a novel recycling heuristic is proposed which facilitates the use of historical solutions when generating initial solutions for the approximate solution approach of the operational phase. The framework and computerised implementations of its components are finally applied to a special case study, involving real-world data, in order to demonstrate the practical applicability of the work reported in this dissertation.

Browse

Recent Submissions