<p> This thesis is an investigation of some of the aspects of costing and machining that are applicable to numerically controlled (N/C) machine tools with particular reference to the facilities at McMaster University,and is divided into two sections. </p> <p> Section A is a brief discussion of the suitability of N/C for simulation methods and a review of the principles of metal cutting and the problems involved in estimating costs. A method is devised to simulate the machining of "typical" parts that have been generated by a random strategy. The results of the simulation were used to find a relationship between the geometric parameters of each part and the time required for all the machining operations to make that part, this relationship was called the "complexity factor" for the part. Sugestions for possible future extensions to the work were made.</p> <p> Section B is a feasibility study for increasing the range of use of a Moog point-to-point N/C machine to contouring, using the computer facilities available at McMaster University. It was proposed to produce a numerical control tape to machine a general oval based on a method of approximate linear interpolations using an on-line, time sharing computer terminal and a PDP "mini computer". The contouring method was tested by machining a circular groove (an oval with equal major and minor axes) and measuring the accuracy. The possibility of extending the work to other contours and three dimensional solids is discussed.</p> / Thesis / Master of Engineering (MEngr)

Less-than-truckload industry has a valuable potential for applications of operations research in two areas, network design and efficiency improvement within existing networks. This thesis focuses on the latter, specifically the less-than-truckload terminals where cross docking operations occur. The assignment of incoming trailers to inbound docks is one of the critical decisions that affect the performance of less-than-truckload terminals. This research reviews existing models in literature and introduces an optimal mixed integer quadratic model with the objective of generating assignments that are robust against variability in system parameters such as truck arrival and service times, terminal characteristics and trailer load content. The computational limitations of the optimal model are discussed. A dock assignment heuristic is developed to overcome the computational difficulties reported with the optimal model to solve realistic size problems. It is concluded that the heuristic is generally applicable and is robust against system variably. A dynamic dock assignment heuristic is later introduced to implement the decision process at real time. It is concluded that the dynamic dock assignment heuristic is also robust against system variability. The last part presents a case study that benchmarks the dynamic dock assignment heuristic and existing static assignments at a real terminal. The results show that the dynamic dock assignment heuristic outperforms the static assignment under system variability. Conclusions and future research areas are finally addressed in the last chapter.

El plan de carga del buque debe adherirse a las instrucciones de estiba del operador del buque. Estas instrucciones especifican las características generales de cada contenedor que habrá de ccargarse. El plan de carga también debe agilizar las operaciones de transporte en la explanada de la terminal. Presentamos dos algoritmos para generar el plan de carga. El primero utiliza el método de descomposición Lagrangeana. El segundo utiliza la metaheurística tabú. Las companías navieras se enfrentan a un problema extremadamente complejo cuando intentan determinar la composición y ruteo óptimo de su flota. Presentamos un modelo y algoritmo para este problema. El modelo representa los costes operativos de una naviera. También permite la respresentación de buques con diferentes propiedades, puntos y costes de transbordo, retrasos en puerto, y la posibilidad de rechazar una solicitud de transporte. Un caso práctico explora la sensitividad de los resultados a cambios en el precio del combustible. / The vessel loading plan must comply with stowage instructions provided by the vessel operator, which specify characteristics of each container to be loaded. Additionally, the vessel loading plan should expedite transport operations in the yard. We present two vessel planning algorithms. In the first model, the vessel planning problem is formulated as a mixed integer programming (MIP) model and solved using Lagrangean relaxation and branch and bound. In the second model, a tabu metaheuristic is employed. Liner companies face a complex decision problem in determining the optimal fleet composition and routing. We present a model that captures the revenues and operating expenses of a liner company. The model allows for vessel types with different cost and operating properties; transhipment hubs; port delays; regional trade imbalances; and the possibility of rejecting transportation demand selectively. A case study explores the sensitivity of optimal fleet composition and routing to bunker costs.

PhD (Business Management) / Department of Business Management / Contemporary cloud-based computing is crucial for the efficient delivery of ICT systems to users, as well as for versatile disaster recovery and business continuity management (BCM) platforms. Based on the need for efficient and fault-tolerant port operations, this study proposes a cloud-based business continuity framework (BCM) for the container terminal operations (CTO) in South Africa. The study examined the impact of ICT systems and services on Container Terminal Operations (CTO); determined Critical Business Functions in operations that must stay operational and analysed the impact of the unavailability of ICT systems and services on critical operations. This empirical study employed both quantitative and qualitative research designs and collecting data through self-constructed and self-administered questionnaires and interviews. Quantitative data was analysed using the IBM Statistical Package of Social Science (SPSS) Version 23 while qualitative data was analysed through thematic analysis of interviews. Results show that to realize effective organizational resilience using BCM, three insights, as underlying mechanisms for assimilating cloud computing in BCM can be inferred: the first is that a digitalized BCM architecture ‘fits’ the cloud computing model. The specific BCM characteristics that befits cloud computing are continuous data processing; continuous data access and delivery; multi-platform data access; always – on ICT Services; and better decision making. Secondly, the cloud computing model is a ‘viable’ model that can contribute to managing complex organizations characterized by business units that are highly differentiated in terms of size, structure, ICT and investment levels. For such complex organizations, viability is visible in terms of how the multi-units can harness collective intelligence (CI) for more effective BCM. Collective intelligence, as a form of universally distributed intelligence that is constantly enhanced, coordinated in real time, and which results in effective mobilization of organizational competence, is a core organizing metaphor for achieving BCM in a complex organization. Lastly, to realize BCM, underpinned by greater digitalization of BCM and harnessing of CI; there is need for rethinking strategy towards adoption of an ‘Intelliport strategy’ or ‘smart’ BCM for ports, currently intertwined with the notion of the 4IR. That the ‘smart’ nature of BCM require the assimilation of 4IR technologies that enable ubiquitous presence and real time information regarding organizational processes. Adopting an ‘Intelliport strategy’ is likely to have two main implications for BCM and practice: the first is linked to the development of a circular economy, in which aggregation of BCM activities can enhance sustainable development of the seaports; and secondly, adoption of a cloud computing model that can result in the enhancement of business growth of the units promote collaborative problem solving and decision making in BCM. / NRF