This thesis is focused on a simulation modelling approach to address the inventory inaccuracy problems in a warehouse operation. The main motivation which led to this research was a desire to investigate the inventory inaccuracy issues that have been highlighted by a logistics company. Previous and current research into inventory inaccuracy issues is largely related to the development of RFID technology as a possible solution to inventory problems. Since the inventory inaccuracy related to RFID technology is focused on the overall measurement of inventory management and retail business, there are differences between this existing research and the research presented in this thesis which is focused on issues of inventory inaccuracy in a warehouse operation. In this thesis, warehouse operation is studied as a detailed sequence of processes that are involved in the flow of items physically in parallel with related information being stored in the computer system. In these processes there are many places where errors can occur in counting or recording details of inventory, or in physically moving, storing or picking items incorrectly. These details of a warehouse operation are used to develop a conceptual model of inventory inaccuracy in warehouse operations. The study also found that typically a product needs to be considered differently at different stages of its progress through a warehouse (and therefore within different sections of the conceptual model). This is because initially batches of a product are likely to be delivered from a supplier, therefore if errors occur soon after the product is delivered to the warehouse, the error might involve the whole batch (for example the batch may be misplaced and put in an incorrect storage location), or the error might involve just part of the batch (for example poor transportation by forklift truck may damage the packaging carton and some of the items within the carton). When the product is stored ready for meeting customer orders, it needs to be considered as individual items (and errors can occur in counting of individual items or individual items may be misplaced or stolen). Finally, when a customer order is received, the product will be picked and grouped to meet the requirements of the order (for example, one order may require 10 of the product whilst another order may require 20 of the product). Errors might again occur to the whole group or to just part of the group. (Continued ...)

This study compares the impact of dimensional modelling and data vault modelling on the performance and maintenance effort of data warehouses. Dimensional modelling is a data warehouse modelling technique pioneered by Ralph Kimball in the 1980s that is much more effective at querying large volumes of data in relational databases than third normal form data models. Data vault modelling is a relatively new modelling technique for data warehouses that, according to its creator Dan Linstedt, was created in order to address the weaknesses of dimensional modelling. To date, no scientific comparison between the two modelling techniques have been conducted. A scientific comparison was achieved in this study, through the implementation of several experiments. The experiments compared the data warehouse implementations based on dimensional modelling techniques with data warehouse implementations based on data vault modelling techniques in terms of load performance, query performance, storage requirements, and flexibility to business requirements changes. An analysis of the results of each of the experiments indicated that the data vault model outperformed the dimensional model in terms of load performance and flexibility. However, the dimensional model required less storage space than the data vault model. With regards to query performance, no statistically significant differences existed between the two modelling techniques. / MSc (Computer Science), North-West University, Potchefstroom Campus, 2014

This study compares the impact of dimensional modelling and data vault modelling on the performance and maintenance effort of data warehouses. Dimensional modelling is a data warehouse modelling technique pioneered by Ralph Kimball in the 1980s that is much more effective at querying large volumes of data in relational databases than third normal form data models. Data vault modelling is a relatively new modelling technique for data warehouses that, according to its creator Dan Linstedt, was created in order to address the weaknesses of dimensional modelling. To date, no scientific comparison between the two modelling techniques have been conducted. A scientific comparison was achieved in this study, through the implementation of several experiments. The experiments compared the data warehouse implementations based on dimensional modelling techniques with data warehouse implementations based on data vault modelling techniques in terms of load performance, query performance, storage requirements, and flexibility to business requirements changes. An analysis of the results of each of the experiments indicated that the data vault model outperformed the dimensional model in terms of load performance and flexibility. However, the dimensional model required less storage space than the data vault model. With regards to query performance, no statistically significant differences existed between the two modelling techniques. / MSc (Computer Science), North-West University, Potchefstroom Campus, 2014

<div>Smart warehouses have become more popular in these days, with Automated Guided Vehicles (AGVs) being used for order pickups. They also allow efficient cost management with optimized storage and retrieval. Moreover, optimization of resources in these warehouses is essential to ensure maximum efficiency. In this thesis, we consider a three dimensional smart warehouse system equipped with heterogeneous AGVs (i.e., having different speeds). We propose scheduling and placement policies that jointly consider all the different design parameters including the scheduling decision probabilities and storage assignment locations. In order to provide differentiated service levels, we propose a prioritized probabilistic scheduling and placement policy to minimize a weighted sum of mean latency and latency tail probability (LTP). Towards this goal, we first derive closed-form expressions for the mean latency and LTP. Then, we formulate an optimization problem to jointly optimize a weighted sum of both the mean latency and LTP. The optimization problem is solved efficiently over the scheduling and decision variables. For a given placement of the products, scheduling decisions of customers’ orders are solved optimally and derived in closed forms. Evaluation results demonstrate a significant improvement of our policy (up to 32%) as compared to the state of other algorithms, such as the Least Work Left policy and Join the Shortest Queue policy, and other competitive baselines.</div>