2024 Diploma Thesis Title: Μοντέλα βασισμένα στην προσομοίωση για τη βελτιστοποίηση του                                           προγράμματισμού των πόρων σε λειτουργίες της αποθήκης

Abstract

Warehouses play a key role in the supply chain, serving as a hub for the management and storage of goods transported from production units to final consumers. The overall performance of a supply chain is significantly influenced by how efficiently a warehouse is organized, both at the spatial and operational level. Resources in warehouses include both labor and material handling equipment and remain a persistent challenge for planning warehouse operations. Since the productivity of any warehouse increases with the speed at which products are moved within it, optimizing resource planning is vital to increase productivity and minimize the operational costs of the warehouse and thus the entire supply chain.

One approach to resource optimization is through simulation models. Indeed, discrete event simulation (DES) is one of the popular modelling techniques that have been used to optimize processes and resource management in warehouses. Some of the advantages of using simulation methods, are the very low cost required to represent the real processes by computers, and the exploration of multiple enactments in a very short time. In this thesis, Flexsim software was chosen for the development of models in a simulation environment.

According to what has been mentioned above, the aim of the thesis is to find the optimal strategy for the management of resources and operations of a warehouse, as well as the appropriate spatial planning, so that all its processes are completed in a specific time horizon, in an optimal way. The structure of the paper is divided into 7 different chapters. Firstly, an extensive literature review on warehouse spatial and operational planning is carried out. This is followed by a literature review on the theory of discrete-event simulation and an explanation of the key elements of the Flexsim simulation software, which will be used to build and analyze the warehouse model. Subsequently, the characteristics of the actual warehouse are presented, and the analysis of these data is presented to better represent the warehouse in the simulation model. In the next phase, a detailed description of the way the simulation model of the warehouse is executed is given. Continuing, experiments consisting of the examination of different scenarios are created in order to optimize resource planning. Each scenario involves modeling a process (e.g., receiving) and finding the optimal number of resources (e.g., workers, unloading docks, handling machines) to fulfill it within a given time horizon. Finally, the results of the simulation scenarios are presented, and the final conclusions are drawn to optimize the allocation of the warehouse resources.  Based on these results, the optimal resource allocation for the inbound processes (receiving, put-away) seems to be 4 workers and 4 material handling machines under typical warehouse workload and 6 workers with 6 material handling machines under high warehouse workload. For the outbound operations (picking, packing, replenishment) the optimal combination of resources is 7 workers and 7 handling machines under standard storage workload and 9 workers with 9 handling machines under high storage workload.

The case study concerns a warehouse that handles pallets of crates with a total of 190 different SKUs. Back-to-back racking is used to store the pallets. The storage system is mixed, and the shelving arrangement is in the form of a spine. Incoming pallets are single code, i.e. each pallet consists of boxes with the same code, while outgoing pallets are multi-code, i.e. they consist of boxes with different codes.

The warehouse operation will be investigated in two storage workload cases, normal and peak. 25 different scenarios will be considered, depending on the number of incoming pallets in each of them. The condition considered in generating the random values for the incoming pallets is that the average number of pallets for all scenarios is equal to 550 under normal storage workload and 855 under peak storage workload. For each warehouse activity, the optimal operational and resource planning design will be investigated for all scenarios.

The results are encouraging, indicating that the use of software and simulation techniques can contribute not only to improving existing processes, but also to making informed decisions, through operational research, for the achievement of any project.