Abstract:
This thesis aims to analyze the advantages and drawbacks of the different heuristics proposed in the past and the modern ones.
To do so, the introductory chapter explains the theoretical background behind the MRP, its evolution, inputs (i.e. with a focus based on BoM, MPS, and inventory), outputs, and the simplest methods to calculate the optimal order or reorder of the basic components (e.g. lot-for-lot and Economic order quantity). After that, the characteristics of the Demand Driven Material Requirements Planning are examined. This has been chosen because the DDMRP is one of the newest approaches to deal with such issues and it is discussed deeply afterwards in this paper.
The traditional methods, which are based on the exact knowledge of short-term demand, are described and implemented in the second chapter. The heuristics that have been chosen are the Part period balancing, the least unit cost, and the Silver-Meal. To do so, real demand data and the Monte Carlo method have been utilized. The latter introduces different possible scenarios on the basis of the real mean and standard deviation which is useful to complete the analysis and compare the different results. In addition, I will include the exact method based on the exact knowledge of the long-term demand for implementing the Wagner-Whitin method.
However, to have an appropriate overview is also important to compare them to the modern tools developed recently (e.g. DDMRP implemented in 2011 and others which take some principles from the traditional heuristics).
In fact, the objective of this thesis is not only to have a deep analysis of different methods for warehouse management but also to get a full comparison of them to understand the improvements and the possible future developments of such techniques.
