Demand driven material requirement planning – A classification based approach for buffer parameters

The rapidly changing customer demands and expectations along with the increasing product variety lead companies to respond with complex and flexible production environments consisting of a wide variety of product trees. In this dynamic production environment required by today's VUCA world, companies compete under continuous pressure to reduce the total costs in their supply chains. Compared to existing planning and control systems, DDMRP (Demand Driven Materials Requirement Planning), which aims to manage material flow efficiently under the "new normal" conditions brought by the VUCA world, has been added as a new approach to existing production planning and control methods. Combining push and pull approaches in material flow, DDMRP uses strategically positioned and specially sized buffers in the BOM structure. This thesis aims to examine the positioning and sizing parameters of DDMRP buffers based on material groups obtained by ABC-XYZ classification analysis and to determine their effects on system performance. For this purpose, A discrete event simulation was created to model DDMRP methodology, and buffer positioning and sizing parameters with ABC-XYZ classifications were analyzed with design of experiments approach. By using the experimental design results, regression trees were created to propose DDMRP buffer and parameter structure to determine the performance indicators based on ABC categories.