Print Email Facebook Twitter Improving the purchase, nesting and order release process of a batch shop production process Title Improving the purchase, nesting and order release process of a batch shop production process Author Sturm, L.C.A. Contributor Beelaerts van Blokland, W.W.A. (mentor) Faculty Mechanical, Maritime and Materials Engineering Department Marine & Transport Technology Programme Transport Engineering and Logistics Date 2015-09-29 Abstract At the sheet metal department at Fokker Aerostructures a high variety of low demand sheet metal products are produced for use in different aerospace assembly programs. The production process operates in a batch shop style of operations in order to cope with the high-variety and low-volume nature of the product mix. In order to ensure that the products are delivered on time, the release of production orders is planned by an ERP system and production order receive a planned start date. It is the task of the production planner to ensure that the different orders are released as close to this planned start date as possible. This task is however made more difficult because of batched release of production orders. Production orders are released in batches in order to be able to cut multiple order from the same sheet at the milling process, which is the first step in the production process. Orders are released in batches in order to reduce the material losses. This batched release does however reduce the effectiveness of the production process by increasing the variation of start of the production process and thereby increasing the level of work in progress. The objective of this master thesis is determine the effects of using different sheet metal dimensions in the milling process on both the incurred material costs and the effectiveness of the downstream production process. Furthermore this research investigate if different functional process designs are able to improve the effectiveness the purchase and order release process. The scope of the research is limited to the process that takes place before the milling process step. The production process is analyzed using a combination of Value Stream Mapping and the ‘PROPER’ model from the Delft Systems Approach. These analyses are performed, in combination with a quantitative and statistical analysis, in order to uncover the operational constrains and limitations of the current order release process. From these analyses it has become apparent that in the current functional design, sheets are purchased based on an expected material usage as calculated by the ERP system. The expected material usage is an overestimation of the actual material requirements. After receiving the different materials a custom milling pattern is created in the nesting process. In the nesting process the actual material requirements are uncovered. Due to the overestimation of material usage additional production orders are released in nesting process to use up the excess ordered sheet material. Therefore two alternative future states are designed, based on lean manufacturing principles, in order to mitigate the negative effects of this material use overestimation. The different future state process designs are tested using a simulated model in order to analyze their effects the performance of the purchase and order release process. As part of the simulation model the nesting process is simulated using a greedy 2D bin packing algorithm. First the current functional design is modeled in which the influence of alternative sets of sheet dimensions are tested. Moreover in this first model the influence of different purchase and order release process parameters is put forward. In the second future state model, named feedback control, the nesting process is performed just before purchase orders are created. Thereby in this future state it is possible to determine the size of the purchase order based on the actual material usage. In the last future state, Kanban control for purchase orders is introduced in combination with using a two bin inventory of raw sheet material. Again in this model the purchase orders are based on the actual material usage instead of the overestimation of material usage as calculated by the ERP system in the current operation. The material efficiency of the different functional designs is judged based on the total material costs as calculated by the simulation model. The effectiveness of the different future states is determined based on the variation in the OTP of the order release process. Additionally the effectiveness is judge based on the average number of production orders that are released in a batch. From the simulation model it can be concluded that by changing the sheet dimensions of 20 out of a total of 86 material groups, a cost reduction of 12.500 euro can be achieved. Furthermore it can be concluded that the purchase and order release process is improved the most when implementing a Kanban control method for the replenishing of sheet material. Both the material costs as well as the effectiveness of the production process is improved the most when implementing this future state design. Variation in the OTP of the order release process can be reduced by an estimated 16%. Furthermore Kanban control reduces the average number of production order released per batch by 10%. Additionally from the simulation model it has become apparent that in the current production process design at least 6.3% of the production volume is comprised out of ineffective work release caused by the overestimation of material usage. This ineffective work release comprises out of the increase of production order batch sizes and the additional excessive early release of production orders. This ineffective work is released in order to use up the sheets that are ordered in excess of the actual material requirements. Both in feedback and Kanban control this overestimation of material usage is remove which in turn reduces the level of overproduction and increases the effectiveness of the downstream production process. The implementation of this control method is however only partially feasible because of the limitation in available storage space. Feedback control offers a good alternative for those material groups that cannot be controlled via Kanban. This functional process designs offers the same material efficiency with a slightly lesser increase of process effectiveness. This future state model offers similar results for the reduction of overproduction and an equal decrease of average number of orders that are released in a batch. The variation in OTP of this future state model is however reduced by 3% in comparison to the current process design. ? Subject lean manufacturingoperational researchcase studyprocess effectivenessmodellingbin packing algorithm To reference this document use: http://resolver.tudelft.nl/uuid:f85976bf-c67e-4243-9184-98c81f2c9d4a Embargo date 2020-08-29 Part of collection Student theses Document type master thesis Rights (c) 2015 Sturm, L.C.A.