Lagerstyrning och materialhantering inom Kraft Foods chokladtillverkning

Abstract

In early 2010 Mattias Andersson, head of internal logistics at Kraft Foods production of chocolate in Upplands Väsby, experienced his factory as messy and unstructured. He suspected that the factory's inventory levels were higher than what was necessary for the production plant; in order meet its commitments. This led to "waste" represented by high cost of capital and a lot of truck transportation. The messy environment also leads to difficulties with the FIFO flow as desired, as it sometimes can be difficult to find the earliest received goods. Currently there are no defined procedures for determining the location of where different goods shall be stored or for how the warehouse layout should look. Neither are there any procedures for determining the inventory control parameters and therefore one does not know whether the inventory is controlled in an optimal way. This thesis main purpose therefore was to reduce waste at the Marabou production unit through determining appropriate safety stock parameters and where various goods should be stored in the factory. Furthermore, the work was delimited to the so called D-factory, which is the part of the factory where Daim products are made. The work was also delimited to include only packing materials.<br>
The problem formulation naturally divided the thesis work into two parts, inventory control and material handling. The inventory control work was based on the theory that all uncertainties can be divided into either time- or quantity uncertainties, and that they should be treated differently; by implementing safety lead time or safety stock. Initially, different processes were mapped through interviews. Then the factory inventory levels for the past year were estimated by gathering historic inventory data, taken from the company's ERP system. To get an idea of what drives the company to keep stock, an uncertainty-identification was performed, where uncertainties that may cause the company to run out of packaging materials were identified. This was done through a brainstorming session where the entire supply chain was examined, based on interviews, observations and casual conversations with staff at the factory. Various problems and uncertainties where highlighted during the work.<br>
To get a feeling of the potential of a reduction in inventory levels, theoretically optimal inventory levels were calculated with the help of a simulation, where no uncertainties existed. This indicated a possible reduction by 59%. Through a more advanced simulation model and subsequent quantitative and qualitative analysis the different materials were divided into appropriate classes where all materials in a given class were given the same level of safety stock and safety lead time. The results of simulation and classification led to a recommended change in the primary protection for the material layer from using safety lead time to using safety stock. Without reduced protection the simulation indicates that inventory levels are lowered by 30-35%.<br>
One problem with the layout of the pallet racks in the D-factory was that they were two spaces deep, which is not preferable in facilitating a FIFO flow of material. Therefore, a new racking layout was drafted. The proposed new inventory control parameters and new inventory levels led to a release in storage space to the degree that a material handling system with fixed storage locations could be used. The benefit for the company, compared to the existing system of partly fixed and partly shared storage spaces, is that goods will be more structured and hence easier to find. With fixed spaces, an optimization problem based on minimizing the time required for handling the material could be constructed. The optimization problem was defined in order to find the most efficient storage spaces for all pallet goods. The problem was then solved with a heuristic solution method for both the existing and a revised rack layout. Since one did not previously work with fixed storage spaces, it is impossible to estimate the reduction in handling time. However, the previous unawareness of problems with rack depth and long lifting time to the upper rack levels clearly indicates reduced handling time with the restructuring of pallet goods. Other positive effects achieved with the recommended changes are that planners will order packaging material based on more solid predictions, that inventory counting will be facilitated and that the planning process in the long run will be less time consuming as fewer manual adjustments will be needed.In early 2010 Mattias Andersson, head of internal logistics at Kraft Foods production of chocolate in Upplands Väsby, experienced his factory as messy and unstructured. He suspected that the factory's inventory levels were higher than what was necessary for the production plant; in order meet its commitments. This led to "waste" represented by high cost of capital and a lot of truck transportation. The messy environment also leads to difficulties with the FIFO flow as desired, as it sometimes can be difficult to find the earliest received goods. Currently there are no defined procedures for determining the location of where different goods shall be stored or for how the warehouse layout should look. Neither are there any procedures for determining the inventory control parameters and therefore one does not know whether the inventory is controlled in an optimal way. This thesis main purpose therefore was to reduce waste at the Marabou production unit through determining appropriate safety stock parameters and where various goods should be stored in the factory. Furthermore, the work was delimited to the so called D-factory, which is the part of the factory where Daim products are made. The work was also delimited to include only packing materials.
The problem formulation naturally divided the thesis work into two parts, inventory control and material handling. The inventory control work was based on the theory that all uncertainties can be divided into either time- or quantity uncertainties, and that they should be treated differently; by implementing safety lead time or safety stock. Initially, different processes were mapped through interviews. Then the factory inventory levels for the past year were estimated by gathering historic inventory data, taken from the company's ERP system. To get an idea of what drives the company to keep stock, an uncertainty-identification was performed, where uncertainties that may cause the company to run out of packaging materials were identified. This was done through a brainstorming session where the entire supply chain was examined, based on interviews, observations and casual conversations with staff at the factory. Various problems and uncertainties where highlighted during the work.
To get a feeling of the potential of a reduction in inventory levels, theoretically optimal inventory levels were calculated with the help of a simulation, where no uncertainties existed. This indicated a possible reduction by 59%. Through a more advanced simulation model and subsequent quantitative and qualitative analysis the different materials were divided into appropriate classes where all materials in a given class were given the same level of safety stock and safety lead time. The results of simulation and classification led to a recommended change in the primary protection for the material layer from using safety lead time to using
v
safety stock. Without reduced protection the simulation indicates that inventory levels are lowered by 30-35%.
One problem with the layout of the pallet racks in the D-factory was that they were two spaces deep, which is not preferable in facilitating a FIFO flow of material. Therefore, a new racking layout was drafted. The proposed new inventory control parameters and new inventory levels led to a release in storage space to the degree that a material handling system with fixed storage locations could be used. The benefit for the company, compared to the existing system of partly fixed and partly shared storage spaces, is that goods will be more structured and hence easier to find. With fixed spaces, an optimization problem based on minimizing the time required for handling the material could be constructed. The optimization problem was defined in order to find the most efficient storage spaces for all pallet goods. The problem was then solved with a heuristic solution method for both the existing and a revised rack layout. Since one did not previously work with fixed storage spaces, it is impossible to estimate the reduction in handling time. However, the previous unawareness of problems with rack depth and long lifting time to the upper rack levels clearly indicates reduced handling time with the restructuring of pallet goods. Other positive effects achieved with the recommended changes are that planners will order packaging material based on more solid predictions, that inventory counting will be facilitated and that the planning process in the long run will be less time consuming as fewer manual adjustments will be needed.

• author: Eriksson, Johan and Norén, Patrik
• supervisor: Johan Lundin
• organization: Engineering Logistics
• course: MTT820
• year: 2011
• type: M1 - University Diploma
• subject: Technology and Engineering
• language: Swedish
• additional info: ISRN LUTMDN/MTP--5714--SE
• id: 1883111
• date added to LUP: 2011-03-31 15:49:47
• date last changed: 2011-03-31 15:49:47