Lagerstyrning av inköpsartiklar på FSL – Dimensionering av säkerhetslager utifrån given servicenivå

Sammanfattning: Background Frigoscandia Logistics AB, FSL, was founded in 1997 as a Fourth Party Logistics provider, 4PL, with the purpose to supply temperature-controlled food and other goods to customers world wide. The main purpose is to provide a complete logistics service package including physical, administrative as well as financial flows. Moreover FSL takes the ownership of the products, which is unique. FSL has taken over the responsibility of the whole supply chain from the supplier to the customer, which results in a complex flow that requires efficient control. FSL strives to be the number one logistic provider for the food industry and to achieve this it is of most importance to balance stock levels while at the same time keep predetermined requirements with regards to service level. The purpose of this master thesis is to, on behalf of FSL, build a model for inventory control that dimensions safety stocks on the basis of a given service level. Problem analysis FSL has chosen to determine optimal stock levels based on a certain service level, SERV2. This is defined as the fraction of demand that can be satisfied immediately from stock on hand. The problem to be investigated in this thesis is formulated as follow: • Determine safety stocks and optimal stock levels on the basis of a given service level, SERV2. Purpose The purpose of this master thesis is to develop a model for stock optimization that enables FSL to determine their safety stock on the basis of a given service level. The model will make it easier for FSL to optimize their stock levels and at the same time satisfy their customers. Method This study has a deductive approach since existing theories has been used as a starting point during all investigations and interpretations. The authors believe that a comprehensive understanding can only be achieved by objectively gathering data from the different parts. This has caused the authors to choose a participant perspective. The authors have chosen to use qualitative and quantitative methods of data congregation to generate solutions to the problem at hand. Most part of the data has been collected through interviews. Other methods that have been conducted in this study are literature studies and observations. Both primary and secondary data has been used throughout the investigation. Conclusion The investigation shows that the forecast for majority of the items are generally too high. This results in high stock levels since all purchase are done according to the forecast. The authors think that the forecast model currently used at FSL is inadequate since it flattens down seasonal variations and does not respond quickly to changes in demand. The authors suggest that FSL considers alternative forecast methods. The authors believe that exponential smoothing with trend would be a more suitable forecasting method to apply at FSL. In addition a seasonal index, which takes seasonal variations into account, for each group of items should be calculated. Investigation reveals positive forecast errors, for the majority of the items, as a result of high forecast. Although FSL believes that MAD is a good way to measure forecast errors, they do not apply it in their work routines. FSL would gain better control of their forecast if they were to measure and analyze the mean average deviation, MAD. FSL uses traditional inventory classification which they call 1, 2, 3 classification. Class 1 defines articles with 99 percent service level requirement and consists of high volume items. Typically class 1 contains 10 percent of the items. At FSL they make up for 36 percent of the stock which could make it complicated for FSL to differentiate the resources. When placing orders neither the total costs, holding and ordering costs, nor the service level requirements are taken into consideration. The designed model takes both parameters into account and is therefore suitable as a complement to existing systems, as well as a support for supply planners in their work. The model also allows changes in parameters such as service level, lead-time, holding costs and ordering costs. All these parameters have an impact on the calculated safety stock. When using the model, results show that current safety stock levels can be decreased for the majority of the articles. The investigation also reveals that current safety stock levels are similar to safety stock levels calculated by the model. If the total holding costs and ordering costs are taken into consideration when placing orders, FSL will be able to reduce capital tied up in inventory and the resources can better be divided by also considering the predetermined service level requirements. This will result in more cost efficient stock control which in turn enables FSL to better serve their customers.