ASSESSMENT OF THE WAREHOUSE AT
SCHOLTEN MEDISCHE GROEP
INCREASING THROUGHPUT
University of Groningen
Faculty of Economics and Business MSc Technology Management Master Thesis [final version] Diagnosis Research Project
Supervisors at University of Groningen dr. ir. H. van de Water
drs. R.A. Rozier Supervisor at Scholten
ing. P.B. de Regt – Operating director Author
ACKNOWLEDGEMENTS
Before everything else I would like to thank Scholten for allowing me to do my final research project at their company, and for providing me with everything I needed to complete the research project. I have enjoyed my time there and it has contributed greatly to this research thesis.
Also I would like to thank all the people at Scholten who, in one way or another, have helped me during the research project. More specifically I would like to thank Peter de Regt for his supervision during the project. In many meetings he helped to form and structure the research project with useful comments and feedback.
Furthermore I would like to thank Hen van de Water for being my supervisor at the University of Groningen and guiding me through the research project with useful comments and feedback. These comments improved the readability and quality of the thesis. And finally I would like to thank Robert Rozier for his comments and reviewing of the thesis.
SUMMARY
Scholten Medical Group operates as a wholesaler of disposable and investment items for the medical sector, more specifically: nursing homes, care homes, municipal healthcare organizations, rehabilitation centres, home care and hospitals. These customers have become larger groups by mergers and budgets are shrinking, which has resulted in a change in demand for Scholten. Local warehouses have been closed and most departments now order their own items, which has increased the amount of separate orders for Scholten and increased the warehouse activities without monetary compensation. These were reasons for the Scholten management to start this research, aimed at assessing the current situation and yield recommendations for improvement.
The current situation at Scholten is explored and includes the warehouse and its related activities. By assessing the warehouse, its layout and activities, recommendations are given to improve the current situation. Focusing on reducing travel distances in the warehouse to improve efficiency and order throughput.
With the usage of literature on warehousing strategies; such as layout design, picking policies, storage policies and routing policies, the warehouse at Scholten is assessed. First the current situation is described by looking at the current layout of the warehouse and its functional areas and the current method of working. Next these aspects are discussed and possible improvements are looked at, resulting in recommendations to improve the current situation.
The layout of the whole warehouse follows the flow of items through the warehouse, which is a good foundation. The bulk-area is close to the entry, where suppliers deliver the items. The picking-area is close to the depot, where picked items are checked and prepared for shipment. The sending-area is next to the exit, where transporters collect the pallets with picked items. The depot is rather small and holds many activities; therefore the depot could be viewed as a bottleneck in the process flow. It is recommended to reduce the activities inside the depot, which can be achieved by moving the sending activities to the sending-area. Another factor which hinders the flow of items is that the depot is used as temporary storage for packed boxes. It is proposed to install a roller conveyor to automatically transport the packed boxes from the depot to the sending-area, which solves two issues: the depot is no longer a temporary storage place and the transportation activity is taken away.
The layout of the bulk-area restricts the routing policies, allowing only one possible routing strategy: the return strategy. According to literature the return strategy often results in the longest routes. To reduce travel distance in the bulk-area it is recommended to create a back-aisle, which would allow for more routing strategies and likely shorter picking routes. The back-aisle can be realised by sacrificing a number of storage locations to create a passage between two adjacent aisles. Another way to reduce travel distances is to store frequently picked items near the depot. For this purpose the bulk-area is divided into three zones (A, B, C), with zone-A locations closest to the depot. Items which are often visited and stored outside zone-A could be relocated to reduce travel distances. Due to the many differences in storage location size it is difficult to relocate many items, so the most savings in travel distance will be achieved by creating a back-aisle and adopt other routing strategies.
all storage locations and their respective items, shows (in)correct usage of the locations. This analysis was done using excel and with the used excel sheet and the warehouse management system incorrectly placed A-items (frequently picked) can be relocated in a zone-A location. With some items picked a number of times a day a move of a couple of metres could yield a high reduction of travel distance.
The current method of working showed possible savings with the replenishing activities of the picking-area. Most items in the picking-area have reserve stock in the bulk-area and are replenished on set times. It does, however, occur that a location is empty. Also the replenishing process is more extensive than it should be. The warehouse management system used at Scholten is able to calculate the replenishing amounts at any time. But to be able to do so extra information about a storage location and stored item is needed. Currently the picking-area is actively replenished every day and also on Tuesdays and Fridays. Each location on the replenishing list is checked by hand and then replenished from the bulk-area. The main cause for this is that all stock is administratively stored at the standard location of an item, which often is the picking-area. By documenting the maximum amount of items to be stored at the picking-area this problem is solved and by adding information about the amount of items in a box at the bulk-area the warehouse management system will be able to calculate replenishing activities. This will save time compared to the current way the picking-area is replenished and could also prevent locations to become empty by replenishing more often.
Scholten already has a list of (top) items based on volume, i.e. the amount of items picked. This list can be used to plan replenishing activities and keep track of stock levels. A list based on pick frequency was made to create a list of (top) items which are most important for the warehouse, because they impact the warehouse (picking) activities the most. This list can be used to check whether an item is stored correctly and should be kept up-to-date (as well as the volume based list). By actively doing so, changes in demand might change the order of the list and justify the relocating of an item.
CONTENTS
ACKNOWLEDGEMENTS... 1
SUMMARY ... 2
CONTENTS... 4
1 INTRODUCTION... 6
1.1BACKGROUND INFORMATION SCHOLTEN... 6
1.2THE WAREHOUSE AND LOGISTICS ACTIVITIES... 6
2 RESEARCH DESIGN ... 7 2.1RESEARCH MOTIVATION... 7 2.2PROBLEM DESCRIPTION... 7 2.3PROBLEM STATEMENT... 10 2.3.1 Research objective... 10 2.3.2 Research question... 10 2.3.3 Sub-questions... 10 2.3.4 Boundary conditions... 11
2.4OUTLINE OF THE RESEARCH... 12
2.5RESEARCH METHODOLOGY... 13
PART I ... 15
3 LITERATURE REVIEW... 16
3.1WAREHOUSE LAYOUT DESIGN... 16
3.2PICKING POLICIES... 17 3.3STORAGE POLICIES... 17 3.4ROUTING POLICIES... 19 3.5LOGISTICS OUTSOURCING... 23 4 THE WAREHOUSE... 24 4.1THE WAREHOUSE LAYOUT... 24 4.2THE BULK-AREA... 25 4.2.1 Bulk-area layout ... 25 4.2.2 Location coding ... 26
4.2.3 Picking and routing policy ... 26
4.3THE PICKING-AREA... 27
4.3.1 Picking-area layout ... 27
4.3.2 Location coding ... 27
4.3.3 Picking and routing policy ... 28
5 CURRENT METHOD OF WORKING... 28
5.1RECEIVING GOODS... 29
5.2STORING ITEMS... 29
5.3REPLENISHING THE PICKING-AREA... 30
5.4PICKING ORDERS... 30
5.5SENDING ORDERS... 31
5.6VALUATING ITEMS AND STORAGE LOCATIONS... 31
5.7THE WAREHOUSE MANAGEMENT SYSTEM... 32
PART II ... 33
6 THE WAREHOUSE LAYOUT... 34
9.1BULK-AREA LAYOUT... 38
9.2BULK-AREA ZONE-LAYOUT... 41
9.3EVALUATING CURRENT LOCATION USAGE... 42
9.4IMPROVING LOCATION USAGE... 43
9.5OTHER IMPROVEMENTS... 44
10 THE PICKING-AREA ... 44
10.1PICKING-AREA LAYOUT... 44
10.2PICKING-AREA ZONE-LAYOUT... 47
10.3EVALUATING CURRENT LOCATION USAGE... 49
10.4IMPROVING LOCATION USAGE... 50
10.4OTHER IMPROVEMENTS... 51
11 EXTERNAL ANALYSIS ... 53
11.1OUTSOURCING... 53
11.2OUTSOURCING ORDER PICKING AND DELIVERY TO HCNON ... 53
11.3EXTENDING BUSINESS AND SERVICE... 54
11.4INTEGRATING IN THE SUPPLY CHAIN... 54
CONCLUSIONS AND RECOMMENDATIONS ... 56
CONCLUSIONS... 56
RECOMMENDATIONS... 57
REFERENCES... 59
1 INTRODUCTION
The introduction will give background information about Scholten, the company that initiated this research project. First some general information about Scholten will be given, next the warehouse and logistics activities will be addressed.
1.1BACKGROUND INFORMATION SCHOLTEN
Scholten is short for Scholten Medical Group and operates as a wholesaler of disposable and investment items for the medical sector. An example of a disposable item is powder free nylon gloves, which are used once by the customer and are then disposed off. Investment items are wheel chairs or special beds for patients. Customers include nursing homes, care homes, municipal healthcare organizations, rehabilitation centres, home care and hospitals. Scholten is brand independent, which allows independent choice of brands and product advice based on the goals of the customers.
Scholten is based in Geldermalsen, The Netherlands, and has customers all over the Netherlands. The warehouse of Scholten consists of three separate areas; a bulk-area, a picking-area and a safety-area for dangerous stock. The bulk-area is 800 square metres large and holds 624 pallet places. Stock is stored on pallets and available in carton-unit-size; a carton with more than one single-sized article inside. The bulk-area is both used as reserve-storage for the picking-area and as storage of large or low-frequency items. The picking-area is 300 square metres large and stores unit-sized items in small storage locations within aisles. Most orders are picked in the picking-area, unless the picking-area is out of stock. At which point the storage location needs to be restocked from the bulk-area. Some orders also have items stored at the bulk-area or the safety-area.
1.2THE WAREHOUSE AND LOGISTICS ACTIVITIES
2 RESEARCH DESIGN
This chapter shows the foundation for this research project. The following will be discussed: the motivation for this research, the problem description, the problem statement; including research objective, research question, sub-questions and boundary conditions, the outline of the research and the research methodology.
2.1RESEARCH MOTIVATION
Customers of Scholten keep growing larger by mergers and with all the necessary savings and budget cuts they need to watch expenditures. The customers need to save costs with the purchasing and storage of inventory items. These effects cause the customers of Scholten to close their local warehouses and order the needed items on departmental levels. This put extra stress on the warehouse employees of Scholten, who now get more orders and need to deliver on a departmental level instead of delivering to a local (central) warehouse. The worst effect of this change in the way customers order is that the extra work this yields is not compensated. Extra transportation costs are paid by Scholten and the extra work also increases the costs. Next to these changes the customers also demand a rapid delivery, preferably within 24 hours. The current time between ordering and delivery is three workdays, which Scholten believes to be too long.
The previous was reason for the Scholten management to start this research project, aimed at assessing the situation and to come up with recommendations to improve the current situation. Problems in the current situation will be explained in the next section.
2.2PROBLEM DESCRIPTION
The problem description describes the problem situations within Scholten. First the problem situations will be listed. Next the problem situations will be explained in detail. The following situations have been discovered:
• quick assignment during move to current location • customer departments now order individually • switch to a new warehouse management system • assigning items to storage locations
• controlling inventory stock-levels
Three years ago Scholten moved to their current location. With the move all items were assigned a location in the new warehouse. While they did assign some specific items to a specific location, most items were placed randomly due to the short time available. The situation was never changed, mostly because the warehouse employees have too little time to determine if items are assigned to an optimal location. Also switching an item to a different location is time-consuming, because the stock needs to be physically and digitally moved to a different location. Another reason the situation was not changed is because there is no method to effectively assign items to an optimal location.
orders they receive increased. Mainly, because each department orders separately and needs to receive their orders separately. While orders are split over the separate departments, the customer (consisting of a couple of departments) still orders about the same as before. Fig. 3 shows the extended path from customer to Scholten and back to the customer. Also the transportation to one address is often combined, yet the order for a separate department is packaged separately. In practice this means an order from an individual department is treated as an individual order, which might be combined with other orders, when the shipping address is the same. The problem for Scholten is the increased amount of orders. Each order will be picked separately, while in many cases the separate departments need the same items. When the orders were still combined pickers would only visit an item-location once. Now, with all the separate orders, the pickers need to visit an item-location more often, increasing the time needed to complete the separate orders of one customer compared to the time needed to complete a combined order of that customer.
Fig. 2. Old order procedure.
Fig. 3. Current order procedure.
The stock levels shown in the WMS for the picking-area are not the actual amounts of stock in the picking-area. This is caused by allocating all the stock of a certain item to the standard location of that item, which often is the picking-area location. However the stock level shown in the WMS is the actual current stock in the warehouse. By not accurately keeping track of the stock levels at the different locations, the WMS is unable to control the replenishing of the picking-area. The result is the replenishing process has to be done manually. The warehouse employee has to check each picking-are location on the replenishing list for available items. The employee also needs to calculate the amount needed to replenish the location, based on actual stock and maximum stock. When the employee has checked all locations on the replenishing list, the needed items have to be collected from the bulk-area. All steps of this process, except collecting the needed items to replenish the picking-area locations, can be done with the WMS.
Keeping track of stock levels is complicated because most picking-area items also have a bulk-area storage location for reserve storage or carton-size storage. Items delivered by suppliers are booked in the WMS at a standard location; if the item has no standard location, a random location is chosen by the warehouse employee. If an item has a storage location in the picking-area, this location is the standard location. The standard location will show up on the picking list as the first location to pick the item from. Most items have a picking-area storage location. Any incoming stock of those items will be booked in the WMS on the standard, picking-area storage location. In practice most of the stock is actually stored at the second, bulk-area, location. In more detail, at Scholten there are three kinds of item storage possibilities, shown in Fig. 4. Item x is the most standard item storage possibility. Most items at Scholten have a pick-area location and reserve stock at a bulk-area location. An example of item x are examining gloves, which are stored in the picking-area as well as in the bulk-area. The bulk-area holds boxes which contain ten small boxes of gloves. The picking-area holds these small boxes. The picking-area location is the standard location and is the first location shown on a pick list. When a quantity of ten or a multiple of ten is ordered, a large box is picked from the bulk-area, otherwise the ordered quantity is picked from the picking-area. Item y is the way “in-and-out” items are stored. These items are not kept on stock, so each of these items will only stay at Scholten for a short time, and are stored at a “cross-dock” location, which is the front picking-aisle in the picking-area or a small area in the bulk-area close to the picking-area. An example of item y is a small part to replace a broken part in equipment Scholten sells or a product that is not kept on stock and is requested by a customer and therefore temporarily stored at the picking-area. The last item storage possibility is item z. These items are only kept on stock in the bulk-area and have no picking-area location. These items are usually too large to fit in the picking-area locations an example is a wheelchair.
2.3PROBLEM STATEMENT
The problem statement consists of three components; the research objective, the research question and the sub-questions and the boundary conditions. The problem statement will be based on the problem description and wishes and goals of Scholten.
2.3.1 Research objective
This research project was initiated by Scholten with the intent to assess a number of points of interest. The current method of working (the way activities in the warehouse are done) in the warehouse will be looked at as well as the way the current method of working fits with the wishes of the customers. Another point of interest that Scholten would like to have assessed is an extension of the offered service. This might include a change in the current method of working, for example by adding activities, or even a totally different method of working, which means the way activities are done currently are changed and will be done differently. These different alternatives could include an own delivery service or outsourcing the warehousing activities.
In further meetings with the general manager and the research supervisor, who is the operating manager, it was decided this research would focus only internally on Scholten, specifically the warehouse and logistics. This was decided to ensure the research project would not be too broad and could be completed within the available time of about five months. Other points of interest might be addressed in another research project. In accordance with the focus on the warehouse and logistics the objective of this research project is defined as follows:
“Analysis of the warehouse and the current method of working within the warehouse will yield recommendations, which lead to improvements in the efficiency of the warehouse.”
2.3.2 Research question
Based on the problem description and the research objective discussed above the main research question is formulated as follows:
Which recommendations, based on analysis of the current situation and relevant literature, will lead to improvements in the warehousing activities and how could Scholten successfully implement these recommendations?
Answering this research question will involve an analysis of the current situation and current method of working. This will be divided over the separate aspects of the warehouse; layout, the items, the storage locations and the method of working. Analysis of the separate aspects will be done according to several sub-questions, which will be discussed next.
2.3.3 Sub-questions
The sub-questions are structured in a way that the questions discuss subjects in a logical fashion. First the relevant literature will be reviewed needed to assess the current situation. Then the current situation will be discussed using the literature and observations made in the warehouse. Next the current situation will be assessed using the literature and information about the current situation. After the assessment of the current situation, recommendations can be given to improve the current situation. After the internal analysis of the warehouse, a strategic analysis of some external factors will be addressed. Finally everything will be combined to answer the research question and give recommendations to improve the current situation.
First an analysis of the current situation is needed. To be able to accurately describe the current situation, relevant literature needs to be discussed first.
1 What relevant literature is needed to assess the current situation?
2 What is the current layout of the bulk-area and the picking-area?
The layout of a warehouse has a great impact on the distances that are travelled by warehouse employees and thus on the time it takes to collect requested items. The layout restricts or enables picking and storing strategies discussed in the literature. Therefore it is important to look at the current layout first.
3 What is the current method of working?
The method of working also impacts the time to it takes collect requested items, directly by the method of picking and more indirect through the way items are stored and how pick locations are replenished. To see the impact of the way various activities in the warehouse are done is looked at.
4 What is the role of the warehouse management system at Scholten?
The management system facilitates the administrative part of the warehouse and is the backbone of all activities, which makes it an important aspect of the warehouse to look at.
After the current situation is described using literature and observations, possible changes and improvements can be pointed out.
5 What changes can be recommended to improve the current situation?
The previous sub-questions should yield a number of possible internal improvements for the current situation. External factors also influence the warehouse activities and a strategic analysis will yield external possibilities for improvements in warehousing efficiency.
6 What possible solutions emerge from analysis of external factors? 7 How could the recommendations be implemented successfully? 2.3.4 Boundary conditions
This research project is bound by a couple of boundary conditions: • Recommendations may not cause increased operating costs. • The research will focus mainly on the warehouse.
• More specifically, the picking-area and the bulk-area.
It was decided to focus mainly on an internal analysis of the warehousing activities at Scholten, because of the limited time available to do the research. The purchasing and sales departments will be outside the scope of this research. While these departments do in fact influence the efficiency of the warehouse in some degree. The purchasing department has to make sure the items are ordered with suppliers. The sales department has contact with the customers and is therefore largely responsible for the actual work in the warehouse. A strategic analysis of some external factors will be done to extent the focus of the research to a broader focus. If only, to show warehouse efficiency is not only affected by internal factors and provide some external points of interest. The limited time available for this research makes it impossible to address all possible external factors. External factors will be addressed very shortly in this section and (some) more extensively in a later section of this research.
In this research the possibility of an automated system is outside the scope of the research. However, very few companies use an automated system, due to the high variability in item size and demand and the high investment costs. So excluding automated systems from the scope of this research does not have a big impact on the completeness of the research. Mainly because Scholten also has a high variability in item size and demand, so the automated system would become very complex and likely very expensive. It is likely an automated system would not benefit Scholten.
Lastly the distribution of orders to customers could be looked at, either by using a different transportation company than the currently used ones or by looking into distributing goods from a local (smaller) warehouse. These other possibilities will be discussed below.
The customers of Scholten have recently changed from ordering for a central warehouse to ordering for smaller departmental warehouses or storage locations. This has caused the amount of orders to increase, without necessarily increasing the volume of the totally ordered amount over all customers. The effect is, however, that the activities in the warehouse increase, because each separate order has to be completed, which means the requested items have to be picked from the warehouse. Another factor which is caused by the customer is that the person responsible for placing the orders is not always qualified to correctly manage their inventory. This might cause mistakes in orders or extra orders being placed. To solve or lower the effects of this problem, Scholten might try to manage their customers better. For example, by using order lists or by keeping track of what a specific customer normally orders. Managing customers is a task for the sales department of Scholten. Internally, the warehouse management system could assist by keeping track of order history, which can be used by the sales employees to assist the customer with their order. Externally, Scholten could provide order lists with the items a customer usually orders listed on it. This way the customer can check the list for items they need. Scholten has many different customers, which makes it hard to assess which would yield improvements in efficiency.
An extension of the current activities, by means of a (small) delivery service for nearby customers, could reduce the amount of work in the warehouse. This concept was mentioned by the director of Scholten and involved a small delivery truck with most commonly used items on board. This truck would visit customers, who are then able to buy items at the truck. The truck is functioning as a small mobile store.
Currently, Scholten has one warehouse. Another possibility to reduce the amount of work done in the warehouse is to create (small) local warehouses near large customers. During this research one possible location was explored in some extend. Another company, HCNON1, supplying other items to much of the same customers as Scholten in the east of the Netherlands (Twente), also has excess warehouse storage space. This is a good opportunity to explore the option of a local warehouse, since the other company has free storage space and knows (most of) the customers that will be ordering the items. HCNON also has its own delivery trucks, which could be used to deliver items to customers in the area on a daily basis. This opportunity could be used as a pilot project. HCNON needs information from Scholten about the products and customers as well as their normal order activity. With this information HCNON will be able to check if the needed storage space is available and also if they have the labour capacity in their warehouse to handle the orders. HCNON and Scholten are currently looking into this pilot project. When this project is a success; i.e. HCNON successfully serves the customers in their area and these customers are (more) satisfied; Scholten might look at other areas in the Netherlands where many customers are clustered and could be served from a local warehouse.
The time available for this research is too short to accurately assess each of these external factors in detail, therefore a more global, strategic analysis of these factors seems more appropriate. On a strategic level the customers, the warehouse (and possibility of local warehouses) will be discussed and may yield possibilities for improvements.
2.4OUTLINE OF THE RESEARCH
situation. Part II will also contain the analysis of the external factors and the strategic recommendations that follow from it.
2.5RESEARCH METHODOLOGY
This paragraph will discuss the used research methodology. The used research method will be explained and the way data is collected and analyzed will be discussed. First a literature review is done. Next the current situation at Scholten will be described based on the literature. Hereafter the current situation will be assessed using the relevant literature. Finally recommendations will be given to improve the current situation.
This research project is focused at improving the current situation at Scholten, so the business problem is an “improvement problem”. Van Aken, Berends and van der Bij (2007) describe a design-focused business problem-solving approach in their book: Problem Solving in Organizations. The design-focused methodology is design-focused on one or more solutions in the form of a redesign of the current situation. This could involve improvements in current business processes or a new business process or information process. There is often not one solution, but rather many (small) improvements which help improving the current situation. This research will adopt this methodology and is aimed at proposing many (small) improvements of the current situation.
Literature addressing relevant subjects will be reviewed and later used to assess the current situation. Relevant subjects are warehouse layout, order picking strategies, storage strategies, processes, resources and organization. Literature was searched for in journals. There are scientific journals and professional journals, both are searched using the (electronic) library of the University of Groningen. Articles from these journals often reference earlier work on the same subject, effectively combining much knowledge and assuring the information is up-to-date. Sometimes the earlier work is more extensive on a certain subject, making the earlier work more suitable to be used than the later work, which might only give a short summary of the subject.
Discussion of the current situation will involve analysis of the layout, the items, the storage locations and the overall method of working. More specifically the layout of the bulk-area and the picking-area will be assessed and checked for improvements based on the current situation and literature. The way items are assigned a storage location will be looked into and checked for improvements. This involves a valuation of the items and storage locations and a correct match between items and storage locations. Further analysis of the method of working will involve the warehouse management system and the way it is used.
To analyze the current situation data will be collected in several ways. The layout of the warehouse will be observed in the warehouse itself. Also blueprints of the warehouse are available. From the observations and the blueprints a detailed overview of the warehouse is obtained. Storage locations are also observed in the warehouse and can be obtained from the warehouse management system. In which information about each location is kept. The warehouse management system also contains all the information needed about the items Scholten has in stock. The current method of working is obtained by interviewing warehouse employees, from detailed work instructions and by experiencing the method of working firsthand. All data collected about the current situation is primary data.
The reliability of the research data can be biased by four potential sources: the researcher, the instrument, the respondents and the situation. The researcher has an impact on the way he/she performs any of the research activities. To avoid harming the reliability of the data much, most data is collected in such a way it is controllable, like the example given earlier. Another way to avoid unreliability of research data is to obtain information from different sources. In this research data is obtained from the warehouse management system, from interviews with warehouse employees, from observations, from third-parties and from literature. In some cases multiple data-sources were used for one subject, increasing the reliability of statements made about that subject.
Validity of research data is the third criterion for evaluating research results. The validity of research data is often judged on its justification; the way the data is generated justifies its validity. For example, when a researcher accurately describes the way the data is generated and this way provides good reasons to believe the generated data is adequate, the validity of that research data is high. This example also shows that for data to be valid, it has to be reliable. Because reliability improves the reasons to believe the data is generated adequately.
The internal validity concerns conclusions about relationships between phenomena. Internal validity increases when there are good reasons to assume the proposed relationships are adequate, which can be done by evaluating the proposed relationships, or in this case solutions to the problem situation. By evaluating proposed scenarios, the effects of these scenarios are checked and discussed.
The external validity of this research will not be high, because it is based on one case; the warehouse at Scholten Medische Groep. With business problem project, such as this research, the external validity is of less importance, since the focus is on solving one specific problem.
PART I
Part I describes the current situation at Scholten. First a literature review will be given. The relevant literature will help in discussing the current situation. Next an overview of the warehouse will be given. In which the bulk-area and the picking-area will be addressed separately, discussing their functionality and layout. After this overview, the method of working in the warehouse will be addressed and will include all warehouse related activities, which are: picking items, sending items, receiving items, replenishing the picking-area, storing items and valuating items and storage locations. Finally the warehouse management system will be looked at. Important aspects are the functionality and how much of the functionality is used in the current situation.
A number of concepts that are used throughout the research will be addressed now for clarity. The warehouse at Scholten is divided into three areas, the bulk-area, the picking-area and the safety-area. The bulk-area is used to store large items and reserve stock for the picking-area. The picking-area is used to store small items and unit- or small-carton-sized items. The safety-area is used to store goods which are flammable or otherwise dangerous. The products Scholten sells are defined as a type of good, for example a box of latex gloves of a certain brand, and are called items in this research. Most products are sold in several packaging sizes; carton, small-carton or unit. The carton-size is the largest size and is stored in the bulk-area. Most carton-sized items contain several small-cartons, which in turn contain unit-sized items, see Fig. 5. The carton-size and small-cartons often have the same product code, while the unit-sized items have the same code with the letter “A” added in front.
Fig. 5. Different item-sizes.
3 LITERATURE REVIEW
Chapter 3 will discuss relevant literature. The literature will help to form clear definitions, which will be used throughout this research. The review addresses literature on warehouse design and control and covers the following subjects: layout, order picking strategies, storage strategies and routing strategies.
There is a lot of literature on warehousing, especially order picking. Mainly, because order picking constitutes 50-75% of the total operating costs for a typical warehouse. See, for example Petersen and Aase (2004) or Le-Duc and de Koster (2005). In this research only manual order picking is considered, because this is the way orders are picked at Scholten. With manual order picking, a picker walks or drives through the warehouse to retrieve ordered items from their storage locations. The picker uses a pick device on which the picked items are placed and transported to the depot. Automated order picking uses an automated system for item retrieval. A lot of research is done to find the most effective picking, storage or routing policies and is aimed at trying to reduce the high costs of order picking which are often directly related to average travel distance. This resulted in many different order picking strategies, storage policies and routing strategies. In this literature review the warehouse layout design is looked at first, since picking, storing and routing strategies are often based upon the layout design of the warehouse. Next picking policies are addressed, followed by storage policies. Finally the different routing policies are addressed.
3.1WAREHOUSE LAYOUT DESIGN
The layout of the warehouse determines for a great part the available picking, storage and routing policies. The layout design concerns two sub-problems: the layout of the warehouse as a whole and the layout within the order-picking areas; de Koster, Le-Duc and Roodbergen (2007). The first problem deals with the departmental or functional layout of the warehouse; i.e. where the different departments are located based on their functionality. The second problem deals with the internal layout of the picking-areas and concerns the placement of the depot, the number of blocks and the number, length and width of the aisles. The solution of the second problem determines which picking, storage and routing policies are possible. The layout can also be set by the choice of a certain picking, storage or routing strategy, because some of these strategies require a certain layout. These strategies are explained in more detail in the coming sections.
The departmental layout of the warehouse is usually determined when a warehouse is built and is hard to alter, because different departments are usually separated by walls or other hard-to-move obstacles. The objective of the departmental layout is to assign warehouse space to certain departments (receiving, picking, storage, sorting/checking, packaging and sending, etc.). This layout design often follows logically from the warehouse activities and product flows. For example, the receiving area will be close to the entrance of the warehouse where suppliers deliver the items.
Fig. 6. Internal layout aspects.
The position of the depot determines the front of the warehouse-area; it can be placed along the front-aisle, which is the first cross-aisle. A cross-aisle is used to reach and access pick-aisles. Pick-aisles contain storage locations and can be divided into sub-aisles by a cross-aisle. A set of picking-aisles with cross-aisles on both sides is a block. See Fig. 6 for a representation of these concepts.
The common goal of these design decisions is to find a “best” layout design within the boundaries of certain constraints, like available space, minimal and maximal storage location size, etc. The objective of the design is to minimize travel distance, which is linearly related to total travel time and with it to a great part of the warehouse operating costs. Since picking activities are a large part of the total warehousing activities and therefore a large part of the warehousing expenses.
3.2PICKING POLICIES
Picking policies determine how ordered items are combined into a pick list and later collected from their respective storage locations; Petersen and Aase (2004); Renaud and Ruiz (2008). It was already stated an order can consist of more than one pick list and a pick list consists of at least one item. Each pick list is assigned to one picker, while one picker can be assigned more than one pick list simultaneously. Four basic procedures for picking orders are discussed: discrete or strict picking; zone picking; batch picking and wave picking. In discrete or strict-order picking, an order picker completes one pick list in one tour; picking each item on the pick list and then returning to the depot. This policy is often preferred and used because it is easily implemented and order integrity is always maintained. In zone picking, the warehouse is divided into several distinct zones, with one picker assigned to one zone. Items in an order are divided over several pick lists and retrieved by the picker in each zone according to their individual zone pick list. In batch picking, one picker is assigned more than one pick list and collects the items on the pick lists simultaneously in one tour. This policy is an extension to the discrete or strict picking policy and has been shown to reduce total picking time significantly. See, for example: Gibson and Sharp (1992) or de Koster, van der Poort and Wolters (1999). In wave picking, each picker picks items from one zone for multiple pick lists at a time. This policy is a combination between batch picking and zone picking. Wave picking requires secondary operations to consolidate the orders from the different zones and pick lists. Any other policies are combinations of these basic policies and may have varying impacts on warehouse efficiency.
3.3STORAGE POLICIES
assigned a random available storage location based solely on needed and available storage space. This could result in the same item changing locations when new stock arrives. With dedicated storage, an item is assigned a fixed location the first time it is stored in the warehouse. This could result in empty storage space waiting for new stock of the assigned item. With class-based storage, each item is assigned a class and each storage location is assigned a class. By assigning storage locations a class, the warehouse is divided in class zones or areas. An item can now be linked based on class, this way an item with class A will be stored in a storage location with class A as well. Generally a three class system, or ABC-classification, is used. The classification of storage locations is usually based on distance from the depot, since this distance is linearly related to travel time of the picker. The classification of the items can be based on any number of criteria. Literature lists many different criteria: inventory cost, part criticality, lead time, commonality, obsolescence, substitutability, number of requests for the item in a year, scarcity, durability, substitutability, repairability, order size requirement, stockability, demand distribution, and stock-out penalty cost; list adopted from Ramanathan, (2006). A classification based on order frequency (i.e. items most frequently ordered and therefore picked) is commonly used in practice, because this strategy is easy to implement and maintain; Le-Duc and de Koster (2005). More complicated models are discussed in more recent literature; Ramanathan (2006) and Ng (2007). These models use multiple criteria in a maximization objective function to determine the item-classification. And state the traditional ABC analysis might not provide a good classification when there is a great variety in items.
The storage policy to be used in a warehouse depends on the layout of the warehouse and the routing policy used in that warehouse. Petersen and Schmenner (1999) describe a couple of different volume-based (or class-volume-based) storage strategies; diagonal, within-aisle, across-aisle and perimeter. These strategies will be discussed briefly.
The diagonal strategy
With diagonal storage items are stored in a diagonal pattern, where the high volume items are stored closest to the depot and low volume items farthest from the depot. This strategy is dependent on the location of the depot, as that location determines where the high volume items have to be stored.
Fig. 7. Diagonal storage. Adopted from Petersen and Schmenner (1999).
The within-aisle strategy
Fig. 8. Within-aisle storage. Adopted from Petersen and Schmenner (1999).
The across-aisle strategy
Across-aisle storage is much like within-aisle storage, but uses the first location of each aisle, instead of a full aisle and then the next. When the depot is in the middle, the first locations used are the middle ones and the next items are stored in the first locations of the aisles out from the middle aisle in alternating order.
Fig. 9. Across-aisle storage. Adopted from Petersen and Schmenner (1999).
The perimeter strategy
In perimeter volume-based storage the high volume items are stored around the perimeter of the warehouse, while the low volume items are stored in the middle of the aisles. The highest volume item is stored closest to the depot and the next items are stored in a clockwise or contraclockwise direction from the depot.
Fig. 10. Perimeter storage. Adopted from Petersen and Schmenner (1999).
Their conclusion is that within-aisle storage performs the best with all the pick list sizes examined, in combination with the depot situated in the middle of the warehouse. They advice to always concentrate high volume items in a few aisles and use a heuristic routing method to form more consistent routes (i.e. use the same strategy, not many different ones). More on routing methods below.
3.4ROUTING POLICIES
Routing policies are methods to sequence or sort the items on a pick list in such a way that a good and efficient tour through the warehouse is ensured. Warehouse routing has similarities with the Travelling Salesman Problem, a standard logistics problem in operations research literature. The Travelling Salesman has to visit a number of places before returning to his/her home base. The problem is to minimize the total travel distance by ordering the sequence in which each place is visited. Warehouse routing is a bit more complicated. A picker can visit the storage locations and depot more than once. Also, the picker is not required to visit all the storage locations in the warehouse, he only visits the ones listed on the pick list. Minimizing warehouse routing is not easily done, although algorithms and heuristics exist to solve the problem for simplified warehouse layouts. With more complex warehouse layouts there might not be an optimal algorithm available; de Koster, Le-Duc and Roodbergen (2007). Another problem of an optimal route calculated with an algorithm, is that the route may seem illogical to the picker, who might deviate from the specified route as a result. Because of these and other problems in practice warehouse routing is often solved using heuristics. The following five heuristics are discussed in the next section: the S-shape (or traversal) strategy; the return strategy; the midpoint strategy; largest gap strategy; the combined strategy. After addressing these heuristics, “Optimal routing” is addressed shortly.
The S-shape strategy
The S-shape heuristic is one of the simplest routing strategies. With the S-shape method order pickers enter aisles that contain at least one requested item and when they enter an aisle, that aisle is traversed completely. When an aisle contains no requested items that aisle is skipped. When the last item is picked, the picker returns to the depot, this might mean the picker stops half-way an aisle and returns to the depot. This strategy requires a warehouse layout with at least two cross-aisles; a front (cross-) aisle and a back (cross-) aisle. See Fig. 11 for a visual example.
Fig. 11. S-shape strategy.
The return strategy
Fig. 12. Return strategy.
The midpoint strategy
The midpoint method divides the warehouse into two equal areas; the front area, connected to the front cross-aisle. And the back area, connected to the back cross-aisle. Requested items in the front area are entered from the front aisle, the back area is accessed from the back aisle. Pickers moves to the back half either from the first or last aisle they visit in the front half. This strategy requires at least two cross-aisles and is in part similar to the S-shape method. The midpoint strategy performs better, in terms of total distance travelled, than the S-shape strategy when the average number of picks per aisle is low; see Hall (1993). See Fig. 13 for a visual example.
Fig. 13. Midpoint strategy.
The largest gap strategy
Fig. 14. Largest gap strategy.
The combined strategy
The combined method combines methods from the other strategies to reduce the total travel distance of the pickers. Aisles with picks are either traversed entirely or entered and left at the same end of the warehouse (i.e. the front or back aisle). The choice to either traverse an aisle or return mid-way is made using dynamic programming; see Roodbergen and de Koster (2001). This strategy also needs multiple cross-aisles or at least two cross-aisles. See Fig. 15 for a visual example. Using dynamic programming to determine possible routes is more complicated than the other strategies, but should result in a lower total travel distance than a single heuristic.
Fig. 15. Combined strategy.
Optimal routing
The before mentioned heuristics provide routing solutions to satisfy a routing problem, the routes resulting from the different heuristics differ in length. Next to these heuristic methods, algorithms can be used to obtain (near) optimal solutions to routing problems. However, as already pointed out, algorithms may not be the best fit in practice. Heuristics are used most in practice; therefore these heuristics were compared with the optimal routing solution by Petersen (1997). He concludes that a heuristic solution is on average 5% over the optimal solution in terms of total travel distance.
illogical to pickers, they might be hesitant to follow the route to the letter. Heuristic methods follow a certain strategy to make a route through the warehouse and are less dependent on a specific pick list. For example the S-shape routing heuristic starts at the depot and enters the warehouse. Next the picker follows the rules of the method to complete the route without difficult calculation. For the S-shape method the picker only enters an aisle if there are items requested from that aisle and when an aisle is entered, it is traversed completely. This process continues until all picks are made, after which the picker returns to the depot via the fastest route (which might involve returning through the current aisle).
Heuristics are preferred over optimal routing to avoid picking mistakes and confusion among pickers because of difficult and (seemingly) illogical routes. The more difficult optimal routes may also require more adjustment and learning time than the heuristic methods, because they cannot familiarize themselves with a certain strategy or way of traversing the warehouse. Instead they need to blindly follow a given route. In practice Optimal routing is more suitable for automated systems, in which the optimal route is simply input for the system. With manual picking, heuristic methods seem more appropriate.
3.5LOGISTICS OUTSOURCING
Bolumole, Frankel and Naslund (2007) developed a framework with which logistics outsourcing decisions can be evaluated. The framework is based on multiple social science theoretical perspectives which are used to explain the reasons and motives for outsourcing as well as the impact on logistics.
Outsourcing is the practice of charging external (third party) service providers with performing in-house activities. With logistics outsourcing these activities could include: transportation, warehousing, inventory management and related lot quantity issues, order processing and customer service. The article lists a number of social science theories, of which several are especially relevant to logistics outsourcing practices and decision strategy. These are: Transaction-Costs Analysis (TCA) and Agency Theory; Resource-Based Theory (RBT); and Network Theory (NT) and General Systems Theory.
Transaction-Costs Analysis proposes that firms exist in order to maximize profits by reducing transaction costs through three different forms of governance structures: market (arms-length, one-off transactions for standard investments); hierarchical (vertical integration through direct ownership); and hybrid structures (combining elements from the market and hierarchical mechanisms). Agency Theory refers to the methods by which one party ensures the development and maintenance/monitoring of beneficial relationships with another party. TCA strives to reduce transaction costs, which often can be achieved by outsourcing activities to service providers at a cost lower than the cost of doing the activity yourself. Agency Theory justifies the establishment of relationships between organisations and service providers when outsourcing activities and/or integrating in the supply chain.
The Resource-Based Theory views the firm as a bundle of resources. The organisation must secure an efficient bundle and flow of the right type of resources from its environment in order to survive and improve its operational performance. RBT views resources and access to resources as sources of competitive advantage. Resources are in this light seen as core competencies of a company. Acquiring and maintaining these core competencies is important for the company’s survival. Through outsourcing companies acquire external resources (transportation, warehousing, inventory management and related lot quantity issues, order processing, customer services) which could add value and competitive advantage for the company.
into two categories: cost savings and service improvements. Outsourcing, and with it working (closely) together with third parties, requires the removal of traditional contractual safeguards and the sharing of (internal) information.
According to the article outsourcing has two underlying principles behind organizations’ decisions for outsourcing; the internal and the external. The internal principle is the recognition of a lacking in-house capability. By using the externally available resources to aid the in-in-house lack of that resource the organization creates more value and competitive advantage. The external principle deals with the organizations need to reduce external (transaction) costs. This can be achieved by integrating in the value chain or gaining control over external resources.
4 THE WAREHOUSE
Chapter 4 will address the layout of the warehouse. The bulk- and picking-area will be addressed separately. General information will be given about each specific area as well as a detailed overview of the layout of that area. First the warehouse as a whole will be discussed. Next the bulk- and picking-area are discussed.
4.1THE WAREHOUSE LAYOUT
Literature suggested the layout of the warehouse as a whole is usually based on different departments which are assigned space according to their functionality. This is also the case at Scholten and can be seen in Fig. 16.
Fig. 16. Layout of the warehouse.
area as well as carton-sized goods. Dangerous and medical goods are stored at the safety-area. These involve flammable items and medicines, which need to be behind closed (and locked) doors. Unit-sized and/or small-carton goods are stored in the picking-area. Because most goods are stored in the bulk-area, the bulk-area is closest to the entrance. Pick lists are obtained from the depot, which is placed near the picking-area and more or less in the middle of the warehouse. The picked orders are checked and packaged at the depot as well, so the depot is also close to the exit. Complete pallets with orders are stored close to the exit for easy access by the transporter.
4.2THE BULK-AREA
The bulk-area of the warehouse at Scholten is 625 m2 and has room for over 600 pallet storage places. The ground floor is used to store pallets of carton-sized items; this can be seen in Fig. 17, which is an overview map of the bulk-area of the warehouse. There are two different pallet sizes, large pallets on which items are higher stacked than on small pallets. Not all storage locations shown in Fig. 17 can contain a large or full pallet. Each pallet place shown on the map has several levels, varying from three to six. However, not all levels are used for pallet storage; some are used to store carton-sized items. Usually the second and/or third level is used for this purpose, when these levels are used like this they are often less high. The first two to three levels are mostly reachable manually, the higher levels need a fork truck or other mechanical warehouse device to be reached.
Fig. 17. Bulk-area layout.
4.2.1 Bulk-area layout
and is the only possibility to change picking-aisles. The picking-aisles are about 240 centimetres wide, which allows pickers to turn around in the picking-aisle. A pick location, as shown in Fig. 17. by the black square, is 90 by 90 centimetres, which is enough to store one large pallet. At some higher levels wooden platforms facilitate the storage of carton-size items and other not-pallet items.
4.2.2 Location coding
Each storage location is given a code by which it is easily found inside the bulk-area. The code is based on the storage location’s place in the warehouse and consists of four parts; area, aisle, rack, position. The area code for the bulk-area is “A”. The aisle part of the code is based on the distance from the depot and is coded using “A” to “E” for the bulk-area. “A” being furthest from the depot. The rack part of the code is based on the storage location’s position within an aisle and is coded using “01” to “12” alternating from the left and right side of the aisle. “01” and “02” being furthest away from the aisle entry. The last part of the code is the location within a rack and is based on the left to right position and height. The bulk-area uses “A” to “E” to code the height, starting with “A” always at ground level. Aisles “AE” and “AD” have extra storage rack levels to store small pallets and single carton-sized items. These storage locations are coded in height, starting with “A” at the bottom and each level above “A” is coded with “B”, “C”, etc. Aisles “AC” to “AA” usually lack the level “B” height and instead go from “A” to “C”. These aisles are mostly used to store pallets stacked high with items and therefore need more space. Next the left to right position is coded, using “A” to “D”, with “A” being furthest away from the aisle entry. An example of a bulk-area code is: [AC05AD], and is shown in Fig. 17 and Fig. 18 with the black square.
Fig. 18. Bulk-area location coding.
4.2.3 Picking and routing policy
The picking policy currently applied at Scholten is a mix between discrete and batch picking. Discrete picking is used when the pick list is sufficiently large, i.e. large enough to fill one picking cart or pallet. Batching is used when a couple of small pick lists are available. The combined small pick lists also fill one pick cart. The choice is made logically by the picker; there is no influence of the warehouse management system in combining pick lists into batches.
4.3THE PICKING-AREA
The picking-area is 300 square metres and is used to pick specific customer orders. Clients of Scholten can order any amount of a certain item. The picking-area holds a lot of different items, some of which are also stored in the bulk-area as reserve stock and/or in larger packing size. Most orders consist of picks in the picking-area, which is why the picking-area is located close to the depot and the exit of the warehouse. All picking locations are reachable manually. An overview of the picking-area is shown in Fig. 19 below.
Fig. 19. Picking-area layout.
4.3.1 Picking-area layout
The picking-area has two blocks of picking-aisles, with one cross-aisle, which is used to enter the picking-aisles. There are eighteen picking-aisles, with six racks located on either side of the aisle. The last (Q and R) and first (A and B) aisles have less racks, because there is only one side with racks. Also the front aisle (Q and R) is used as a cross-dock to (temporarily) store items which are ordered specifically for a certain customer order. Each rack has seven height levels, each level has 185 racks which totals to 1295 racks in the picking-area. Also each rack can be divided into three sections. A rack is 1 metre wide and about 25 centimetres high. A visual example of a rack is shown in Fig. 20. More on how these locations are coded in the next section. The picking-aisles are narrow, preventing the picker to take the picking cart along in the aisle. The cross-aisle is 2.5 metres wide, allowing the pickers to cross each other and park their carts in the cross-aisle while they visit a picking-location. The cross-aisle is in line with the entry to the depot for easy access. Both empty and full picking carts are stored near the depot; the latter are stored closest to the depot entry. Within the depot are two desks, one on either side of the room. One side holds the computers and printers. The other side is used to check the picked items and, if correctly picked, pack the items in boxes.
4.3.2 Location coding
pick-aisle and is coded using “01” to “12” alternating from the left and right side of the aisle for the lower half and alternating from right to left for the upper half. “01” and “02” being furthest away from the aisle entry. The last part of the code is the location within a rack and is based on the left to right position and height. The picking-area uses “A”, “B”, “C”, “G”, “H”, “I” and “J” to code the height, starting with “A” always at ground level and going up to the highest location which is coded “J”. Next the left to right position is coded, using “A” to “C”, starting with “A” on the left hand side when facing a storage location to “C” on the right hand side. An example of a picking-area code is: [DN07AA], and is shown in Fig. 19 with the black coloured pick location and below in Fig. 20 with the black coloured location.
Fig. 20. Picking-area location-coding.
4.3.3 Picking and routing policy
The picking-area uses batched picking more often than the bulk-area, because requested items and pick lists are small. This allows many pick lists on one picking cart, while decreasing the average distance travelled. However, some pick lists are very large and are picked individually, which fits with the discrete picking strategy.
Pickers in the picking-area are routed using a variant of the return picking strategy (see Fig. 12). Pickers sort the pick lists, with the farthest location on top, and move to this farthest location near the back of the picking-area. From this location the picker moves towards the depot making picks in aisles on either side of the cross-aisle. When all picks are completed the picking cart is placed near the entry of the depot and the picker retrieves new pick lists and starts a new picking tour.
5 CURRENT METHOD OF WORKING
Fig. 21. Warehouse activities (same as Fig. 1. in Chapter 1: Introduction).
5.1RECEIVING GOODS
Scholten is a wholesaler with a warehouse and does not produce any items themselves, so everything Scholten sends to clients has to come to Scholten first. The purchasing department places orders with suppliers who deliver the goods to Scholten, usually during the morning. Deliveries from suppliers are stored in the warehouse based on the packing size. Items stored at the bulk-area are delivered either by whole pallet or on a pallet with other carton-sized items. The smaller-sized items are usually delivered in a box with other small-sized items. These are stored in the picking-area. All incoming deliveries have to be signed off with the transporter. Next a warehouse employee has to check the delivery for completeness, which is done using the list of contents and a delivery order list together. When the delivery is correct the receipt can be booked in the warehouse management system. This involves typing in the received amount of items (when differing from ordered amount) next to the ordered amounts; any differences will be looked into by the purchasing department. Also some items have an expiry date and LOT number; these have to be entered into the warehouse management system along with the amount received of that LOT, before the receipt of the items can be booked. When all the needed information is entered the receipt can be booked. However, the warehouse employee often stores the items at their assigned locations first. While storing the items he/she fills in the required information on the delivery order list. When all items are stored and the needed information is collected the employee books the delivery in the system and fills in the required information. Hereafter the booked delivery is registered in the warehouse management system, which concludes the process of receiving goods.
The last action, registering the booked delivery, may seem unnecessary because it seems to have no function when the process is done in this order (registering right after booking the delivery). This is true, but is caused by the warehouse employees who switched the booking of the delivery further back. The working instructions state that the receipt has to be booked first and stored later, if possible, in combination with other booked receipts. This was chosen because when received items are booked, they are automatically stored in the location “ONTV”, which is short for the Dutch word for received. Also when the receipt is booked the warehouse management system knows these items are in stock and orders containing these items can be accepted and released. When the receipts are booked later, the system also releases any pick lists containing those items later.
5.2STORING ITEMS
Items are stored at Scholten on two occasions, either when items are delivered by suppliers or when items are returned from clients. On both occasions items enter the warehouse at the same place, but are temporarily stored under a different name, respectively “ONTV” and “RETOUR”. The storing of the items is the same for both and will be described next.
When storing received items a clear difference is found between items that have got an assigned storage location and items that are new to Scholten and do not have an assigned storage location. The latter will be assigned a new storage location, which can either be a temporary location or a standard location, on which the item will then be stored each time it needs to be stored.
