1. INTRODUCTION
1.2 Outline of the report
The structure of the report is as follows: Chapter 2 provides an overview of the research context of the master thesis project. Chapter 3 analyses the initial problem: underperformance of the factories and suppliers. Furthermore, Chapter 3 introduces a specific type of equipment that is produced by Vanderlande factories and on which the research will focus. Chapter 4 elaborates on the initial research design, which focusses on inventory control. In Chapter 5, the sub-questions defined in Chapter 4 are translated into three conceptual inventory control policies. To goal was to analyse the performance of the control policies by a simulation and compare the results. Therefore, valuable input data about the demand in purchase items, lead-times, and costs was essential. However, Vanderlande only had information about systems, and this information had to be translated into demand in purchase items.
3
Chapter 5 elaborates on this demand analysis. At the end of the analysis, it is concluded that most information required to define purchase item demand is available at an earlier stage than what is currently known. Consequently, a purchase item analysis is initialised in Chapter 6 to define what information is required and when this information is available. The purchase item analysis led to the conclusion that redesigning the control structure has a bigger positive impact on factory performance and as a result, Chapter 6 is finalised with a redefined research question. Chapter 7 answers the second research question by redesigning the control structure of Vanderlande. As the redesign is not applicable on all items, and as the lead-times provided by suppliers can be different from actual lead-times, Chapter 8 considers a central inventory control policy in which the effect of lead-time uncertainty is investigated.
Chapter 9 contains the conclusion of the redesign of the control structure and inventory control policy, and the recommendations for Vanderlande, and for future research. In Appendix A, all abbreviations and their definitions are given. Appendix B contains the list of figures, tables and graphs, and Appendix C explains the assumptions made for the research.
4 2. RESEARCH CONTEXT
This chapter starts with a company introduction, followed by an explanation of the supply chain, and product hierarchy. Moreover, Chapter 2 elaborates on the processes, control, and information of supply chain and manufacturing and the chapter finalises with the performance measures used by Vanderlande.
2.1 Company introduction
Vanderlande is an Engineer-To-Order (ETO) organisation and market leader in designing and installing baggage handling systems in the major international hubs and regional airports, and sorting systems for the parcel and postal industry. Next to that, it is a leading supplier of warehouse automation systems. It distinguishes itself from its competitors with high-quality, reliable products, and fast deliveries.
Vanderlande is geographically divided into three supply chain centres: Northern America, Europe, and Asia Pacific. This thesis is executed at both supply chain centre Europe (SCCE) and the manufacturing department. Manufacturing originally only consisted of a factory in Veghel. Over a year ago, Vanderlande bought two factories: one in Spain and one in America. The factory in Spain, located in Santpedor, is already producing equipment, whereas the factory in America, located in Calhoun, is not.
Currently, several business cases are executed to find out what equipment can be produced in America and how much it will cost. When referring to the factories, the following abbreviations are used: the factory in Veghel is called VIM (Vanderlande Industries Manufacturing); the factory in Santpedor, Spain is called VIS (Vanderlande Industries Spain), and the factory in Calhoun, America is called VIA (Vanderlande Industries America). The headquarters of Vanderlande are located in Veghel. Next to the three factories, Vanderlande has a number of subcontractors producing Vanderlande equipment.
2.2 Supply chain
The processes in the supply chain can be non-physical and physical. The non-physical stage concerns sales engineering, engineering, design, supply chain, and planning activities activities and the physical stage concerns the manufacturing and assembly activities (Bertrand and Muntslag, 1993).
The non-physical stage starts with the activities of sales engineering, consisting of the conceptual design development. After the contract between Vanderlande and a customer has been signed, the planning department defines a schedule for the project and the engineering department starts designing the layout and functional requirements. Hence, engineers translate the high level characteristics into a detailed system. At the same time, high level information about a project is send to a supply chain centre and when engineering is finished, the supply chain centres decides upon what factory or subcontractor will produce what sub-component and sends orders to the factories and subcontractors. The factories and subcontractors are in charge of the production and make sure material is delivered to the warehouse within the lead-time given by the supply chain centre. Subsequently the supply chain centre arranges the distribution of the sub-components to site, and at site, the site department is in charge of the installation, testing and commissioning of the equipment.
2.3 Product hierarchy
Vanderlande uses different terms for the material of the systems in the different phases of the supply chain. This hierarchy is displayed in Figure 3. The highest level is the equipment level, representing the
5
type of project that is sold. This project is split up into specifications, referred to as SPEC’s. One SPEC contains detailed information about the layout of a component, and consists of multiple orders, representing sub-components. The orders can be production orders (PO), if a sub-component is ordered from VIS, VIA, or a subcontractor. They can be orders from inventory (ON), if it involves sub-components picked from inventory, and they can be factory orders (OF), if they are produced by VIM. Vanderlande strives to use VIM as a flexible location, producing the most innovative equipment. VIS, VIA, and the subcontractor produce more standardized products, as those factories are not equipped for the production of advanced equipment. An OF is split up into different work kits (WK), depending on the routing through the factory and colours in which the assembly will be coated. One WK represents multiple work orders (WO) and within the WO, the Bill of Materials (BOM) is defined. A WO is created for a specific group within the factory, producing the assembly. The WO contains purchase item orders (PIO) involving sub-assemblies or raw materials. The suppliers from which sub-assemblies or raw materials are ordered are referred to as second tier suppliers. Hence, materials that enter a factory are intermediate products or raw materials, and materials that leave the factory are sub-components.
A sub-component can be engineered and manufactured in multiple ways, resulting in different types.
Each type has a master-item number assigned to it. Moreover, a type can also be manufactured in multiple ways, resulting in instances. An instance has a dash number assigned to it. The master-item number consists of 5 to 8 numbers or letters and a dash number is defined with 5 additional dash numbers. An example of a master-item and dash explanation is given in Appendix D.
2.4 SCCE and Manufacturing
This section elaborates on the processes, control and information flow between SCCE and manufacturing, as those are under investigation in this master thesis project.
Figure 3, Product hierarchy
6 2.4.1 Processes
SCCE is responsible for the coordination of the movement of goods from suppliers and manufacturers to customers within Europe. This involves the production, purchasing, warehousing, distribution, and forecasting of material and projects. It gathers information about incoming projects from sales engineering, and adds the information to a forecast of that specific type of equipment to monitor the predicted demand. Meanwhile, engineering specifies the lay out and requirements of the project and splits it in SPECs. After an engineer has finished its activities, a supply chain coordinator divides the SPECs into ONs, OFs and POs, and sends these to the factories that produce the sub-component. Thereafter, the order is further specified by the manufacturing department. Order specification includes breaking the order up in WKs and WOs and defining the routing through the factory. The final action of manufacturing is to attach a BOM to a WO, and order the required intermediate products and raw materials. At this point, the physical stage starts. The material and invoice are received by a factory, which then starts its manufacturing activities. Appendix E illustrates the manufacturing activities.
Finished assemblies are send to the European distribution centre (EDC) for all European factories. At the EDC, the assemblies are allocated and send to site. The EDC is also the responsibility of SCCE. The invoice is extended with the costs for production and holding costs of the factory and send to SCCE. The material and information flow is illustrated in Figure 4. Purchase items are abbreviated to PI and finished goods are referred to as FG.
Figure 4, Material and information flow
2.4.2 Control structure
This sub-section elaborates on the control structure of SCCE and manufacturing.
Demand forecast
An estimation of the expected workload is created by SCCE. SCCE weekly gathers information about sold projects and developes a forecast per type of equipment that shows expected workload and the due date of projects.
Order release
SCCE releases the orders and sends them to a factory. Thus, order release includes order allocation.
Order allocation is based on historic production: for every sub-component, the information system knows whether the order is a make (OF or ON), switch (ON, PO or OF), or buy (PO) order. If the order contains a new sub-component, SCCE and manufacturing together decide who should produce it. This decision depends on the capacity of the factories and the newness of the equipment.
7 Order acceptance
Before a factory starts with its activities, it first has to accept the OF. This decision is based on the capacity of the factory and the planning attached to an OF. If the production planning department concludes that the worload is too high or the planning too tight, it contacts SCCE and together they decide whether the planning is changed or the order is outsourced.
Production control
The given lead-time from OF or PO release until OF or PO finish is eight weeks, consequently the requested finish date for the factories is eight weeks later than the release date (Figure 5). The factory has two weeks to create WKs and WOs and before it places PIO at suppliers. Thus, material procurement happens after all uncertainty about the product and processes is eliminated. Within four weeks, Vanderlande expects to receive the materials from the second tier suppliers, resulting in two weeks for production and distribution to the EDC. The control structure is similar for all factories.
Figure 5, Lead-time from factory order release until finish date
Purchase item control
Three different definitions for the lead-time are mentioned in this thesis:
The four weeks in which second tier suppliers are supposed to deliver their material are referred to as the planned lead-time;
The lead-time given by second tier suppliers is referred to as the standard lead-time;
The lead-time it took for material to arrive at the factory is referred to as the actual lead-time.
Different factories order material from the same second tier suppliers. Every factory orders its own material and ordering is done via an MRP-system. Subsequently the factories receive an order confirmation stating when they will receive their material. If the receipte date is later than the request date, the factory tries to expedite the order. The lead-time of PIs consists of two parts: a production and a transportation part. The production part is similar to each factory. However, the transportation part differs between the factories. All second tier suppliers are located in Europe, resulting in a transportation time of one week to a European factory. Shipping material to America takes at least three more weeks, resulting in a transportation time of more than four weeks. In this report, whenever there is referred to the standard lead-time given by second tier suppliers, it involves European transportation times.
Inventory control
PIS with a standard lead-time longer than four weeks are stocked in order to decrease the risk of material unavailability. The safety stocks are determined in different ways: the level can be equal to one or two weeks of historic demand, or the expected demand during a part of the standard lead-time. The computations for the expected demand are based on gut feeling and updated every 3 to 6 months.
8
Second tier suppliers also stock material due to demand fluctuations. However, these inventory levels are unknown at the factories.
Project cost control
During the selling stage, the price of a project is determined. The price consists of direct and indirect costs and profit. The direct costs are the costs for material and hours of direct work allocated to a project.
Both the indirect costs and the profit are a percentage over the direct costs. The percentage of indirect costs includes a fixed part for indirect and overhead costs. Similarly, the percentage of profit consists of a fixed percentage of EBIT (Earnings Before Interest and Tax) computed over the direct costs. There is often an extra disclosure for penalty costs in the contract signed by the customer. The penalty costs are a weekly fine that should be paid whenever a project is delayed. This fine can vary per project and can either be a fixed number or a percentage of the total price.
FO or PO cost control
The costs are controlled differently for the three factories. VIM does not have to create actual invoices for its production, as it is part of the same private limited company. However, the invoice of material does arrive at the factory and this invoice, together with direct and indirect costs required for the production of a sub-component, is send to SCCE. VIS and VIA are not part of this private limited company and therefore they do have to create invoices that are send to SCCE. All three factories strive to have a 0% EBIT. Subcontractors do want to make a profit, resulting in a different invoice for similar orders.
PIO cost control
The purchase price for the material ordered from second tier suppliers is determined in cooperation with the suppliers and captured in a contract. The costs made by the second tier supplier include amongst others raw material ordered by the supplier, production and transportation. The purchase price may fluctuate, due to market and economic changes. A PIO consists of the daily demand of purchase items required for multiple projects. Small batches cost more than large batches, and expediting orders results in a higher purchase price.
Holding cost control
The fluctuating purchase price for purchase items is not considered when computing the holding cost.
The purchase price used to compute the holding costs is a fixed price, adjusted every six months. The holding cost, as computed by Vanderlande, is a fixed percentage of this fixed purchase price, and is time independent. This fixed percentage is determined at the end of a year by computing the total costs for utilities, rent, insurance, handling, and dividing it by the total costs for purchase items of that year. Only when a major increase or decrease occurs, the purchase price is recomputed. Currently, Vanderlande computes its holding costs in the following way:
ℎ𝑜𝑙𝑑𝑖𝑛𝑔 𝑐𝑜𝑠𝑡𝑠 𝑝𝑒𝑟 𝑖𝑡𝑒𝑚 = 𝑝𝑢𝑟𝑐ℎ𝑎𝑠𝑒 𝑝𝑟𝑖𝑐𝑒 𝑝𝑒𝑟 𝑖𝑡𝑒𝑚 ∗ 0.049
As this holding costs computation is time independent, the computation is not used in the project.
Instead, a holding cost computation found in literature is used.
The control structure proves that Vanderlande has an ETO strategy, as PIO are based on customer orders, inducing a pull strategy.
9 2.4.3 Information
There are a number of information systems, and information sharing techniques used by Vanderlande.
This report exclusively elaborates on the ones used by the SCCE and the factories, and that are interesting for the research.
ERP-system
J.D. Edwards, abbreviated JDE, is the Enterprise Resource Planning system that is used company-wide.
One can integrate other applications with it, and collect, store and manage data from all sorts of activities in this system. Departments can work in different J.D. Edwards work-branches. However, VIS uses a different ERP system to manage its data.
MRP-system
The factory daily uses two MRP-systems. The first one is used to compute the amount of assembly hours required in a WO and the second one is used to compute the time required for parts production, the amount of PIs required in WOs set that day. Per PI, the MRP-system generates a PIO is, which specifies the requested quantity and request date. This quantity is based on information about the lead-time, minimum order quantity, safety stock and multiple order quantity of the PI. The factory daily sends these PIO to its second tier suppliers and the confirmed date is saved in the second MRP-program. The second MRP-program also keeps track of the inventory position of each PI. The PIO send to the second tier supplier is thus based on the total amount of items required for the production activities that start four weeks later. The material is ordered anonymously, which means that the items within the PIO are not assigned to a specific project.
2.5 Performance measures
The performance of the factories is measured on either a monthly or weekly basis by a number of key performance measures (KPIs). A KPI is a performance measure that evaluates the success of an organisation. Together with SCCE, the factories have defined four KPI-groups that consist of a number of KPIs. These groups are: employees, process, customers, and financial. Within the customers-group, there are two critical KPIs: on time and complete performance of the factories and on time and complete performance of the suppliers:
𝑂𝑛 𝑡𝑖𝑚𝑒 𝑎𝑛𝑑 𝑐𝑜𝑚𝑝𝑙𝑒𝑡𝑒 =# 𝑜𝑟𝑑𝑒𝑟𝑙𝑖𝑛𝑒𝑠 𝑑𝑒𝑙𝑖𝑣𝑒𝑟𝑒𝑑 𝑜𝑛 𝑡𝑖𝑚𝑒 𝑎𝑛𝑑 𝑐𝑜𝑚𝑝𝑙𝑒𝑡𝑒
# 𝑜𝑟𝑑𝑒𝑟𝑙𝑖𝑛𝑒𝑠 ∗ 100%
SSCE expects the factories to have a 95% on time and complete performance and the factories expect their second tier suppliers to have a 98% on time and complete performance.
10 3. RESEARCH SCOPE
In Chapter 2, the research context is discussed. The chapter finalises with performance measures of factories. This chapter analyses the actual performance of this performance measure. Moreover, performance of second tier suppliers is analysed and a specific type of equipment that is used as a business case for the research project is discussed.
3.1 Factory performance
The on time and complete performance target for the three factories is 95%, and is measured by SCCE.
Graph 2 illustrates the performance of VIM in 2016, and Graph 3 the performance of VIS in 2016. VIS did not have production before week 10, therefore there is no data concerning weeks 2-10. The records of VIA are not tracked yet, which is why the performance of this factory cannot be measured. The average percentage for on time and complete orders of VIM is 83%, with a minimum of 68% and a maximum of 94% over on average 64 orders per week. VIS finishes on average 65% of its orders on time and complete.
In week 35, VIS reached a minimum of 33% and in week 32 a maximum of 100%. It can be concluded that the average performance of both factories is a lot lower than the threshold of 95%. VIM is, based
In week 35, VIS reached a minimum of 33% and in week 32 a maximum of 100%. It can be concluded that the average performance of both factories is a lot lower than the threshold of 95%. VIM is, based