By: ing. L.C. Dijk
l.c.dijk@student.rug.nl | s1588877
March 24, 2009
Supervisor University: Dr. Nicky van Foreest Supervisor Sara Lee: Ir. Marcel Venema
Leeuwarderweg 1
MSc Technology Management 501 ZD Joure
Faculty of Economics and Business Tel. 0513 48 89 11
Preface
‘I did it, I did it, is on my mind when the graduation date is determined.’
Origin.
This paper is the result of the Master Thesis at Sara Lee Corporation for the study Technology Management. Three and a half years ago I started with a bridging programme of two years, to get the bachelor of Technology Management. Now, three and a half years later I am surprised and joyful that I finished the master of Technology Management as well. I managed to combine my study with my passion for horse riding and my busy social life. My experiences with the academic world were great.
Acknowledgments.
Six months ago I got in contact with Sara Lee due to an interview for a business course. It was a great interview and my interest for this company grew. I sent an application to Sara Lee in Joure and this was well received. I started my thesis in August and at the end of September my research proposal was approved by my supervisors and the initiator of the project. From the University of Groningen I requested professor Nicky van Foreest to be my supervisor. Nicky van Foreest is an Assistant Professor at the University of Groningen in the operations section of the Faculty of Economics and Business. Wiebe Jongsma has the function of Manager of the Technology Department at Sara Lee Joure and initiated my research project. My supervisor at Sara Lee was Marcel Venema, who works as a Packaging Engineer. Warse Klingenberg, Associate Professor at the University of Groningen, has been the co-assessor of my thesis. My first gratitude goes to them.
Further I want to thank the stakeholders of the ‘B2 tester’ project and the ‘B2 strategy’ project for getting the input I needed. Especially Joost Balder, who helped me to shape the control part of my research and gave me a lot of statistical insight.
Social contact is important for me and I wouldn’t have graduated if this wasn’t so great. At the start of this study I came in contact with a group of people and now I am grateful to have them as friends: Pieter Bas, Hans, Derk and Caroline! During the last year there were some people who genuinely took an interest in me and this project and I like to thank them for that: Herman, Ruurd and especially Igor for the help during my Master Thesis. The people who were always there for me are Agnieta, Aletta, Joanne, Minke and my sister and one person who became a special friend the last three years is Joost. My special thanks to them!
Summary
The objective of this research is to get insight how a reliable control procedure can be realized for the B2 dosers and how the buyer-supplier relationship between Helvoet and Sara Lee can be migrated into a strategic partnership enabling reduction or even abolishing incoming control eventually.
According to the Methodology ‘Strategic Analysis’, the effect of different strategies for both the control problem as well as the strategic problem is systematically analyzed and alternatives are given for a reliable control procedure and a strategic partnership.
Realizing a reliable control procedure
Outgoing control. A reliable outgoing control is realized when the production process is in
control and capable. The production process is control when the process is stable, which means that the remaining samples all lie within control limits. The production process is capable when the Cp= around 1.67 in combination with a Cpk=1.33 using USP=0,1221 cc/pulse and LSP=0.0907 cc/pulse. This information has to be reported in the PQA.
Incoming control. A reliable incoming control (with a batch of 180.000 B2 dosers, the
reduced level of inspection, and double sampling plan) is realized when in the Military Standard table IIIC the sampling plan ‘M’ is used. The first and second sampling takes eighty B2 dosers. The batch should be accepted when ≤0 B2 dosers in AQL 0.01 and ≤3 in AQL 2.5 and rejected when ≥2 in AQL 0.01 and ≥8 in AQL 2.5 in
the first sampling. Between these quantities the second sampling should be taken. If in the second sampling ≥2 B2 dosers in AQL 0.01
and ≥12 in AQL 2.5 the batch should be rejected. The dosing level specifications for AQL 0.01 and AQL 2.5 are shown in the figure. This has to be reported in the PQA.
B2 measuring system. A reliable B2 measuring system can be realized by eliminating the
variation source ‘(un)loading process of B2 doser’. The cause of this factor could be the B2 measuring system but also the B2 dosing system.
Organizing a strategic partnership
Content Preface ...2 Summary...3 1. Introduction ...5 1.1 Area of concern ... 6 1.1 Research objective... 7 1.2 Research questions ... 7 1.3 Stakeholders ... 8 2. Methodology...9 2.1 Framework of ideas... 9 2.2 Choice of Methodology... 9
3. Status quo outgoing and incoming control ...11
3.1 Data collection... 11
3.2 Product Quality Agreement between buyer-supplier ... 11
3.3 Outgoing control ... 12
3.4 Incoming control... 13
4. Control procedure ...14
4.1 Data collection... 14
4.2 Analysis of measurement approaches ... 15
4.3 Conclusion of measurement approaches ... 19
4.4 Analysis of measuring system... 20
4.5 Conclusion of measuring system ... 21
5. Buyer-supplier relationship ...22
5.1 Data collection... 22
5.2 Analysis of the relationship between Helvoet and Sara Lee... 22
5.3 Conclusion of the relationship between Helvoet and Sara Lee ... 23
5.4 Analysis of expectations of a strategic partnership ... 24
5.5 Conclusion of expectations of a strategic partnership... 25
6. Conclusion and recommendations: a reliable control procedure and a strategic partnership ...26
6.1 Realizing a reliable control procedure ... 26
6.2 Organizing a strategic partnership... 28
References ...29
Appendix 1. Feasibility study...30
Appendix 2. Case Study Results Sara Lee ...33
Figure 1, Cafitesse coffee machine, Liquid coffee pack, B2 doser and B2 measuring system
1. Introduction
Company.
Sara Lee Corporation is a global manufacturer and marketer of high-quality products for consumers throughout the world. One of the main brands of Sara Lee is Douwe Egberts. Douwe Egberts is the famous brand name for a variety of coffee flavours.
Production liquid coffee.
The production of liquid coffee from green beans takes place at the Sara Lee ‘coffee treatment and supply’ (CT&S) plant in Joure. The liquid coffee concentrate follows two routes in the packaging process; (1) the concentrate will be freeze dried en packed in glass jars or (2) directly packed in a bag where a dosing system (B2 doser) will be sealed on and completely folded into a carton box. The first finished product is used for the brand Moccona and the second is used within the Cafitesse machines in the market. This report is focused on the B2 doser in the finished product used within the Cafitesse machines.
Component of liquid coffee pack.
Figure 2, Process of outgoing and incoming control of B2 dosers
Production process and controlling process of component.
The production of the B2 dosers is outsourced. The B2 dosers are currently supplied by two suppliers, namely Helvoet and Forteq. Every two weeks Forteq delivers a batch of B2 dosers for production (210.000 B2 dosers) and Helvoet delivers weekly a batch of B2 dosers (180.000 B2 dosers). There are two B2 measuring systems in operation for outgoing control (Helvoet and Forteq) and one for incoming control (Sara Lee Joure). The control procedure between Helvoet and Sara Lee, outlined in figure 2, and the buyer-supplier relationship between Helvoet and Sara Lee are what matters in this research.
1.1 Area of concern
The Technology Department of CT&S, who initiated the research project, has a structural problem with the results of the B2 measuring system. There is an offset between the outgoing control and the incoming control. The deviation of the measurement results leads to a lot of discussion, especially when the measured dosing volumes are around reject level. Rejecting batches of B2 dosers for a few weeks will lead to ‘out of stock’ of the B2 dosers. The discussion occurs most often between Helvoet and Sara Lee and entails the causes of deviation for the B2 dosers and the measurement results of the B2 measuring system. Helvoet mentions that they can’t bring the quality of the B2 doser to the desired level and refer to the reliability of the B2 measuring system. The question is whether or not the B2 measuring system or/and the B2 doser have effect on the measurement variability.
To get insight into the problem, a few conversations where held with different stakeholders. Concluding this, Sara Lee wants to have a reliable control procedure and form a strategic partnership with the supplier Helvoet in the future enabling reduction or even abolishing incoming control eventually.
A reliable control procedure means that the results of the B2 measuring systems are reliable as well as the measurement approaches used for outgoing and incoming control are clear and used in a correct way. According to Mottonen, Belt, Harkonen, Haapasalo and Kess (2008) measurement variability are unavoidable due to the variations in operators and/or devices regardless how well measurement procedures are designed or maintained. Acceptance sampling is used for incoming control and is the process of determining the acceptability of a quantity of items based upon inspection of a sample of items (Nicholas, 2001). According to Nicholas (2001) acceptance sampling runs counter
Production Forteq and outgoing control
Production Helvoet and outgoing control
Incoming control at Sara Lee Joure
Sealed on liquid coffee bag Sending control information
Sending control information
Delivering B2 dosers Transport to Suffolk
Fulfillment procedure QA Accepted
to modern manufacturing practices and JIT/TQM philosophy of now a day, which is to build in quality, eliminate waste and continuously improvement. Acceptance sampling is done after a batch is produced or received. As a consequence if defects are found, they must be reworked or scrapped. Such kind of control provides little information about the cause of problems, because sources of defects are often difficult to determine. Besides this, acceptance sampling cannot guarantee that the sample items chosen for inspection are accurate representations of the larger batch. On the contrary, Statistical Process Control (SPC) focuses on the production process of the items and seeks to identify potential defect-production problems and resolves them before damage occurs (Nicholas, 2001).
Enabling reduction or even abolishing incoming control also means that Sara Lee has a reliable feeling about the results of outgoing control of Helvoet, which means that a good relationship is necessary. Followed from the conversations one cannot speak of a good relationship between Sara Lee and Helvoet. With strategic partnership in mind, we want to know what both parties expect from one another to achieve an optimal migration from a buyer-supplier relationship to a strategic partnership.
1.1 Research objective
The objective of this research is to give Sara Lee insight into the way the incoming control of the dosing volume of B2 dosers can be reduced or abolished eventually. This objective contains two goals:
- On the short term getting insight in a reliable control procedure between Helvoet and Sara Lee.
- On the long term getting insight in the way the buyer-supplier relationship between Helvoet and Sara Lee can be migrated into a strategic partnership.
1.2 Research questions
From the given area of concern and in order to realize the research objective, seven research questions have been formulated.
Chapter three ‘status quo of outgoing and incoming control’ deals with the question:
1 How is the current control procedure of the B2 dosers organized?
Chapter four ‘control procedure’ deals with the questions:
2 Which factors influence the establishment of a reliable measurement approach to
control the dosing volume of B2 dosers?
3 Which factors influence the establishment of a reliable B2 measuring system to
Chapter five ‘Buyer-supplier relationship’ deals with the questions:
4 What are the current perspectives on the buyer-supplier relationship between
Helvoet and Sara Lee?
5 What are the expectations of the strategic partnership between Helvoet and Sara
Lee?
Chapter 6 ‘Conclusion and recommendations: a reliable control procedure and a strategic partnership’ deals with the questions:
6 How can a reliable control procedure for the B2 doser be realized?
7 How can the buyer-supplier relationship between Sara Lee and Helvoet migrated
in a strategic partnership?
1.3 Stakeholders
The initiator and the supervisor of this research project are: - Wiebe Jongsma (Manager Technology, CT&S Joure)
- Marcel venema (Packaging Engineer at Technology, CT&S Joure)
The stakeholders of this research project are:
Stakeholders Interest
Evert Meindertsma
(Quality manager, CT&S Joure) Marcel Lok
(Quality Assurance, CT&S Joure) Laurent Le Quevel
(Manager Quality & Environment, SLIF Utrecht)
Interest in project because of the measurement requirements for the B2 doser. They inserted the requirements about the measurement procedure in the quality agreement between Sara Lee and the suppliers
Ernst Kruithof
(Laboratory Manager, CT&S Joure)
The control of the B2 dosers on the B2 measuring system takes place at the laboratory. He is interested in a reliable B2 measuring system.
Jack Eberson
(Technical Development Manager, SLIF Utrecht)
Sara Lee International Foodservice (SLIF) sells the Cafitesse machines and is the customer of the coffee packages produced at CT&S. This department has an interest in the research project because they give the requirements, comparable with the coffee machines in the market, for the dosing volume of the B2 dosers. Diederik ten Hoeven
(senior Account manager, Helvoet representative)
Interested in a reliable B2 measuring system at Helvoet and Sara Lee in order to improve the quality of the B2 dosers.
Joost Balder
(external Six Sigma Consultant, Care for Quality)
Performer of the feasibility study to improve the B2 measuring systems.
Martine van Meeteren
(Category Specialist co-packing /co-manufacturing, SLIF Utrecht)
2. Methodology
The research project is set-up using the framework adapted from Checkland and Howell (1998) which shows elements relevant to any piece of research (see figure 3).
2.1 Framework of ideas
The research objective, as outlined in the previous chapter, is in the area of analysing uncertain problems with the control procedure and improving the relationship between Helvoet and Sara Lee to build up trust and establish a reliable control procedure and a strategic partnership between the two companies to reduce or abolish incoming control eventually. Considering the area of concern, a solution is required regarding the discussions about the variability in measurements and the relationship between Helvoet and Sara Lee needs further improvement. Nicholas (2001) mentioned in his book that relying on suppliers can have negative effects and positive effects. To eliminate the negative effects of outsourcing and maximize the positive effects, management must begin by taking a new look at how the customers and suppliers relate to each other.
2.2 Choice of Methodology
The problem situation, as discussed in the area of concern, has both hard and soft characteristics. The causes of deviation in B2 measuring system and the use of measurement approaches are a ‘hard’ problem with modelling languages of mathematics. The potential partnership between Helvoet and Sara Lee is more a ‘soft’ problem where people, who are concerned within a partnership between Sara Lee and Helvoet, can have different perceptions and attitudes. This indicates a socio-technical system.
‘ Systems Analysis’ was defined as (Miser and Quade, 1984): ‘analysis to suggest a course of action by systematically examining the cost, effectiveness and risks of alternative policies or strategies, and design additional ones if those examined are found wanting’. This approach is a way to look at a complex problem of choice under uncertainty. ‘System Analysis’ aims to help decision makers resolve problems arising in complex socio-technical systems. This makes ‘System Analysis’ very capable for this research. In this research the effectiveness of different strategies for both the control problem and the strategic problem will be systematically analyzed and alternatives will be given for a reliable control procedure and to organize the strategic partnership.
In itiatio n
O bjective
A ltern ative s
C o nsequ en ces
C o m m u n ica tin g the resu lts Fo rm u lating th e problem
Iden tifyin g, design in g and screen in g th e altern atives
B u ildin g and u sin g m o del fo r pre dicting co n sequ e n ce s
C o m parin g an d ran kin g alte rna tive s V alu es e n
criteria an d co n strain tsB o u n darie s
Fore castin g fu tu re co nte xt
Evalu atin g th e an aly sis
D ecisio n an d im plem e ntatio n
Eva lu atin g th e o utco m e
Fo rmul at in g Re se arch Evalu a ti o n an d pr es ent at ion
Figure 4, The Systems Analysis Methodology (Reproduced from Miser, 1995)
The Methodology of System Analysis consists of seven major steps, represented in figure 4 (Jackson, 2000). This Methodology is comparable with the DOV-model (diagnose, design and change) of de Leeuw (2002), but the author of this research finds this methodology more extensive which gives more support to expound the different phases in this research.
Formulating the problem. The author started the master thesis with the formulation of
the real problem in the area of concern (chapter one, introduction).
Identifying, designing and screening the alternatives. The author identified all the
aspects of outgoing and incoming control and designed these aspects in tables and models to outline the current outgoing and incoming control (chapter three). Different conclusions about the identification and design are made and used to screen the literature for the alternatives.
Building and using models for predicting consequences. The literature and results
of the case study are used to build models and tables for the control procedure and buyer-supplier relationship (chapter four and five). The design in chapter three and the models and tables of chapter four and five are used to design a reliable control procedure and a strategic partnership (chapter six).
Comparing and ranking the alternatives. The results of the identification and design
and the models and tables for the control procedure are presented and analyzed to the stakeholders of the ‘B2 measuring system’ project. After the presentation the author organized a discussion to get the assumptions needed for writing chapter six.
Evaluating the analyses. The case study results are returned to the interviewers and
presented and analysed with the stakeholders of the ‘B2 Strategy’ project who are involved in a partnership between Sara Lee and Helvoet.
The sixth step ‘decision and implementation and the seventh step ‘evaluating the
outcome’ are the last two steps and were proposed to be out of scope for this research.
Figure 5, Process of outgoing and incoming control of B2 dosers
3. Status quo outgoing and incoming control
In this chapter the first research question is explored: ‘How is the current control
procedure of the B2 dosers organized?
The process of outgoing and incoming control is expressed in figure 5.
This chapter describes the following topics and the most important results are:
1. Product Quality Agreements (PQA) between the suppliers and Sara Lee; the general opinion was that the classifications of defects are not compatible with the classification according to the Military Standard. This will be further explained in chapter four. Beside that, the dosing level specifications of outgoing control are not correctly mentioned in the PQA between the suppliers and Sara Lee.
1. Outgoing control of B2 dosers at the suppliers; the general opinion was that the suppliers use different dosing level specifications for outgoing control and sending different kind of information on the SPC report about outgoing control.
2. Incoming control of B2 dosers at Sara Lee Joure; the general opinion was that the procedure of incoming control is comparable with the reduced control procedure according to the Military Standard, but the tolerances that Sara Lee uses, are not the same. This will be further explained in chapter four. Beside that the dosing level specifications used for outgoing control at Helvoet and incoming control at Sara Lee are not the same.
3.1 Data collection
To describe the current outgoing and incoming control, interviews are made with Quality Assurance, Quality Control, and Technology. Beside that, the PQA and SPC reports of outgoing control at the suppliers and incoming control at Sara Lee are investigated.
3.2 Product Quality Agreement between buyer-supplier
Between the suppliers and Sara Lee a product quality agreement (PQA) will be maintained to ensure a constant quality level as required for the B2 dosers, which records the quality agreements made between the suppliers and Sara Lee. The classifications of possible defects are defined in the PQA as main defect ‘A’ and ‘B’:
SPC outgoing control at Forteq
SPC outgoing control at Helvoet
AQL Incoming control at Sara Lee Joure Sending SPC report
Sending SPC report Delivering B2 dosers
Table 2, Suppliers Information about outgoing control on SPC Report
Table 1, Specifications in PQA
- Main defects ‘A’: Defects which may lead to unusable packages and result in losses, for these defects an AQL value of 0,01 % - 0,15 % is applicable.
- Main defects ‘B’: Defects which may lead to reduced applicability or process ability, for these defects an AQL value of 1,5 % - 2,5 % is applicable.
These classifications are not compatible with the classification according the Military Standard, which will be explained later in chapter four. Beside this, the description of the defects is very general and the dosing level defects are not specified. Due to generality and lack of specification, the reason for choosing AQL 2.5 and AQL 0.01 for any particular kind of defect is not quite clear.
The B2 dosing level specifications are also added in the PQA between Sara Lee and the suppliers. These specifications are outlined in table 1.
It is clear, that there are no consistent agreements about the dosing level specifications of B2 dosers:
- The incoming control of Sara Lee is the same as the outgoing control at Forteq, but these values are not correctly mentioned in the PQA.
– The incoming control of Sara Lee is not the same as the outgoing control at Helvoet and in PQA other specifications are mentioned.
3.3 Outgoing control
Both suppliers use Statistical Process Control (SPC) as outgoing control and communicate the results of SPC to Sara Lee using a report. According to Nicholas (2001) SPC is a periodic process inspection, which can determine whether a process is under control or is potentially moving out of control.
The SPC report of Forteq and Helvoet gives the following information about the outgoing results of the dosing volume of B2 dosers:
Forteq (dosing volume in grams) Helvoet (dosing volume in cc/pulse)
- Test results each tested dosing system
- Product lot identification - Probability plot
- Run chart of dosing trend and BT pressure - Run chart of dosing trend - Histogram of process capability - Histogram of process capability - B2 dosing volume specifications - B2 dosing volume specifications
- Results based on Acceptance Quality Level - Results based on Acceptance Quality Level B2 dosing level specification
PQA between Sara Lee-Helvoet
B2 dosing level specification PQA between Sara Lee-Forteq Average 0,1013 - 0,1114 cc/pulse 0,101-0,111 cc/pulse
Table 3,
Suppliers specifications used for outgoing control
Figure 6, Dosing level specifications Table 4, Specifications used for
incoming control at Sara Lee
Both suppliers mention in their report the B2 dosing volume specifications, which are shown in table 3.
It is clear that, Sara Lee did not make consistent agreements with the suppliers regarding the kind of information Sara Lee wants to receive on the reports. The manner of communication about the outgoing control on the reports is different as well as the report of the dosing level specifications, which should be used for outgoing control.
3.4 Incoming control
From every received batch, Sara Lee uses hundred B2 dosers for incoming control. Every week a double sampling incoming control takes places with hundred B2 dosers from Helvoet and hundred from Forteq, based on the acceptance sampling procedure Acceptance Quality Level (AQL). AQL is an acceptance sampling procedure which states the maximum allowable percentage of defective items in a batch for it to be considered definitely ‘good’ (Nicholas, 2001). The B2 dosing level specifications used for the incoming control at Sara Lee are shown in figure 6 and table 4. The control procedure is described in table 5. Sara Lee takes a first sample of fifty B2 dosers and accepts the batch when two are less B2 dosers are AQL 2.5. If there are more in AQL 2.5 or there are B2 dosers in AQL 0.01, then second a sample of fifty B2 dosers will be measured. The batch will be rejected when in total six or more B2 dosers are AQL 2.5 and/or one B2 dosers is AQL 0.01 in the second sample.
This procedure is comparable with the reduced control procedure according to the Military Standard, but the tolerance Sara Lee uses, is not the same. This is further explained in chapter four.
∗ Ac= Accepted and Re= Rejected
B2 dosing level specifications Helvoet B2 dosing level specifications Forteq AQL 2.5 0,90-0,095 cc/pulse 0,115-0,120 cc/pulse 0,0907-0,0957 cc/pulse 0,1169-0,1221 cc/pulse AQL 0.01 < 0,090 cc/pulse > 0,120 cc/pulse < 0,0907 cc/pulse > 0,1221 cc/pulse AQL AQL 2.5 0.01
Sample Sample Size Ac Re
∗ Ac Re First Second 50 50 2 >2 ≥6 0 1 AQL 0.01 0,0907 cc/pulse AQL 0.01 AQL 2.5 AQL 2.5 0,0957
cc/pulse cc/pulse0,1063 cc/pulse0,1169 cc/pulse0,1221
4. Control procedure
This chapter consists of two phases; (1) the measurement approaches to control the B2 dosing volume and (2) the measurement problem with the B2 measuring system. For both phases the conclusions are written.
4.1 Data collection
Measurement approaches. To outline the measurement approaches in the first phase
of this chapter, the literature is scanned on information about statistical quality control. We will use Nicholas (2001) as the main reference for the measurement approaches.
Measurement system. To outline the measurement problem with the B2 measuring
system a feasibility study is done to identify sources, which can lead to variation in the measurement results. The feasibility study contains a zero measurement to describe the current situation, a brainstorm session to identify all the variation sources of the B2 measuring system, and finally the experiments with the selected sources.
There are five main concepts that are used to describe the performance of a measurement system: accuracy, repeatability, reproducibility, stability and linearity (Mottonen et al., 2008). The zero measurement is assessed through a repeatability and reproducibility study (gage R&R), because there were two important aspects for this research:
− Repeatability is the ability of the measurement system to provide consistent results. The same B2 dozers will be tested a few times on the same system. − Reproducibility is the variability induced by the measurement systems. The same
product will be tested on different systems.
According to Woodall and Borror (2007) gage R&R studies are widely used to assess measurement system variation relative to process variation and tolerance limits. Gage R&R studies provide information on measurement system performance by analyzing measurement error from various sources1. As confirmed by Appele and Raffaldi (2006), who describes a gage R&R study as a study which provides information about the reliability of a measurement system and what needs to be improved in an unreliable system by reviewing the components that make up the percentage gage R&R.
The brainstorm session is done with all stakeholders of the ‘B2 project’. We identified the sources, which can possible lead to variation in and between the B2 measuring systems. A Failure Mode Effect Analysis is used to select the most important ones.
The experiments with the selected sources are described in appendix one.
4.2 Analysis of measurement approaches
This section answers the research question: ‘Which factors influence the establishment of
a reliable measurement approach to control the dosing volume of B2 dosers?’
According to Nicholas (2001) in statistical quality control (SQC), samples from a population are inspected for the purpose of making statistical inferences about the total population. Because of this statistical procedure there is always some risk of making the incorrect conclusions (see figure 7):
- Mistakenly rejecting a good batch is a type I error. - Mistakenly accepting a bad batch is a type II error.
The two main approaches in SQC are statistical process control and acceptance sampling. Statistical process control is used for outgoing control at Helvoet and acceptance sampling is used for incoming control at Sara Lee. We will deal with statistical process control in paragraph 4.2.1 and with acceptance sampling in paragraph 4.2.2.
4.2.1 Statistical Process Control
Every production process is variable in some way. Statistical process control (SPC) is a way to monitor this process to remain within a small variability. The implementation of the SPC procedure at the production company to track and control the process is enacted through the use of control charts. With control charts the process can be determined to be stable, which means that the remaining samples all lie within control limits. A control chart has two limits, an upper control limit (UCL) and a lower control limit (LCL). A value that does not lie between these limits indicates that a process is possibly going out of control. A process should be in control and should be capable. A process is capable when the output falls within the specification limits of the customer, an upper specification limit (USL) and a lower specifications limit (LSL). Control limits and specification limits are not the same. The SPC procedure with UCL and LCL and USL and LSL are shown in figure 8.
The procedures of SPC are used for controlling the process and used to assess the capability of a process. According to Kureková (2001) capability of the process means ability of the process to meet technological or other requirements specified by the customer, in this case Sara Lee. Therefore, this chapter is focused on assessing the capability of the process, called process capability.
Figure 7, Different Test Cases
Figure 8, Control limits and specification limits
Table 6,
Process capability indices Process capability indices. Process capability indices establish the relationship
between the actual process performance and the manufacturing specifications (Chen, Pearn and Lin, 2003). Several types of capability indices exist (Kureková, 2001 and Pearn and Yang, 2001) Capability indices of any process can be divided into the capability indices of the first and second generation (Kureková, 2001). The first generation capability indices (Cp, Cpk) is based on classical philosophy of the SPC. According to that philosophy all measurement results within required tolerance interval are intended to be good and measurement results outside the tolerance interval are considered to be bad. The second generation capability indices (Cpm, Cpmk) are rising from the new approach to the quality improvement (Taguchi approach) and gives knowledge on how good the process is. According to Pearn et al. (2001) Cpm is a loss-based index and Cpk is a yield-loss-based index. The Cp and Cpk indices are appropriate measures of progress for quality improvement in which reduction of variability is the guiding principle and process yield in the primary measures of success (Wu and Pearn, 2008). For this reason and because Helvoet is has just started with the SPC procedure, the Cp and Cpk values will be used to asses the process.
Process capability index Cp is determined by comparing the natural variability of a process or product to customer specifications (see table 6). Cp expresses only the potential process capability. It does not represent the position of the natural tolerance interval considering the position of the required tolerance interval, which means that it does not give a clear answer whether the measured value fits within the tolerance interval (Kureková, 2001). So, correct application of Cp presumes that the process mean is the target value, that is, the process is squarely centred between the USL and LSL. If this is not the case, the Cpk should be measured, which is the minimum of the upper process capability and the lower process capability (see table 6). The design of the Cpk index is visible in figure 9. For the production company it is important to meet the specifications of the product and to get as close as possible to the target value with the process mean. In figure 9 this is the interval between X and X0. This can be measured by the scale distance k (see table 6), which can be interpreted as follows:
- 0 ≤ k ≥ 1 The process mean lies between USL and LSL.
- k = 1 The process mean lies on target, which should be the goal. - k ≥ 1 The process mean falls on or outside the specifications.
Indices Cp = USL-LSL 6σ Cpk = min (Cpu, Cpl) Cpu = USL - μ 3σ Cpl = μ – LSL 3σ k= 2 |μ – target| USL-LSL where Cp (1-k) = Cpk
Figure 10, Sampling Procedure
General remarks on Cp and Cpk. If the process is centred, the measurement process
is always capable when capability index value exceeds 1 (Kureková, 2001, Hopp and Spearman, 2008). Practical recommendations consider capability index of 1.33 for companies who starts with SPC (Nicholas, 2001 and Kureková, 2001). This is also a general rule in statistic2. Extremely precise measurements should have a value of 1.67 (Kureková, 2001 and Chen et al., 2003). According to Mottonen et al. (2008) the data used for calculating Cp and Cpk are based on samples and a relatively high quantity of samples is required for the values of Cpk en Cp to be reliable (100 to 200, at least).
For the production process of B2 dosers on this moment at Helvoet, we propose a value of Cp around 1.67 in combination with a value of Cpk=1.33. These values are realistic when the specification limits are the same as used at the AQL 0.01 for incoming control, which means the USP = 0,1221 and the LSL = 0,0907. With these values the natural tolerance of the process includes 99,9967% of the process distribution, which means three hundred twenty parts per million fall out of range. In the statistic a rough-and-ready rule is that the difference between the control limits and specification limits are 2:3. In this case we suggest that Helvoet should work with the UCL = 0,1169 and LCL = 0,0959.
4.2.2 Acceptance sampling
Acceptance sampling is implemented by using sampling plans, which make inferences about a larger population. We proposed a four step procedure to get the right sampling plan, which is outlined in figure 10.
Step 1. Sampling plans appear in published
tables, the most common used are Military Standard (MIL-STD) and Dodge-Romig (Nicholas, 2001). The MIL-STD creates a
sampling plan based upon a desired Acceptable Quality Level (AQL) and Dodge-Romig creates a sampling plan based on Lot Tolerance Percent Defective (LTPD) and Average Outgoing Quality Level (AOQL). Nicholas (2001) defined AQL as the maximum allowable percentage of defective items in a batch for it to be considered definitely as ‘good’. LTPD is the poorest level of quality that the customer is willing to accept (Wu and Pearn, 2006) Sara Lee specifies the quality level of a batch according to the AQL procedure, which makes the MIL-STD suitable for the B2 dosing controls. There are different MIL-STD reports for different kinds of inspections. The MIL-STD has been almost universally
2 http://www.isixsigma.com/
Acceptance Sampling
Kind of published tables (Military Standard, Dodge-Romig)
Level of inspection
(I, II, III) Batch size
Kind of sampling plan (single, double, mulitple)
Risk Level (classification defects and
AQL)
Final sample plan (sample size, acceptance number)
Selecting Code letter
Selecting Table Level of inspection
(I, II, III)
Code letter Step 1
Step 2
Step 3
adopted by government and private consumers for the lot-by-lot sampling inspection of products (Brown and Rutemiller, 2006). The MIL-STD-105 is useful for sampling procedures and tables for inspection by attributes, which makes this the one for the B2 dosing control.
Step 2. The MIL-STD tables include sampling plans for reduced (I), normal (II), and
tight (III) levels of inspection. Normal inspection will be used at the start of the inspection, the tight level of inspection can be used when recent quality deteriorated and the reduced level of inspection when recent quality improved. Switching procedures of the MIL-STD can be used to require changes in inspection levels. When there are relatively small sample sizes necessary and large sample risks can be tolerated, four special inspections levels are given by the MIL-STD, But at the selection of the four special inspections levels care must be taken to avoid AQL’s which are inconsistent with these inspection levels, so it is of no use to choose S-1 if the AQL is 0,1.
Because several sampling plans are given, the one which has to be selected in one of the tables of the MIL-STD depends on the batch size and desired level of inspection, which is formulated as a code letter. The sampling plan for the B2 dosers is described in chapter three. The selected sampling plans for a delivered batch under/above 150000 B2 dosers at the inspection levels ‘reduced’ control and ‘normal’ control can be found under the specific code letters in the tables of the MIL-STD 105 (see table 7).
Special inspection levels General inspection levels Lot or batch size
S-1 S-2 S-3 S-4 I II III
150000 to 500000 D E G J M P Q
Table 7, Sample Size Code Letter
Step 3. The MIL-STD tables include sampling plans for single, double and multiple
sampling of batches. Suppliers prefer double or multiple sampling, because they are given two or more chances to acceptance. In chapter three is explained that Sara Lee uses double sampling plans. The table which should be used to find the sampling plan is the combination of the kind of sampling plan with the desired level of inspection.
Step 4. Part of choosing a sampling plan is to specify the level of acceptable risk in using
the plan, this is the defined AQL. The MIL-STD classified defects in three classes, where different AQL’s can be chosen for different kinds of defects:
1. Critical defect: defect that results in hazardous and unsafe situations or defect that is likely to prevent performance of the tactical function of major end item. 2. Major defect: defect that is likely to result in failure, or to reduce materially the
use of the unit of product for its intended purpose
According to the Hong Kong Quality Centre Ltd, critical defects will not be accepted and the AQL 0.01 is suitable, for major defects the AQL 2.5 is suitable and for minor defects the AQL 4.0 unless otherwise in structured by the customer3.
With the specified level of risk in combination with the code letter, the sampling plan is given in the selected table and consists of two parameters: (1) sample size and (2) acceptance number. Sara Lee uses two specifications of risk levels; AQL 2.5 and AQL 0.01. According to the MIL-STD 105 the two different kind of sampling plans can be found in the table for double sampling, reduced control under the code letter L and double sampling normal control under the code letter N. The specified sampling plan for L and N in combination with the AQL 2.5 and AQL 0.01, which Sara Lee uses for the different kind of defects, are shown in table 8 and 9 respectively.
4.3 Conclusion of measurement approaches
Within paragraph 4.2 the question ‘Which factors influence the establishment of a reliable
measurement approach to control the dosing volume of B2 dosers?’ was explored.
It is shown, that for outgoing control of the B2 dosers the capability of the production process at Helvoet is interesting for Sara Lee. The capability of the production process at Helvoet is determined by comparing the natural variation of the B2 doser to Sara Lee specifications, with the measure Cp. Because there is almost always a variation between the process mean and the target value, the minimum of Cpu and Cpl should be taken, which is the Cpk.
It is shown that for the acceptance sampling control, which is the incoming control at Sara Lee, a few steps should be made to come to a final sampling plan. The factors which influence the incoming control, when the Military Standard is used for acceptance sampling, are shown in figure 8. Combining these factors results in a final sampling plan, where the sample size and the acceptance numbers are mentioned. In chapter six we have written a proposal for the measurement approaches of the B2 dosers.
3 http://www.hkqcc.com/quality_control_standards/acceptable_quality_level.html Table 8, Sampling plan double, reduced control
= Takes first sampling plan under the arrow = Takes first sampling plan above the arrow
4.4 Analysis of measuring system
This section gives answer to the research question: ‘Which factors influence the
establishment of a reliable B2 measuring system to control the dosing volume of B2 dosers?’
To solve the discussion about the measurement results, we performed a feasibility study. The objective of the feasibility study was to recognize the causes of deviation in the B2 measuring systems. This section gives the analysis of the feasibility study, which is divided in (1) the zero measurement and (2) the experiments with the variation sources. The detailed analysis of the feasibility study is presented in appendix one.
4.4.1 Zero measurement
According to Woodall and Borror (2007) a value of 10% gage R&R is acceptable and a value between 10% and 30% might be acceptable, depending on factors such as the importance of the application, the cost of the measurement system, and the cost of repair. Values above 30% are unacceptable, because they indicate that there are measurement errors from various sources, which means that the measurement system is unreliable. It is recommended that every effort should be made to improve the measurement system, when the gage R&R is above the 30%.
The total gage R&R between the measurement results at Helvoet and Sara Lee was in one measurement 68,53% and in another one 60,45% (see appendix one). The differences are limited, but the gage R&R percentage is unacceptable.
4.4.2 Experiments with variation sources
The experiments are done to find out what the effects of sources are on dosing level and dosing stability. We define the effect on dosing level as the volume output of an individual B2 doser and the effect on dosing stability as the stability of volume output of different B2 dosers. We did different kinds of experiments with the under mentioned sources. The most important results for each source are given.
1. Temperature effect of the coffee extract: we discovered that there were no significant differences with respect to the dosing level and dosing stability for different temperature categories.
2. Liquid effect: comparing coffee extract as a test liquid with the glycerol/aqua solution, the coffee extract seems the best test liquid for B2 dosers.
Figure 11, within and between variation
that the current design of the B2 measuring system gives the most stable measurement result and seems the optimum one.
4. Endurance test with B2 measuring system: the cause by the stops during the endurance test had an effect on the dosing volume and was on average 1.5%. 5. Multi-scale∗ versus B2 measuring system: comparing the multi-scale with the B2
measuring system, both systems had very stable measurement results. 6. Control electronics: there was no instability with the control electronics.
7. (Un)Loading process of the B2 doser: the unloading process of a B2 doser had a significant effect on dosing stability and dosing level. Looking at the results of the zero measurement the unloading process gives a ‘within variation’ of 1% till 1,3%, while the unloading process gives a ‘between variation’ of 3% till 12%. These variations are shown in figure 11. The ‘within variation’ is the short term measured variation within one test sequence. One test sequence starts with five dosages, next ten dosages and each will be measured, and finally two deflation dosages. The ‘between variation’ is the long term measured variation between several sequences.
4.5 Conclusion of measuring system
In paragraph 4.4 the question ‘Which factors influence the establishment of a reliable B2
measuring system to control the dosing volume of B2 dosers?’ was explored.
A reliable B2 measuring system can be realized by eliminating the variation source. We discovered from the feasibility study that measurement instability can be explained by ‘(un)loading of B2 doser’. It is clear that during one test sequence, the B2 doser is very stable. But when the B2 doser is unloaded and placed for a second time in the measurement system, the dosing level of the next run of ten dosages is again very stable but on a different level (see figure 11). The cause of ‘(un)loading B2 doser’ could be the B2 measuring system but also the B2 doser. The gage R&R of the zero measurement is done with the assumption that the B2 doser was a none-variable factor. We should now consider the interpretation of the gage R&R, because the B2 doser can have an impact on the variation in the measurement results. The source ‘(un)loading B2 doser’ should be further researched to know the effect of the B2 measuring system and B2 doser on the gage R&R.
5. Buyer-supplier relationship
This chapter consists of two phases; (1) the analysis of the relationship between Helvoet and Sara Lee and (2) the analysis of the expectations of a strategic partnership between Helvoet and Sara Lee. For both phases the conclusions are written.
5.1 Data collection
The literature is scanned on information about buyer-supplier relationships. We use Han, Wilson and Dant (1993) and Walter, Müller, Helfert and Ritter (2003) as main references. Han et al. (1993) investigated the major factors, which are related to moving companies toward using fewer suppliers along with characteristics of a good relationship and the advantages and problems with long-term relationships. Walter et al. (2003) found empirical evidence that the supplier’s fulfilment of direct and indirect relationship functions for the customer definitely matters for the customer’s perception of relationship quality.
The factors regarding the buyer-supplier relationship today mentioned by Han et al. (1993) and the functions of industrial supplier relationships and their impact on relationship quality mentioned by Walter et al. (2003) are used to formulate the research variables in the case study. We carried out the case study at operational and tactical level at Sara Lee and Helvoet. All results and recommendations of the departments of Sara Lee and Helvoet about the variables influencing the quality of the buyer-supplier relationship and the expectations about the strategic partnership are mentioned respectively in appendix two and appendix three.
5.2 Analysis of the relationship between Helvoet and Sara Lee
This section answers the research question: ‘What are the current perspectives on the
buyer-supplier relationship between Helvoet and Sara Lee?’
5.2.1 Factors influencing the quality of a buyer-supplier relationship
Han et al. (1993) mentioned that the characteristics of a good relationship are based on mutual trust and satisfaction. Mutual trust in the relationship means that the trust in one’s partner is developed by constant and detailed exchange of information, which reduces uncertainty of performance, and the continued exhibition of commitment, which nurtures and maintains the exchange relationship. Satisfactory exchange means that each partner involved in the exchange relationship should be happy and satisfied with the performance of another. Walter et al. (2003) mentioned these factors as essential concepts to describe relationship quality and added the factor commitment. Commitment means an agreement or promise to do something in the future.∗ Beside these factors, Walter et al. mention direct functions and indirect functions which matter the perception of relationship.
5.2.2 Perspective on the buyer-supplier relationship
The most important perspectives of Sara Lee and Helvoet, with regard to the factors that influence the quality of a buyer-supplier relationship, are mentioned in this paragraph.
Most important perspectives of Sara Lee. The last two years there are more quality
problems with Helvoet. Apart from that, the communication of quality problems from Helvoet is not proactive. Helvoet is currently making improvements.
Most important perspectives of Helvoet. The relationship between Helvoet and Sara
Lee can be formulated as a good and familiar long-term relationship, but since the beginning of the relationship there are still points for discussion which needs to be solved. These points are correct dosing level specifications, AQL or Cpk value, test procedures and the communication. It is improving, but there is still a lack of communication due to functional problems. Helvoet has no marges on the B2 dosers.
5.3 Conclusion of the relationship between Helvoet and Sara Lee
In paragraph 5.2 the question ‘What are the current perspectives on the buyer-supplier
relationship between Helvoet and Sara Lee?’ was explored.
It is shown in paragraph 5.2.1 that different factors have an influence on the relationship quality between a buyer and a supplier. Regarding these factors the following conclusions can be made about the relationship between Helvoet and Sara Lee:
− Direct functions; the perception on relationship quality can be improved by the direct functions quality and cost of the B2 dosers.
− Indirect functions; the perception on relationship quality can be improved by the indirect functions innovative development and social support.
− Commitment; commitment to focus on a long-term relationship is present.
− Trust; communication, according to constant and detailed information exchange between both companies, needs further improvement for a more stable relationship and building up trust.
− Satisfaction; due to the direct functions, indirect functions and the lack of communication, the satisfaction about supplying and buying is not optimal.
5.4 Analysis of expectations of a strategic partnership
This section answers the research question: ‘What are the expectations of the strategic
partnership between Helvoet and Sara Lee?’
Besides B2 dosers, Helvoet currently delivers the dosers for aqua, milk, tea and choco as well. Helvoet and Sara Lee invested a lot in one another regarding the production of these dosers. With the expected quantities for coming years, Sara Lee has decided to move from two suppliers to one (Helvoet) and considered to have this supplier on two locations (USA and Europe) and form a strategic partnership with Helvoet.
Moving to fewer suppliers. In spite of this decision, it is shown that from the different
points of view of the Sara Lee departments, there is a need for dual sourcing to avoid a sole dependence on one supplier. Especially for such a critical strategic product and keeping in view that the quality performance and quality agreements of Helvoet are not optimal.
According to Han et al. (1993) there are three latent factors for companies to move toward fewer relationships:
− Enhanced performance, which refers to the ease of interacting with the exchange partner, reducing anxiety and uncertainty associated with doing business with an unknown customer or supplier, and making it easier to resolve and settle conflicts that might arise in the course of business relation
− Purchasing cost reduction, which refers to the savings accrue from establishing a close relationship with the exchange partner
− Increased technical cooperation, which refers to the access of the partner’s technology. The firm has a strong partner which helps in product development. Comparing these factors with the results of the case study only the purchasing cost reduction and the increased technical cooperation factors are the factors relevant for Sara Lee to move toward one strategic partner. However, the supplier Helvoet is assessed as a less strong partner than Forteq in product development.
5.4.1 Factors influencing a strategic partnership
Advantages of a close long-term relationship. The research of Han et al. (1993)
suggests that industrial buyers and suppliers differ in their perception of the main advantages of close buyer-supplier relationships. The primary advantages for buyers relate to the purchasing efficiency that are realized when they maintain close ties with their suppliers. The supplier considered factors (1) price/production stability, (2) enhanced marketing efficiency, (3) optimal capacity planning, and (4) customer-orientated as primary advantages.
Problems of a close long-term relationship. The latent problems of long-term
relationships for buyers as well as suppliers are pending the future competition (1) better exchange alternatives and (2) overdependence. The wealth of the each partner becomes inexorably linked to the performance of the other.
5.4.2 Expectations of a strategic partnership
The most remarkable expectations of the strategic partnership for both companies are quality improvement, improved scheduling and stronger partnership. On the other hand, Sara Lee also expects that a long-term relationship with Helvoet can lead to more uncertainty about relationship working, opportunistic behaviour by Helvoet, and overdependence on one supplier. Helvoet expects better knowledge of customer needs, joint product development and price stability, but no cost reduction. A strategic partnership is according to Helvoet very beneficial for both parties and Helvoet is willing to enter into a strategic partnership, but it is necessary for Helvoet and Sara Lee to work together on a better partnership relation. From Helvoets’ point of view Sara Lee can use the dosing competence of Helvoet for further improvement of the B2 doser for a new life cycle and Sara Lee will benefit in functional quality and supply stability.
5.5 Conclusion of expectations of a strategic partnership
In paragraph 5.4 the question ‘What are the expectations of the strategic partnership
between Helvoet and Sara Lee?’ was explored.
It is shown in paragraph 5.3.1 that there are different advantages and problems with a close long-term relationship between buyer and supplier.
Advantages of a close long-term relationship. As a result of comparing the
advantages mentioned by Han et al. (1993) with the case study results, the factor purchasing efficiency is relevant for Sara Lee to form a strategic partnership. The factors price/production stability and optimal capacity planning are relevant for Helvoet.
Problems of a close long-term relationship. As a result of comparing the problems
6. Conclusion and recommendations: a reliable control procedure and a strategic partnership
The objective of this research is to give Sara Lee insight into the way the incoming control of the dosing volume of B2 dosers can be reduced or even abolished.
- On the short term the incoming control can be reduced when B2 measuring system is reliable and the outgoing control of Helvoet is stable within the values of Cp around 1.67 in combination with Cpk≥1.33. The reliability of the B2 measuring system is dependent on the results of further research; the cause of the variation of the (un)loading process of the B2 dosers.
- On the long term the incoming control can be abolished when the production process of Helvoet stays in control and remains capable and when there is stability and satisfaction regarding the strategic partnership and both companies trust each other.
6.1 Realizing a reliable control procedure
This section answer the research question: ‘How can a reliable control procedure for the
B2 doser be realized?’
A reliable control procedure is based on the factors influencing the outgoing control, incoming control and the B2 measuring system, which are shown in figure 12. This section describes the topics (1) a reliable outgoing control, (2) a reliable incoming control, and (3) a reliable B2 measuring system.
For the establishment of a reliable control procedure we took a few assumptions into account:
− Sara Lee and Helvoet use the dosing specification levels according to the current incoming control.
− Sara Lee uses Acceptance Quality Levels for the acceptance sampling incoming control.
− Sara Lee uses the double sampling, reduced inspection level for acceptance sampling incoming control.
− Sara Lee classifies the defects in critical and major classification. Outgoing control Incoming control B2 test system
Process capability Level of inspection (Un) loading process Batch size
Kind of sampling plan Classification of defects
in combination with risk level (AQL)
Table 10, information for a reliable outgoing control
6.1.1 A reliable outgoing control
Chapter three made clear that the suppliers use different specification levels for outgoing control and mark different information on the SPC report. The establishment of a reliable outgoing control is realized when Helvoet uses the USL and LSL in table 10 and has a result of Cp≥1.5 in combination with
Cpk≥1.33. If, in the long term, the production process stays in control and remains capable, the Cpm should be considered for decreasing the variability around the target.
The information in table 10 should be reported in the PQA between Helvoet and Sara Lee and every week Helvoet should send the SPC report per assembly line and before the delivery of the batch. The information on the SPC report should be agreed between Helvoet and Sara Lee and contain at least the following information: (1) run chart, with the observations in cc/pulse and upper and lower specifications, (2) probability plot with the p-value (should be lower than <0.05; means it is a normal deviation) and on the x-axis the cc/pulse and on the y-x-axis the percentage, (3) process capability values, (Cp, Cpk and Cpm) and (4) performance.
6.1.2 A reliable incoming control
Chapter three made clear that the procedure of incoming control is comparable with the reduced control procedure according to the Military Standard, but the tolerance Sara Lee uses, was not the same. Besides, the dosing level specifications used for outgoing at Helvoet and incoming control at Sara Lee were not the same. The establishment of a reliable incoming control at Sara Lee is realized when the sampling procedure in figure 13 is followed. This result in a final sampling plan under the code letter M in the MIL-ST-105 table double sampling, reduced control inspection level. This sampling plan has a double sample quantity of eighty B2 dosers and the accept en reject numbers are red encircled. If in the first sample batch the quantity of defective items is between accept and reject values, a second sample is required. The specifications used for incoming control (see figure 13) should be reported in the PQA between Helvoet and Sara Lee.
B2 dosing level specifications Helvoet
Process capability indices
USL= 0,1221 cc/pulse Cp around 1.67
LSL= 0,0907 cc/pulse Cpk = 1.33
Figure 13, Reliable incoming control Acceptance Sam pling
Military Standard
Reduced
level of inspection 180000
Double sam pling plan
AQL 0.01 and AQL 2.5 (sam ple size, acceptance num ber)Final sam ple plan
Figure 14, Factors influencing a strategic partnership
6.2.3 A reliable B2 measuring system
A reliable B2 measuring system can be realized by eliminating the variation source ‘(un)loading process of B2 doser’. The cause of this factor could be the B2 measuring system but also the B2 doser. In further research the (un)loading process of the B2 dosers should investigated to get a reliable outcome of the measurement of the B2 test results.
6.2 Organizing a strategic partnership
This section answers the research question: ‘How can the buyer-supplier relationship
between Sara Lee and Helvoet migrate in a strategic partnership?’
A strategic partnership can be achieved by (1) improving the factors which influence the quality of the buyer-supplier relationship and by (2) getting the expectations on the same level. These factors and expectations are shown in figure 14.
6.2.1 Improving relationship quality between Helvoet and Sara Lee
To organize the strategic partnership, the relationship quality should be improved by the following factors:
− Cost of B2 dosers should be lower and quality of B2 dosers should be improved. − Sara Lee and Helvoet should work together on the development of the B2 dosers. − Social support between Helvoet and Sara Lee should be improved.
− Constant and detailed information exchange needs further improvement for a stable relationship.
− By improving the direct, indirect functions and communication between Helvoet and Sara Lee both companies should be more satisfied with each other.
6.2.2 Getting the expectations of the strategic partnership on the same level
Both companies are committed to form a strategic partnership and expect quality improvement, improved scheduling, and stronger partnership. Helvoet also expects joint product development and price stability. Helvoet and Sara Lee should work together and use each other as sparring partners, which challenges Helvoet to be more proactive. Helvoet wants to be responsible for the functional development of the B2 doser. On the other hand, Sara Lee is afraid for opportunistic behaviour by Helvoet and become overdependence on one supplier.
Factors influencing
quality of relationship Expectations Cost Functional development Price/production stabiliy Quality Communication Delivery volumes Trust Satisfaction Commitment Quality
References
Brown, G.G. and H.C. Rutemiller, 2006, ‘A cost analysis of sampling inspection under Military Standard 105D’, Navall Research Logistics Quarterly, Vol. 20, Issue 1, pg. 181-199
Checkland P., Holwell S., 1998, Action Research: Its Nature and Validity, Systemic practice and action research, Vol. 11, Issue 1, pp. 9-21
Chen, K.S., Pearn, W.L. and P.C. Lin, ‘Capability measures for process with multiple characteristics’, Quality and Reliability EngineeringIinternational, Vol. 19, pg. 101-110
Han, S., Wilson D.T. and S.P. Dant, 1993, ‘Buyer-Supplier Relationships Today’, Industrial Marketing Management, Vol. 22, pg. 331-338
Hopp, W.J. and M.l. Spearman, 2008, Factory physics, McGraw-Hill, Singapore
Jackson, M.C., 2000, Systems approaches to management, Kluwer academic/plenum publishers, New York. Kappele, W.D. and J. Raffaldi, 2006, ‘Quality software and analysis: Gage R&R Improves Quality and Profitability’,Quality magazine, Vol. 6, pg. 50-51
Kureková, E., 2001, ‘Measurement process capability; trends and approaches’, Measurement Science Review, Vol. 1, Issue 1, pg. 43- 46
Leeuw, de A.C.J., 2003, Bedrijfskundige procedureologie; management van onderzoek, Van Gorcum, Assen. Miser, H.J. and E.S. Quade, 1984, Handbook of system analysis, John Wiley & Sons Ltd., New York.
Mottonen, M., Belt, P., Harkonen, J., Haapasalo, H., and P. Kess, 2008, ’Manufacturing process capability and specification limits’, The Open Industrial and Manufacturing Engineering Journal, Vol. 1, pg. 29-36
Nicholas, J.M., 2001, Competitive manufacturing management, Tata McGraw-Hill, New Delhi
Pearn, W. L. and S.L. Yang, 2001, ‘Testing process capability using the index Cpmk with an application’, International Journal of reliability, Quality and Safety Engineering’, Vol. 8, Issue 1, pg. 15-34
Ronen, B., 1994, ‘An information-economics approach to quality control attribute sampling’, European Journal of Operational Research, Vol. 73, pg. 430-442
Walter, A., Müller, T.A., Helfert, G. and T. Ritter, 2003, ‘Functions of industrial supplier relationships and their impact in relationship quality’, Industrial Marketing Management, Vol. 32, pg. 159-169
Wilson, B., 1990, Systems: concepts, procedureologies and applications, John Wiley & Sons Ltd., New York. Woodall, W.H. and C.M Borror, 2007, ‘Some relationship between Gage R&R Criteria’, Quality and Reliability Engineering International’, Vol. 24, pg. 99-106
Appendix 1. Feasibility study
2.1 Analysis Zero measurement
Goal. Define the total gage R&R between Helvoet and Sara Lee
Set up. Sixteen Helvoet B2 dosers are tested the first time at Helvoet, a second time at
Sara Lee and a third time at Helvoet. Another sixteen B2 dosers (selected of a delivered batch of Helvoet) are tested the first time at Sara Lee, a second time at Helvoet and a third time at Sara Lee.
Results. The results are presented in figure 10.
Gage R&R Helvoet Æ CT&S: (range van 0,019 cc/puls)
Study Var %Study Var %Tolerance Source StdDev (SD) (6 * SD) (%SV) (SV/Toler) Total Gage R&R 0,0034264 0,0205586 72,88 68,53 Repeatability 0,0024331 0,0145989 51,75 48,66 Reproducibility 0,0024125 0,0144751 51,31 48,25 Tester 0,0024125 0,0144751 51,31 48,25 Part-To-Part 0,0032194 0,0193165 68,47 64,39 Total Variation 0,0047016 0,0282096 100,00 94,03
Very little differences However GAGE R&R percentage unacceptable
Gage R&R CT&S Æ Helvoet: (range van 0, 012 cc/puls)
Study Var %Study Var %Tolerance
Source StdDev (SD) (6 * SD) (%SV) (SV/Toler) Total Gage R&R 0,0030224 0,0181347 45,12 60,45 Repeatability 0,0013005 0,0078032 19,42 26,01 Reproducibility 0,0027283 0,0163699 40,73 54,57 Tester 0,0022709 0,0136251 33,90 45,42 Tester*Product 0,0015123 0,0090736 22,58 30,25 Part-To-Part 0,0059777 0,0358660 89,24 119,55 Total Variation 0,0066983 0,0401900 100,00 133,97
2.2 Analysis temperature effect
Goal. Quantify the temperature effect of coffee on the B2 dosing level.
Set up. The temperature is tested with two dosers of Helvoet and two of Forteq. All four
dosers are tested four times at each temperature category (6 °C, 9°C, 11°C, 13°C and 15°C.), which gives sixteen measurements.
Results. There were no significant differences with respect to individual dosing-level
between the five temperature-categories. There were no significant differences with respect to the spread between the five temperature-categories.
2.3 Analysis Liquid effect
Goal. Quantify the liquid effect on dosing stability and dosing level and to compare the
results of glycerol/aqua solution (50%) with the results of coffee extract.
Set up. Twenty five B2 dosers of Helvoet and Forteq are tested with the glycerol/aqua
solution. Sixteen B2 dosers of Helvoet and Forteq are tested with coffee.
Figure 16,
Different Height Liquid Column Figure 17, Different Position of Inlet Figure 18, Different Diameter of Inlet Results. The average dosing level of the liquid glycerol/aqua is greater than the average
dosing level of coffee. The stability of coffee extract is much higher than testing with glycerol/aqua.
2.4 Analysis factors related to transport and flow of coffee extract
Goal. Quantify the effect of the factors and interaction of the factors on dosing level and
dosing stability.
Set up. The factors related to the transport and coffee flow are the height of the liquid
column, the position of coffee inlet and the internal diameter of coffee inlet. For this ‘Design of Experiment’ the ‘Two Level Full Factoral’ was chosen to investigate these factors including interactions on different settings, which are designed in figure 16,17, and 18. The experiment is set up with two B2 dosers of Helvoet and of Forteq. The number of test runs was 128 tests.
Results. Regarding the dosing level of a B2 doser, the height of the liquid column was the most significant factor. The higher liquid column gives higher dosage level. The other factors had no significant effect. Regarding the stability of the dosage, the height of the liquid column is the most significant factor. The higher liquid column gives more variation between the measurements. The other factors had no significant effect.
2.5 Analysis endurance test with B2 measuring system
Goal. Quantify the dosing behaviour with an endurance test.
Set up. Two litre coffee extract is used for the endurance test, with stops at 50, 100,
150, 250, and 350 dosage. One doser of Helvoet and one of Forteq were tested and during these stops the B2 dosers are not unloaded.
