Understanding Supply Chain
Complexity within the Food Processing
Industry: a Multiple Case Study
Master Thesis
MSc Technology & Operations Management
University of Groningen, Faculty of Economics and Business
Keywords: Supply Chain Complexity, Food Processing Industry.
Author: K. Dun
Student number: S2394308
Supervisor: Prof. Dr. Dirk Pieter van Don k
Co-assessor: Hendryk Dittfeld
Date: June 22, 2015
2 Table of Content
Preface ... 3
1. Introduction ... 4
2. Theoretical Background ... 6
2.1. Supply Chain Complexity (SCC) ... 6
2.2. Supply Chain Complexity Drivers ... 7
2.3. Characteristics Food Processing Industry (FPI) ... 7
2.4. Relationship FPI Characteristics and SCC ... 9
3. Methodology ... 11
3.1. Case Selection ... 11
3.2. Measuring SCC Influenced due to FPI Characteristics ... 13
3.3. Data Collection ... 14
3.4. Data Analysis ... 14
3.5. Research Quality ... 15
4. Results ... 16
4.1. Variability of Supply ... 16
4.2. Variability of Quality of the Raw Materials ... 18
4.3. Perishability and Divergent Product Structure ... 19
4.4. Food Safety Regulations ... 21
4.4. Several Recipes and Setup Times ... 22
4.5. Overview Results ... 22
5. Discussion ... 24
5.1. Discussion Findings ... 24
5.2. Discussion Overview and Measurements ... 26
6. Conclusion ... 28
3 Preface
4 1. Introduction
Supply chains of the food processing industry (FPI) are becoming more and more complex due to globalisation and growing expectations of the customers. Characteristics of the FPI such as a divergent product structure, perishability of raw materials and a highly coupled production process can make the supply chain even more complex. The specific characteristics limits the possibilities for supply chain integration in the FPI (Van Donk et al., 2008) and are of vital importance for supply chain performance (Rong et al., 2011). Analysing and understanding the effect of these specific characteristics on the supply chains in the FPI, should be the first step to develop a clear strategy to deal with supply chain complexity (SCC). Companies need to understand the source of complexity in order to manage it (Manuj and Sahin, 2011).
Literature shows an increase in studies that investigate SCC, where authors identified the negative effect of SCC on performance of a company. For instance Vachon & Klassen (2002) found the negative effect of SCC on delivery performance and Bozarth et al. (2009) found the negative impact of SCC on manufacturing plant performance. Several authors identified drivers of SCC (e.g. Bozarth et al.,2009, Isik, 2010, de Leeuw et al., 2013), a SCC driver is any property of a SC that increases complexity (Serdarasan et al., 2013). The SCC drivers play a varying role in the different types of SCs (Serdarasan et al., 2013), however the FPI has received only little attention in recent literature. According to Rong et al. (2011) is managing SCs of the FPI complicated due to the specific product and process characteristics of the FPI. SCC drivers do not always have negative impact on SCC and therefore, need not always be eliminated or reduced (Bozarth et al., 2009). Increased SCC is often accepted to achieve corporate profitability goals (Manuj and Sahin, 2011) or competitive goals (Bozarth et al., 2009). For instance a characteristic of the FPI is the divergent product structure, that leads to increased SCC. However a divergent product structure is needed because of the consumer-driven market in the FPI (Van Donk et al., 2008). Companies within the FPI should manage SCC instead of only reducing SCC to the lowest possible levels.
5 to understand the source of complexity in order to manage it (Manuj and Sahin, 2011). According to Serdarasan et al. (2013), analysing and understanding the SCC drivers first, will be an effective way to proceed to develop a clear strategy to deal with SCC. This research will take the first step to manage the SCC by understanding the SCC drivers within the FPI, taken into account the characteristics of the FPI. Thereafter, the influence of the SCC drivers on SCC will be investigated. The research question of this paper is as follows:
How do specific food processing industry characteristics influence supply chain complexity? This research used a multiple case study to conduct interviews with internal and external managers of FPI plants. A case study is in this research very suitable, because a case study can determine the link between cause and effect. This paper can be used as an overview to get an understanding of the SCC within the FPI. This can support managers to improve their SC management and for instance increase performance. As stated by Perona and Miragliotti (2004) the way companies handle their operations system complexity has a deep effect on how well they perform.
6 2. Theoretical Background
This chapter discusses the theoretical background of the research. Sections 2.1 and 2.2 explain SCC and SCC drivers. Section 2.3 discusses the characteristics of the FPI found in literature. 2.4 describes the relationship between the FPI characteristics and SCC and the conceptual model of this research.
2.1.Supply Chain Complexity (SCC)
7 2.2.Supply Chain Complexity Drivers
A SCC driver is any property of a SC that increases complexity (Serdarasan, 2013). The drivers determine SCC and are common used in literature. Understanding the SCC drivers is critical when determining strategies to manage SCC (Manuj and Sahin, 2011; Serdarasan et al., 2013). In literature there are several authors who have considered drivers. For instance, De Leeuw et al. (2013) identified eight drivers based on recent literature that increase SCC: Uncertainty, Diversity, Size, Variability, Structure, Speed, Lack of information synchronization and Lack of cooperation. These drivers can be seen as categories that include several smaller drivers. Due to the expectation that the specific FPI characteristics affect SCC, we consider them as separate complexity drivers. The SCC drivers can drive SCC within the SC at several places: Upstream, internal and downstream complexity (Blecker et al., 2005; Bozarth et al., 2009; Childerhouse & Towill, 2004; Isik, 2011; Mason-Jones & Towill, 1998). Internal complexity is complexity that occurs within the manufacturing facility’s products, processes and planning and control systems. Complexity that occur at the supply base of a manufacturing facility is called upstream complexity. Downstream complexity occurs at the demand side of the manufacturing facilities. In this research the SCC drivers and the classification of drivers according to their origin will be used to identify SCC and will be further elaborated in section 2.4.
2.3.Characteristics Food Processing Industry (FPI)
The FPI has been considered generally as a part of the process industry and characterised as being large, capital-intensive, mass producers of bulk products in large batches for low costs (Van Donk et al., 2008). However the scenery in the FPI is changed due to the FPI is experiencing growing logistical demands, growing variety in products and intensive competition. These experiences and the specific FPI characteristics make the SCs within the FPI even more complex (Van Donk et al., 2008) . Van Donk (2001) identified and divided the characteristics of the FPI into three groups:
1. Plant characteristics 2. Product characteristics
3. Production process characteristics
8 coupled with small product variety and high volumes. Characteristics of the food products are the variability of the supply, quality and price of the raw materials due to unstable yield of farmers. For instance, the raw materials can deviate from aimed specification to such an extent that raw materials must be classified differently (Van Dorp, 2002). Within the FPI are volume or weight used different from discrete manufacturing. One of the product characteristics that is always present is perishability of the raw materials, semi-manufactured products, and end products, if products are stored too long they become obsolete. The production process in the food processing industry is characterised by the several recipes that are available and the divergent product structure due to the growing expectations of customers. The processing stages within the FPI are not labour intensive, where plants that produce consumer goods have typically an extensive labour intensive packaging stage. other production process characteristics within the FPI are a variable yield and processing time, at least one process that deals with homogeneous products and that the production rate is mainly determined by the capacity.
9
Plant characteristics Product characteristics Production process
characteristics Expensive-single purpose
capacity coupled with small product variety and high volumes.
Variability of supply, quality and price.
Variable yield and processing time.
Long set-up time (sequence-dependent).
Use of volume and weights. Homogeneous product step.
Perishability. Processing stages not labour
intensive.
Food safety regulations. Divergent product structure.
Labour-intensive packaging line.
Several recipes are available Table 2.1: Overview FPI characteristics
2.4.Relationship FPI Characteristics and SCC
10 suppliers and the variability of supply. For example in the dairy industry, the raw material (milk) is supplied by an enormous number of farmers. The above mentioned similarities between drivers and FPI characteristics confirm that using the FPI characteristics as drivers is a suitable manner to research if the FPI characteristics influence SCC.
As earlier mentioned, SCC drivers can influence SCC at several levels, in this research SCC will be invested at the manufacturing plant level. At the manufacturing plant level, SCC can arise due to the SCC drivers from within the company (Internal complexity) or via the company’s connections with downstream and upstream partners (Bozarth et al., 2009). Knowing where the FPI characteristics drive exactly SCC will help to developing a clear strategy in efforts to manage the SCC and increase performance.
To identify possible relationships between FPI characteristics and SCC a conceptual model is designed. The conceptual model visualizes the expected relationship between the FPI characteristics (SCC drivers) and upstream, internal and downstream complexity. The conceptual model is illustrated in figure 2.1.
11 3. Methodology
This research is based on a multiple case study to get an understanding of the influence of FPI characteristics on SCC. The advantages of a case study are (1) meaningful relevant theory can be generated from the understanding gained through observing actual practice, (2) allows the question how, to be answered with a relatively full understanding of the nature and complexity of the complete phenomenon and (3) lends itself to early, exploratory investigations where the variables are still unknown and the phenomenon not at all understood (Meredith, 1988). In this research the FPI characteristics are identified and the influence on SCC is found based on quantitative and qualitative data .
3.1.Case Selection
12 Plant Raw materials Semi-produced End products Turnover (yearly) Employees Perishability end products Divergent product structure
A Potatoes Starch Starch
derivatives
650.000 tons of starch and derivatives
900 Long High
B Potatoes Flakes Deep frozen
fries
95.000 tons fries, 5.000 tons flakes
110 Long Low
C Grain Flour 180.000 tons
flour 110 (together with D) Middle Middle D Grain Breading products 20.000 tons breading products 110 (together with C) Middle High
E Milk Milk powder Baby milk
powder 130.000 tons of milk powder 900 Long High F Whey Ingredients for baby milk powder 52.000 tons of ingredients 200-250 Long Low
G Milk Milk powder 65 million
litres of milk processed
45 Long Low
H Chicken Bulk chicken Meals and
fresh chicken
360 million chickens
550-600 Short High
I Milk Baby milk
powder
10 million cans of milk powder
450-500 Long High
J Flour Bread 50 million
breads
850-900 Short High
K Ingredients Energy drink 15 million
cans of energy drink - Middle Middle L Eggs Chicken (breeded) 78 million chickens 50 - Low
13 3.2.Measuring SCC Influenced due to FPI Characteristics
Measuring complexity is in recent literature commonly used and different approaches are available. The approach of De Leeuw et al. (2013) measures the SCC based on the complexity level of the SCC drivers. According to De Leeuw et al. (2013) is SCC a multi-dimensional construct. The eight drivers of De Leeuw et al. (2013) are also multi-dimensional in terms of several measurements per driver. In this research are the FPI characteristics used as SCC drivers to measure the complexity and contain only one measurement. Certain FPI characteristics can be measured with hard data that can be easily gathered. However, several FPI characteristics cannot be measured with hard data and are measured by the perception of the interviewees. For these FPI characteristics are statements formulated and the interviewees had to fill in if the FPI characteristic is very low(1), low (2), neutral (3), high (4) or very high (5) compared to other plants within the FPI. The interviewees should not compare their plants with other industries, because the characteristics are specific for the FPI and comparing it with other industries could give only high and very high outcomes of SCC. The FPI characteristics measured with hard data are after collecting data categorized in very low (1), low (2), neutral (3), high (4), very high (5). The measurements are illustrated in table 3.2.
Possible Driver Measurement
Expensive-single purpose capacity coupled with small product variety and high volumes
- Level of complexity
Long set-up time (sequence-dependent) - Length setups
Variability of supply of raw materials - Level of complexity
Variability of quality of raw materials - Level of complexity
Variability of Price of raw materials - Level of complexity
Use of volume and weights - Level of complexity
Perishability - Shelf life products
Food safety regulations - Level of complexity
Variable yield and processing time - Level of complexity
Homogeneous product step - Number of homogeneous production
steps
Processing stages not labour intensive - Number of employees
Divergent product structure - Number of end products
Labour-intensive packaging line - Number of employees
Several recipes are available - Number of recipes
14 3.3.Data Collection
To collect the data, two semi-structured interviews of approximately 60 minutes are conducted at each plant. A research protocol is designed as prompt for the semi-structured interviews and as a checklist to make sure that each topic was covered (Karlsson, 2009). Two persons from different levels in the plant are interviewed; an internal person (e.g. planner or production manager) to gather information on the manufacturing plant level of complexity; and a supply or sales person (e.g. SC manager or sales manager) to gather information on the downstream level and upstream level of complexity. In an open manner is the understanding of the influence of the FPI characteristics on SCC asked. The data of the measurements is collected with a survey.
At each plant, two researchers have conduct the interviews were one has lead the interview, while the other has lead the data collection. According to Karlsson (2009), multiple investigators can enhance the creative potential of the teams and convergence of observations increases confidence in the findings. The semi-structured interviews are recorded and transcripts are made to provide accurate rendition of what has been said.
3.4.Data Analysis
15 3.5.Research Quality
16 4. Results
Analysis of the data from twelve cases revealed the influence of seven FPI characteristics on SCC. Each interviewee is asked if the FPI characteristics drive SCC in their plant. An overview of the results is given in table 4.1. The results show that only seven of the fifteen FPI characteristics found in literature are driving SCC.
Overview results
A B C D E F G H I J K L
Variability of supply X X
Variability of quality X X X X X X
Perishability X X X X X X X
Divergent product structure X X X X X X X
Food safety regulations X X X X X X X X
Several recipes X X X X X
Long setup times X X X X X
Table 4.1: Overview SCC per plant
In order to explain the results, the relations found between FPI characteristics and SCC are explained and presented in the following subsections. The measurements of SCC influenced due to the FPI characteristics are used to indicate the difference between the level of SCC of the plants. Lastly an overview is developed based on the results to illustrate the influence of the FPI characteristics on SCC more clearly. The measurements of plant L are taken out because the interviewee did not filled in the survey correctly.
4.1.Variability of Supply
Variability of supply is indicated at plants A and B as a SCC driver. Plant A is a cooperation where the suppliers own the plant and is dependent of these suppliers. In terms of supply this means that the amount that is produced by the supplier determines the amount of supply a plant receives that year of those suppliers. The plant should process all the received supply due to dependability of each other.
17 cause internal complexity due to low yield. If a new season starts and the inventory is too high, storage problems occur. Plant A indicate that their variability of supply is high (4) compared to other FPI plants (table 4.2).
A B C D E F G H I J K L
X X
4 2 3 3 4 4 3 2 5 2 3 -
Table 4.2: Results variability of supply
Seasonality of agricultural production is the cause of driving internal complexity due to variability of supply at plant B. The agricultural supply that the plant processes is harvested in a short time frame and is stored until the plant needs supply. Due to storing is the quality of supply decreasing and causes variability of the quality of raw materials. The variability of the quality of raw materials is further explained in section 4.2. Plant B indicates that the variability of supply is low (2) compared to other FPI plants. This is rather remarkable due to fact that plant B clearly indicates in the interview that the high variability of the supply increases quality problems. It seems to be that the interviewee did not see seasonality of agricultural production as variability of supply over a certain time.
18 4.2.Variability of Quality of the Raw Materials
Plants B, C, D, H, I and L indicate that the variability of quality of the raw materials is related to SCC. The causes of the SCC at each plant show similarities and differences. Each plant indicate that the quality of the raw materials is dependent of the supplier itself. For instance, a good cultivation and harvesting method will help to increase the quality of supply and decrease variability of the quality to gain homogeneous raw materials.
Plants B,C and D indicate that the variability of the quality of the raw materials is also highly influenced by the weather. The same as with the variability of supply causes the weather the variability. Each time if raw materials are delivered it is uncertain for the plant how the quality exactly is. The weather does not only fluctuate by time but also by location. Not each location where supplies are cultivated will have the same weather circumstances. These three FPI plants process agricultural raw materials, were company H, I and L process raw materials from the cattle breeding. This result is similar as with the variability of supply, were agricultural suppliers are weather dependent and the cattle breeding are not. Plant B indicated the best situation for plants with agricultural raw materials "The best situation is a high quality of cultivation and a good growing season which causes less or the least issues".
Another finding is that Plants B,C and D manage the complexity by blending different qualities with each other to obtain homogeneous raw materials. As earlier mentioned are homogeneous production steps a characteristic of the FPI, but is not driving complexity. In this case is it a method to manage the variability of quality of the raw materials. All companies use intensive collaboration with their suppliers to minimize the variability of the quality causes by the suppliers themselves. Knowing what is coming is here the key issue to manage the complexity.
19 Variability of quality of raw materials
A B C D E F G H I J K L
X X X X X X
1 4 4 4 3 3 3 2 4 3 4 -
Table 4.3: Results variability of quality of the raw materials 4.3.Perishability and Divergent Product Structure
Plants A, D, E, H, I, J and K mention that the perishability of their end products cause complexity in the SC. Perishability of raw materials and semi-produced products of the plants are never mentioned as drivers of SCC. Plants H and J clearly face complexity due to the short shelf life of the end products (table 4.4). A shelf life of maximum two weeks means that the forecast has to be as accurate as possible, each overproduction will obsolete and will cost money. The Plants A,D,E,I and K have products with a shelf life between the six months and three years. This indicates a much longer shelf life then the products of the other two plants. There are two main reasons of the complexity that occur at these companies.
Perishability A B C D E F G H I J K L X X X X X X X 1-2 years 2-3 years 6-12 months 6-12 months 2-3 years 2-3 years 2-3 years 1-2 week 2-3 years 3 days, 1-3 months 1,5-2 years - Table 4.4: Results perishability
The first reason is that a customer does not except products that are already 1/3 over their shelf life. Plants D,E,I and K indicate this as the reason of driving complexity due to perishability. As mentioned by one of the interviewees “what is the first thing what a consumer does? Looking at the bottom of the can, when is it made?”. Consumers do not allow old end products and they literally shorten the shelf life of these end products.
20 divergent product structure make it difficult to determine how much should be in stock for each product.
Divergent product structure
A B C D E F G H I J K L
X X X X X X X
1400 30 200 600 300 15 10 1000 300 - 100 1
Table 4.5:Results divergent product structure
These findings indicates that there is a relation between the perishability of end products and the level of diversity of the product structure. The plants who did not mentioned perishabilty as a complexity driver give an even more clear explanation of the complexity that occur. The plants B,F and G indicate that they are not facing complexity due to perishability. This is in line with the long shelf life of the end products, 2-3 years, and the small amount of end products, 10-15 products. Figure 4.1 indicates the relation between perishability of end products and a divergent product structure. 9 of 12 plants are plotted in this graph. Plant J and L are missing in the graph due to no exact number of end products of plant J and no perishability of end products due to living animals at plant L. The graph is designed based on the measurements of perishabilty (table 4.4) and divergent product structure (table 4.5).
21 Plant C is slightly different from the other plants, with 200 end products and a shelf life between 6-12 months it is expected that perishability can cause complexity. The plant indicates that perishability is completely no problem due to the high turnover rate of the end products. Products that are produced today or tomorrow are also send off today or tomorrow. This is also mentioned at plant K " That sounds pretty much and for the ***** as our main product it is not that critical anymore due to high turnover rates''.
4.4.Food Safety Regulations
The food safety regulations characteristic is the most mentioned characteristic that drives complexity in the SC of the FPI. The plants A,C,D,E,F,I and J indicate that the national food safety regulations cause complexity in their SC. The regulations are becoming stricter and stricter and the plants are facing internal complexity and upstream complexity. The plants have to adjust their working procedures and re-evaluate their suppliers constantly. Plants F and I even indicate that due to the food safety regulations setup times are getting longer due to the need of carefully cleaning. As mentioned by plant I " Well, the setup time is quite long because there is a high hygiene standard for the food processing industry".
Plants A,E and F mention that the food safety requirements of the customers themselves are driving complexity in the SC. Several customers want their own working procedures to produce their products. This drives internal complexity due to the different working procedures and requirements that are needed. The plants are audited by the customers self, and the regulations are becoming stricter and stricter.
The last finding is the relation between food safety regulations and divergent product structure. Plants E and K are facing complexity due to the difference between the food safety regulations of the countries were their customers are located. Each country has his own food safety regulation in terms of what is allowed to put on the package and which ingredients are allowed. Each country is unique and increases the amount of end products.
22 Food safety regulations
A B C D E F G H I J K L
X X X X X X X X
4 4 4 4 5 4 2 4 4 5 5 -
Table 4.6: Results food safety regulations 4.4.Several Recipes and Setup Times
The plants A, D, E, I and J indicate that they are facing complexity due to the amount of recipes they process. The quantitative data gathered from the plants concerning the amount of recipes and the setup times is not complete , however it does show some patterns.
Plants A, D, E and J are able to indicate the number of different recipes they process (table 4.7). These plants have between the 15 and 200 different recipes, and are facing complexity due to an increase of setups. Their production process is sequence dependent and makes it even more complex to plan their production (internal complexity). Other plants who indicate between the 1-2 recipes are not facing any SCC due to the amount of recipes. Plant I was not able to indicate the amount of different recipes, however they clearly mentioned that that the several recipes are the cause of driving SCC due to the amount of setups. With an optimized setup and/or cleaning framework the plants minimize the amount of setups and the setup times as possible. The amount of setups that will increase due to an increase of recipes is driving complexity internally in the plants. Additionally, it means more recipes more end products and increases the complexity due to a divergent product structure.
Recipes
A B C D E F G H I J K L
X X X X X
200 2 160 200 10-15 2 1 1 - 120 - 1
Table 4.7: Results recipes 4.5.Overview Results
23 indicate the indirect influence of FPI characteristics on SCC. The red line indicate a interrelationship between FPI characteristics, were they together drive SCC even stronger.
24 5. Discussion
In this chapter are the results in relation with the research question "How do specific food processing industry characteristics influence supply chain complexity?" discussed and compared with the expectations from literature. In section 5.1 are the findings discussed in section 5.2 the overview and measurements.
5.1.Discussion Findings
The influence of variability of supply on internal complexity is caused by the seasonality of agricultural production. In a short time frame are the raw materials of the plants harvested and brought to the plant or stored. The seasonality of agricultural production forces the plants to produce the supply as soon as possible, and storing the supply will decrease the quality of supply. These findings are in line with the expectations based on literature. Salin (1998) indicate that seasonality of agricultural can affect SC approaches. The downstream complexity influenced by the variability of supply is caused by the weather circumstances. The uncertainty of the weather causes uncertainty of sales management, due to not knowing the amount of products to sell. Salin (1998) mention that the dependence on climate conditions with agricultural production make the SC within the FPI more complex than other SC and is confirmed due to these findings. A finding what not was expected is the need of processing the supply. A plant who is part of a cooperation is forced to process all the raw materials that is harvested, if they need it or not. The variability of the amount of supply will only occur if a plant is part of a cooperation, plants who are not part of a cooperation will simply purchase the amount of supply they need. However, only plants who are part of a cooperation active in the agriculture are facing these problems. This is strongly confirmed by the plants B,C and D were they process raw materials that are also influenced by the weather but are not facing SCC. The relation between being a cooperation and processing raw materials were the supply is influenced by the weather does make it complex for plants. These plants have long contracts with suppliers for a certain amount of supply, if they do not have enough they can buy supply on the free market or the opposite. These plants can ensure each year the amount of supply they need, not more not less.
25 communicate with their suppliers (upstream complexity) and constantly adjust their production process (internal complexity) to gain a homogeneous product. The variability of the quality of the raw materials is also depended on the suppliers themselves. Each supplier has his own cultivation and harvesting technique and the differences between them will cause variability of the quality of raw materials.
Perishability within the FPI is in earlier research already indicated as a complexity driver (Shukla et al., 2013; Van Donk et al., 2008). However, the perishability of raw materials and semi-produced products is driving less SCC than expected. It is the perishability of the end products that drives SCC. A short shelf life of end products will increase the SCC in plants within the FPI. The place where the complexity occur and the interrelation with a divergent product structure is also surprising. Plants are not facing complexity internal in their SC, but at the downstream side of the SC. A short shelf life makes it downstream complex due to the fact that every overproduction should be thrown away and the trade-off should be made between throwing products away or not meeting customer demands. Forecasting these kind of products is extremely complex. The complexity will increase if there is a divergent product structure, more end products means complex forecasting management. The strong interrelation between perishability and a divergent product structure is not earlier mentioned in literature. All the plants try to have an accurate forecasting method as possible to manage the complexity caused by the perishability of the end products and divergent product structure of the plants. An easy way should be to decrease the amount of products, however the plants indicate that the divergent product structure is necessary to be competitive and earn more money. The influence of a divergent product structure itself is expected, Van Donk (2001) indicate that the divergent product structure increases the amount of set-ups.
26 requirements of customers is the internal complexity even making more complex due to different working procedures are necessary. This is in line with the remark of Van der Vorst and Beulens (2002) that consumers concerns are increasing for food safety. The last finding concern food safety regulations indicate a relationship of between the amount of countries a plant is active and the divergent product structure. The difference between the food safety regulations increases the amount of end products due to different label regulations of end products. This is a remarkable finding that the food safety regulations influence the level of diversity.
The last finding in this research is the relationship between several recipes and set-ups. More recipes means more set-ups, this is in line with a divergent product structure. It is rather logic that if the amount of recipes increase also the amount of end products increase. However set-ups at the packaging stage of a plant are much shorter and not sequence depended then at the processing stage. More recipes means a more complex production and planning management due to cleaning and set-up times. These complex set-ups are driven by the sequence dependent production and the national food safety regulations. The amount of recipes should be considered separately with the amount of end products to understand the effect of recipes on internal complexity.
5.2.Discussion Overview and Measurements
The overview illustrated in figure 4.2 should be used in combination with the explanation of the FPI characteristics and gives an understanding of the FPI characteristics as SCC drivers. Literature identifies that FPI characteristics make it more complex in the SC, however they do not explain how. The overview is an effective first step to proceed to develop a clear strategy to manage SCC. Only seven of the fifteen FPI characteristics found in literature are identified in this research as driving SCC. It seems to be that the other eight are characteristics that do not fluctuate. These characteristics are present or not and if they are present the plants just have to deal with it. How and where the FPI characteristics influence SCC is not for each plant the same. There is a high diversity between the plants in terms of influence of FPI characteristics on SCC.
28 6. Conclusion
This research is the first step to develop a framework for managers to manage SCC within the FPI. This first step gives an answer on the research question ''How do specific food processing industry characteristics influence supply chain complexity?". However additional research is needed to cover all the influences of the FPI characteristics on SCC.
The seven FPI characteristics found in this research influence SCC in different ways and are not for each plant the same. The characteristics influence not only SCC itself but also other FPI characteristics. There are even FPI characteristics that are interrelated and together increase SCC even more. This research give a clear answer and overview of how these FPI influence SCC what in literature not is mentioned. The FPI characteristics influence SCC upstream, internal and downstream and in terms of complex supply management, production and planning and sales management.
29 Another limitation of this research is the perception of the interviewees. The interviewees are asked if the FPI characteristics are driving complexity in their plant and how this influence SCC. Interviewees are asked how they experience SCC in their plant, and indicate complexity often in only in their field of knowledge and do not cover the complete plant. Furthermore interviewees experience complexity differently. For instance, plant G indicate that they are not facing any kind of SCC due to FPI characteristics. However this plant has to deal with high national food safety regulations and customer food requirements. Interviewees will experience the FPI characteristics on their own way.
30 References
Ahumada, O., & Villalobos, R. J. 2009. Application of planning models in the agri-food supply chain: A review. European Journal of Operational Research. 196 (1): 1-20. Blecker, T., Kersten, W., & Meyer, C. M. 2005. Development of an approach for analyzing
supply chain complexity. In Mass Customization: Concepts–Tools–Realization. Proceedings of the International Mass Customization Meeting, pp. 47-59.
Bozarth, C.C., Flynn, B.B., Flynn, E.J., Warsing, D.P. 2009. The impact of supply chain complexity on manufacturing plant performance. Journal of Operations Management, 27 (1): 78-93.
Childerhouse, P., &Towill, D. R. 2004. Reducing uncertainty in European supply chains. Journal of Manufacturing Technology Management, 15 (7), 585-598.
Closs, David J., et al. 2008. "Toward a theory of competencies for the management of product complexity: six case studies." Journal of Operations Management 26 (5): 590-610. Davis, T. 1993. Effective supply chain management. Sloan management review, 34 (1):
35-35.
De Leeuw, S., Grotenhuis, R., & Van Goor, R. 2013. Assessing complexity of supply chains: evidence from wholesalers. International Journal of Operations & Production Management, 33 (8): 960-980.
Frizelle, G.,& Woodcock, E. 1995. Measuring complexity as an aid to developing operational strategy. International Journal of Operations and Production Management, 15 (5): 26–39.
Ganeshan, R., & Harrison, T. P. 1995. An introduction to supply chain management. Department of Management Science and Information Systems, Penn State University. Isik, F. 2010. "An entropy-based approach for measuring complexity in supply chains."
31 Manuj, I., Sahin, F. 2011 "A model of supply chain and supply chain decision-making complexity." International Journal of Physical Distribution & Logistics Management 41 (5): 511-549.
Mason-Jones, R., & Towill, D. R. 1998. Shrinking the supply chain uncertainty circle. IOM control, 24 (7): 17-22.
Meredith, J. 1998. Building operations management theory through case and field research. Journal of operations management, 16 (4), 441-454.
Miragliotta, G., Perona, M. 2004. Complexity management and supply chain performance assessment. A field study and a conceptual framework. International Journal of Production Economics, 90 (1): 103-115.
Rong, A., Akkerman, R., & Grunow, M. 2011. "An optimization approach for managing fresh food quality throughout the supply chain." International Journal of Production Economics 131 (1): 421-429.
Salin, V. 1998. Information technology in agri-food supply chains. The International Food and Agribusiness Management Review, 1 (3): 329-334.
Serdarasan, S. 2013. A review of supply chain complexity drivers. Computers and Industrial Engineering, 66 (3): 533–540.
Sivadasan, S., et al. 2002. An information-theoretic methodology for measuring the operational complexity of supplier–customer systems. International Journal of Operations and Production Management, 22 (1): 80-102.
Sivadasan, S., et al. 2006. Advances on measuring the operational complexity of supplier– customer systems. European Journal of Operational Research, 171 (1): 208–226. Shukla, M., & Jharkharia, S. 2013. Agri‐fresh produce supply chain management: a
32
Vachon, S., & Klassen, R. 2002. An Exploratory Investigation of the Effects of Supply Chain Complexity on Delivery Performance. IEEE Transactions on Engineering Management, 49 (3): 218-230.
Van Donk, D.P. 2001. "Make to stock or make to order: The decoupling point in the food processing industries." International Journal of Production Economics, 69 (3): 297-306.
van Donk, D.P., Akkerman, R., & Van der Vaart, T. 2008. "Opportunities and realities of supply chain integration: the case of food manufacturers." British Food Journal 110 (2): 218-235.
Van Dorp, K. 2002. "Tracking and tracing: a structure for development and contemporary practices." Logistics Information Management 15 (1): 24-33.
33 Appendix A: Coding schemes
Table A.1: Coding scheme variability of supply
Table A.2: Coding scheme variability of quality of raw materials
Cooperation (need ) weather dependent quality loss upstream complexity Internal complexity Downstream complexity
A
If the weather is good for about 3 weeks, then our harvest-forecast will increase, which means that we have to sell more. What happens, looking back at the 2 different processes, the first process is a campaign-process (August-February). The second process, derivatives, is throughout the year. So sales is throughout the year. The first half a year the plant produces type 1, then you produce more than demand. Our capacity at the plant is fully utilized then and at the end of February we have got starch (stocks) for the rest of the year. That needs to be planned well. So what happens is that we sell during a campaign but in the end we have to store. We have a limited storage capacity. All silos located in Groningen and Drenthe and we store around 20.000 Tons starch. But, if the inventory level of starch in August is too high and a new campaign period starts, you cannot
store anymore.
x x x x
B
and there is a lot of distribution/spread/dispersion in the quality, length, and defects in The Netherlands.We are
constantly managing that, getting the process right, which makes it complex.
x x x C D E F G H I J K L Categories Variability of supply
Since the potatoes are already planted, are in the ground, and if there will be sales forecast or not, those potatoes will be distributed to our plants and will be available for processing. So, if the sun is shining and it will rain, we will get a lot of
potatoes and thus starch. If it doesn’t rain and there won’t be any sunshine, at the end we have
got very little starch. (cooperation)
Our S&OP-proces is essential to cope with the uncertainty. There is a
variation in sales, we operate worldwide, and our supply is also uncertain, since it goes the opposite way. If the harvest is well, the sales drops. Then, we need to push the sales. “You will ensure that more will
be sold”. Operations will force sales to sell more.
However, the growth of potatoes is very dependent on climate,
Still you can’t get it completely reliable since storing potatoes is not beneficial for the quality what causes an increase in defects. Storage disorders like thick potato skins for example also affect the quality of the
input.
weather dependentsupplier dependentupstream complexityinternal complexitydownstream complexityagriculure cattle breeding
A
B
look, with potatoes it is about the fact that it is a natural product. This makes it complex, as not one year is the same. You plant a potato plant, that is happening at this moment, in April, the planting month, then it all depends on the growing season. How is the growing season, does it stay cold for a long period of time? Is there enough water? Can they
be flooded by rain? Is it cold during the summer of rather tropical? All these factors influence the final product. And the final is harvested from August until October, and that indicates the result for the rest of
the season.
The best situation is a high quality of cultivation, a good growing season which causes less or the least issues. However, the growth of potatoes is very dependent on climate, and
there is a lot of distribution/spread/dispersion in the
quality, length, and defects in The Netherlands.
We are constantly managing that, getting the process right, which makes it
complex.
So, yes, with a potato anything can happen, we require a potato that is long and big, but the size can either disappoint or be better than expected, they product can also be contaminated and decay, all kinds of ailments and diseases, and in the end we have to create a homogeneous product from a natural product that differs on a yearly basis. Because the final product has to meet various specifications. And it is therefore that we produce three different quality types of products, standard, LF and PR and we also use different breeds of
potatoes, but in the end we have to make a product that meets all the specification. This is
one of the challenges within our factory.
x x x x x
C Then we have uncertainty of the harvest, last year we got a harvest with a very fluctuating quality.
Yes. Look, when you are dealing with raw materials, with a natural product, you have to deal with climate, where does it grow, what is
the quality and how the weather is like. You have to deal with various
quality aspects
That means that when a certain seller has sold a batch of grain with
a certain percentage of protein, than he will sort that from his big pile and store this to be shipped
during the season.
That gives us quite some troubles this year to, in particular, keep the baking specifications right. Next year it could be better, that is also a part of
uncertainty.
x x x x
D
We have a … the harvest is always quality depended, harvests always differ from one another. You discuss a certain quality with your suppliers and you try to focus on that to your best abilities in order to deliver
that.
That means that when a certain seller has sold a batch of grain with
a certain percentage of protein, than he will sort that from his big pile and store this to be shipped
during the season.
Quality, if we look at Lano Crumbs, of which I now more, we are talking about granulation, which means the size of the grain flocks, which various in millimetres. So this ranges from 0-0.5mm, 0.25-0.5, 0.3-0.6, and that keeps on going.
x x x x
E F G
H
It is also important for poultry farmers to visit our plant and to talk about issues we experience and how
they can influence and help us solving those issues. For example, a time ago, a few poultry farmers visited us and I guided them through our location. An
issue at this site is the weight of the chickens, reported by the poultry farmers, a couple of days before they enter our plant. What their weight is at this moment and how much they will grow. Because of that, our planner can plan when to slaughter those chickens in order to get the weight we want, to reach the ideal weight for to meet specifications.
We know how many chickens there are and how many there will be slaughtered. However, a chicken is a
living being and grows on its own manner, it could be fat,
So I explained to those poultry farmers the importance of communicating the right weight. If their weight don’t meet our requirements, we cannot pack them etc. So the supplier are unsaleable in case the poultry farmers don’t take effort to inform us in time, really measures the chickens, and to give a
reliable result.
which means that we have to deal with a
heterogeneous product at the slaughterhouse x x x x
I
Yes. At the beginning of the business, there are less suppliers than now. In addition, the size and capability of the suppliers are also relatively low at the beginning. With the development of the business, there are more requirements on the products such as the time to delivery and quality. Therefore, the technique level of the suppliers will influence the product. For example, the quality of the products may be assessed as unqualified. This will further
influence the production.
Yes, of course. As we talked before, if the quality of a batch of products is below standard, the
whole production schedule will be disorganized. x
J K
L
Yes. And also because there is a difference in batches between companies. In one chicken farm it might do perfectly well and in the other it might not work at all. And they might both be very good, so you tell me
why it happens. But it happens …
Well for us, as we are responsible for the first 5 days, if there is a loss over 1% then we have to reimburse our client. If this shows, they have to show medicine, so the chickens are no longer ambiotica-free, and therefore they get less money for their chickens. So if something happens in the chain it goes on to next stadia. So it is of great interest that we
deliver that …
x x x
Category
34 Table A.3: Coding scheme Perishability and Divergent product structure
Short shelf lifePerishability due to dpsMindsetUpstream complexityinternal complexityDownstream complexity A
In general we produce on stock, M ake-To-Stock. We don’t want our products to get too old, about hitting date. At Avebe, the products which are hitting date, are the derivatives.That is where the products became unique. We don’t have any native starch that
is hitting date, we can always sell that.
So, the more the products are unique/specific, the higher the risk on obsolescence.
So, we have a lot of different products, and we have to deal with setup times
if we are talking about size, the smaller the volumes, the more complex it gets and the bigger the change of obsolescence. You need to clearly agree upon with sales which specific agreements to make with the customer.
x x x
B
C
There is never a product left on site. One part is bags, which has an enormously fast turnover, so also no
perishability there.
D
Well, you are stuck with, let’s take THT as an example. On every package you can find a THT, maximum guarantee. 1/3 or the lifespan is for production, the other 2/3 is for the customer. So that is a risk, you cannot put anything on stock. That is a risk. Then you have customers saying they don’t want the product anymore. So that is something you have to
keep in mind.
But also here you have other complexity, then you have to think about; how much can I put
in stock? I expected it be way more, we have a rather short inventory position, the factories run seven days a week.
Because at some point the customer will say … and supermarkets are way stricter, that they don’t want the product. And then you have to sell against bottom
prices.
x x x
E
Yes, we should not make too much, the advantage of baby food is the long shelf life. It has a shelf life of 2-3 years but, the problem is, what is the first think that a consumer does? Looking at the bottom of the can, when is it made?. They don’t want a product of a year
old. So there is uncertainty
Diversity does also has to do with that. Diversity, what I already said, we have cans that for every country is different in terms of
looks.
Yes, a lot of set-ups.
What we are trying, if you speck about our product portfolio, to have a more central management. That of us is in Singapore, it had also being placed in the Netherland or with
country you want. So if we talk about our top brand Friso, cannot every customer Friso has
to look this way and a other customer want it that way. So
x x x x
F G
H
It depends on the type of meat, on average around 9-10 days. From the moment it is packed which makes sense since if lead time is 3 days, the shelve life for the end customer would be 3 days less. Another challenge is the order. If orders are fixed, you know what to produce the day after tomorrow. Because of that, you can produce on the last possible moment and there is more time to organize the production in that way. However, the retailer also don’t know what he/she is going to sell the day after tomorrow. So if he needs to order 3 days in advance, you know for sure he/she
orders the wrong.
Well, it is the combination of perishability, a short delivery lead time and a divergent
product structure
That is a really big issue. We produce on forecast, so it means
that we think we will sell 100 products based on sales history which will be between 80 till 120. We also take factors as seasonality etc. into account. It will be an issue if we only sell
60-40.
It relies on systems. Forecasting is very difficult, since the sales patterns are fluctuating. It’s hard to predict. We also have to take the dynamics of our assortment into account, sales. We are continuously contacting sales about product development, changes in packaging etc., all we need to react on in about half a year/year. We have specific recipes and preprocessed packages, and retailers are changing all the time, so we cannot store too much. Despite the fact we control
that well, every year we have 200.000 Euros unsaleable raw materials and packages, due to a change in assortment. It’s a huge lost, if we didn’t put
that much effort to it, it would be much more.
x x x
I
Yes. This factor do has an influence. The wholesalers will not accept the products if the products already over quarter of their shelf life. However, we still need the safety stock. Therefore, we always replace the old safety stocks with the new one to keep the safety stock available. Although this still cause some wastes, it is
better than out-of-stock.
Yes. This factor do has an influence. The wholesalers will
not accept the products if the products already over quarter of their shelf life. However, we still need the safety stock. Therefore, we always replace the old safety stocks with the new one to keep the safety stock available.
Although this still cause some wastes, it is better than
out-of-stock.
x x x x
J
Well, two factors may cause this uncertainty. The first one is because the weather. The moist weather may reduce the shelf life of the products. Sometimes, we did not notice that some products are already influenced by the moist weather and still keep these
products as inventories.
The diversity is quite large in our supply chain. First, as I said before, we provide several types of foods. Each type contains many kinds of different product. Therefore, we have hundreds
of final products. Although the main production process is similar, there are still
some divergences when producing these products. This causes a large number of suppliers and activities. In addition, because our products need to be consumed in a short time, the demand of raw materials is determined by the customer demand.
In addition, the facilities needed to produce these products are used in different way. The set up time of these facilities are also distinct. All these factors may increase the complexity in the
supply chain.
x x x x
K
So high uncertainty is connected to the question whether we produced too much and face problems
with the perishability of our final product. For example, considering 28 Black we do have a shelf-life of two years. We guarantee our customers a minimum shelf-life of one year. That sounds pretty much and for the 28 Black Acai as our main product it is not that critical anymore due to high turnover rates, but for newly introduced products the risk is much higher. Our production is based on assumptions which we try to have as accurate as possible. Of course that goal is
not easy to reach.
On the one hand, we have to produce a minimum amount, at least 150.000 cans per production run, at one time and need to sell these stocks within a certain time (for example: pink grapefruit has a 15 months
shelf-life with guaranteed minimum six months for the customer), on the other hand, if we do not produce enough we cannot react fast enough due to shortages in stock and lose our flexibility. I cannot spontaneously produce more products as I have to deal with certain lead times (up to 4,6 or 8 weeks). At that moment I am not able to act anymore, that is
probably one of our biggest fears.
x
L
35 Table A.4: Coding scheme Food safety regulations
National food safety regulationsCustomer food regulationsDifference countries Upstream complexityInternal complexityDownstream complexityIncrease dvp
A
Rules and regulations, very important for us.In addition, for example Unilever,
demands/requires us to manage our supplier base well. In the future we also
need to add some sort of CO2 footprint/energy footprint. That’s something we need to translate upstream, also to the farmers.The CO2 footprint of the potato fields need to be reported as well, as a kind of license to operate. If we
don’t do that, Unilever moves to our competitor.
All suppliers need to be rated well, need to be periodically audited and
documented. x x x
B
C
But this week we were talking about, BSC6 will expire in a while, I don’t know if you know BSC but it is also a safety protocol, and it will be replaced with BSC7 which is already stricter than its predecessor. Every time it becomes a little stricter,
and eventually you will also notice this as the purchasing department. Because also you suppliers have to comply to stricter rules and
regulations.
x x
D
in the last couple of years, has to become stricter is that also a problem for your organization? You have to deal with … we
have an audit recently. They will look at whether or not we control if particles enter the product that shouldn’t be there. You can think about pieces of metal or rock, I mean you retrieve a product directly from land, then stuff can enter your product.
All processes have been designed in such a way that these particles are taken out, because of course you don’t want to be to blame for it ending up with a customer. So there are very strict rules in regard to food
safety.
x x
E
Yes, that’s also one of the issues, the label claims, what do you put on your label.
That’s very important for baby food, because how can you seduce mothers to buy our product Friso and not one of Danone or Mead Johnson. You can do this by claiming things in commercials, In some countries you are not allowed to make commercials for baby food for the youngest group, in other countries is more allowed. Often you have support this with publications. So in this respect is each country unique. So those
things we can’t change.
Is food Safety Regulations becoming more and more strict, and makes this it more and more complex? Yes, correct, you also see that, not only the Dutch Food safety but other authorities are getting stricter. Sometimes also, for
instance Russia
So those things we can’t change, but you can look into your product portfolio to get the same cans, cover. We just changed our simple cover for are better looking cover. So we are trying to keep that diversity moderate
But you can see that the requirements of our customers are getting higher and higher, so we
are not easy for our suppliers.
x x x x x
F
All procedures that we have here are coming from audits from customers. Constantly that Danone and Nestle you
name it, Chinese authorities Russian authorities doing audits here. Looking if we
have our process correctly. We show them our logbooks, track and tracing, looking at the factory if everything is correct. We have to be prepared to receive at any time
customers.
Yes, continuously they say, look for what we make ingredients, for baby food, In fact most critical product what is available. For instance as a baby makes a misstep then we have a big problem. Then you are trending topic in the newspapers,
when a product of FrieslandCampina is in there.
Then you have a really big problem.
That’s also because of the control needs, if the production start on day 1
then it is ready to be delivered to the customers on 21st day. Then you
already 3 weeks further, from entering the raw materials here until it is packaged and ready for transport to the customer. That’s because what I already because of the controls on
bacterial.
x x x
G
H
I
Well, the setup time is quite long because
there is a high hygiene standard for the food processing industry.
This has a very huge influence. In China, there are lots of organization take the charge of food safety. When a safety problem
occurs in other companies, these organizations will increase some standards accidentally.
These changes of the safety standards impact the whole supply chain dramatically.In addition, there are
also some requirements on the packages and these requirements
change each year. T
x x
J
In China, the food safety regulations are very strict. For example, there is a food organization production permission (QS), which contains many rules in terms of the quality of products. If you want start business in the food processing industry in
China, the permission is required. In addition, because of some safety problems,
the safety regulations are updated frequently.
Ofcourse.If the standards of an element improved, we need to reevaluate the suppliers and shift in a
short time to react to the new regulations. This requires a lot of
efforts.
x x
K
But of course there are different food safety regulations in every country. For example, certain ingredients are not legalised in some
countries and other not.
Important to us is that we were able to develop a specific a basic material which only needs to be completed by small additives. To illustrate this, i.e. palette A and B require the add-on for Germany, while palette C requires the
add-on for Australia.
x x x
L
Categories
