BATCH SIZE DETERMINATION IN HYBRID FOOD
PROCESSING INDUSTRIES
Master Thesis, MSc Supply Chain Management University of Groningen, Faculty of Economics and Business
January 29, 2018
Martijn van Boven S2786680
m.a.van.boven.1@student.rug.nl
Supervisor University of Groningen: prof. dr. D.P. van Donk
Co-assessor University of Groningen: dr. O.A. Kilic
ABSTRACT:
Purpose – The combination of different production strategies in one production facility is more and more common and even understudied. This so called hybrid production is even more interesting in the food processing industry, because of the specific characteristics. This study aims to get a better understanding of how batch sizes are determined in such an environment. Design – A multiple case study is conducted in six different companies by semi-structured interviews and by collecting additional data regarding planning and scheduling issues.
Findings – We found that the perishability is the most dominant factors in determining batch sizes. Next, the expensive capacity and mixing processes increases batches. The batch size in less formalized companies is determined by the size of the production entity. The quality of raw materials increases or decreases the batch size. The difference in yield in extracting processes and processing time in mixing processes decreases or increases batch sizes.
Key words: batch size, hybrid production strategy, food processing industry
LIST OF CONTENT
1. INTRODUCTION 4
2. RESEARCH FRAMEWORK 7
2.1 Food processing industry (FPI) 7
2.2 Hybrid production system 8
2.3 Link FPI characteristics to batch sizes in hybrid production systems 9
2.4 Conceptual framework 12
3. METHODOLOGY 14
3.1 Research method 14
3.2 Case selection & description 15
3.3 Data collection 17 3.4 Data analysis 18 4. FINDINGS 21 4.1 Production strategy 21 4.2 Expected findings 22 4.3 Planning method 23 4.4 Production capacity 24
4.5 Materials before production 25
4.6 Materials during production 27
4.7 Workforce 28
4.8 Differences production strategies 29
5. DISCUSSION 30
5.1 Link results to literature 30
5.2 Limitations & Further Research 33
6. CONCLUSIONS 34
6.1 Conclusions 34
6.2 Theoretical implications 35
6.3 Managerial implications 35
7. APPENDICIES 43
Appendix A: Interview protocol 43
Appendix B: Examples coding 44
Appendix C: Quotes production strategy 46
Appendix D: Quotes Food Processing Industry 48
Appendix E: Quotes size companies 50
Appendix F: Quotes variation raw materials before production 51 Appendix G: Variation in (raw) materials during production 52
1. INTRODUCTION
In this research, we investigate how the batch size is determined in the hybrid food processing industry. The hybrid production strategy exists of the make-to-order (MTO) strategy, and the make-to-stock (MTS) strategy in one production facility (Beemsterboer et al., 2017). MTS and MTO-strategies are two extremes of production strategies and so have other Key Performance Indicators (KPI’s) (Almehdawe & Jewkes, 2013; Nagib, Adnan, Ismail, Halim, and Khusaini, 2016). The MTS KPI’s focuses for more on the product for which the goal is to maximize the fill rate (demand satisfied from stock on hand). The MTO KPI’s focuses on the process where the order lead time should be limited. In a fully MTS system, everything is aligned smoothly to have the most efficient system, and thus batch sizes are determined properly (Pang, Shen, & Cheng, 2014). On the other hand, in a fully MTO system, batch sizes are determined by the customer and so it is not an issue: just produce what is asked. However, in a hybrid system, it is not possible to have a strategy that operates as it does in a fully MTO or MTS system, since two different production strategies are present in the hybrid production system. So, MTO-items and MTS-items share the same resources of the facility, but do not share the same strategy. We yet do not know how batch sizes are determined in such a hybrid production system.
We know from literature that hybrid production systems exists in all different kinds of businesses (Linn and Zhang 1999; Wu, Jiang, Chang, 2008; Fernandes et al., 2015; Kanda, Takahashi, & Morikawa, 2015). Moreover, studies have shown that hybrid production systems are more and more common due to market developments over the last years (Van Donk, 2001) Romsdal, 2014; Claassen, Gerdessen, Hendrix, & van der Vorst, 2016; Kim & Park, 2016). So, companies nowadays and in the future have to deal with hybrid production strategies in some way. The mix of MTO and MTS products and the interaction with the shared resources can open possibilities but also problems in production planning. The food processing industry has to deal with specific characteristics (seasonality, variability of raw materials, long throughput times of raw materials, and perishability of raw materials, intermediates and finished goods). These characteristics make the batch size determination puzzling and more complex than other businesses (Claassen, Gerdessen, Hendrix, & van der Vorst, 2016). Existing theories of for example “postponement strategies” are even not applicable to the food processing industry due to the perishability issues.
does not consider the real-life complexities. They most often stress mathematical models over empirical insights and practical applicability. The deterministic nature of these models are often based on assumptions far from reality (e.g. considering randomness and variability) (Rabta & Reiner, 2012) and so is not a good reflection of how companies actually deal with it. Like Suri et al. (1995) argued, it is better to find approximate solutions to models that are close to reality rather than find exact solutions to models that are rough approximations of real life. Besides this available literature is limited in its applicability for real-life situations, it also does not explain fundamental interactions between the two types of strategies that share the same resources, time (due dates), and quantity buffers (inventory).
Soman et al. (2004) found that several decisions are made before batch sizes are determined in hybrid food processing industries. They even found several factors that have an influence on planning and scheduling decisions in food processing industries, and so we are able to derive effects from those factors on batches. However, we do not know how these factors play a (combined) role and to what extent these factors determine the batch size in hybrid food production environments. Soman et al. (2004) therefore stated that the scientific community could benefit from more empirical studies about this subject. Khakdaman et al. (2015) stated that “As empirical evidence in the area of MTS–MTO production planning is scarce, more research is required”.
Besides the scientific interest, it is in interest of companies as well to discover this subject more in depth, since it is proven that determining the batch size properly can have a major impact on the associated costs and performances of the company (Sawaqed & Foote, 1989; Hopp, 2011; Gamberini et al, 2011; Rabta & Reiner, 2012; Beemsterboer, Land, and Teunter, 2016).
By answering the question “How do hybrid food processing companies determine their batch sizes?” this paper aims to get a better understanding of which, how, and to what extent these factors influence the batch size in hybrid food processing industries.
A multiple case study in the hybrid food processing industry will be performed at six different companies. A case study fits this research the best, since we know that some factors play a role, but we do not know how. How-questions can be answered the best by a case study (Karlssen, 2016).
2. RESEARCH FRAMEWORK
In this section, existing literature will be explored about food processing industries and hybrid production strategies. Eventually our framework is derived from the existing literature. After this part it is clear what studies are conducted in this research area and how they can
contribute to our study.
2.1 Food processing industry (FPI)
Soman et al. (2004) showed a typical food processing process in which two different stages can be distinguished: the processing stage and a packaging stage (Figure 2.1).
Figure 2.1. Typical food processing process (Soman et al., 2004).
By reviewing the papers of Nakhla, (1995), Meulenberg and Viaene (1998), Van der Vorst et al. (2001), Van Wezel (2001), Soman et al. (2004), Aramyan et al. (2007), Romsdal (2014), Nagib et al. (2016) and Claassen et al. (2016), the characteristics of the market, plant, product, and production process, regarding hybrid food processing industry (FPI) can be established. These are listed in Table 2.1.
Group Characteristic
Market Customers asks for more quickly delivered and more varied products Retailers and wholesalers expect small deliveries with a short lead time Consumer behavior is more erratic
Plant Expensive capacity with flow shop oriented design because of conventional small product variety and high volumes
Extensive sequence-dependent set-up and cleaning times between different products or product types
Product Limited shelf life for raw materials, semi-finished and finished products Variation in supply, and quality of raw material.
Volume or weight as the unit of measure in the processing part, and when the product is packed the products become discrete ones
Long lead times of raw materials.
Multiple recipes for a product.
Packaging stage is labor intensive where the processing stage is not Production rate is mainly determined by the capacity
Table 2.1. Food processing industry characteristics.
2.2 Hybrid production system
Having both the MTS and MTO strategy in one production facility, is named hybrid production (Beemsterboer et al., 2017). From common sense and previous studies, we know that large batches with less change overs may reduce setup costs and so increase machine utilization, which is more in line with the make-to-stock (MTS) strategy. While small batch sizes may reduce lead times which is more a goal of the make-to-order (MTO) strategy (Almehdawe & Jewkes, 2013; Nagib, Adnan, Ismail, Halim, and Khusaini, 2016).
The reason why companies do this hybrid production is because markets are changing: consumers ask for more product variety, customers want a short lead times, and demand is uncertain (Romsdal, 2014; Claassen, Gerdessen, Hendrix, & van der Vorst, 2016; Kim & Park, 2016). These changes require manufacturers to shift their customer order decoupling point (CODP) more upstream, and so shift a part of their production systems from a MTS to a more MTO one (Soman et al., 2006).
Perona et al. (2009) state that planners tend to base their decisions on experience and common sense, which calls for approaches that are simple and illustrative without giving up decision-making in a rational way. O’Reilly et al. (2015) found that unclear MTO/MTS policies can cause issues in the production plan resulting in ad hoc decision making. Having no well-defined decision rules can cause that the sales department follows customer requests too easily. Therefore, we have to look for a model that could help us in defining batch sizes.
product families are formed, how the segregation between MTO and MTS is, and what the target service levels are. However, we do not know how this model works out in the food processing industry that exists of specific characteristics.
Like said earlier, existing literature does not consider the real-life complexities. Even Soman et al. (2004) assumed for their model that intermediate storage is not possible and that all products have limited shelf life which is a characteristic of the food processing industry. However, there are also food plants that processes products that are for instance frozen at the end and so do not have a short shelf life.
Figure 2.2. Hierarchical approach to MTO-MTS problem (Soman et al., 2004).
2.3 Link FPI characteristics to batch sizes in hybrid production systems
We will now explain the single effects of the food processing characteristics on the batch sizes in the hybrid production strategy. These effects as said are just single effects, and could change when another characteristic is present as well.
1. Market characteristics (often holds for only a part of the products)
In order to serve customers quickly a broad range of products, the products should always be available. This asks for small batches in order to serve quickly a broad range of products to the customers with the best best-before-date (BBD). When perishability is not an issue, products can be stored to cover fluctuations in demand.
b. Retailers and wholesalers expect small deliveries with a short lead time. This prefers also small batches, since the retailers and wholesalers wants small deliveries as well with the best BBD. Again, when perishability is not an issue, products can be stored to cover fluctuations in demand.
c. Consumer behavior is more erratic.
Since customer behavior is unstable and unpredictable, small batches needs to be produced in order to respond quickly to customer demands. An interesting insight of customer behavior was given by Rajagopalan (2002). He stated that demand for an item might be higher if it is made to order because of the potential for customization. When perishability however is not an issue, products can be stored to cover erratic customer behavior.
2. Plant characteristics
a. Expensive capacity with flow shop oriented design because of conventional small product variety and high volumes.
This asks for a high capacity utilization and thus large batches. Since the design is capital expensive you always will keep your production running.
b. Extensive sequence-dependent set-up and cleaning times between different products or product types.
3. Product characteristics
a. Limited shelf life for raw materials, semi-finished and finished products. This asks for small batches, since the products cannot be stored for a long time. Sometimes products are perishable until some point in the chain. When products are for instance frozen at the end of the chain. If that is the case, batch sizes should be small until that time and after that point in the chain (packaging for example), batch sizes could increase.
b. Variation in supply, and quality of raw material.
When the variation in quantity and quality of raw material is high, it is preferable to have raw materials on stock. However, due to the perishability of raw materials, most often this is not an option. This characteristic does not especially asks for small or large batches, but due to this uncertainty, batch sizes should be changed ad hoc in practice when some products does not meet the requirements. Rajaram and Karmarkar (2004) analyzed the planning and scheduling of multiproduct batch operations in the food processing industry. They noticed that if products are subject to “campaigns” like sugar beets or potatoes, multiple batches of the same product are produced sequentially to minimize set up and quality costs associated with rework costs. Regarding the quality costs: the risk of having large batches is that once the quality of a batch does not meet the requirements, the whole batch should be considered as a loss for the actual purpose it was meant for.
4. Production process characteristics
a. Processes having variable yield and processing time.
Batches can be changed during production when this rate deviates from what was expected. The same holds when a production order deviates in processing time: batches in this case can be decreased or the entire production order is cancelled or delayed. As yield and processing times can be moved in any direction, batch sizes can be both increased and decreased.
b. A divergent flow structure.
unpredictable demand. The divergent flow structure enables the production process to produce some intermediate products where after they are customized. In that case, these intermediate products are produced on stock in large batches and wait in a buffer to be customized in the packaging stage in rather small batches. This is however only possible when perishability is not an issue. Depending on this structure, batch sizes tend to be either large or small.
c. Packaging stage is labor intensive where the processing stage is not.
The packaging stage seems to be more flexible than the processing stage, since planning human capital is easier than planning machines. Another remark about the difference between the processing and the packaging stage is that the processing stage is subjected to extensive change overs where the packaging stage is less. This note can be translated to batch sizes in a sense that the batch sizes before the packaging stage are allowed to be larger than batch sizes in the packaging stage.
d. Production rate is mainly determined by the capacity.
It can be obtained that batches are partly determined by the capacity of the processes. It is for example more efficient to have a full silo or container than just half of it thanks to economies of scale. So, batches cannot always perfectly be matched by the customer order.
2.4 Conceptual framework
Now we know what the single effect of the food processing characteristics is on the batch size. The direction of the characteristics on batch sizes is presented in Table 2.2.
Group Characteristic Batch size
Market Customers asks for more quickly delivered and more varied products
Decrease Retailers and wholesalers expect small deliveries with a short
lead time
Decrease
Consumer behavior is more erratic Decrease
Plant Expensive capacity with flow shop oriented design because of conventional small product variety and high volumes
Increase Extensive sequence-dependent set-up and cleaning times
between different products or product types
Increase Product Limited shelf life for raw materials, semi-finished and finished
products
Decrease Variation in supply, and quality of raw material. Increase &
Decrease Process
Processes having variable yield and processing time
Increase & Decrease A divergent flow structure
Packaging stage is labor intensive where the processing stage is not
Increase & Decrease Production rate is mainly determined by the capacity
Increase & Decrease Table 2.2. Food processing industry characteristics related to batch sizes.
From Table 2.2 we can obtain that the market characteristics prefer smaller batch sizes. The plant characteristics tends towards larger batches. The product characteristics tends to decrease batches. Depending on the context of the process, batches can be both increased and decreased based on the process characteristics.
3. METHODOLOGY
In this section, the methodology will be explained by treating the following subjects respectively: research method, case selection and description, data collection, data analysis, and a time path. After this part, it is clear how we execute this research.
3.1 Research method
From literature we know some single effects of the food processing characteristics on the batch size. However, we do not know how these factors affect the batch size in real life. A case study is the perfect way to investigate “how” these factors play a role in real life. Karlssen (2016) stated as well that “how” questions can be answered the best by a case study, so for this reason a case study will be performed in this research. Case study analysis focuses on a small number of cases that are expected to provide insight into a causal relationship across a larger population of cases (Gerring 2007). To be able to generalize our findings, it is preferable to conduct multiple cases. Multiple cases allows us as well to explore which factors plays a role in which environments, and so we eventually can draw meaningful conclusions. Multiple cases also allow for comparisons that clarify whether an emergent finding is simply idiosyncratic to a single case or consistently replicated by several cases (Eisenhardt 1989). Next to that, emerging insights from multiple case studies are better grounded, more accurate and theoretically transferable when compared to a single case study design (Eisenhardt 1989). Using multiple sources to investigate the same phenomenon is a good thing and strengthens the construct validity (Voss, 2002). Besides, Yin (2013) stated that multiple case studies provides more grounded and reliable results than single case studies, so therefore multiple companies will be investigated in this study.
This detailed and focused subject is to the best of our knowledge not examined before, so in that sense you could typify this study as an explorative and inductive study. We however do know something about previous studies and so we have a sort of direction in which we could think. So in that way we could describe it as a descriptive and deductive study. That is why we see this research more as a mixture of exploratory/inductive and descriptive/deductive study.
Next to that, the interviewees were asked how they would describe the major factors in the interviews like “What is a batch?”, and “How would you describe MTS?” for example. Once we noticed contradicting answers to any question in the interview, we asked probing questions to examine what the interviewee really meant. After the interviews were conducted, we worked out the total interviews and sent those transcripts back to the interviewee to be checked. In one case, the transcript was modified by the interviewee, because we interpret something wrongly. After the transcript was approved, we coded the interview. We conducted the interviews with two interviewers. Besides, as we asked open questions, we enabled the interviewees to explain certain phenomena and cause-effect relations. Since we incorporated six companies that differ on several aspects and even have some characteristics in common, we are able to generalize our findings to a certain level. We used a case study protocol for the interviews. Despite all the cases were different on several aspects, we found the right persons of the company willing to participate in the interviews. We explained the companies before what we were investigating, and so companies assigned the right people to our interviews that are the most knowledgeable in our field of research. All interviewees were responsible for the planning and/or supply chain departments. Next to that, we secured our transcripts of interviews and other data on a database and in the cloud.
3.2 Case selection & description
First, the companies should process food obviously. The second criteria is that the company should exist of a planning department. Since we can obtain from the theory section that the food processing industry exists of some unique characteristics, we have to find companies that differ on market, plant, product, and production process characteristics. A few cases should exist of the hybrid production strategy. To avoid cultural biases, the companies must be located in the same country. In Table 3.1 the preferred characteristics for this study are given. Most criteria are related to the food processing characteristics.
Characteristics Different environments Market Mix of location in supply chain
Plant Mix of small and large companies (<50 employees, > 500 employees). Mix of planning methods (based on experience, based on statistics). Mix of production strategies (more towards MTO, more towards MTS). Mix of extensively and smaller set-up and cleaning times.
Product Mix of perishability Mix of quality variation Process Mix of variation in processes
Mix of divergence in processes Mix of labor intensive processes Table 3.1. Case criteria.
Eisenhardt (1989) stated that there is no ideal number of cases, but a number between 4 and 10 cases usually works. We asked around in our personal circuit and fellow students to acquire the most interesting companies. We even searched at www.regiobedrijf.nl and requested a list of food related companies by the Chamber of Commerce to gain more potential cases. Next to that, we made use of the network of professors of the University of Groningen. In order to include the most interesting companies for our research, we established a list of 48 potential companies. These potential companies in the list all met the preferences based on our common sense. We sent these companies a mail to examine which companies fit in our research and to discover the willingness of companies. Due to time constraints and the willingness of companies, six cases were incorporated in our study. Five of them were acquired by the network of professors of the University of Groningen, one was acquired via the list of the Chamber of Commerce. This selection is a good mix of companies that differ on all kinds of characteristics and so fits our preferences perfectly. In Table 3.2 the cases are explained in general facts.
Case A Case B Case C Case D Case E Case F
Annual volume end product (tons) 6,600 275,000 14,500 600,000 230,000 235,000 Annual turnover (mln euros) 7 150 4,000 550 80 871,000 # FTE 22 600 350 1,000 120 700
# shifts 3 3 2&3 5 5 3&4
# SKU’s RM 1 25 1,000 1 100 15
# SKU’s EP 18 150 200 1,000 1,000 495
Location in SC
Upstream Upstream & Middle
Upstream & Middle
Upstream Upstream & Middle
Planning method
Experience Tool Tool Tool Experience Tool
Driving force(s) Quality, Flexibility High consistent quality, Service, R&D Flexibility, process optimization High consistent quality, reliability High consistent quality, R&D Reliability, flexibility
Table 3.2. General data of cases.
3.3 Data collection
Semi-structured interviews are conducted. Next to that, relevant other data is acquired: an overview of their production process, a list of SKU’s (raw materials and end products), a list of product groups, an overview of a production planning schedule, an overview of production cycles, and some companies provided a tour through their facility. In Table 3.3, an overview is provided about the interviewees and the data that was provided to us. Unfortunately the additional documents were confidential and so may not be shared. Even during the facility tours, it was not allowed to take photos and videos.
An interview protocol is established based on the literature (see Appendix A). Besides the general part about the basic numbers and processes of the company, the structure is based on the model of Soman et al. (2004). Since their study argued that the determination of batch sizes in hybrid production systems is a consequence of prior decisions, the structure of the interview relates to that model. We work top down from the Production Plan State towards the Process State. At the highest level of Production Plan State, questions regarding product groups and performance measures were asked. In the second level, Capacity Coordination, questions about the production planning were raised. At the most operational level, the Detailed Scheduling State, questions regarding batch sizes and sequences were addressed.
Case A Case B Case C Case D Case E Case F Interviewee(s) Supply Chain Manager Forecast Manager Europe + Planning Coordinator Netherlands Planning & Logistics Manager + Team Leader Planning Supply Chain Manager + S&OP Planner Supply Chain Manager Supply Planning Coordinator + Continuous Improvement Manager Length of interview (min) 60 80 70 70 70 70 Transcripts of interview (pages) 8 7 12 11 11 13 Overview production process X X X X Overview SKU’s X X X X X X Overview product groups X X X X X X Overview production planning X X X X X X Overview production cycles X X X X X X Tour X X
Table 3.3. Data gathered.
3.4 Data analysis
presented in Table 3.4. In Appendix B, examples can be found on how these codes are linked to quotes of the interviews.
Group First-order codes
Production strategy MTO, MTS, Hybrid
General Driving force, Outperforming competitors, General facts
Forecasting Contract based, Customer contact, Experience based, Forecast by customer, Historical data forecast
Market Due dates, Variability, Complexity, Contracts, Customer contact, Customer specific, Fast & slow movers, Flexibility, Profit margin, Rush order
Plant Capacity, Variability, Change overs, Complexity, Dedicated resources, Flexibility, Minimum run length, Mixed products, Production cycle, Production sequence
Product Variability, Complexity, Mixed products, Perishability, Quality supply
Process Workforce, Variability, Capacity, Complexity, Dedicated resources, Divergent flow structure, Flexibility, Inventory, Minimum run length
Table 3.4. Coding scheme.
After the cases are coded individually, a within-case analysis was done by linking quotes to one of the first-order codes. By doing it this way, we were able to link single phenomena to each of the determinants of batch sizes. After the within-case analysis, a cross-case analysis was performed in which patterns will be investigated among all cases.
Crucial in this study is the batch size. What is a big batch for one company, could be a very small batch for another company. The batch sizes should therefore be related to a company characteristic in order to compare the batch sizes among all the different cases. In this study the relative size of the batch is expressed as a percentage of the total monthly production volume of the plant. Since every company expresses their output in tons, and the batch sizes are expressed in tons as well, this is a valid and reliable measurement. The relative batch sizes are presented in Table 3.5.
Case A Case B Case C Case D Case E Case F
Average batch size
2.94% 3.45% 0.53% 2.02% 2.87% 5.08%
To determine the presence of the food processing characteristics and the extent to which a company is MTO or MTS driven, the interviews and additional data were analyzed. In table 3.6 an overview is provided how the different created groups are analyzed.
Group Characteristic Measurement
Production strategy Production strategy Interviews, overviews production processes Market Customers asks for more quickly
delivered and more varied products
Interviews
Consumer behavior is more erratic Interviews Plant Expensive capacity (small variety,
high volumes)
Interviews, overviews production processes, lists of SKU’s, production planning schedules Change overs Interviews, product groups, production
planning schedules, production cycles
Product Perishability Interviews
Variation in supply raw material. Interviews Variation in quality raw materials Interviews Long lead times of raw materials. Interviews Process Processes having variable yield and
processing time
Interviews, production planning schedules
A divergent flow structure Interviews, overviews production processes, lists of SKU’s, production planning schedules Packaging stage is labor intensive
where the processing stage is not
Interviews, production planning schedules
Production rate is mainly determined by the capacity
Interviews, overviews production processes
4. FINDINGS
First, the findings regarding the production strategy of the case companies are given. After that, the observed phenomena are explained. In this section of observed phenomena we make a distinction between the expected rather simple findings and the unexpected more complex findings.
4.1 Production strategy
The company in which the MTO strategy was the most dominant, was case C. They literally stated: “We are demand driven”, and this demand is derived from the actual point of sale data: “It all starts with the customer demand. The customer communicates their forecast with us a year ahead. This forecast is updated monthly and daily and we see that the closer the actual demand is, the more correct is the demand”. The companies that are closely related to each
other are Company A and E. Those companies exist of both a MTO-strategy and a MTS-strategy and thus can be defined as hybrid production companies. “If the customer does not need a specification it can also be removed from the stock, but if a customer orders something in advance, for example, bio beans, the entire process is dedicated for that customer and his beans.” (Case A), and “Most of the products we have on stock, but it occurs that we do not have. In that case you have to bake, mill, and pack first” (Case E). Case D was more MTS-driven, but even allows MTO-orders: “Some products we sell to more customers, so we have those ones on stock, and so we deliver from stock. But we have as well products that we produce especially for certain customers, but that is around 10% of all the total production”. Companies
B and F are fully MTS oriented: “We only produce on stock” (Case B), and “We always produce on stock, because we are able to always sell it” (Case F). In Figure 4.1 the cases are placed on a spectrum from MTO to MTS.
MTO Hybrid MTS
A
B
C E D F
Figure 4.1. Strategy of cases.
All the quotes regarding the production strategy of companies can be found in Appendix C.
4.2 Expected findings
The most dominant factor that affects the batch size is the perishability of products. When the shelf life of (raw) materials is short, products will not be produced in large batches or produced to be stocked, regardless of any other factor that could influence the batch size. So when perishability is not an issue, other characteristics play a role in the batch size determination. The perishability among all cases is depicted in Figure 4.2 and can be related to Figure 4.1 in which the production strategy of all cases is showed. The place of Case F in Figure 4.2 should be nuanced, since the raw materials are perishable, but the end product is not.
Low Perishability High A B C D E F Figure 4.2. Perishability in cases.
The second dominant factor is the capital: the more expensive the capital is (and so the higher the costs when the machine is not running during cleaning sessions and set-ups), the larger the batch sizes. The production sequence that is leading in the planning of all cases, is based on these cleaning, and change over times. It can be seen that perishability is the most dominant factor, by the fact that the capital in Case C is expensive, but still has small orders that are produced on order. The capital expensive and change overs among all cases is shown in Figure 4.3 and 4.4 respectively. Low Expensive capacity (small variety, high volumes) High A B C D E F Figure 4.3. Expensive capacity.
Low Change overs High
A C D B
E
F
Figure 4.4. Change overs.
had one case in which perishability and seasonality played a role together (Case C). In that case they just increased the workforce and operate with 4 shifts in stead of 3 shifts so that batches could be increased in both processing and packaging, since they had no possibilities to store their finished products and they had some capacity left. The variability of customer behavior among the cases is depicted in Figure 4.5.
Low Variabilty customer behavior High A B C D E F Figure 4.5. Variability customer behavior.
We discovered also that when products are not perishable and lead times of raw materials are long or subjected to campaigns, raw materials were stored. However, when raw materials are perishable like in Case F, contracts with their suppliers are established in which they agreed that the supplier should deliver the manufacturer within 24 hours. By doing this, Company F shifted the risk of obsolete products to their suppliers.
However, there are several interesting phenomena that we discovered in the cases when a combination of characteristics play a combined role. These phenomena deserve attention, and further in-depth explanation.
4.3 Planning method
planning tool as well, but this tool is owned by their customers: “It all starts with the demand of the customer. The customer fills in the products that he needs for the upcoming year, updated weekly. However, in Case F, before the batch planning is finished, it is checked by humans: “When the tool plans something optimal but your capacity is not sufficient, you need extra capacity. At that moment, the supply planner can modify the plan and decreases orders, replace orders, call customers et cetera”.
Case A and E are the smaller companies in terms of turnover and employees, and as a result the operations in that companies are less formalized. These cases are the most hybrid companies as well. Based on our obtained data, we are not able to relate this planning method to either a MTO or a MTS strategy, because both the case company that produces fully on order (Case C), as well as the companies that produce on stock (Case B, D, and F) used advanced planning tools for their production planning. So, it may be that the absence of a planning tool is a characteristic of hybrid production systems, but we do not have reasons for that. Since we cannot relate this planning method to any production strategy, it seems that the formality of the company is the reason why companies just plan based on experience and contracts. This could be argued by the fact that smaller companies have less money available to fund an advanced planning tool. Besides, all the different orders and customers can be overseen more easily in smaller companies. To sum up, there is not an incentive to have such a tool in smaller companies.
4.4 Production capacity
These cases are the less formalized, and are even the most hybrid as well. Case A and E partly produce on order, but produce also partly on stock, since they produce more than what is asked. So, this could be a sign that producing always a plural of full capacity is typical for production systems in which products can be put on stock. However, this is not the case when we analyze other companies that are able to put finished goods on stock (Case B, D and F). They do not plan full entities of their capacity, so this conclusion cannot be drawn from our interviews. Since not all companies that are able to stock products produce a plural of their full capacity, it could be a characteristic of hybrid production strategies. However, there is no strong evidence for that. Another reason why these more informal companies always plan full silos and containers, could be that they do not use a tool that calculates the optimal batch size. Since the batch sizes are not calculated by a tool, the planner thinks that planning a full silo is more efficient based on common sense.
All quotes regarding the “size of companies” can be found in Appendix E. The extent to which the production planning is determined by capacity, is depicted in Figure 4.6.
Low Production rate is determined by capacity High A B C D E F
Figure 4.6. Production rate is determined by capacity.
4.5 Materials before production
B). When the company is more MTO driven, the planning should be adjusted. This is because the planned production orders are allocated to a customer, and so the batch cannot be produced when the quality does not meet the standard of that particular customer. As a result, the total production batch is delayed, or another batch with a higher quality is increased in order to be able to still deliver a smaller part of the total order to a customer. “We communicate to the Central Planning whether we should produce this order and how many of that order. In the meantime we try to arrange something with our suppliers” (Case C).
Companies that only have a few raw materials that are responsible for a relative broad range of products (divergent flow structure) have an internal process that tries to avoid any quality issues on the those many end products. “We have to deal with a natural product that differs in quality and protein content. So we have a mixer and that person makes recipes so that the quality is constant at the end of the process” (Case E), “We use no fresh products, but paste. That paste has a certain level of brix value. Depending on this value, we adopt our water supply to the product” (Case F). Whereas it makes sense that companies always do quality checks, it seemed that the more the end products rely on the quality of raw materials, the more companies do on quality consistency. As a result, these companies are able to use their raw materials always, no matter whether they produce on stock or on order.
All quotes regarding the “variation of raw materials before production” can be found in Appendix F. In Figure 4.7 and 4.8 we presented the variation in quality of raw materials and the divergence of the production process among the cases respectively.
A B C D E F Low Variation in quality raw materials High Figure 4.7. Variation in quality raw materials.
4.6 Materials during production
During production the materials plays a major role as well. As obtained from the interviews, this is more complicated than the variation in quality of materials before production. There is a remarkable difference between the processes that extract something out of the material and the processes that mix some materials for their end products.
Companies that extract something out of a material (Cases A, C in the first phase of production, D in the first phase of production, and E) adjust their batch size according to the yield of the process. This is illustrated by Company C for example: “It is a natural product, and your extraction process has a certain return what you can get out of the raw material. So it differs from product to product how many return you get”. When the company is producing on order like Case C, it could be that the supply of raw materials should be adjusted in order to still achieve the amount of end product asked by the customer. When the yield of the process is not as high as expected, and the production orders are not that strict as in an MTO-environment (Case A, D, and E), and products are not on stock, the customer is contacted. These companies are then still able to deliver the desired amount over a longer period by increasing batches that will be produced later “We first ask the customer whether they really need the order. If so, we look if we can put in that order in one of the other batches, or split the order in two separate orders so that next week the rest of the order is delivered” (Case E). When more than expected is extracted from the material, is will be put in stock or post-processed in a fast-moving product: “This overproduction is most often the fast movers, since the products can only stay for 3 months in our stock.” (Case A).
materials and check whether it meets our quality requirements, and so you have to recirculate a lot with the mixing tank. Depending on the product you need 6,000 to 9,000 kilos of your product” (Case F). Raw materials are added to the product when the quality does not meet the standards, since the raw materials cannot be extracted from the mix anymore. So the batch size only can increase in such cases. Again, these extra products are used for the fast moving products and stocked afterwards. In processes where chemical reactions play a role rather than mixing for the taste, safety becomes an issue. In such cases, a maximum batch size is determined as well: “The reactors cannot handle the large amount, or the chemical reaction time is too long to produce more of that particular product … when you produce a product too long, you have a change that it explodes or will be combusted” (Case D).
All quotes regarding the “variation of (raw) materials during production” can be found in Appendix G. To what extent companies experience variable yield in their production process, is depicted in Figure 4.9. Low Process has variable yield and processing time High A D C E F B Figure 4.9 Process has variable yield and processing time in first phase of production.
4.7 Workforce
All quotes regarding “workforce” can be found in Appendix H. To what extent companies had issues with their workforce, and how labor is divided, is shown in Figure 4.10 and 4.11 respectively. Low Workforce High A B C D E F Figure 4.10. Workforce Low Labor intensive packaging compared to processing High A B C D E F Figure 4.11. Labor intensive packing compared to processing.
4.8 Differences production strategies
5. DISCUSSION
5.1 Link results to literature
The food processing characteristics leads to certain batch sizes according to literature. Both tables 5.1 and 5.2 can help us in relating real life data with the literature. In Cases B, C, and F, it is clear which factors are the most dominant. In Case B and F, it can be seen that the plant characteristics are more dominant than the market and product characteristics. This is supported by reality, because their production is not based on actual customer orders, but rather on their own capacity. They even do not have to deal with products that become obsolete after a certain time. In Case C, the market and product characteristics play a more dominant role. This can be confirmed by our findings, since the products in their process are subjected to perishability. Besides, Company C produces on order, and thus market characteristics determine in their batch sizes. The batches of company A and E are even bigger than D what seems to be strange compared to Figure 4.1. We would expect that the batch size of A and E is smaller than Case D, since D is more directed to MTS. However, as we discussed, these companies like to plan full silos and so do not stick to the real orders what causes the bigger batch sizes.
Characteristics Batch size
Market Decrease
Plant Increase
Product Decrease
Production process Increase & Decrease Table 5.1. Batch sizes according to literature.
Case Batch size (% of monthly output) A 2.94 B 3.45 C 0.53 D 2.02 E 2.87 F 5.08
Table 5.2. Actual batch size.
case out of six. Therefore, we are not sure whether it is such a big issue in the total food processing industry as suggested.
Now, we know that the single effects of the food processing industry characteristics on the batch sizes are confirmed by our study. The more interesting is the effect of the combined characteristics.
We found that the smaller the company is, the less they use a planning tool to determine their batch sizes. At the same time, these companies are the most hybrid as well. This characteristic about production strategy could play a role in the use of planning tools, since Aslan, Stevenson, and Hendry (2015) proved that MTS companies make more use of planning tools like ERP systems. They argued that “production strategy is an important contextual factor affecting both applicability and impact. Their findings however does not fully supports our results, since Case C also uses a planning system and is a MTO company. Thus, we are not able to justify the link between production strategy and the use of planning tools. The combination of O’Reilly et al. (2015) and Perona et al. (2009) however does support our results. O’Reilly et al. (2015) found that unclear policies of MTO/MTS decisions results in ad hoc decision making, and Perona et al. (2009) stated that this ad hoc decision making results in experience and common sense based decisions of planners. This is exactly what happens in these less formalized companies. Moreover, O’Reilly et al. (2015) stated that not having well-defined decision rules can cause that the sales department follows customer requests too easily, what is illustrated by Company E: “When nothing is arranged, sales has the most power in this organization”.
We know from literature that the supply of raw materials plays a role in the food processing industry, and so that it probably plays a role in determining the batch size. Rafiei (2016) stated that the uncontrollable supply of raw materials is a difficult issue in MTO production systems. This is supported by our study, since the supply of raw materials is even a bigger issue in Case C.
This adjustment of the batch size is done on the lowest operational level in the model (Process state). The deviation between “extracting” and “mixing” processes in this context is never made before in literature. Nevertheless, we discovered differences in batch size determination in these processes. Where the batch size in extracting processes was determined on the yield of the processes, was the batch size in mixing processes determined by the quality of the mixed product. Besides, in the mixing process the batch size can only increase, since ingredients cannot be extracted from the mixed product anymore. Even the maximum batch size in chemical reactions of mixing ingredients was not taken into consideration before. When we would like to relate this mixing issue to a broader sense, we can see it as an assembling process. However, existing literature about assembling issues only discusses assembling processes in other industries as the food processing industry. In other businesses where materials are assembled, these parts can be separated from the end product after it is assembled which is even not possible with food that is mixed.
According to Soman et al. (2004) there is a difference in labor intensity in the production process. This workforce spread partly holds for the food processing industry, since we found some exceptions. We found that in mixing processes, the workforce is more concentrated on the processing stage rather than the packaging stage.
Almehdawe & Jewkes, (2013) and Nagib et al. (2016) argued that the more the strategy is towards MTS, the more efficiency benefits are achieved by fully use the production capacity. This is proven by our study in a sense that the case companies that produce more towards MTS “over planned” their production planning. The other way around, the companies that produce more on order had some capacity left to cope with incoming orders.
Soman et al. (2006), Beemsterboer, Land, and Teunter (2016), and Beemsterboer, Land, and Teunter (2017) suggested different options regarding batch sizing in hybrid production environments. These are summed up in Table 5.3.
Authors Suggestions
Soman et al. (2006) Four options are suggested: a) treat MTO-items as if belonging to a separate product family, b) produce additional stock in MTS-items, c) reserve capacity for MTO in each cycle, or d) use the idle time in pure MTS situation.
Beemsterboer, Land, and Teunter (2016)
Beemsterboer Land Teunter (2017)
Three options: 1) Fully flexible, 2) partly flexible in which switching after MTO item is not allowed, 3) no flexible at all
Table 5.3. Suggestions on batch sizing in hybrid production systems.
From our findings, Case C did reserve capacity for MTO orders in their production, every company produces MTS items when there were no MTO items, and none of the companies prioritized either MTS or MTO orders. So we can see that none of the cases fully adopt one of the suggestions of these studies, but only parts of these methods.
5.2 Limitations & Further Research
6. CONCLUSIONS
6.1 Conclusions
In this study we attempt to examine how the batch sizes are determined in the hybrid food processing industries.
The batch sizes in the food processing industry as a whole has two dominant determinants in common. First, the perishability of items is always leading in determining batch sizes and decreases the batch size. When the shelf life of products is not an issue, other factors become important in determining the batch size. The expensive capacity is after perishability the most common determinant for batch sizes and increases the batch size. The corresponding effect of this expensive capacity is the fixed production sequences in the production cycles.
Based on our cases, we can conclude that the smaller and less formalized companies do not use an advanced planning tool to determine their batch sizes. They determine the batch sizes based on experience and contracts. The capacity of production entities is leading in these kind of companies. This can affect the batch size in two ways: increasing and decreasing.
The variation in supply of raw materials has a huge influence on batch sizes in MTO environments. The influence can be in two directions: increasing and decreasing. When finished products can be stored as in the hybrid and MTS production, this is not an issue. The more companies have a divergent flow structure, the more companies do on quality consistency and asks for either increasing or decreasing the batch size.
The batch size of companies that extract something out of the raw material are depending heavily of the quality of the raw material. When the yield is less than expected, MTO companies should process more raw materials to achieve the same amount of finished goods and so the batch size should be increased. For MTS companies this is not an issue when there are still enough products on stock. But when the product is not on stock, the total order is mostly spread over a longer period. When the yield is more than expected, both MTO and MTS companies use the extra products for the fast moving products.
subsequently determined by the added materials and can only become bigger, since materials cannot be extracted anymore from the mixed product. When the end product is subjected to a chemical reaction, a maximum batch size is set as well due to safety issues.
The workforce is influencing the batch size as well in companies that depends heavily on mixing processes that is done by humans.
6.2 Theoretical implications
To the best of our knowledge, the issue of batch sizing in hybrid food production systems is not examined before.
Literature did not suggest that the supply of raw materials had such a big influence on the batch size, both before and during production. Moreover, the relation between the divergence of the process and the quality of raw materials was not discovered before. The separation between extracting and mixing processes is never made before regarding batch size determination. However, we showed that the batch size is determined differently in these two processes. We further showed that the labor intensity is depending on the kind of process: extracting or mixing.
Besides, we opened the discussion whether perishability is really an issue for the total food processing industry or for a small part of the industry.
6.3 Managerial implications
REFERENCES
Adan, I. J., & Van der Wal, J. (1998). Combining make to order and make to stock. OR Spectrum, 20(2), 73-81.
Akkerman, R., Van Der Meer, D., & Van Donk, D. P. (2010). Make to stock and mix to order: Choosing intermediate products in the food-processing industry. International Journal of Production Research, 48(12), 3475-3492.
Almehdawe, E., & Jewkes, E. (2013). Performance analysis and optimization of hybrid manufacturing systems under a batch ordering policy. International Journal of Production Economics, 144(1), 200-208.
Aramyan, L.H., Lansink, A.G.J.M.O., van der Vorst, J.G.A.J., and van Kooten, O. (2007). Performance measurement in agri-food supply chains: A case study. Supply Chain Management: An International Journal, 12 (4): 304-315.
Aslan, B., Stevenson, M., & Hendry, L. C. (2015). The applicability and impact of Enterprise Resource Planning (ERP) systems: Results from a mixed method study on Make-To-Order (MTO) companies. Computers in Industry, 70, 127-143.
Beemsterboer, B., Land, M., & Teunter, R. (2016). Hybrid MTO-MTS production planning: An explorative study. European Journal of Operational Research, 248(2), 453-461.
Beemsterboer, B., Land, M., & Teunter, R. (2017). Flexible lot sizing in hybrid make-to-order/make-to-stock production planning. European Journal of Operational Research, 260(3), 1014-1023.
Chang, K. H., & Lu, Y. S. (2010). Queueing analysis on a single-station make-to-stock/make-to-order inventory-production system. Applied Mathematical Modelling, 34(4), 978-991. Choi, H. S., & Lee, D. H. (2009). Scheduling algorithms to minimize the number of tardy jobs in two-stage hybrid flow shops. Computers & Industrial Engineering, 56(1), 113-120.
Claassen, G. D. H., Gerdessen, J. C., Hendrix, E. M., & van der Vorst, J. G. (2016). On production planning and scheduling in food processing industry: Modelling non-triangular setups andproduct decay. Computers & Operations Research, 76, 147-154.
Eisenhardt, K. M. (1989). Building theories from case study research. Academy of management review, 14(4), 532-550.
Federgruen, A., & Katalan, Z. (1999). The impact of adding a to-order item to a make-to-stock production system. Management Science, 45(7), 980-994.
Fernandes, N. O., Silva, C., & Carmo-Silva, S. (2015). Order release in the hybrid MTO–FTO production. International Journal of Production Economics, 170, 513-520.
Gamberini, R., Lolli, F., Rimini, B., Torelli, M., & Castagnetti, E. (2011). An innovative approach for job pre-allocation to parallel unrelated machines in the case of a batch sequence dependent manufacturing environment. International Journal of Production Research, 49(14), 4353-4376.
Gerring, J. (2007). The case study: what it is and what it does. In The Oxford handbook of comparative politics.
Gupta, D., & Benjaafar, S. (2004). Make-to-order, make-to-stock, or delay product differentiation? A common framework for modeling and analysis. IIE transactions, 36(6), 529-546.
Iravani, S. M., Liu, T., & Simchi-Levi, D. (2012). Optimal Production and Admission Policies in Make-to-Stock/Make-to-Order Manufacturing Systems. Production and operations management, 21(2), 224-235.
Ivert, L.K., Dukovska-Popovska, I., Kaipia, R., Fredriksson, A., Dreyer, H.C., Johansson, M.I., Chabada, L., Damgaard, C.M., and Tuoikangas, N. (2014). Sales and operations planning: responding to the needs of industrial food producers. Production Planning & Control, 2014 Kallrath, J. (2002). Planning and scheduling in the process industry. OR spectrum, 24(3), 219-250.
Kanda, S., Takahashi, K., & Morikawa, K. (2015). A flexible service rule for the dynamic make-to-stock/make-to-order hybrid production system. Procedia Manufacturing, 2, 46-50. Karlsson, C. (Ed.). (2016). Research Methods for Operations Management. 2nd edition. New York: Routledge.
Khakdaman, M., Wong, K. Y., Zohoori, B., Tiwari, M. K., & Merkert, R. (2015). Tactical production planning in a hybrid Make-to-Stock–Make-to-Order environment under supply, process and demand uncertainties: a robust optimisation model. International Journal of Production Research, 53(5), 1358-1386.
Kim, E., & Park, T. (2016). Admission and inventory control of a single-component make-to-order production system with replenishment setup cost and lead time. European Journal of Operational Research, 255(1), 91-102.
Linn, R., & Zhang, W. (1999). Hybrid flow shop scheduling: a survey. Computers & industrial engineering, 37(1-2), 57-61.
Meulenberg, M. T. G., & Viaene, J. (1998). Changing food marketing systems in western countries. In Innovation of food marketing systems (pp. 5-36). Wageningen Pers.
Nagib, A. N. M., Adnan, A. N., Ismail, A., Halim, N. H. A., & Khusaini, N. S. (2016, November). The Role of Hybrid Make-to-Stock (MTS)-Make-to-Order (MTO) and Economic Order Quantity (EOQ) Inventory Control Models in Food and Beverage Processing Industry. In IOP Conference Series: Materials Science and Engineering (Vol. 160, No. 1, p. 012003). IOP Publishing.
Nakhla, M. (1995). Production control in the food processing industry: the need for flexibility in operations scheduling. International Journal of Operations & Production Management, 15 (8): 73-88.
Oakland, J. S., Sohal, A. S., & Wirth, A. (1991). A decision network application to batch size determination for single-stage systems. Engineering costs and production economics, 21(2), 143-153.
O'Reilly, S., Kumar, A., & Adam, F. (2015). The role of hierarchical production planning in food manufacturing SMEs. International Journal of Operations & Production Management, 35(10), 1362-1385.
Olhager, J. (2003). Strategic positioning of the order penetration point. International journal of production economics, 85(3), 319-329.
Pang, Z., Shen, H., & Cheng, T. C. E. (2014). Inventory Rationing in a Make-to-Stock System with Batch Production and Lost Sales. Production and Operations Management, 23(7), 1243-1257.
Rabta, B., & Reiner, G. (2012). Batch sizes optimisation by means of queueing network decomposition and genetic algorithm. International journal of Production research, 50(10), 2720-2731.
Rafiei, R., Nourelfath, M., Gaudreault, J., De Santa-Eulalia, L. A., & Bouchard, M. (2016). Analysis of uncontrollable supply effects on a co-production demand-driven wood remanufacturing mill with alternative processes. INFOR: Information Systems and Operational Research, 54(3), 234-261.
Rajagopalan, S. (2002). Make to order or make to stock: model and application. Management Science, 48(2), 241-256.
Rajaram, K., & Karmarkar, U. S. (2004). Campaign planning and scheduling for multiproduct batch operations with applications to the food-processing industry. Manufacturing & Service Operations Management, 6(3), 253-269.
Reinertsen, D. G. (2009). Principles of Product Development Flow (Vol. 1). Redondo Beach: Celeritas Publishing.
Romsdal, A. (2014). Differentiated production planning and control in food supply chains. Doctoral thesis at Norwegian Univeristy of Technology and Science (NTNU), 2012:16. Sawaqed, N., & Foote, B. L. (1989). The effect of batch sizing and customer pressure (order modification) on a hybrid assembly job shop. International journal of production research, 27(8), 1317-1334.
Soman, C. A., Van Donk, D. P., & Gaalman, G. (2004). Combined to-order and make-to-stock in a food production system. International Journal of Production Economics, 90(2), 223-235.
Soman, C. A., van Donk, D. P., & Gaalman, G. J. (2006). Capacitated planning and scheduling for combined make-to-order and make-to-stock production in the food industry: An illustrative case study. International Journal of Production Economics, 108(1), 191-199.
Van der Vorst, J., van Dijk, S.J., and Beulens, A.J.M. (2001). Supply chain design in the food industry. International Journal of Logistics Management, 12 (2): 73-86.
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 Wezel, W., D.P. Van Donk, G. Gaalman. (2006). The planning flexibility bottleneck in food processing industries. Journal of Operations Management. Vol. 24, pp 287-300.
Van Wezel, W. (2001). Tasks, hierarchies, and flexibility. Planning in food processing industries. Labyrint Publication, Capelle a/d IJssel.
Voss, C., Tsikriktsis, N., & Frohlich, M. (2002). Case research in operations management. International. Journal of Operations & Production Management, 22(2), 195–219.
Wagner, H. M., & Whitin, T. M. (1958). Dynamic version of the economic lot size model. Management science, 5(1), 89-96.
Williams, T. M. (1984). Special products and uncertainty in production/inventory systems. European Journal of Operational Research, 15(1), 46-54.
Wu, M. C., Jiang, J. H., & Chang, W. J. (2008). Scheduling a hybrid MTO/MTS semiconductor fab with machine-dedication features. International Journal of Production Economics, 112(1), 416-426.
Yu, J. M., Huang, R., & Lee, D. H. (2017). Iterative algorithms for batching and scheduling to minimise the total job tardiness in two-stage hybrid flow shops. International Journal of Production Research, 55(11), 3266-3282.
7. APPENDICIES
Appendix A: Interview protocol
Interview: xx-xx-2017
Place: Opening
• Introduction of interviewer and interviewee • Confidentiality assurance
• Permission to audiotape General
• Job-title of participant • Company description:
o Characteristics such as; size, market share and geographical locations o Number of employees (plant and organization)
o Turnover plant in terms of money and volume o Location in entire supply chain
• Could you give a rough description of the manufacturing process from raw material to final products?
• At what point is the product group customer specific? Product groups/families
1. Do you subdivide products in families/groups? 2. On which characteristics are these products grouped?
a. On what specific product characteristics are these products grouped? - Perishability
- Profit margin
- Variability in quality (raw materials and finished goods) b. On what specific market characteristics
- Actual customer demand/orders - Forecastability of orders
- Processing time
- Dedicated process steps - Variability in yield
3. What are the main performance measures of your customers? a. Based on which factors do you win or lose customers? 4. What makes your plant more successful than your competitors?
a. What is the driving force in your planning? Production planning
5. Which steps do you take to create your production planning per product goup? a. What is the timeline
b. Who are involved?
c. How do you deal with capacity constraints? d. What are your process constraints?
e. How do you deal with these constraints? f. What is limiting the output of the factory?
6. What is your planning horizon and how do you translate your long-term planning to a detailed scheduling?
Detailed scheduling
7. How do you make your detailed schedule?
a. How do you determine the production sequence? b. How do you involve cleaning times and set-up times? c. How do you determine batch/charge sizes?
d. What do you schedule (e.g. machines, products, people)? e. How do you deal with unexpected orders/stock outs? f. How control this process?
g. How do you deal with variation?
Appendix B: Examples coding
Group First order code Quote
Production strategy
MTO “We only start production when there is a customer order”
MTS “We only produce to stock, because we are sure we will sell it”
