Supporting agility by Production Planning and Control in Make-To-Order companies A single case study

(1)

Supporting agility by Production Planning and Control

in Make-To-Order companies

A single case study

7

th

_{of December 2020}

Master Thesis

MSc. Dual Degree Operations & Supply Chain Management

Faculty of Economics and Business

University of Groningen

Newcastle University Business School

Supervisor:

Dr. M. J. Land (University of Groningen)

Second supervisor:

Dr. Y. Yang (Newcastle University)

Daniël Pieter Deelstra

S2892367

B190572794

Dr. C. Hofstede de Grootkade 11-11

9718KA Groningen

(2)

2 Abstract

Many studies have been conducted on the concept of Agile Manufacturing (AM). Frameworks are developed to enable ‘agility’ for manufacturing companies, in order to react responsively to changing circumstances within the environment. The literature on AM is consistently lacking to incorporate planning related activities. Land et al. (2015) showed by means of simulation that adjusting certain Production Planning and Control (PPC) decisions to changing demand levels, has a significant impact on the delivery performance. The aim of this research is to bring these two concepts together and to provide a practical extension to the work of Land et al. (2015). This research focuses on the impact of a changing production mix on delivery and throughput performances in the practical context of a Make-To-Order (MTO) company. A diagnosis is performed by means of Throughput Diagrams (TDs) that is based on historical order progress data. The results of the research reveal that delivery and throughput performance are negatively affected by a changing production mix. In order to reduce the performance impact a generic design is developed to adjust PPC decisions. The order acceptance and release decisions are considered as the two most important decisions to reduce the negative impact of a changing production mix.

Keywords: Agile Manufacturing; Make-To-Order (MTO) companies; Production Planning & Control

(3)

3 Abstract ... 2

Preface ... 4

List of abbreviations ... 5

1. Introduction ... 6

2. Theoretical background ... 8

2.1 Agile Manufacturing ... 8

2.2 Production Planning and Control and Performance in the MTO context. ... 9

2.3 Production Planning and Control in the context of Agile Manufacturing ... 11

3. Methodology ... 12

3.1 Research design ... 12

3.1.1 Selection of the case company ... 13

3.2 Data collection ... 14

3.3 Data analysis ... 14

3.3.1 Diagnostic phase ... 15

3.3.2 Detailed explanation of the Throughput Diagrams ... 15

3.3.3 Interpretation of the availability and completion curves ... 20

3.3.4 Curves with an extreme ... 21

3.3.5 Design phase of the research ... 21

4. Results ... 22

4.1 Results of the diagnostic phase ... 22

4.1.1 Changing variance of the mix over time ... 22

4.1.2 Irregular promising pattern ... 26

4.1.3 Arbitrary release patterns for the forming operation ... 31

4.1.4 Irregular availability of forming hours ... 34

5. Discussion and design implications... 36

5.1 Reflection on diagnosis findings ... 36

5.1.1 Improvement potential at the delivery time promising process ... 36

5.1.2 Improvement potential at the realisation process ... 37

5.2 Design implications for PPC in MTO companies in general ... 38

5.2.1 Generic design outline focused on the delivery time promising process ... 38

5.2.2 Generic design outline focused on the realisation process ... 39

5.3 Limitations and future research ... 40

6. Conclusion ... 41

7. References ... 43

8. Appendices ... 45

Appendix A – Process description of the case company ... 45

Appendix B – Throughput Diagrams 2017 ... 47

Appendix C – Throughput Diagrams 2018 ... 51

Appendix D – Throughput Diagrams 2019 ... 55

(4)

4 Preface

I would like to express my appreciation and gratitude to all the people involved who made it possible to conduct this research.

At first, I like to write my gratitude to Dr. M.J. Land (University of Groningen), my supervisor during this project. He provided me feedback and guidance whenever it was needed. He was able to motivate me after every feedback discussion. Without his guidance I could not have come to this result. Additionally, I would like to thank my second supervisor Dr. Y. Yang (Newcastle University). She provided me useful feedback and support when it was needed during the project.

Secondly, I would like to thank the people involved from the case company. I would like to thank Theodoor who made everything possible. I would like to thank Allard and Kjeld who provided me all the data required for this project. In particular, Allard was my designated mentor at the case company, and he put a lot of effort in helping me with the data collection and answering my questions.

Despite the tough working conditions due to the Coronavirus, I was able to conduct this research and I am glad about the final result.

Daan Deelstra

(5)

5 List of abbreviations

AM Agile Manufacturing DP Delivery-time-promising process MTO Make-To-Order MTS Make-To-Stock

OPP Order Penetration Point

PPC Production Planning and Control

RBC Repeat Business Customisers

RP Realisation process

TD Throughput Diagram

VMC Versatile Manufacturing Companies

WIP Work-In-Progress

(6)

6 1. Introduction

This thesis studies the impact of a changing production mix and how this impact can be reduced by adjusting Production Planning and Control (PPC) decisions in the practical context of a Make-To-Order (MTO) company. Understanding the impact of a changing production mix will help MTO companies to better respond to customer requirements and needs. Responding to changing circumstances relates to the concept of ‘agility’. In the early ’90s, this concept first appeared in the literature related to production and manufacturing. Yusuf et al. (1999, p. 37) define the term ‘agility’ as: “The successful exploration of competitive bases (speed, flexibility, innovation proactivity, quality and profitability) through the integration of reconfigurable resources and best practices in a knowledge-rich environment to provide customer-driven products and services in a fast-changing market environment.”. An important concept in order to meet these challenges is Agile Manufacturing (AM). AM could be described as the adaptivity to respond to future changes, being flexible and responsive to current demands (Gunasekaran, 1998). Albeit AM consists of many concepts in the literature, the overarching purpose of AM is to respond adaptively to future changes, being flexible and responsive to current demands, which aligns with the term ‘agility’ (Gunasekaran, 1998; Ramesh and Devadasan, 2007).

(7)

7

company will omit the full potential to exploit these resources. The concept of PPC is underexposed within the concept of ‘agility’.

This research will focus on how such adaptations of PPC decisions to a changing production mix can be realised within a practical context based on the full capabilities of a planning system. This leads to a subset of two research questions. The first research question that will be addressed in this research is posed as follows:

“How does a changing production mix in an MTO company affect its delivery and throughput performance?”

The second concept of interest in this research is to identify how PPC decisions can be adjusted to reduce the negative performance impact of a changing production mix. This leads to the second research question:

“How can adjustments of PPC decisions reduce the performance impact of production mix changes?”

A case study will provide a contribution to the literature in terms of getting a better understanding and insights of an MTO company who is adapting its PPC approach in reality. This will contribute to the concepts of AM and PPC.

To answer these questions, it is necessary to identify where in the past a changing production mix has affected delivery and throughput performance. A diagnosis was conducted based on historical order progress data. Soepenberg et al. (2012) developed a framework to diagnose delivery performance particularly for MTO companies. Parts of this framework are used to perform the diagnosis to reveal the performance impact of a changing production mix. This knowledge serves as input to determine where in their planning system MTO companies can accommodate adjustments to PPC decisions to reduce the performance impact. The adjustments of PPC decisions in this research relate to the concept of Workload Control (WLC). The concept of WLC is considered as one the most appropriate PPC concepts for MTO companies (Kingsman and Hendry, 2002; Stevenson et al., 2005).

(8)

8 2. Theoretical background

In this section, the relevant literature on AM and PPC in relation to this research has been reviewed. The first sub-section discusses the concept of AM, followed by a review of relevant theory on PPC in the context of MTO companies. In the last sub-section, the concept of PPC is discussed within the context of AM followed by the identification of the relevant research gap.

2.1 Agile Manufacturing

The aim of AM is to present customers a solution tailored to the customer’s needs, rather than just a product (Gunasekaran, 1998). A lot of research has been conducted on the concept AM and many definitions are enunciated, but the definitions of different scholars have as a commonality that AM is seen as the capability of the manufacturing enterprise to quickly respond to changing market requirements (Ramesh and Devadasan, 2007). It is imperative for companies who want to become agile to adopt agility enablers. These enablers pursue the integration of three basic elements of the company: (1) people, (2) technology and (3) organisation (Vernadat, 1999). AM will result from the integration of these three elements (Goldman and Nagel, 1993). Vázquez-Bustelo et al. (2007) identified several groups of agility enablers based on a review of the literature on AM: agile human resources, agile technologies, internal and external organisation (integration of the value chain), concurrent engineering and knowledge management and learning. These agility enablers are mainly driven by unpredictable changes in the environment (dynamism) and competition (Vázquez-Bustelo et al., 2007). Nabass and Abdallah (2019) explained how the concept of AM positively affects delivery performances. They used survey data collected from 282 manufacturing companies. All companies who adopted the concepts of AM identified by Vázquez-Bustelo et al. (2007), significantly improved their delivery performance on three different aspects (1) on-time delivery, (2) deliver faster than competitors and (3) delivery speed. This aligns with other studies that proved concepts of AM positively influences delivery performances (e.g., Vázquez-Bustelo et al. (2007) and Hallgren and Olhager (2009)).

(9)

9 2.2 Production Planning and Control and Performance in the MTO context.

In order to investigate how PPC should be adapted in response to changing circumstances, it is necessary to determine which aspects of PPC are relevant and adaptable to improve performances for MTO companies.

MTO companies are generally operating in an environment that is less predictable than the environment of a Make-To-Stock (MTS) company (Stevenson et al., 2005). This is caused by the degree of repetitive business. The MTS sector has a high degree of standardization, which makes it easier to apply a universalistic PPC approach in this sector. In terms of Olhager (2003), MTO companies have an early Order Penetration Point (OPP), which is the point in the manufacturing value chain where the product is linked to a customer order. Customization is a term that can be linked to an early OPP and this often leads to a high degree of variability on the shop-floor. Therefore, it is considered as an important determinant for the PPC approach. Amaro et al. (1999) made a distinction between two types of MTO companies based on the degree of customization (1) Repeat Business Customisers (RBC) and (2) Versatile Manufacturing Companies (VMC). The former relies on a repetitive base, the products are unique, but ordered more than once. The latter relies on a high degree of variability, low volume batches and few repeats. This high degree of customization makes it hard to predict demand, which makes delivery reliability one of the most important performance indicators for MTO companies (Stevenson et al., 2005; Soepenberg et al., 2008).

(10)

10

capacity groups where congestion emerges. Output decisions are often made at the same time as input control decisions (Land and Gaalman, 1996). An overview of these input and output control decisions is depicted in Figure 1.

Figure 1: Input and output control decisions (Soepenberg et al., 2008).

(11)

11

Table 1: The effect of PPC decisions on the average lateness and variance of lateness (Soepenberg et al., 2012).

2.3 Production Planning and Control in the context of Agile Manufacturing

(12)

12 3. Methodology

This section explains how this research was designed, performed and how this design was suitable to answer the two research questions. In addition, a motivation of the selection of the case company is given. The final two sub-sections explain how the required data was collected and analysed, respectively.

3.1 Research design

The aim of this research is to generate new knowledge on the concepts of PPC and AM. As the literature assumes, it is vital to adopt agility enablers in order to respond better to changing circumstances in the environment, especially for MTO companies. The role of PPC is underexposed within the concept of AM. This research contributes to the literature by bridging the gap between these two concepts. In addition, the study provides a more practical extension to the study of Land et al. (2015), who looked at the effect of adjusting PPC decisions to changing circumstances by means of simulation. In order to investigate this topic, the research aims to answer the following related research questions:

1) How does a changing production mix in an MTO company affect its throughput performance? 2) How can adjustments of PPC decisions reduce the performance impact of production mix

changes?

Both questions can be considered as questions guiding design. The former question is answered by means of a diagnosis based on historical order progress data. The findings of the diagnosis serve as input for the latter research question to develop a generic design outline. The contribution of this research will be twofold: (1) a solution-oriented design for a practical problem and (2) using the findings of the diagnosis to indicate where the theory on PPC could have provided more clarity (i.e., new theory development).

(13)

13

to obtain a profound understanding of a real-life phenomenon that can be observed in the real-life case (Yin, 2017). A case study makes the findings more realistic and will give a proper indication of how other MTO companies would probably operate (i.e., increases generalisability).

This research consists of two different phases. In the first phase, an in-depth diagnosis based on historical order progress was conducted to examine where in the past there would have been opportunities to realise better delivery and throughput performances under a changing production mix. In the second phase, the design phase, the capabilities within the planning system of the case company were identified in order to learn what kind of adjustments can be accommodated in the PPC decisions to improve the delivery and throughput performances.

3.1.1 Selection of the case company

The company that suits the purpose of this study is an MTO company that is active in the metal industry. The company produces high quality (bent) metal sheets for a wide variety of applications. Their main focus is on the production of kits to build the hulls of large ships. These kits could be customised to the wishes and requirements of the client. The building kits will ensure a shorter construction time for the client. The company continuously wants to improve shipbuilding efficiency. The processes that are involved in the production process are engineering (involving the customer), cutting, forming, sorting, distribution and delivery (a full description of this process is enclosed in Appendix A). Currently, the company is developing new tailor-made planning software. One of the requirements for this software is to better respond to changing circumstances, as the company had problems with this in the past. The findings of this research should serve as direct input for the new planning software.

(14)

14 3.2 Data collection

According to Yin (2017) case study evidence comes from different sources. The sources that were used for the purpose of this research were documentation, meetings and direct observations. Quantitative data documented in the company’s databases were required to perform the diagnosis. The diagnosis is mainly quantitative and was validated by means of meetings and discussions. Data needed to be obtained from different departments. In order to perform a proper diagnosis, the following data were obtained for each order in the period from 01/01/2017 up to 31/08/2018:

● The pre-calculated production time in hours for each operation of the order. ● Dates of order acceptance, release to the shop floor and operation completion. ● Data of the promised and realised delivery date.

The data that have been used goes three years and eight months back in time, which allowed for the identification of changing circumstances. These quantitative data were extracted for 1022 orders from the company’s Enterprise Resource Planning (ERP) system and served as the main data source for the diagnosis. Validation meetings with domain experts were held to clarify the quantitative data, thereby strengthening construct validity. For 105 out of 1022 orders, operation completion data was missing. For these orders, the average throughput time of the 30 days prior to that period was used. For these orders, the results are neater than in reality. Furthermore, archived research documents of former studies at the company were used, and direct observations were made by attending meetings with regard to the production planning of the company. People from different departments attended these meetings, several aspects regarding planning and orders were discussed during these meetings. These meetings and archived documents helped to obtain a proper understanding of the whole production process and to identify information about the possibilities within the PPC system of the company.

3.3 Data analysis

(15)

15 3.3.1 Diagnostic phase

At first, an in-depth diagnosis based on the historical order progress data was conducted to inventory what kind of agility needs there were in the past in terms of PPC. The diagnosis aided to reveal where in the past there would have been opportunities to realise a better delivery and throughput performance under a changing mix in production hours for different operations. The focus is based on the two operations of the company’s production process that are considered as the most significant operations by the case company. These two operations are cutting and forming. Within the production process, the metal plates first need cutting followed by forming. The bulk of the work is in these two operations, which is why they are considered as the bottlenecks. There are therefore many PPC related opportunities within this area of focus. The performance aspects of this research are delivery performance and throughput performance.

The diagnosis is guided by parts of the framework developed by Soepenberg et al. (2012). This framework is specially developed to diagnose the delivery reliability performance of an MTO company. The framework of Soepenberg et al. (2012) is separated into two parts. The first part consists of four steps to determine to focus the diagnosis average-oriented or variance-oriented. As discussed in section 2.2, TDs are particularly used for average-oriented diagnoses and order progress diagrams for variance-oriented diagnoses. In other words, the first part of the framework is used to determine which diagnosis tool to use. Order progress diagrams aid to explain the variance of lateness by relating the lateness of an individual order to the progress of this individual order (Soepenberg et al., 2008). As the aim of the research is on diagnosing the average delivery and throughput performance of the case company, it justifies the decision to use TDs as diagnosis tool. Focusing on reducing the variability of individual orders by means of order progress diagrams could lead to even more late deliveries when the average delivery dates are tight. In addition, using the order progress diagram would have made it difficult to identify a changing production mix, as the focus will be on individual orders. In contrast, TDs will give a broader perspective of the whole process, which makes it easier to identify change over time.

3.3.2 Detailed explanation of the Throughput Diagrams

(16)

(17)

17

Figure 2: Example of a stylized throughput diagram based on planned cutting hours

Figure 3: Example of a stylized throughput diagram based on planned forming hours

0 500 1000 1500 2000 2500 3000 01/0 1/20 17 06/0 1/20 17 11/0 1/20 17 16/0 1/20 17 21/0 1/20 17 26/0 1/20 17 31/0 1/20 17 05/0 2/20 17 10/0 2/20 17 15/0 2/20 17 20/0 2/20 17 25/0 2/20 17 02/0 3/20 17 07/0 3/20 17 12/0 3/20 17 17/0 3/20 17 22/0 3/20 17 27/0 3/20 17 H o u rs Date

Example of TD cutting operation

Accepted Released Completion first part Completion last part Best case delivery Worst case delivery Promised delivery

0 500 1000 1500 2000 2500 3000 01/0 1/20 17 06/0 1/20 17 11/0 1/20 17 16/0 1/20 17 21/0 1/20 17 26/0 1/20 17 31/0 1/20 17 05/0 2/20 17 10/0 2/20 17 15/0 2/20 17 20/0 2/20 17 25/0 2/20 17 02/0 3/20 17 07/0 3/20 17 12/0 3/20 17 17/0 3/20 17 22/0 3/20 17 27/0 3/20 17 H o u rs Date

Example of TD forming operation

(18)

18

As the graphs depict, different curves are incorporated for both of the operations. These curves will now be defined and explained.

The following curves are incorporated for the cutting operation:

1. Acceptance curve: this curve displays how many cutting hours are accepted until a certain date.

2. Release curve: this curve displays how many cutting hours are released to the shop-floor for production until a certain date.

3. Completion first part curve: this curve displays how many cutting hours have been reported as completed for cutting until a certain date, based on the first part of an order1_.

4. Completion last part curve: this curve displays how many cutting hours have been reported as completed for cutting, based on the last part of an order.

5. Best case delivery curve: this curve displays how many cutting hours would have been delivered until a certain date when all parts of an order would have been delivered at the first registered delivery date related to the order.

6. Worst case delivery curve: this curve displays how many cutting hours would have been delivered until a certain date when all parts of an order would have been delivered at the last registered delivery date related to the order.

7. Promised delivery curve: this curve displays how many cutting hours have been promised until a certain date. This curve is based on the agreed delivery date of the orders.

The following curves are incorporated for the forming operation:

1. Acceptance curve: this curve displays the number of forming hours accepted until a certain date.

2. Release curve: this curve displays the number of forming hours released to the shop-floor for production until a certain date.

3. Best case availability curve: this curve displays how many forming hours have been reported as completed for cutting until a certain date, based on the first part of an order. The forming hours that have passed cutting are then available for the forming operation to start.

4. Worst case availability curve: this curve displays how many forming hours have been reported as completed for cutting until a certain date, based on the last part of an order. The forming hours that have passed cutting are then available for the forming operation to start.

(19)

19

5. Best case delivery curve: this curve displays how many forming hours would have been delivered until a certain date when all parts of an order would have been delivered at the first registered delivery date related to the order.

6. Worst case delivery curve: this curve displays how many forming hours would have been delivered until a certain date when all parts of an order would have been delivered at the last registered delivery date related to the order.

7. Promised delivery curve: this curve displays how many forming hours have been promised until a certain date. This curve is based on the agreed delivery date of the orders.

The curves represent the cumulative input or output in the number of hours for the way PPC has been executed. For example, when an order is released that needs 50 hours of cutting work on a certain date, the release curve increases with 50 hours on this particular date. When the order completed cutting three days later, the completion curve increases with 50 hours on that certain date. The vertical distance between two curves represents the number of hours between the curves. The vertical distance between the acceptance curve and an output curve represents the total number of hours in operation on the corresponding date, i.e., the Work-In-Process (WIP). The horizontal distance between two curves represents the throughput time of the work for a certain period, a simple example is depicted in Figure 4. When the delivery date sequence is equal to the arrival sequence of an order, the distance reflects the actual throughput time of an order. In practice, the sequence will often change. In that case the horizontal distance between two curves represents the average throughput time, which is still a proper indication of the throughput time of processed orders within that certain period. When the number of promised delivery hours on a particular day is also realized, it does not automatically mean that there are no late deliveries on that day. The curves do not show which orders are exactly delivered and promised at that time. There may therefore be late deliveries, even though the number of promised delivery hours for that day has been met.

(20)

20

Figure 4: Basic explanation of the distances between an input and output curve (Soepenberg et al., 2008).

3.3.3 Interpretation of the availability and completion curves

Evidently, both operations contain different curves. The availability and completion curves display the interaction of the mix in planned production hours for both of the operations and how this mix is handled. Understanding these curves are important to understand the outcomes of the research. The best and worst case availability curves in forming hours can be considered as the input from the cutting operation to forming. These curves only contain orders that need both operations cutting and forming or only require the operation forming. For those orders that require both operations, the curve displays the moment when the cutting operation is completed, and the materials become available for the operation forming. For those orders that only need the operation forming, this moment is the same as the release moment. These orders are immediately available for the operation forming after the release moment. These curves are valuable, because they display how input was provided from the cutting operation to the forming operation. When the curves have a smooth pattern, it means that there is a consistent inflow from the cutting operation to the forming operation. When the curves display irregularities at the same time as the release curve, forming hours become available on an irregular basis. This indicates the process is not in control and can be a major cause of poor performances.

(21)

21 3.3.4 Curves with an extreme

The completion curves at cutting, the availability curves at forming and the best and worst case delivery curves at both cutting and forming show two extreme options. These curves assume that all parts became available or were delivered at the earliest registered time or latest registered time for a single part of the order, respectively. This means that the remaining parts of an order are completed, made available or delivered between the two extremes (e.g., forming operations were started after the best case availability curve and before the worst case availability curve).

3.3.5 Design phase of the research

(22)

22 4. Results

This section will discuss the results that are based on the generated TDs. The TDs are based on the planned production hours of the two most significant operations, cutting and forming. As mentioned before, the TDs make a distinction between the planned cutting hours and the planned forming hours. The TDs of each quarter for each of the operations are enclosed in Appendix B, C, D and E. The mix between these hours has a significant impact on the performances of the company, which makes it interesting to focus on these two operations. The variation over time in the mix of these two operations affect the delivery performances, throughput times and the overall control of the flow within the production process. By getting a proper understanding of the impact of the variance over time in the mix, a translation can be made to how this can be intercepted in future situations.

4.1 Results of the diagnostic phase

A number of particularities have been noted from the TDs. It is revealed that each year contains a different mix pattern, hours are promised on an irregular basis, forming hours are released arbitrarily and the forming hours are becoming available on an irregular basis. These findings will now be addressed in the following sub-sections.

4.1.1 Changing variance of the mix over time

Remarkable is that all years that were examined have a different composition of the mix. This is valuable to identify different situations that should be intercepted in the future. In the first year 2017, both the operations cutting and forming contain a high number of planned hours, which are close to each other. In the year 2018, the number of planned hours for both operations is significantly lower than in the previous year, and there are more forming hours planned than cutting hours. In 2019, the number of hours for both operations were high, but especially the number of forming hours. The hours for each year and operation are depicted in Table 2.

Table 2: Overview of the planned production hours for each operation per year

Year 2017 2018 2019 2020 (up to August)

Cutting hours 7.800 4.200 7.100 5.150

Forming hours 8.450 6.850 10.350 5.050

(23)

(24)

24

0 500 1000 1500 2000 2500 3000 3500 01/0 1/20 17 06/0 1/20 17 11/0 1/20 17 16/0 1/20 17 21/0 1/20 17 26/0 1/20 17 31/0 1/20 17 05/0 2/20 17 10/0 2/20 17 15/0 2/20 17 20/0 2/20 17 25/0 2/20 17 02/0 3/20 17 07/0 3/20 17 12/0 3/20 17 17/0 3/20 17 22/0 3/20 17 27/0 3/20 17 H o u rs Date Forming Q1 2017

Accepted Released Availability best case Availability worst case Best case delivery Worst case delivery Promised delivery

Accepted Released Availability best case Availability worst case Best case delivery Worst case delivery Promised delivery Figure 5: TD of the forming hours of the first quarter of 2017

Figure 6: TD of the forming hours of the first quarter of 2018

Irregular availability of forming hours (green and dark blue curves)

(25)

25

As mentioned before, it is desirable to have a consistent smooth inflow in forming hours from the cutting operation. When the inflow is irregular it could occur that at one moment much of forming work is available and at the other moment not enough work is available. To have a controlled and predictable process a smooth inflow in forming hours is required. In both 2017 and 2019, where there is a lot of cutting and forming work, it is noticeable that there are many irregularities in the availability curves when looking at the TDs of the forming operation, this is visible in Figure 5 highlighted with the red arrows. This is caused by the manner in which orders were released in combination with the cutting sequences used. The company’s process managers state that orders were released when it was technically possible to release them. Orders are released without any guidance from the planning. When cutting is not the majority of the work, the forming work is automatically delivered to the forming operation in a smoother way. This is clearly visible in the availability curves of the forming operation in 2018, which are smoother (highlighted with the red arrows in Figure 6 and visible in Figure 7, for the remaining quarters of 2018, see Appendix C). However, it is still visible that the availability curves slightly follow the pattern of the release curve when there is a small increase (highlighted with the red arrow in Figure 7). This is not a concern during 2018, because the variance in the mix is relatively stable during this year. But it indicates that when there will be significant increase in release, the availability curves will directly follow.

(26)

26

9200 9700 10200 10700 11200 11700 12200 12700 01/0 4/20 18 06/0 4/20 18 11/0 4/20 18 16/0 4/20 18 21/0 4/20 18 26/0 4/20 18 01/0 5/20 18 06/0 5/20 18 11/0 5/20 18 16/0 5/20 18 21/0 5/20 18 26/0 5/20 18 31/0 5/20 18 05/0 6/20 18 10/0 6/20 18 15/0 6/20 18 20/0 6/20 18 25/0 6/20 18 30/0 6/20 18 H o u rs Date Forming Q2 2018

Figure 7: TD of the forming hours of the second quarter of 2018

4.1.2 Irregular promising pattern

The promising pattern is undesirable in some periods. In the ideal case the hours are promised in a constant way, in a straight line from the beginning to the end. In the second quarter of 2019 it can be observed that the promised number of forming hours increases in response to the number of hours accepted. The promised cutting hours remain stable, this is highlighted in Figures 8 and 9. It occurs that the best and worst case curve in forming hours start to diverge further. This implies that it will become more difficult to deliver orders on time and that there will be more delays between the first and last delivery of an order. It will be too tight to deliver the promised number of hours, the company often delivers the forming work within the last delivery, since it is easier to deliver the parts that do not need the forming operation first. As a result, there will be a delay in forming hours due to too tight delivery times.

Slight increase in release directly causes an increase in

(27)

27

12300 13300 14300 15300 16300 17300 18300 01/0 4/20 19 06/0 4/20 19 11/0 4/20 19 16/0 4/20 19 21/0 4/20 19 26/0 4/20 19 01/0 5/20 19 06/0 5/20 19 11/0 5/20 19 16/0 5/20 19 21/0 5/20 19 26/0 5/20 19 31/0 5/20 19 05/0 6/20 19 10/0 6/20 19 15/0 6/20 19 20/0 6/20 19 25/0 6/20 19 30/0 6/20 19 H o u rs Date Cutting steel Q2 2019

16400 17400 18400 19400 20400 21400 22400 01/0 4/20 19 06/0 4/20 19 11/0 4/20 19 16/0 4/20 19 21/0 4/20 19 26/0 4/20 19 01/0 5/20 19 06/0 5/20 19 11/0 5/20 19 16/0 5/20 19 21/0 5/20 19 26/0 5/20 19 31/0 5/20 19 05/0 6/20 19 10/0 6/20 19 15/0 6/20 19 20/0 6/20 19 25/0 6/20 19 30/0 6/20 19 H o u rs Date Forming Q2 2019

Accepted Released Availability best case Availability worst case Best case delivery Worst case delivery Promised delivery Figure 8: TD of the cutting hours of the second quarter of 2019

Increase in accepted cutting hours.

Promised cutting hours remain stable.

Increase in accepted cutting hours.

Increase in promised cutting hours.

(28)

(29)

29

0 500 1000 1500 2000 2500 3000 3500 01/0 1/20 17 06/0 1/20 17 11/0 1/20 17 16/0 1/20 17 21/0 1/20 17 26/0 1/20 17 31/0 1/20 17 05/0 2/20 17 10/0 2/20 17 15/0 2/20 17 20/0 2/20 17 25/0 2/20 17 02/0 3/20 17 07/0 3/20 17 12/0 3/20 17 17/0 3/20 17 22/0 3/20 17 27/0 3/20 17 H o u rs Date Cutting steel Q1 2017

0 500 1000 1500 2000 2500 3000 3500 01/0 1/20 17 06/0 1/20 17 11/0 1/20 17 16/0 1/20 17 21/0 1/20 17 26/0 1/20 17 31/0 1/20 17 05/0 2/20 17 10/0 2/20 17 15/0 2/20 17 20/0 2/20 17 25/0 2/20 17 02/0 3/20 17 07/0 3/20 17 12/0 3/20 17 17/0 3/20 17 22/0 3/20 17 27/0 3/20 17 H o ur s Date Forming Q1 2017

Accepted Released Availability best case Availability worst case Best case delivery Worst case delivery Promised delivery Figure 10: TD of the cutting hours of the first quarter of 2017

Distance between the best and worst case

curve are not diverging yet.

Diverging distance between best case and worst case delivery curve.

Increase in accepted and released forming

(30)

30

1500 2000 2500 3000 3500 4000 4500 5000 5500 01/0 4/20 17 06/0 4/20 17 11/0 4/20 17 16/0 4/20 17 21/0 4/20 17 26/0 4/20 17 01/0 5/20 17 06/0 5/20 17 11/0 5/20 17 16/0 5/20 17 21/0 5/20 17 26/0 5/20 17 31/0 5/20 17 05/0 6/20 17 10/0 6/20 17 15/0 6/20 17 20/0 6/20 17 25/0 6/20 17 30/0 6/20 17 H o u rs Date Cutting steel Q2 2017

1700 2200 2700 3200 3700 4200 4700 5200 5700 01/0 4/20 17 06/0 4/20 17 11/0 4/20 17 16/0 4/20 17 21/0 4/20 17 26/0 4/20 17 01/0 5/20 17 06/0 5/20 17 11/0 5/20 17 16/0 5/20 17 21/0 5/20 17 26/0 5/20 17 31/0 5/20 17 05/0 6/20 17 10/0 6/20 17 15/0 6/20 17 20/0 6/20 17 25/0 6/20 17 30/0 6/20 17 H o u rs Date Forming Q2 2017

Figure 12: TD of the cutting hours of the second quarter of 2017

In Q2 the distance between the best and worst case curve start to

diverge in cutting hours as well.

In Q2 the distance between the best and worst case curve remain

(31)

31 4.1.3 Arbitrary release patterns for the forming operation

(32)

32

16400 17400 18400 19400 20400 21400 22400 01/0 4/20 19 06/0 4/20 19 11/0 4/20 19 16/0 4/20 19 21/0 4/20 19 26/0 4/20 19 01/0 5/20 19 06/0 5/20 19 11/0 5/20 19 16/0 5/20 19 21/0 5/20 19 26/0 5/20 19 31/0 5/20 19 05/0 6/20 19 10/0 6/20 19 15/0 6/20 19 20/0 6/20 19 25/0 6/20 19 30/0 6/20 19 H o u rs Date Forming Q2 2019

Accepted Released Availability best case Availability worst case Best case delivery Worst case delivery Promised delivery 12300 13300 14300 15300 16300 17300 18300 01/0 4/20 19 06/0 4/20 19 11/0 4/20 19 16/0 4/20 19 21/0 4/20 19 26/0 4/20 19 01/0 5/20 19 06/0 5/20 19 11/0 5/20 19 16/0 5/20 19 21/0 5/20 19 26/0 5/20 19 31/0 5/20 19 05/0 6/20 19 10/0 6/20 19 15/0 6/20 19 20/0 6/20 19 25/0 6/20 19 30/0 6/20 19 H o u rs Date Cutting steel Q2 2019

Accepted Released Completion first part Completion last part Best case delivery Worst case delivery Promised delivery Figure 14: TD of the cutting hours of the second quarter of 2019

Release used as stabilizing factor to avoid fluctuations on the shop-floor. The release curve remains smooth, despite an increase

in accepted hours.

Release pattern is arbitrary. An increase in accepted hours

occurred and this is followed by an increase in released

(33)

33

17800 18800 19800 20800 21800 22800 01/0 1/20 20 06/0 1/20 20 11/0 1/20 20 16/0 1/20 20 21/0 1/20 20 26/0 1/20 20 31/0 1/20 20 05/0 2/20 20 10/0 2/20 20 15/0 2/20 20 20/0 2/20 20 25/0 2/20 20 01/0 3/20 20 06/0 3/20 20 11/0 3/20 20 16/0 3/20 20 21/0 3/20 20 26/0 3/20 20 31/0 3/20 20 H o u rs Date Cutting steel Q1 2020

23700 24700 25700 26700 27700 28700 01/0 1/20 20 06/0 1/20 20 11/0 1/20 20 16/0 1/20 20 21/0 1/20 20 26/0 1/20 20 31/0 1/20 20 05/0 2/20 20 10/0 2/20 20 15/0 2/20 20 20/0 2/20 20 25/0 2/20 20 01/0 3/20 20 06/0 3/20 20 11/0 3/20 20 16/0 3/20 20 21/0 3/20 20 26/0 3/20 20 31/0 3/20 20 H o u rs Date

Forming Q1 2020

Accepted Released Availability best case Availability worst case Best case delivery Worst case delivery Promised delivery Figure 16: TD of the cutting hours of the first of 2020

Figure 17: TD of the forming hours of the first quarter of 2020

Smooth acceptance and release pattern.

Release pattern is arbitrary. An increase in accepted hours

occurred and this is followed by an increase in released

(34)

34 4.1.4 Irregular availability of forming hours

As previously mentioned, the availability curves are important curves to look at. If the curves correspond to the release pattern, it implies that no attention is given to the number of forming hours that an order contains when released. When the number of forming hours is stable in proportion to the number of cutting hours, the release decision plays a less important role. The release curves in cutting and forming hours will then be smooth anyway (discussed in sub-section 4.1.1), which means that the availability curves at the forming operation will automatically be smooth as well. This would imply that a constant amount of forming work would get available from the cutting operation.

(35)

35

12300 12800 13300 13800 14300 14800 15300 15800 16300 16800 17300 17800 01/0 4/20 19 06/0 4/20 19 11/0 4/20 19 16/0 4/20 19 21/0 4/20 19 26/0 4/20 19 01/0 5/20 19 06/0 5/20 19 11/0 5/20 19 16/0 5/20 19 21/0 5/20 19 26/0 5/20 19 31/0 5/20 19 05/0 6/20 19 10/0 6/20 19 15/0 6/20 19 20/0 6/20 19 25/0 6/20 19 30/0 6/20 19 H o u rs Date Cutting steel Q2 2019

16400 16900 17400 17900 18400 18900 19400 19900 20400 20900 21400 21900 01/0 4/20 19 06/0 4/20 19 11/0 4/20 19 16/0 4/20 19 21/0 4/20 19 26/0 4/20 19 01/0 5/20 19 06/0 5/20 19 11/0 5/20 19 16/0 5/20 19 21/0 5/20 19 26/0 5/20 19 31/0 5/20 19 05/0 6/20 19 10/0 6/20 19 15/0 6/20 19 20/0 6/20 19 25/0 6/20 19 30/0 6/20 19 H o u rs Date Forming Q2 2019

Accepted Released Availability best case Availability worst case Best case delivery Worst case delivery Promised delivery Figure 18: TD of the cutting hours of the second of 2019

Figure 19: TD of the cutting hours of the second of 2019

Completion curves partly follow the release pattern.

They still look smooth, as release is used as a stabilizing factor at the

cutting operation.

Availability curves follow the release pattern. The forming hours are coming available in

(36)

36 5. Discussion and design implications

This section discusses how the findings of the diagnosis contribute to both the literature on PPC and AM. A reflection on the diagnosis findings is given to understand the new insights that can be added to the theory on PPC and AM. In the subsequent sub-section these findings are translated into PPC design implications for MTO companies in general. The final sub-section discusses the limitations of the researches and future research recommendations.

5.1 Reflection on diagnosis findings

The findings of the diagnosis are primarily useful to answer the first research question of this study. The findings of the diagnosis clearly demonstrate how a changing production mix negatively affects the case company's delivery and throughput performance. As stated by Soepenberg et al. (2012) the TDs help to determine in which problem areas (DP or RP) the adjustments to PPC decisions can be made, to realise better delivery and throughput performances. The TDs revealed that the case company could have achieved better performances within both problem areas. In addition, the TDs also aid to link the performance within a certain period to the PPC decisions made. The impact of these PPC decisions within the DP and RP of the case company is addressed in the next two sub-sections, respectively.

5.1.1 Improvement potential at the delivery time promising process

The diagnosis at the case company revealed that creating the right delivery promising pattern can be considered as an important PPC decision to still guarantee a desirable delivery performance in the event of a changing production mix. The case company sets a delivery date in coordination with the customer shortly after an order is accepted (for a detailed explanation see Appendix A). This deviates from the literature, where it is assumed that the delivery date is set in the customer enquiry stage, before an order is accepted. For this reason, the acceptance decision is incorporated in the DP within the design implications.

(37)

37

acceptance decision is already an important steering moment when it comes to changing mix situations. The case company has the tendency to promise more hours as a response to the high number of accepted hours. And as the TDs display, it often results in the promised delivery curve surpassing the best and worst case delivery curves for both operations, which is a proper indication of poor delivery performance. This is a problem that emerges at an early stage as a result of disregarding the changing production mix at the acceptance stage. This finding demonstrates that the literature on PPC should incorporate decision criteria related to specific mix situations.

5.1.2 Improvement potential at the realisation process

By means of the TDs it was possible to link the causes of poor throughput performances to the release decision within the RP. The case company has problems to cope with the variance of the mix over time. The main finding here is that the case company is experiencing difficulties to smoothen the flow of work through the factory when the mix changes. It is primarily difficult to provide a smooth inflow for the downstream operation, which depends on the outflow of the first operation. This does not cause problems at the first operation, as almost every order must undergo this operation. The completion and availability curves follow the pattern of the release curve in almost all of the TDs. The release and completion curves in cutting hours are relatively smooth. This is due to the fact that the vast majority of the orders need this operation. The release and availability curves in forming hours are rather irregular. From this it can be concluded that the mix of production is disregarded at the release decision. This is not a threat to the first operation itself, but it is a concern for the downstream operation. In theory, this could lead to two phenomena at the downstream operation: (1) needless overloading or (2) needless underloading. The former will mainly affect delivery performances and the latter will affect productivity. As the diagnosis has demonstrated, in practice overloading is more likely. When this occurs, the downstream operation has to decide which orders to start up, which could lead to the wrong decisions in practice. Additionally, the variable inflow at the downstream operation causes unpredictability of the throughput times. Predictable throughput times are desirable to coordinate proper delivery dates and possible outsource decisions. The theory on PPC is lacking to incorporate aspects of agility to better respond to changing circumstances.

(38)

38 5.2 Design implications for PPC in MTO companies in general

The findings of the diagnosis can be considered as useful for the case company, but also for other MTO companies. Other MTO companies are likely encountering similar types of mix problems. To reduce the performance impact of a changing production mix, a generic design is developed. The design can be considered as the first step into the development of new theory in this area of research and as an answer to the second research question. The design implications focus on adjustments to PPC decisions to reduce the performance impact in both the DP and the RP. The next sub-sections discuss the improvement potential of the DP and on the improvement potential of the RP, respectively.

5.2.1 Generic design outline focused on the delivery time promising process

In this sub-section a design outline is described to reduce the performance impact of accepting an excessive amount work for a certain operation. To reduce the negative impact, MTO companies should constantly monitor the number of production hours accepted for all of the operations at any moment in time. Each operation has its own predetermined threshold for the number of production hours it could accept in a particular period. When the number of accepted production hours surpasses the threshold, the staff responsible for the acceptance of an order receives a signal. The signal indicates an excessive amount of work is accepted for a particular operation. The signal serves as input for a follow-up decision on the acceptance decision. When too much work is accepted for a particular operation, it becomes more difficult to deliver orders on time. Three different follow-up decisions can be made:

1. Outsourcing orders: it is important to make outsourcing decisions timely to relieve the shop-floor from too work. The outsourced order will presumably be late, as outsourcing an order is more time consuming for that particular order. But by outsourcing that particular order, space for other orders is created to be on time.

2. Adapting capacity: the capacity decision can ensure to increase capacity to the operations where congestion is expected. Considering capacity adjustment at an early stage can ensure a proper utilization to smoothen the flow.

3. Setting looser delivery dates: looser delivery dates can be set in coordination with the customer to smoothen the planned outflow.

(39)

39 5.2.2 Generic design outline focused on the realisation process

In this sub-section two approaches are outlined that should generate a smooth flow between two subsequent operations. The generic design outline focused on the RP entails the following two approaches:

1. The switching policy: the first operation can ensure a smooth inflow to the downstream operation by using the appropriate sequences of work while keeping the hours of the downstream operation in consideration. At the release stage, the amount of work is released when it is technically possible. Proper sequences will ensure that the variable input will be transformed into smooth output. When the mix is stable, the first operation can optimise their own process without taking hours of the downstream operation into account. When the number of hours for the downstream operation surpasses a predetermined threshold, then the first operation receives a signal. After the signal, the first operation should consider the hours of the downstream operations by determining the right sequences. When the number of hours for the downstream operation returns below the threshold, the policy switches back to the former situation. There will be a switch in focus based on the changing production mix. 2. Controlled release: decide very consciously which orders should be released from the pool. Each operation has its own standard norm for the number of production hours it should process. When an order arrives at the pool that contains an amount of production hours that exceeds the norm for either one of the operations, the order is not allowed to be released. Releasing an order that exceeds the norm causes overloading and releasing an insufficient number of hours causes underloading. At release, it can be determined whether an order fits the norm of the operations. This can then be a decision moment for whether or not to release an order. When this principle is used there is no need to take into account sequences at the first operation. The orders that fit within the predetermined norm are released at the start of the first operation to maintain an equal smooth inflow to the downstream operation at any moment in time. This policy is based on principles of the WLC concept as discussed by Land (2004).

(40)

40 5.3 Limitations and future research

(41)

41 6. Conclusion

This research aimed to provide a contribution to both the literature on PPC and AM. An in-depth diagnosis based on historical order progress data at a case company was conducted to inventory what kind of agility needs there were in the past in terms of PPC. The findings of the diagnosis close the gap between PPC and AM by stressing the importance of PPC decisions to react better to changing circumstances within the environment. These findings can be considered as a practical extension to the work of Land et al. (2015) who only looked at a limited set of PPC decisions by means of simulation. By collecting a wide range of empirical data at the case company, it was possible to focus the diagnosis on all significant PPC decisions that are stated in the literature by Soepenberg et al. (2012). In addition, the diagnosis focused specifically on a changings production mix, as this aspect is underexposed within the literature of PPC. From the perspective of the literature on AM, the concept of PPC can be considered as a new 'agility enabler', as proper PPC decisions enable to better respond to changing circumstances in the environment. In addition, the findings serve as input for the first steps of the development of a generic PPC design to reduce the impact of a changing production mix on delivery and throughput performance. Moreover, it provides a twofold contribution to both the practical context of MTO companies and the gap in the literature between AM and PPC. The following research questions that have been formed to achieve this can now be answered:

1) How does a changing production mix in an MTO company affect their throughput and delivery performance?

2) How can adjustments of PPC decisions be made to reduce the performance impact of production mix changes?

The diagnosis was primarily useful to answer the first research question. It appeared to be important to steer the throughput of orders as early as the acceptance stage. Disregarding the changing production mix at the acceptance stage negatively affects the delivery and throughput performance. In addition, at the release stage the number of production hours of all operations need to be considered when an order will be released. This is an important steering moment to generate a smooth flow between the subsequent operations. It also increases the predictability of when an order will leave the factory and the throughput times themselves. These findings served as input for the development of a generic PPC design and simultaneously answering the second research question.

(42)

42

accepted for each operation. From this, three follow-up decision can be made based on outsourcing, capacity or setting looser delivery dates. These follow-up decisions help to steer the throughput of orders at the earliest stage possible and improve delivery performance.

In order to generate a smooth flow between two subsequent operations, the design suggests that there are two approaches to achieve this. The first approach is a focus switching policy, that uses particular sequences at the first operation when the amount of production hours of the downstream operation surpasses a particular threshold. The second approach is based on elements of the WLC concept and focuses on controlled release with a norm for each operation. It is expected that both approaches will create a smooth predictable flow between the two subsequent operations.

(43)

43 7. References

Amaro, G., Hendry, L. and Kingsman, B. (1999) 'Competitive advantage, customisation and a new taxonomy for nonmake-to-stock companies', International Journal of Operations & Production Management, 19(4), pp. 349-371.

Goldman, S.L. and Nagel, R.N. (1993) 'Management, technology and agility: the emergence of a new era in manufacturing', International Journal of Technology Management, 8(1-2), pp. 18-38.

Gunasekaran, A. (1998) 'Agile manufacturing: enablers and an implementation framework', international journal of production research, 36(5), pp. 1223-1247.

Hallgren, M. and Olhager, J. (2009) 'Lean and agile manufacturing: external and internal drivers and performance outcomes', International Journal of Operations & Production Management, 29(10), pp. 976-999.

Hopp, W.J. (2011) Supply Chain Science. Long Grove, Ill.: Waveland Press.

Kingsman, B. and Hendry, L. (2002) 'The relative contributions of input and output controls on the performance of a workload control system in make-to-order companies', Production planning & control, 13(7), pp. 579-590.

Kingsman, B.G. (2000) 'Modelling input–output workload control for dynamic capacity planning in production planning systems', International journal of production economics, 68(1), pp. 73-93. Land, M.J. (2004) 'Workload control in job shops, grasping the tap'.

Land, M.J. and Gaalman, G. (1996) 'Workload control concepts in job shops a critical assessment', International journal of production economics, 46(47), pp. 535-548.

Land, M.J., Stevenson, M., Thürer, M. and Gaalman, G.J. (2015) 'Job shop control: In search of the key to delivery improvements', International Journal of Production Economics, 168, pp. 257-266.

Nabass, E.H. and Abdallah, A.B. (2019) 'Agile manufacturing and business performance: The indirect effects of operational performance dimensions', Business Process Management Journal, 25(4), pp. 647-666.

Olhager, J. (2003) 'Strategic positioning of the order penetration point', International journal of production economics, 85(3), pp. 319-329.

Ramesh, G. and Devadasan, S. (2007) 'Literature review on the agile manufacturing criteria', Journal of Manufacturing Technology Management, 18(2), pp. 182-201.

Slack, N., Chambers, S. and Johnston, R. (2010) Operations management. Pearson education. Soepenberg, G., Land, M. and Gaalman, G. (2008) 'The order progress diagram: A supportive tool for diagnosing delivery reliability performance in make-to-order companies', International Journal of Production Economics, 112(1), pp. 495-503.

(44)

44

Stevenson, M., Hendry, L.C. and Kingsman, B.G. (2005) 'A review of production planning and control: the applicability of key concepts to the make-to-order industry', International journal of production research, 43(5), pp. 869-898.

Thürer, M., Land, M.J., Stevenson, M., Fredendall, L.D. and Godinho Filho, M. (2015) 'Concerning Workload Control and Order Release: The Pre-Shop Pool Sequencing Decision', Production and Operations Management, 24(7), pp. 1179-1192.

Thürer, M., Stevenson, M., Land, M.J. and Fredendall, L.D. (2019) 'On the combined effect of due date setting, order release, and output control: an assessment by simulation', International Journal of Production Research, 57(6), pp. 1741-1755.

Vázquez-Bustelo, D., Avella, L.a. and Fernández, E. (2007) 'Agility drivers, enablers and outcomes: Empirical test of an integrated agile manufacturing model', International Journal of Operations & Production Management, 27(12), pp. 1303-1332.

Vernadat, F.B. (1999) 'Research agenda for agile manufacturing', International Journal of Agile Management Systems, 1(1), pp. 37-40.

(45)

45 8. Appendices

Appendix A – Process description of the case company

In this appendix a description of the case company’s delivery-time-promising process and the realisation process is given.

Delivery-time-promising process (DP)

This is the process where the company set the promised delivery time in coordination with the customer. The sales department accepts a project when the order is ± 80% certain, the company then reserves reserve capacity for production in their planning. When the project is 100% certain, the remaining capacity will be reserved. The capacity will be reserved based on a pre-calculation of planned production hours for each operation. The pre-calculation will be made after the order is accepted (starts when the order is 80% certain). This pre-calculation is based on experience and previous projects. When the quality requirements are high, the pre-calculation often deviates significantly from the post-calculation. For repeat orders, pre-calculations can be made more accurate. The company often accepts every customer enquiry, they have a high strike rate.

Realisation process (RP)

The process where the promised delivery time needs to be achieved. This process can be subdivided into two sub-processes, the release decision separates these two sub-processes. They are:

1. Pre-shop-floor processes: encompasses work preparation, nesting and purchasing the customer-specific materials. Orders will be released to the shop floor when the materials are received and when capacity is available on the shop-floor. At this moment an order will be released whenever it is technically possible to release the order to the shop-floor.

(46)

46

available on the shop-floor. When this occurs, the plates will first be cut and after this operation the parts will be outsourced. The remaining operations are incorporated as ‘other operations’, this is because these operations have no significant impact on the controllability of the production process. After the operations are completed, the parts are sorted and delivered to the customer. While planning is done for full orders, orders are delivered to customers as partial orders. Therefore, there are multiple delivery moments for an order. An overview of the process is depicted in Figure A.

(47)

47

0 1000 2000 3000 4000 01/0 1/20 17 06/0 1/20 17 11/0 1/20 17 16/0 1/20 17 21/0 1/20 17 26/0 1/20 17 31/0 1/20 17 05/0 2/20 17 10/0 2/20 17 15/0 2/20 17 20/0 2/20 17 25/0 2/20 17 02/0 3/20 17 07/0 3/20 17 12/0 3/20 17 17/0 3/20 17 22/0 3/20 17 27/0 3/20 17 H o u rs Date Cutting steel Q1 2017

0 1000 2000 3000 4000 01/0 1/20 17 06/0 1/20 17 11/0 1/20 17 16/0 1/20 17 21/0 1/20 17 26/0 1/20 17 31/0 1/20 17 05/0 2/20 17 10/0 2/20 17 15/0 2/20 17 20/0 2/20 17 25/0 2/20 17 02/0 3/20 17 07/0 3/20 17 12/0 3/20 17 17/0 3/20 17 22/0 3/20 17 27/0 3/20 17 H o u rs Date Forming Q1 2017

(48)

48

1500 2500 3500 4500 5500 01/0 4/20 17 06/0 4/20 17 11/0 4/20 17 16/0 4/20 17 21/0 4/20 17 26/0 4/20 17 01/0 5/20 17 06/0 5/20 17 11/0 5/20 17 16/0 5/20 17 21/0 5/20 17 26/0 5/20 17 31/0 5/20 17 05/0 6/20 17 10/0 6/20 17 15/0 6/20 17 20/0 6/20 17 25/0 6/20 17 30/0 6/20 17 H o u rs Date Cutting steel Q2 2017

1700 2700 3700 4700 5700 01/0 4/20 17 06/0 4/20 17 11/0 4/20 17 16/0 4/20 17 21/0 4/20 17 26/0 4/20 17 01/0 5/20 17 06/0 5/20 17 11/0 5/20 17 16/0 5/20 17 21/0 5/20 17 26/0 5/20 17 31/0 5/20 17 05/0 6/20 17 10/0 6/20 17 15/0 6/20 17 20/0 6/20 17 25/0 6/20 17 30/0 6/20 17 H o u rs Date Forming Q2 2017

(49)

49

(50)

50

(51)

51

6800 7300 7800 8300 8800 9300 9800 10300 01/0 1/20 18 06/0 1/20 18 11/0 1/20 18 16/0 1/20 18 21/0 1/20 18 26/0 1/20 18 31/0 1/20 18 05/0 2/20 18 10/0 2/20 18 15/0 2/20 18 20/0 2/20 18 25/0 2/20 18 02/0 3/20 18 07/0 3/20 18 12/0 3/20 18 17/0 3/20 18 22/0 3/20 18 27/0 3/20 18 H o u rs Date Cutting steel Q1 2018

(52)

52

8100 8600 9100 9600 10100 10600 11100 11600 01/0 4/20 18 06/0 4/20 18 11/0 4/20 18 16/0 4/20 18 21/0 4/20 18 26/0 4/20 18 01/0 5/20 18 06/0 5/20 18 11/0 5/20 18 16/0 5/20 18 21/0 5/20 18 26/0 5/20 18 31/0 5/20 18 05/0 6/20 18 10/0 6/20 18 15/0 6/20 18 20/0 6/20 18 25/0 6/20 18 30/0 6/20 18 H o u rs Date Cutting steel Q2 2018

(53)

53

9400 9900 10400 10900 11400 11900 12400 12900 01/0 7/20 18 06/0 7/20 18 11/0 7/20 18 16/0 7/20 18 21/0 7/20 18 26/0 7/20 18 31/0 7/20 18 05/0 8/20 18 10/0 8/20 18 15/0 8/20 18 20/0 8/20 18 25/0 8/20 18 30/0 8/20 18 04/0 9/20 18 09/0 9/20 18 14/0 9/20 18 19/0 9/20 18 24/0 9/20 18 29/0 9/20 18 H o u rs Date Cutting steel Q3 2018

Supporting agility by Production Planning and Control in Make-To-Order companies A single case study