Page 1 of 61
Assessing the integration of quick response manufacturing considerations in the ERP systems of SMEs
“Manufacturing Critical-path Time mapping is the starting point for improvement”
Dr. Rajan Suri, 2015
Master Thesis
by
Niels Hofman
University of Groningen Faculty of Economics and Business MSc Technology and Operations Management
Newcastle University Business School
MSc Operations and Supply Chain Management ABSTRACT
Small and medium-sized enterprises (SME) are more than ever being forced to compete on a global scale. This requires SMEs to respond quickly to customers’ expectations and thus be responsive.
Lean manufacturing and quick response manufacturing (QRM) as well as the integration of an enterprise resource planning (ERP) system can improve organizational processes and lead to competitive advantage. However, a combination of these two practices is difficult to realise and lacks academic literature. By means of an explorative multiple case study and interviews with experts, this dissertation investigated the integration of Manufacturing critical-path time mapping (MCT), a QRM metric used as an indicator of potential improvement opportunities, in ERP systems.
The results show that this current support of ERP systems for MCT mapping deals with a wide variety between SMEs. A distinction is made between different customer order decoupling points and complexity of companies. In addition, this dissertation shows possibilities to acquire the potential missing support of ERP systems for MCT mapping and what trade-offs companies have to make in order to strengthen their ERP systems with MCT mapping. Guidance for SMEs on how to combine QRM and ERP and how this can increase their competitive advantage has been given. Where this research focused on SMEs in general, further research should go more in depth in specific industries to create an industry specific solution.
Keywords: Enterprise resource planning, Manufacturing critical-path time, Quick response
manufacturing, Lean manufacturing
Page 2 of 61 Master Thesis
N.J. (Niels) Hofman
Paterswoldseweg 77a, 9727 BB Groningen
N.J.Hofman@student.rug.nl / N.J.Hofman@gmail.com Tel: +31 (0)612714709
Student number RUG: S2021269 Student number NUBS: B4061601
DD-MSc. Technology & Operations Management
Supervisors: prof. dr. ir. J. (Jannes) Slomp and prof C. (Chris) Hicks
Universities
University of Groningen, Faculty of Economics and Business Nettelbosje 2, 9747 AE Groningen
Tel: +31(0)50 3633741
Newcastle University Business School
5 Barrack Road, Newcastle Upon Tyne, NE1 4SE
Tel: +44(0)191 2081500
Company
Contact person: Nico Bijl A.B. Software & Consultancy
Groningerweg 13b, 9765 ZJ Paterswolde
Tel: +31 (0)50 3092230
Page 3 of 61 Preface
This dissertation is written to conclude my Dual Degree Master program in Operations Management at the University of Groningen and the Newcastle University Business School. It has been a challenging yet interesting and rewarding journey which has made this dissertation pleasant to work on.
I would like to express my gratitude to my supervisors Dr. Ir. J. Slomp and Prof. C. Hicks for their feedback and constructive criticisms which significantly improved the quality of this thesis.
Furthermore, I would like to thank Nico Bijl by giving me guidance but also the freedom to do my own research, by providing me his wide range of contacts to address case companies and by giving me the opportunity of being an intern at his company to do my research.
Of course, this dissertation would not have been possible without the input from the industry and therefore, I would like to thank the employees of A.B. Software & Consultancy, the HAN and the case company representatives who have been involved in this research by providing me with interesting data, interviews and helpful discussions. In addition, the attended masterclass of Dr. Rajan Suri about MCT mapping was insightful and inspiring and, therefore, dr. Suri was quoted on the title page.
I hope that you will enjoy reading this dissertation and this will give you insight in the combination of
quick response manufacturing and ERP systems.
Page 4 of 61
Table of contents
1 Introduction ... 6
2 Theoretical background ... 9
2.1 Enterprise Resource Planning systems ... 9
2.2 Quick Response Manufacturing ...10
2.3 Small and medium-sized enterprises ...13
2.4 ERP vs. lean/QRM ...13
3 Methodology ...14
3.1 Research design ...15
3.2 Case selection ...16
3.3 Data collection ...16
3.4 Data analysis ...17
3.5 Quality criteria...18
3.6 Case Companies ...18
4 Results ...20
4.1 Necessary data ...21
4.2 Present information ...23
4.2.1 Interview with experts ...23
4.2.2 In-depth study ...27
4.2.3 Additional cases ...31
4.3 Missing data ...39
5 Discussion ...45
5.1 General discussion ...45
5.2 Trade-offs ...46
5.3 Implications for practice ...48
5.4 Implications for theory ...49
5.5 Limitations ...50
6 Conclusion ...50
6.1 Future research ...52
References ...53
Appendix A: Interview protocol ...57
Appendix B: List of interviewees and job description ... 61
Page 5 of 61 List of abbreviations
ATO – Assemble To Order BOM – Bill Of Materials
CODP – Customer Order Decoupling Point ERP – Enterprise Resource Planning RFID – Radio-Frequency Identification FGI – Finished Goods Inventory
FMTS – Focused Market Target Segment MCT – Manufacturing Critical-path Time MES – Manufacturing Execution System MTO – Make To Order
MTS – Make To Stock
QRM – Quick Response Manufacturing SME – Small and Medium-sized Enterprises TBC - Time Based Competition
WIP – Work In Progress
Page 6 of 61
1 Introduction
Manufacturing companies are facing increasing global competition from low-cost countries and have to deal with rising customer expectations (Powell, 2013). This forces them to look for ways to increase their competitiveness in the global marketplace. Implementing an Enterprise resource planning (ERP) system to facilitate decision-making through the integration of business processes and the deployment of lean manufacturing practices are both consistently rated as the most important strategies of achieving competitive advantage (Caroll, 2007). A synergy of both strategies is interesting, but difficult to accomplish since both practices stem from fundamentally different approaches to production. ERP based production systems are predominantly push oriented, thus, based on forecasted demand whereas lean oriented systems are typically using pull-based production, thus, based on actual demand. With the large enterprise ERP market nearly saturated, ERP vendors are focusing on small and medium-sized enterprises by adjusting ERP systems to the specific needs of SMEs (Zach, Munkvold & Olsen, 2014).
The lean practice is designed to preserve a continuous flow of materials in order to be able to deal with changes in demand in a flexible manner by quickly anticipating to the customers’ need. An important aspect of this flow is just-in-time (JIT) production in which, with the use of different techniques, all forms of waste are minimized thus leading to a higher productivity, efficiency, quality and a reduction in costs (Bhuiyan & Baghel, 2005). A key element of lean is having the mindset to continuously improve processes. ERP systems are online interactive business systems that support the integration of processes from different functional areas with the use of a common database to improve the workflow, standard practices and access real-time data (Mabert, Soni & Venkataramanan, 2003).
Time-based competition (TBC) which can be seen as “The recognition and careful management of time as a limited resource while continuously removing waste or non-value-added activities” (Sim &
Curatola, 1999, p. 659) has laid the foundation of quick response manufacturing (QRM). QRM is defined as “A company-wide strategy for reducing lead times throughout the enterprise” (Suri, 2010, p. 2) and focuses on manufacturing companies rather than the approach of TBC which is quite general.
Results have shown that it is possible for companies to achieve a 94% lead time reduction and 25%
cost reduction by applying a QRM strategy (Tubino & Suri 2000). QRM shares many similarities with
lean manufacturing and can be seen as a lean enhancement (Powell & Strandhagen, 2012). However,
Page 7 of 61
where lean manufacturing originates from high volume, repetitive production with stable demand, QRM focuses on low-volume with high variability. The focus of QRM suits the rising customer expectations and, therefore, the use of QRM becomes increasingly interesting.
ERP systems have been criticized and classified as sources of waste by lean practitioners due to slow production and large inventories (Bell, 2006). However, several authors tend to think differently due to developments in IT and the enhanced capabilities and configurations of ERP systems (Riezebos, Klingenberg & Hicks, 2009; Cunningham & Jones, 2007). Powell (2013) also stressed the fact that it is interesting to investigate how ERP systems can support lean production principles in order to move closer to the “ERP-enabled lean” paradigm.
Even though several early proponents of lean argue against the use of IT-systems, Powell, Alfnes, Strandhagen & Dreyer (2013) placed the ERP implementation process as an essential part of the lean implementation process and they suggested that future perspectives of lean manufacturing should consider using the ERP system as one of the tools in the lean toolbox. Moreover, the area of lean enabled ERP software has been indicated as a subject of numerous recent and important developments (Cruz-Cunha, 2009). However, very few researchers discuss the availability of lean enabled ERP systems in academic literature (Halgeri, Pei, Iyer, Bishop & Shehadeh, 2008).
Looking for ways to improve the current processes of companies is essential in order to stay competitive. Suri (2010) developed a metric that clarifies the focus of the QRM strategy which is called Manufacturing critical-path time (MCT) and it is defined as “The typical amount of calendar time from when a customer creates an order, through the critical path, until the first piece of that order is delivered to the customer” (Suri, 2010, p. 12). It distinguishes itself from other lead time calculations by differentiating between touch and non-touch time, extending it to office tasks and considering the influence of other jobs. This metric can be visualised in an MCT map where different lead times of processes a product follows can be seen with the purpose of identifying areas where improvement efforts should be aimed at.
Setting up an MCT map is time-consuming and sometimes difficult to realize since often data is missing
to perform analyses where to focus improvement efforts on. Even though ERP systems offer an
enormous amount of data, a vast amount of potentially valuable data that can be utilized for these
process improvement purposes is not used or not available. When implementing an ERP system, the
main focus is on the planning and control of the current situation. However, what is often overlooked
Page 8 of 61
is utilizing this available data for further improving processes. ERP systems might be useful in generating these MCT maps and thus, this offers an interesting area to conduct research. Since these MCT maps represent a snapshot of a certain time span at a certain point in time, it is more interesting to investigate the possibility of automatically generating these maps from ERP systems. These observations lead to the following research question:
RQ: How can Manufacturing Critical-path Time mapping be integrated into current ERP systems in SMEs?
To support this main research question, three supporting sub research questions are formulated. First MCT mapping has to be investigated, what is needed to construct an MCT map, to what extent do current ERP systems withhold valuable data to construct an MCT map and finally an assessment is made of how potential missing support can be gathered.
SQ1: What information is needed in order to construct a Manufacturing Critical-path Time map?
With the answer to the first sub-question, the following step is to compare this information with the current availability of this data.
SQ2: How do current ERP systems in SMEs offer support for Manufacturing Critical-path Time mapping?
When the current support for MCT mapping is investigated the next phase is to consider what the missing support is and how to acquire this missing information.
SQ3: How can the missing support of ERP systems for Manufacturing Critical-path Time mapping be acquired?
With the answer of the three sub-questions, an assessment can be made regarding the integration of MCT mapping in ERP systems of SMEs.
The next section contains theoretical background and will cover a more extensive background of the
relevant literature. Subsequently, the methodology is presented in chapter three. Section four deals
with the results and is divided per sub-question. Section five is concerned with the discussion and
interpretation of the results. Finally, section six finalizes this thesis with a conclusion and opportunities
for future research.
Page 9 of 61
2 Theoretical background
This section starts with discussing ERP systems where after QRM and is explained. Next, SME characteristics with a focus on ERP implementation are outlined. Finally, the ERP versus lean/QRM discussion will be highlighted.
2.1 Enterprise Resource Planning systems
Zhang, Lee, Huang, Zhang & Huang (2005) considered the implementation of ERP systems as one of the most broadly used choices for manufacturing companies to achieve competitive advantage. ERP systems are (online) interactive business systems that support the integration of processes from different functional areas with the use of a common database to improve the workflow, standardize practices and access real-time data (Mabert et al., 2003). An ERP system is built to manage all organizational activities through independent software modules, such as product planning, inventory control, accounting and human resources (Quiescenti, Bruccoleri, La Commare, Noto La Diega &
Perrone, 2006). These modules are all linked to a common database. In this way, all the activities are connected and have the possibility to interact. Their ability to process transactions in an efficient manner and provide organised record-keeping structures for these transactions can be seen as the core benefit of an ERP system (Jacobs & Bendoly, 2003). Figure 1 depicts a simplified overview of an ERP system. Mabert, Soni & Venkataramanan (2001) stated that an ERP system has three key properties: multifunctionality, integrability and modularity. As can be seen in figure 1, ERP systems are multifunctional, they track a range of activities such as inventory, manufacturing, sales, distribution and human resources. Furthermore, they are integrated in nature, by changing data in one of the functions, information in all the related functions also changes by the use of the central database. In addition, their structure is modular, a firm can choose to implement all the modules or a subset of these modules, as well as a connection to other support systems which are called ‘bolt-ons’.
This dissertation focuses on the complete ERP landscape in the manner as they appear in SMEs in the
Netherlands.
Page 10 of 61
Figure 1: An overview of an ERP system (adapted from Mabert, Soni & Venkataramanan, 2001)
2.2 Quick Response Manufacturing
QRM is defined as “a company-wide strategy for reducing lead times throughout the enterprise” (Suri,
2010, p2). These lead time reductions are specifically focused from the customer’s point of view. The
core lean techniques such as Takt times and Kanban are aimed to eliminate variability in
manufacturing processes (Muda, Muri and Mura) and work well in stable demand situations, but may
not be the best solution for low-volume, high-variety environments (Suri, 2010). Suri (2010) claimed
that QRM enhances the lean program by distinguishing two types of variability. Dysfunctional
variability which deals with errors, ineffective systems and poor organization (e.g. rework, constantly
changing due dates and priorities) and strategic variability which is used to maintain the competitive
position of an organization (e.g. ability to cope with unexpected changes in demand, offering a wide
variety of products to the customer). Lean techniques aim to eliminate all variability, including
strategic variability, whereas QRM tries to exploit it (figure 2).
Page 11 of 61
One of the most distinctive differences between QRM and lean is the positioning of the customer order decoupling point (CODP). This CODP specifies the point in the value stream up to which a customer order penetrates. It distinguishes between the part of a supply chain that is driven by forecasts and the part that is driven by orders. This relates back to the typical product attributes, for example, a make to order company (MTO) mostly deals with high variety, low volume products and is order driven. Whereas a make to stock firm (MTS) is forecast driven and concerned with low variety and high volume. QRM focuses on MTO and assemble to order (ATO) while lean focuses on MTS and ATO (Powell & Strandhagen, 2012) (figure 3).
Figure 3: Positioning of QRM and lean with respect to CODP (adapted from Powell & Strandhagen, 2012) Figure 2: QRM enhancement of Lean (Suri, 2010)
Page 12 of 61
Along with the QRM approach, a metric based on the standard critical path method has been developed which clarifies the focus of the QRM strategy which is called Manufacturing Critical-path Time (MCT). This metric is measured in the number of calendar days it takes from the moment a customer creates an order, via the critical path, until the order delivery at the customer. This represents the cumulative lead time of the processes on the critical path. Calendar time is used, opposed to several other lead time calculating methods since this metric is customer focused and is not focused on the time a company is operational. MCT focuses on the distinction between value adding and non-value adding lead times and distinguishes between touch time and non-touch time in the process flow of a product (figure 4). Touch time represents the time that an employee or machine is actually working on the job and often accounts for less than 5% of the total lead time (Suri, 2010).
Figure 4: An MCT map (Suri, 2010)
An MCT map gives a visual representation of the amount of time a product is actually worked on, is easy to understand and proportionally shows the waiting time. This makes the process of a product clear-cut and selecting an area to focus improvement efforts on becomes more transparent and easy.
For one end product, all the processes and storage points it passes from receiving the order to the
delivery of the product at the customer are mapped which results in a critical path time. Managing
lead time effectively can be a competitive advantage (Tersine & Hummingbird, 1995). In addition,
Wieters (1979) found that costly expediting, scheduling problems, lengthened planning horizons and
Page 13 of 61
inflated inventories are common problems that are the result of having lengthy lead times. Reduced lead time may lead to exploiting new markets, more flexibility and reduces the potential of creating waste (Tersine & Hummingbird, 1995). QRM efforts that started with the generation of this MCT have proven to be successful by achieving up to 94% reduction in lead time and saving up to 25% in associated costs on numerous occasions (Tubino & Suri 2000).
The level of detail that is necessary to draw an MCT map depends on the individual using it. A high- level manager is not interested in all the specific steps that are taken at a machine but rather as a higher level overview for management such as figure 4. An MCT map can also be focused on a single department to find an area for improvement in that scope.
2.3 Small and medium-sized enterprises
Small and medium-sized enterprises (SMEs) are important for the economic structure of the European Union (EU). However, they are facing fierce competition from larger enterprises. Nevertheless, 99%
of the businesses in the EU are SMEs providing 66% of the jobs and generating 50% of the total value- added created by organizations in the European Union (European commission, 2013). Due to their limited resources and knowledge, SMEs often have difficulties applying both ERP and lean (Achanga, Shehab, Roy & Nelder, 2006; Snider, Silveira & Balakrishnan, 2009). According to Bell (2006), most of the SMEs are dealing with a higher degree of uncertainty than bigger firms and, therefore, are not able to reach high levels of lean or QRM integration. The same holds for the implementation of ERP systems. Let alone having the funds to fully exploit both practices to increase their competitive advantage.
2.4 ERP vs. lean/QRM
Philosophies related to run manufacturing operations in an effective manner have been heavily debated (Nakashima, 2000). In addition to this debate, increased competition and expectations have put pressure on production engineers and managers to develop innovative and improved solutions.
Two main camps exist regarding production control. On one side, lean manufacturing is embraced by
numerous of firms which are particularly SMEs (Bartholomew, 1999). The other side consists of
proponents of ERP systems, systems on which organizations have spent billions of dollars in the past
years (Bradford, Mayfield & Toney, 2001). Both practices stem from fundamentally different
approaches to production. Where ERP systems are mainly forecast driven and inflexible, lean and QRM
Page 14 of 61
practices are demand driven. Halgeri et al. (2008) state that when ERP systems were emergent, authors argued that ERP systems were the antithesis of lean manufacturing and resulted in forms of waste, high inventories and slower production. However, the growth of e-business methodologies combined with increased competition has made authors rethink their statements. Bradford et al.
(2001) stated that it is evident that ERP systems gather valuable data and that this data should be used
in conjunction with lean practices. Furthermore, lean practitioners have started to adopt ERP software
in order to enhance their lean initiatives and thus reach a higher level of operational excellence (Bragg,
2004). Due to this ongoing debate, the emergence of new technologies and the lack of literature that
combines QRM and ERP systems it is interesting to explore the current state of ERP systems how they
can support MCT mapping and give new insights in light of this debate.
Page 15 of 61
3 Methodology
This section provides an overview of the methods that are used in this research. It constitutes as a blueprint for the collection, measurement and analysis of data. First, the research design and case selection are addressed followed by the data collection and the data analysis. Subsequently, the quality criteria are discussed and this section ends with a description of the case companies.
3.1 Research design
Mainly a qualitative approach was used for this study as the aim was to provide an in-depth analysis of the integration of MCT mapping in ERP systems in SMEs. Due to time and resource constraints in this research and the fact that there exists a wide range of different ERP systems, there is chosen to study one of the most widely used ERP system for SMEs in the Netherlands which is called Exact ERP software. In addition, there was chosen to do a multiple case study since it is argued by Silverman (2001) that qualitative research has the ability to gain a more comprehensive understanding than quantitative research. A case study is the most relevant type of research if ‘how’, ‘what’ or ‘why’
questions are addressed, when the researcher has little control over events and when the focus is on
studying a contemporary phenomenon in a real-life context (Yin, 2009). The research questions in this
thesis are of ‘how’- and ‘what’-type questions and a contemporary phenomenon is being studied
(support of ERP systems for generating an MCT map). Very little academic research is available of the
combination between QRM and ERP systems. Therefore, this research was explorative. Firstly,
literature is addressed to create an overview of the information which is needed to construct an MCT
map. Secondly, an investigation regarding the extent to which an MCT map can be composed at
different manufacturing companies was conducted. This included outlining the possibilities and issues
that arise when evaluating the current support of ERP while constructing a complete MCT map at a
company. This in-depth analysis is done at one single company. Three other companies provided ERP
data and clarifying interviews regarding the information they registered were held to get an
understanding of what is currently present in ERP systems. Subsequently, employees of the
gatekeeping company were interviewed who have a great understanding of the use and possibilities
of ERP systems. Finally, missing information to construct an MCT map was analysed and by the means
of interviews, literature and observations an assessment was made how to cope with this missing
data. The company from where this dissertation is written, in the form of an internship, is an ERP
vendor and implementer in the Netherlands and was the gatekeeper in this research (more details
Page 16 of 61
can be found in section 3.6). This company works with one of the most widely used ERP systems in SMEs in the Netherlands and was able to provide contacts with their customers to select case companies to investigate the second research question (SQ2). The other four companies are elaborated in section 3.6.
3.2 Case selection
The unit of analysis for this research was ERP systems in SMEs. The cases to conduct the second part of the research question (SQ2) were manufacturing companies that are situated in the Netherlands.
The cases had to make use of an ERP system and they should fit the definition of an SME set by the European Commission (2013) (between 10 and 250 employees and an annual turnover between €2 million and €50 million). By means of addressing cases in different industries with different CODPs (MTO vs. MTS), differences between these types of companies are highlighted. Eisenhardt (1989) claimed that the optimal number of cases for a multiple case study lies between four and ten cases, therefore, this research aimed to use at least four case companies. The gatekeeper in this research was able to provide access to these companies. The four cases that were addressed operate in the furniture, packaging, construction and fabric manufacturing industry.
3.3 Data collection
Interviews
A variety of people was interviewed, amongst them were employees of the selected case companies
in order to understand the way they are using their ERP system and to investigate what benefits an
MCT map could have for them. These employees were in a higher function since they had to be aware
of every aspect of the company to give a complete overview. In addition, they had to be aware of how
they were using their ERP system. Furthermore, employees of the gatekeeping company who are
experts in implementing ERP systems were interviewed in order to gain an understanding in what ERP
systems currently offer in terms of information. Both the interviews were semi-structured in order to
allow new ideas to be brought up by either the interviewee or the interviewer. The interviews were
all conducted face-to-face. The length of the interviews varied depending on the interviewee’s
expertise, availability and the creation of new ideas due to the semi-structured nature of the
interviews. Interviews were planned for at least one and a half hour in order to have time to gain an
in-depth understanding of the answers given. Each interview started with a pre-determined set of
questions related to the topics that were of interest in this research, which would aid in answering the
Page 17 of 61
research questions. This set of questions is refined based on the results from the first sub-question.
Given responses by interviewees were summarized during the interview to validate their correctness.
Furthermore, the interviews were recorded to make sure no relevant information was missed. The interview protocol with a consent form which was given to each interviewee and the list of pre- determined questions can be found in appendix A. Furthermore, a list of interviewed people and their job description can be seen in appendix B. Due to confidentiality the company names and names of interviewees are left out.
Informal conversation
As this research was conducted in the form of an internship at the gatekeeping company and other selected case companies a lot of knowledge was gathered in the form of informal conversations and discussions. Notes were taken at discussions that aided in answering the sub-questions.
Archival research
A large base of information about current ERP output of the case companies was available at the gatekeeping company and at could be accessed via a software program. This data was a recent copy of the ERP database of the case companies.
Direct observations
By accompanying employees of the gatekeeping company performing their work and exploring their processes, direct observations were made. Furthermore, at every case company a tour throughout the company was given by the interviewee to get a better understanding of the processes. These observations were used to gain more in-depth understanding of the practical environment in which the gatekeeping company and the case companies conduct their field of work. During these observations, questions were asked for clarification and notes were made. Observations with potentially major consequences were validated by involving experts to agree on the interpretation.
3.4 Data analysis
The data gathered during the interviews was first drafted into transcripts. The output from the interviews at the different companies was compared in order to seek familiarities or contrary findings.
The received output of ERP systems from the selected companies was accessible via a software
programme in which a copy of the real data was inserted. This data is analysed with the help of
employees who are experts in using this ERP system and this analysis aided in answering the second
Page 18 of 61
sub-question. Furthermore, the data was analysed with the use of the book of Suri (2010). This book helped to create MCT maps since it offers guidance in calculating and creating MCT maps which was helpful in answering the first sub-question. In addition, the Wisconsin University QRM centre has developed an MCT-tool in Excel which was further improved by the Hogeschool van Arnhem en Nijmegen (HAN) and was utilized to visualize an MCT map.
3.5 Quality criteria
This research mainly consisted of qualitative data, it, therefore, insists on validating the gathered data and assess quality criteria. Firstly, construct validity was supported by using triangulation, this involved the use of multiple sources (literature, interviews, observations and archival analysis) to study the same phenomenon. Secondly, external validity was dealt with through the use of a multiple case analysis such that generalizable insights can be derived. Finally, reliability concerns the repeatability of the research with the same results and is secured by using interview protocols. Data from interviews, open discussions and observations existed of directly taken field notes, expanded typed notes and a running record of analysis. Furthermore, the same way of analysis was used throughout the different cases for answering the second sub-question.
3.6 Case Companies
This thesis research was mainly conducted from A.B. Software & Consultancy in Paterswolde in the form of a 5-month internship. A.B. Software & Consultancy sells and implements an ERP system called Exact in a wide variety of SMEs and is located in Paterswolde, Deventer, Uden and Bleiswijk in the Netherlands. Exact is the most widely applied ERP software for SMEs in Holland. Besides implementing ERP systems, they also use lean and quick response manufacturing (QRM) methods and techniques in order to simplify, accelerate and improve business processes of their customers.
Data was analysed at four different customers of A.B. Software & Consultancy which are outlined below. In order to be discrete and ensure confidentiality, these companies are stated as companies A to D.
Company A
Company A is a bedroom furniture manufacturer in the eastern part of the Netherlands. It mainly
produces a standard collection of bedroom furniture. Furthermore, they also produce customized
Page 19 of 61
furniture on a project base. It has 36 employees and a yearly turnover of €4.2 million. It is using Exact ERP software since September 2014.
Company B
Company B is a candy packaging company is the eastern part of the Netherlands with 200 mentally disabled employees. It has implemented Exact ERP software in July 2012. A big candy producer delivers bulks of candy which then is packaged in plastic bags, sealed and then put into boxes. The customer is the same the candy manufacturer, which then delivers the candy to stores all over Europe.
Company C
The third company is a fabric manufacturing company which produces indoor blinds and is situated in the eastern part of the Netherlands. It has a wide variety of different end products (over 500) and mainly operates in an MTS fashion. They have 50 employees, a yearly turnover of €17.5 million and they use Exact software since 2006. This company purchases raw fabrics and outsources the treatment such as colouring to a third party where after they produce folded fabrics which is then sold to assemblers. These assemblers are mainly situated in Germany the rest of the customer base is spread throughout Europe and a few outside Europe.
Company D
The fourth company produces lightweight trailers and is situated in the southern part of the Netherlands. It has 165 employees and mainly operates in Netherlands, Germany, Great Britain and Belgium, but also in the rest of Europe and Russia. Besides producing trailers, it also has a service centre where trailers are repaired. They are currently implementing Exact ERP software and this is scheduled to be finished by the end of 2015.
An overview of these companies are given in table 1, additionally, their position on the
product/process matrix is indicated in figure 5 (Hayes & Wheelwright, 1979). Companies with a
different complexity, size, turnover, CODP, industry and degree to which they use lean practices are
addressed in this research which makes it able to draw comparisons between different characteristics.
Page 20 of 61
Table 1: Summary of the case company characteristics
Company A Company B Company C Company D
Industry Furniture Packaging Fabric Steel
Main business Producing bedroom furniture
Packaging candy
Treatment of fabrics for blinds
Producing lightweight trailers for transport
# Employees 36 200 50 165
Annual turnover
€4.2M Unknown €17.5M Unknown
Primary CODP MTS MTO MTS MTO
Lean/QRM practices
Kanban, 5S 5S None Kanban, 5S, Takt time,
cellular manufacturing Uses Exact ERP
system since
September 2014 January 2013 July 2006 November 2014
Figure 5: Product/process matrix with the position of the case companies adapted from Hayes & Wheelwright (1979)
Page 21 of 61
4 Results
This section covers the results from the data analysis. This chapter is divided into three sections, following the three sub-questions.
4.1 Necessary data
To answer the question how current ERP systems in SMEs offer support for manufacturing critical- path time mapping, it is imperative to first understand what data is needed to construct such a map.
This section tries to answer the first sub-question: “What information is needed in order to construct a Manufacturing Critical-path Time map?”.
Based on literature (Suri, 2010), a list of required data is composed. It is chosen to restrict the MCT map to internal processes of a company. For instance delivery of raw- and/or semi-finished materials and the transport of finished goods from the factory to the customer by a third party can be seen as external processes and are not addressed since these external processes are more challenging for companies to control compared to internal processes.
One of the most important aspects of an MCT map consists of the route a service or product follows throughout an organisation. This route is the essential foundation for the creation of an MCT map. It serves as a backbone since it withholds data concerning which different sequential process steps a product or product family has to follow in order to be ready to be transported to the customer. This also includes data regarding the subcomponents it consists of within the scope of analysis. In addition, this route should include all the processes a product has to follow to be made from scratch within a company.
In addition, the nature of the aforementioned process steps needs to be known. Whether the process
step of the product or service is categorized as non-touch time or touch time. With the use of Suri
(2010) and an MCT-Excel tool created by the University of Wisconsin and further developed by the
HAN, five categories that are commonly applicable to processes in most organisations to cover the
nature of the process are distinguished. These categories are processing, setup, in transport, in queue
and in stock and are elaborated in the next paragraph. This categorization is useful to distinguish non-
touch time (in transport, in queue and in stock) from touch time processes (processing and set-up)
and characterize different sorts of touch and non-touch times. This knowledge will aid in a more
detailed breakdown of the processes in an MCT map.
Page 22 of 61
Besides the nature of these processes, the time it takes for a product or service to pass each of these steps need to be known. Processing times concern the time it takes employees or machines to handle a product or order either in production, warehouse or in the office. Setup times involve the period which is needed to prepare a machine, device or system to handle a certain product, product family or order. Transport times consist of the movement of goods. Queueing time includes the time a job spends in queue waiting to be processed. Time spent in stock deals with the period a product spends in either raw material, work in progress (WIP) or finished good inventory (FGI) and can be derived by the use of Little’s Law (Little, 2011). This law states that the time spent in inventory, or any system W, is equal to the average number of customers in a queueing system L, divided by the rate at which products or orders arrive or depart at a system λ: 𝑊 =
𝐿λ
Thus, in order to use Little’s Law to estimate the duration of the time spent in inventory, both the inventory level and the amount of products consumed by the following process step per unit of time must be known.
Furthermore, when a product or service does not flow through the process in a one-piece manner, batch sizes need to be known in order to be able to derive information that accounts for a single product or service. This makes it possible to, by using process times, derive the time it will take to process one product or order spent in queue waiting on a batch.
Lastly, the amount of operational hours per week need to be known in order to convert duration of processes that are not yet measured in calendar time to calendar time. Operational time is the time an organisation is operational and actually producing products or offering their service. An overview of all the necessary information is given in table 2.
Table 2: Overview of necessary information needed to construct an MCT map
Type of information Description
Product route All process steps required to produce a final product from scratch starting when an order arrives
Nature of the process steps Distinction between processing, setup, in transport, in queue or in stock Duration of the process steps Time it takes to successfully complete each separate process step Batch size Amount of products in a certain batch
Operational time Total amount of time an organisation is actually producing products or
offering their service per week
Page 23 of 61
All the necessary data should concern only one end product or average product of a product family, which is the first useable end product for the customer. Furthermore, it should exclude non-critical items an end-item consists of. Suri (2010) describes non-critical items as following: (i) the cost of this item should be relatively low to the end-item; (ii) there is an abundant supply of this item in the market; (iii) the specifications are not eligible to change in the near future. This section gives a comprehensive overview of the necessary data which is required for an MCT map, hence the first sub- question is answered.
4.2 Present information
To assess the presence of the list of data which is mentioned in the previous section and to answer the second sub-question “How do current ERP systems in SMEs offer support for Manufacturing Critical-path Time mapping?”, triangulation is used. First, several semi-structured interviews with consultants are held whose daily job it is to work with ERP systems in terms of implementing these systems or giving advice relating to them. Secondly, an MCT map was made at a furniture manufacturer. One product was selected and followed from the arrival of the order until the shipment of the product through the whole process. By having interviews with the employees who are operating the ERP system as much information as possible was collected from the ERP system that is needed to construct an MCT map. The missing information was gathered by conducting interviews with the employees, who are actually executing the tasks. In addition, their managers or foremen were also questioned to validate the given answers. Lastly, actual ERP data was analysed from three other companies strengthened with observations and conducting interviews with employees. In addition, the extent to whether the available information is reliable and actually valuable to use in a real-time MCT map is addressed at all three sections. Regarding the fact that three different kinds of sources are used, this section is divided into three sections and finalized with an overall summary.
4.2.1 Interview with experts
This section will describe the main outcome of the conducted interviews with experts how they assess the presence of the required information in current ERP systems at SMEs. The interviewees are experts in handling, designing and operating ERP systems. The outcomes of these interviews were compared and led to the results that can be found below.
Product route
Page 24 of 61
Regarding the product route, all interviewees agree that a high-level route is present at all times. It is clear that, in an MTO environment, the process starts with an order from the customer which then is followed by the creation of a production order, the release of the production order, the manufacturing of the product and finally the shipment of the product. In an MTS environment, the order mostly comes from the warehouse that signals when a stock is below a certain level. In addition most products in an MTS environment follow the same route. Furthermore, the bill of materials is almost always included and gives an indication of the components the final product comprises of. However, a detailed routing of the (sub)components that lead to a final product is often not present since it is seen as not relevant to import this data. Specific steps that a product must undergo at a department or machine is not registered but mostly just known by the employees operating at that department.
This applies mainly to an MTS environment since here process routes are often more standard and there is no sense of urgency to register. In an MTO environment, these routes are less standard and the added value is more apparent.
Nature of process steps
The nature of the different steps can be derived with quite ease. When a product spends a week in a warehouse and it has not changed its location it is evident it can be seen as non-touch time. Whether it is touch or non-touch time is not registered, however, it can be derived by sound judgement of employees.
Duration per process step Transport and queue
Transport of arriving and departing trucks containing raw materials or finished goods are becoming easier to trace, nevertheless, information regarding transport of products inside a company is almost never registered. A plan containing the location of warehouses, machines offices etc. is present.
However, internal transportation does often not take much time into account and is thus not worth registering. When a company also takes care of the transport of the finished goods, this is registered on a high level. Tracing or registering time spent in queue is not seen as interesting and thus not currently registered. On several occasions, these times can be derived by applying queueing theory.
Setup
Page 25 of 61
Setup times are only registered when the planning department takes the setup times in consideration in creating batches and releasing orders to the production floor or when high costs are involved in the setup of machines. When these times are available, they are not very accurate and entered once and barely adjusted. To account for setup in the cost price calculation, a certain percentage is taken rather than recalculating setup times.
Stock
The stock levels of finished goods and raw materials are known in basically every company. Averages of the amount of stock that is present can be derived per time period. In addition, the amount of products that are shipped to customers is known. The stock of non-critical items such as screws and bolts are often Kanban-driven since these inventories do not involve high costs and are easy to control.
Stock levels regarding semi-finished goods within the production process are not always registered. In addition, WIP or semi-finished goods are often also Kanban-driven. This is due to the fact that for instance the production is Kanban-driven or no registration possibilities are present on the shop floor.
Furthermore, the more a company operates in an MTO fashion, the more likely it is they will register WIP.
Processing
The time it takes for each process step is mostly not present in SMEs. When these are present, the standard times (time it cost an average skilled employee to complete a predefined task) are assigned to production tasks and do not involve office related tasks. These standard times are steady state values and entered in the ERP system when it was implemented. This means these values are not eligible to change and therefore not useful for a real-time MCT map. Nevertheless, this information can be used as an indication of an ‘initial’ MCT map and as a comparison with the current state.
Derivations and estimates based on available data, that will be mentioned more in-depth in the next section, can be made to retrieve certain data.
Batch size and operational time
When a company produces in batches, the batch sizes are registered. This is mostly done at the
planning department but may be subject to change on the production floor to the discretion of the
foreman. The time that a company is operational can be derived from the work schedules.
Page 26 of 61 Additional findings
One interesting fact that came up was that in the early 2000s a large number of SMEs in the Netherlands in manufacturing performed registration in the production area. These SMEs had the idea it was interesting and valuable to store this information. However, eventually in most SMEs this data was not used and they stopped registering. Though, current trends show that due to the increasing ease of registration possibilities and the acceptance of automation in an organization by employees that registration becomes increasingly prevalent. From the interviews, it became clear that process industries are ahead of discrete industries in this field. In addition, the more MTO oriented a company is, the higher the degree of registration. Since in these MTO environments the used materials, labour hours and lead time vary to a much greater extent compared to MTS companies. Likewise, since MTO companies deal with topics like the customizability of products, specific materials, labour and machine hours that have to be registered to charge the correct price to the customer. In addition, when a certain process takes an additional week to complete, this week often directly results in an additional week of waiting for the customer. In an MTS environment, this would lead to an increase in stock. This direct increase in lead time triggers a higher urge to register and improve processes when compared to an increase in stock. Registering this data in an MTS environment is, therefore, less interesting for these companies. Also, since SMEs have a clearer overview of their company compared to bigger firms so the urgency to register is less. An argument which is often heard is: “Why should I register my inventories when I just can walk through my warehouse and see them?”.
Moreover, an overall issue that arises is the accuracy of the data entered. Wrong data could be entered, orders could unintentionally be released to the production or orders received by mail, fax or phone are not registered in the ERP system immediately but could take a lot of time, these issues cannot be derived from the system. Wrongly entered data can have a huge impact on processes in the company and should be monitored to a high degree when important decisions are made based on this data. Therefore, reliability is also addressed when investigating the different companies.
Summarizing, a variation between different companies can be seen. Differences between complex and less complex type of operations and different CODPs. SMEs operating in a high-tech industry with expensive parts or industries dealing with severe safety regulations such as pharmaceuticals, aerospace and medical equipment, significantly register more than the average manufacturing SME.
An overview can be seen in table 3.
Page 27 of 61
Table 3: Summary of availability and reliability of necessary data based on interviews
Type of information Availability Reliability
Product route Medium Medium
Nature of the process steps - - Duration
of the process steps
Transport Very low Very low
Queueing Very low Very low
Setup Medium Low
Stock Medium High
Processing Low/medium Medium
Batch size Very high Medium
Operational hours Very high High 4.2.2 In-depth study
One product, a specific single person bed, which is produced on a regular basis and of which the employees had a good understanding and thus could make well-founded estimates was chosen to be the subject of an MCT map at the furniture manufacturer (company A). The components of this bed that are fabricated by this company are two side panels, a headboard and a footboard. An overview of the process can be found in figure 6. An order is handled by the customer service where after this order is sent to the planning department. Here, the bed is split up in the three separate components and, when possible, released to the production which consists of sawing, woodworking and assembly with WIP inventories in between. After assembly, the components are sent to the FGI where the components wait to be packed in trucks and shipped.
Figure 6: Single person bed process overview
The complete MCT map can be found in figure 7. The fact that this product consists of three different
main parts which this company produces and the fact that these parts are used in other end products
as well leads to different process times. The head- and footboard are both used for the same products
Page 28 of 61
and, therefore, have the same warehouse and logistics time. As can be seen, all three components pass the process steps customer service, planning, production, warehousing and logistics. At step 8 the head- and footboard are packed together and placed in the warehouse. At the final step, the packed head- and footboard and side panels are combined to get shipped to the customer. Data from the ERP system in the period of 1 January until 1 October 2015 was used, but any time span could have been chosen.
Figure 7: MCT map of a single person bed
Product route
The specific product route, as presented in figure 6, was not present in the ERP system. However, a
bill of materials (BOM) of this specific bed was present and, therefore, all three subcomponents could
be identified. Furthermore, the ERP system had a record of which machines in production the
components have to pass in order to be completely finished albeit not in a sequential manner. The
ERP system itself works with a high-level sequential process which consists of the moment an order is
Page 29 of 61
entered in the system, released to the production, received at production, finished by production and sent to the customer. The exact route was verified by means of an interview with employees. The 22 steps mentioned in figure 7 all consist of 2-5 sub-process steps where at least one is categorised as touch time and one as non-touch time. These sub-process steps consisted of processing, setup, queuing, stock process or transport time and are left out to secure readability and clarity.
Nature of process steps
Whether the duration of a process step can be classified as touch time or non-touch time is not registered but can be derived to the aforementioned categories with quite ease. These categories are commonly applicable to most organisations to cover the nature of the process.
Duration per process step Transport
Information regarding the layout of the production area and warehouse was available. However, no predefined pathways nor time registration about transport were present. The total internal transport time appeared to be relatively small compared to the total MCT time.
Queuing & setup
No information regarding the time jobs spent in queue or specific setup times per product, product family or machine was available. However, employees were able to give accurate indications of average setup times.
Stock
The inventory level of finished components was updated in real-time and, therefore, an average over
a certain period can be derived. Furthermore, the throughput rate (number of products which is sold
per unit of time) is known. With these two variables, the average time a product spends in inventory
could be calculated with the use of Little’s Law (Little, 2011). These times spent in inventory for side
panels and head- and footboards account for a substantial amount of time in the drawn MCT,
respectively 27 and 59 days. These figures are based on a 9 month period. When divided into periods
of 3 months, the 59 days is subject to variation between 21 and 84 days. The 27 days remained the
same. Information about the stock level of the WIP is not registered and thus not available, this data
Page 30 of 61
is also gathered by interviews. Furthermore, the stock level of raw materials is registered and could be used in an expanded MCT map. Stock levels are counted regularly whether they still correspond with the data in the ERP system, this ensured the reliability of the data.
Processing
A number of specific process step durations, solely in production, are listed in the ERP system. These values, however, are entered when the ERP system was implemented in 2014 and thus static figures.
These figures do not incorporate system dynamics. Static processing times are present for machine operations. These times might be accurate at the time of implementation but could be entirely different at this moment. To get an accurate and real-time estimate of processing, setup, transport, queueing and planning times, employees and their managers were interviewed and this data was used to make the MCT map that can be seen in figure 7.
The number of orders by customers that the customer service handles in a specific time period, at what time an order is entered and when it is sent to the planning department can be derived from the ERP system. Production orders are released once a week by the planning department and the number of released production orders are known. Rough estimates about the touch time per product can be made since the planning department stated that the planning of each product takes roughly the same time. Furthermore, the planning department indicated that orders arrive in an equally distributed manner and since production orders are released once a week they spent on average 3.5 days before being released to the production. The time when orders are released to the production is known and also, the time when a product enters the finished goods inventory is known. The time in between can, therefore, be allocated to production. Products released to the production floor were not tracked, however, Little’s Law can be applied for the overall production since the amount of products that have been released to the production department is known, as well as the number of specific products that arrives at the FGI per time period. Furthermore, no additional information during production can be found in the ERP system and, therefore, no in-depth information in the production area can be found.
The reliability of a certain amount of process time durations that were found has to deal with the
judgement capabilities of the employees.
Page 31 of 61 Batch size & operational time
Batch sizes are known and are determined by the planning department. Therefore, this information is available in the ERP system. This is, however, sometimes subject to change by the judgement of the foremen on the production floor. Also, the amount of time the company is operational (36 hours a week) was be derived from the ERP system.
The ERP system at this company withholds data which can be used to construct an MCT map, although not fully reliable. However, there is still a vast amount of missing information which was not derivable without human interaction. An overview is given in table 4. Regarding the MCT map, a touch time of only 1.44% over a period of 73 days was found for this specific bed. This is mainly due to the fact that a high level of stock for the parts of this bed is maintained. In addition, since every department in production (sawing, woodworking and assembly) and planning is given a week to complete their orders, even though it costs a few hours to finish these tasks, the percentage of touch time is low.
Table 4: Summary of availability and reliability of necessary information in company A
Type of information Availability Reliability
Product route Medium Medium
Nature of the process steps - -
Duration of the process steps
Transport - -
Queueing - -
Setup - -
Stock Low High
Processing Low Medium
Batch size Very high Medium
Operational hours Very high High 4.2.3 Additional cases
The three other companies B (candy packaging company), C (fabric manufacturer) and D (trailer
manufacturer) are analysed by interviewing higher management employees regarding registration
and observing the processes within the company. The results can be found below. The simplified
overview of the general processes a product passes in these companies is visualised in figures 8, 9 and
10.
Page 32 of 61
Figure 8: Process of packaging candy at company B
Figure 9: Process of fabric manufacturing at company C
Figure 10: Process of producing a trailer at company D
