Realize high delivery performance by organizing the
Pre-production stage
Master thesis, MscTOM, Technology Operation Management
University of Groningen, Faculty of Economics and Business
June 17, 2013
Lipeng Jing
Student number: 2222973
E-mail: ljing881004@gmail.com
Supervisor/ university
Erik Soepenberg
Co-assessor/ university
Stuart Zhu
Supervisor/ field of study
Ronald Vinke
2
Abstract
Content
Introduction ... 4
Theoretical background ... 5
Characteristic of ETO Company ... 5
Characteristic of pre-production stage... 6
Delivery performance ... 7
Potential causes of problems ... 7
Conceptual model ... 10 Methodology ... 11 Case selection ... 12 Research structure ... 12 Data collection ... 16 Results ... 17
Delivery performance problems ... 18
Causes of the problems ... 19
Work preparation department ... 19
Sales department ... 21
Engineering department ... 21
Deep insight into the causes ... 23
Work preparation department ... 23
Sales department ... 25
Engineering department ... 26
Special late orders ... 27
Discussion... 31
Conclusion and future research ... 32
Acknowledgement ... 34
Reference ... 35
4
Introduction
Many Engineer to Order (ETO) /Make to Order (MTO) companies are faced with the problem that delivery performance cannot be performed as well as they promise, due to the uncertainties related to the specific customer orders are not easy to be managed (M. J. Land & G. J. C. Gaalman 2009). Existing production planning and control (PPC) concepts can provide solutions to this kind of problems. While previous research has discussed many deficiencies related to PPC in practice related to the shop floor. Furthermore, Land & Gaalman (2009) illustrated that poor planning of the pre-production stage could be the root of delivery performance problems. Therefore, delivery reliability in ETO/MTO companies has been the initial trigger for this research.
Many delivery performance problems have already been identified and analyzed in previous researches, which are all related to the production stage of Small and Medium sized Enterprises (SMEs). SMEs is an important part of the ETO/MTO sector. These companies have to deal with the customer ordering uncertainties accurately (e.g. Specification and due date), otherwise it will be difficult to survive in the fiercely competitive environment. Only limited empirical evidence of the specific PPC requirements of SMEs is provided by scientific research, which illustrates existing PPC concepts are suitable for MTO/ETO SMEs (M. J. Land & G. J. C. Gaalman 2009).
Workload Control (WLC) is an essential PPC solution designed for the requirements of MTO/ETO Industry (Brian Kingsman & Linda Hendry 2002). WLC provides a simple framework to control the decisions of general order flows from order acceptance to release phase. (B. G. Kingsman 2000). Furthermore, supportive tools developed by Soepenberg et al (2008) can be applied to MTO/ETO companies to detect the root of the logistics performance problems (G. D. Soepenberg et al. 2008). On the other hand, Ebadian, Rabbani, Torabi, & Jolai, (2009) have developed a hierarchical production planning (HPP) structure, which prioritize the accepted order based on different customer groups to guarantee delivery reliability. Therefore, this structure effectively relieved the congesting orders on the shop floor through setting correct priority. However, relatively scarce literatures address the problem in pre-production stage. This research strives to fill the gap by finding and solving the delivery problems in the pre - production stage of ETO companies. Therefore, the research question is:
We will take a deep insight into the pre-production stage to identify the question after this chapter. Where previous research only pays more attention to control shop floor such as workloads, the novelty of this research is that it focuses on the pre-production stage of ETO Company. Eventually this study tries to discover the main causes of delivery problems in pre-production stage and illustrates how to organize this stage better to improve delivery performance. In practice, the results try to help companies in improving their delivery reliability.
The rest of the paper is arranged as follows. Section 2 presents the theoretical background and conceptual model, which elaborates the characteristics of pre-production stage in an ETO company and potential causes of poor delivery performance. Section 3 introduces the methodology, relevant cases and interviewees for research. Section 4 explains the results of this research. Section 5 is discussion phase, which takes a deeper insight into the results and illustrates suggestions to the problems. Section 6 draws the conclusion and future research avenues.
Theoretical background
Characteristic of ETO Company
The characteristics of Engineer To Order (ETO) companies are described in terms of their market (customers oriented), products (customized order specifications) and the internal processes (flexible) of their organization (C Hicks et al. 2000). These companies focus on customized, complex products and who are struggling with the uncertainties of the markets, which leads to delivery reliability to be the most significant order winning factor. In order to realize the high delivery performance, managers must learn to control the ETO production situation. As Bertrand & Muntslag, (1993) mentioned, the typical control characteristics of the ETO company can be classified into 3 aspects: dynamics, uncertainty and complexity.
As BG Kingsman et al. mentioned dynamics means the capacity of ETO Company is impossible to cope with the dynamic market fluctuations (BG Kingsman et al. 1989). The dynamics of the situation need to be taken into account due to its characteristic of randomness.
The second factor uncertainty concerns the product specifications, mix and volume uncertainty of future demands, and process uncertainty which is an important feature of ETO companies’ products. Different orders come randomly with diverse order specifications, which forces companies to enjoy capabilities to fulfill all kinds of requirements. On the other hand, they have to be able to adjust the products with the changing of the order specifications.
6
physical stage. Since ETO companies produce customized products, the good flow and preparation also depend on the order specification. Different orders take different routings and durations to accomplish, which leads to underestimation or overestimation of the time spent on some orders. Underestimations or overestimations are caused by the uncertainties of parameters or raw materials delivery performances. These factors result in the problems of delivery performance, thus, in the problem detecting phase all these relevant factors were taken into consideration.
Characteristic of pre-production stage
In ETO companies, all production activities are carried out around specific customers’ orders. An order process cycle includes five stages, which are customer inquiry stage, design & engineering stage, order entry stage, job release stage, and shop floor stage (Sun et al. 2012). The sales department should decide whether to accept an order or not and make a due date quotations based on the capacities and workloads in both pre-production stage and production stage. In the design & engineering stage, if the orders are accepted, this determines the company must own the technical feasibility and manufacturability, which means the engineering department must enjoy the capability to realize the customer requirements. The order entry stage determines the due date and design & engineering stage is in charge of preparation including material purchasing, outsourcing, and shop floor routing. On the job release stage, the key decisions are when and where should an order be released, which could affect the workload and Works-in-Progress. Shop floor stage can be divided into scheduling and controlling to monitor and control the progress to ensure the due date (B. G. Kingsman 2000).
Delivery performance
Delivery performance is measured by four variables, which are delivery lead-time, throughput time, percentage of customer orders delivered late, and inconvenienced by any late delivery (Milgate 2001). The first two factors measure speed and last two factors measure reliability. Delivery lead-time was defined as the actual time that elapses from the placement of an order until its shipment to the customer, while throughput time was defined as the time to finish an order from start to its completion. However, in this study, only two factors which are bound up with pre-production stage will be analyzed: delivery lead-time and percentage late deliveries. In other words, processing speed of pre-production stage and reliability of order deliveries are the main variables.
As a key indicator of delivery reliability, lateness is defined as the difference between the promised delivery date and the real delivery date. Positive lateness means an order delivered late, while negative lateness means an order delivered early. The seminal work of Baker (1974) divided the lateness into average lateness and the variance of lateness, which required different approaches to detect the causes of problems. Both indicators can result in a poor delivery performance (figure 2.1).
Figure 2.1 influence of the average lateness and the variance of lateness
The goal is to identify why some orders waste more time to prepare in the pre-production stage, why some customers gain unrealistic promising due date.
Potential causes of problems
8
After order acceptance, engineering requirements vary from small adjustments for catalogue products to extensive engineering processes depending on order specification (M. J. Land & G. J. C. Gaalman 2009). In this phase, the engineering department must make accurate pre - calculations based on the historical experience and correct routing according to a different order specification. In addition, procurement activities in the company are driven by customer orders. Therefore, if the specification of customer orders changes, there is no doubt that longer preparation is needed in these two phases. However, customer specifications must be translated in feasible parts and constructions (Konijnendijk 1994). The better an ETO company can do this, the better it will perform. The changes of order specification could be another potential cause of the delivery performance problem.
On the other hand, this factor will also affect the next step and result in order release late. In the order release phase, because capacity is often finite, it is of concern to release orders to shop floor with a balanced load, which will support the control of average lateness (MJ Land 2004). Priority dispatching rules can improve the delivery reliability, however, it has a limited influence on average and variance lateness (G. D. Soepenberg et al. 2008)
Accurate forecasting of the duration of activities at the contract negotiation and planning stage is an effective factor to improve the reliability of lead-time estimates (Chris Hicks, McGovern, & Earl, 2001). According to Land & Gaalman, 2009, they pointed the delivery performance problems into 4 aspects: 1) promising too much, 2) promising incorrectly, 3) realizing too slowly, 4) realizing wrong orders. Therefore, in order to predict due date accurately, every department should work in an integrated manner with each other to keep information exchange efficient. Considering the sequence of pre-production stage, potential causes of delivery problems could be discovered step by step from order acceptance to order release.
With respect to workload control, pre-production is also limited by the finite capacity of each department. Meanwhile, every department does not work independently, which means the efficiency of order delivery relies on whether the workloads of departments are balanced. Whether there is a better communication between departments affects the efficiency of information exchange, which provides an overview of the whole plan to involve all departments. The priority dispatching rule could also be an important factor to keep delivering on time.
On one hand, if the companies do not take actions to internal integration between departments in the pre-production stage, there could be some potential communication problem between departments (e.g. Engineering department gives the right due date quotation, however, sales department only takes historical information and promises much shorter delivery time). On the other hand, less communication and knowledge between each department will also result in misunderstanding each other (Rahim & Baksh 2003). Therefore, considering the characteristics of ETO, more complexity and uncertainties could make every stage take much more time to make decisions and plans. In fact, some unreliable deliveries are caused by the latest raw materials, which can also be attributable to the poor external integration.
10
Conceptual model
Figure 2.2: Conceptual model
This figure illustrated my conceptual model, which is built based on the theoretical background. The elements in left volume are the independent variables, which can affect the intermediate variable in the middle of the model. Furthermore, the intermediate variables can result in high or low delivery performance, which is defined into two dependent variables (preparing speed and delivery reliability). The independent variables are defined with respect of stages in pre-production stage. For instance, sometimes sales department accepts a complex project. The engineering department has to spend much longer processing time than the other orders to make the drawing; especially for those who did not enjoy enough capability and experience
+ + - + - + + + - - + - + + + + - + - + + + - - + - + +
to accomplish the project. On the other hand, this factor could also have a positive impact on the 2 independent variables (uncertainties of customers and suppliers), which will further influence the waiting time. Therefore, long processing time and long waiting time imply slow preparing speed, which will cause poor delivery reliability.
In order to provide accurate due date quotation, the capacity of all departments should be known clearly by each other, which is helpful to make efficient plan and allocate jobs. However, this significantly relies on the information reliabilities and whether do they share planning. Therefore, the better these two factors perform, the more accurate due date quotation can be provided. Finally, a more accurate due date quotation implies high delivery reliability.
On the other hand, combining with the uncertainties of order specifications, supplier, and capacities had a positive influence on waiting time and poor communication. Incorrect due date quotation can be the consequence of poor communications. Furthermore, since the uncertainty could be affected by the complexity of orders, more uncertainties also imply more modification and negotiation. Hence, this could force ETO companies waiting a long time, which can also negatively affect the preparing speed. In a word, accurate due date quotation ensures high delivery performance. Long processing time and waiting time lead to slow preparing speed; therefore, which turns out poor delivery reliability.
Methodology
The goal of this research is to gain insight into the causes the delivery performance problems of typical ETO companies. In order to answer the research question, a case study approach was taken to analyze the research problem based on the framework of my conceptual model. The advantages of this approach are the richness of data which ensues from data and informal information gathered by visiting the company and having face-to-face meetings with representatives of the company. Seeking triangulation could strengthen construct validity of a research (Leonard-Barton 1990). Quantitative data analysis and qualitative data analysis were applied to discover the solutions, which means mixing qualitative and quantitative methods were built to form triangulation in action to ensure the validity of research. Triangulation is broadly defined by Denzin (1978) as the combination of methodologies in the study of the same phenomenon.
12
used to gather rich quality data. In this study, open questions were provided to the interviewees, since those questions allow the respondent to answer without the limitation of implied choices1 (Steinar 1996). This also took the feature of research aim into consideration, because the goal is to find the causes and solutions instead of comparing the impacts of existing causes (there are no existing causes mentioned in previous research). Finally, conclusions can be drawn from the comparison of answers from different interviewees.
Case selection
All the relevant data of variables will be gathered through a typical company, which must be a SMEs ETO company that is struggling with the delivery performance problem. Afterward, data can be gathered from this company and the researcher will analyze these data to find the major causes of delivery problem. Finally, the main problem will be focused to find whether there is a solution to solve the problem. As Yin mentioned in Case Study Methodology, a single case study is suitable for the typical or longitudinal case study. In this research, a typical SMEs ETO company is required to gather information for deep analysis. Ramix/Romit B.V. (Hereafter referred to as RaRo), is a small to medium sized production company with approximately 65 employees. The company produces and engineers a wide range of moulds for the plastics industry and is a partner in machining projects, almost every product is ETO. Therefore, all these characteristics indicated that RaRo is a typical case for this research. In addition, their current delivery performance is around 75-80%, and they want to improve this to at least 90% or better. However, the utilization rates of machines on the shop floor have already been improved over the past years. Now they are faced with the problem that pre-production stage occupied too much time, which push too much pressure to the production stage to keep delivering on time.
Research structure
As I mentioned before, mixing qualitative and quantitative methods are built to form triangulation, which was the basis of data collection. The structure was divided into two main parts; the first part was aimed at understanding whether the problems were the average lateness or the variance of lateness. Therefore, quantitative data analysis was applied in part one. The aim of the second part was to gain insight into the roots of the causes. The research method was based on previous work by G. Soepenberg, (2012). This method is divided into 4 steps, step 1: analyzing the distribution of
1
lateness, step 2 analyzing difference among order subsets, step 3 analyzing differences across time, and step 4 analyzing the differences between delivery time promising and realization process. In between with step 2 to step 4, lateness is divided into an average-oriented diagnosis and variance-oriented diagnosis. Details and the diagnostic tree were introduced in the next part.
Since both high average lateness and high variance of lateness implied poor delivery reliability, diagnostic tool was applied to analyze different orders to classify the problems into 2 orientations.
In this research, step 1 is to detect where the problems (Average lateness / variance of lateness) are. Drawing the histograms of the lateness of the orders can classify the problem into the average lateness or the variance of lateness. All raw data was classified into 3 types. Specific problems were classified into the average of lateness and variance of lateness based on the method mentioned in theoretical background part (a key performance indicator of the delivery reliability).
Step 2 developed further analysis of the problems. Checking the average waiting time and processing time on orders based on all relevant department gain insight into the causes of problems. Figure 3.1 is the flow diagram of the pre - production stage. On the other hand, the variance of lateness was relatively high implied special late order should be diagnosed. Whether they made the right due date quotation should also be checked in this phase.
Figure 3.1 process of pre-production stage
Step 3 is the phase to detect real causes of the problems. Open questions especially for “why” questions were put forward, which were all related to average processing (or waiting) time too long or the variance of processing (or waiting) time too long. In case the causes turned out to be the faults of customers or suppliers, related questions were also made to be specific to these points. For example, questions related to internal or external integration as I mentioned in the conceptual model. However, these questions also had relationships with accuracy of the due date quotation. With regards to long average lateness, open questions were sent to all relevant departments. For variance of lateness and due date quotation problems, open questions were prepared for specific orders. Therefore, with the help of answers from different departments and different persons, information related to causes can be gathered. Afterwards, within case
14
Figure 3.2: research structure
Quantitative data
Qualitative data
Delivery reliability
Average lateness? Variance of lateness?
A ver ag e pr o ces sin g t im e A ver ag e w ai ting tim e V arian ce o f pr o ces sin g t im e V arian ce o f w ai tin g t im e Analyze special orders
Analyze each step of pre-production stage
Which blocks are the roots of problems?
16
Data collection
A single case study will be used in this research, and the data will be gathered from an ETO company (Ramix/Romit), because this company is a typical ETO company that has delivered performance problem as I mentioned before. Raw quantitative data were collected from the ERP-system of RaRo, which include all information of orders accepted from 2013/02/01 to 2013/04/01. Since the new ERP-system was applied last year, RaRo can only provide the latest 3 months’ data and all relevant data are reliable and credible for the results.
However, there are three kinds of orders, which are Jobbing, Rollepaal, and Moulds. For Moulds, the engineering department of RaRo is responsible for drawing. Afterwards, work preparation department starts purchasing the raw materials and making some regular work preparations. In the next step, planning department and work preparation print the production routing cards and release the order to shop floor. Figure 3.3 shows the flow diagram of the pre - production stage. On the other hand, Jobbing and Rollepaal have a different characteristic with Moulds. The customers rather than engineering department offer the drawing to RaRo.
Figure 3.3 Process description
The qualitative data were collected through interviews with open questions based on the results of quantitative analysis. In order to answer the questions, the interviewees had to consult with their company’s information systems or relevant colleagues. All of the interview conversation was recorded by digital devices. All the interviewee gave permission in advance for the researcher to record all relevant information. The answers of open questions were gathered from 3 different departments (the sales department, the engineering department, and the work preparation department), which also improved the validity of this research (Karlsson 2009). Furthermore, in order to validate the gathered data, the summary and results of information outcomes were sent back to RaRo. Therefore, the interviewees were able to correct any mistakes or misunderstandings and provide answers to the remaining questions, which can also ensure the reliability of data. Finally, in-company consultation can also guarantee the
reliability of research, which means related persons with specific open question should reach an agreement on both the answers and the conclusions of interviews. Table 3.1 shows the methods to gather all relevant data, and the department should be involved, which explained more in details related to Figure 2.2 and Figure 3.2. All details information about open questions can be found in the appendix.
Table 3.1 data classification
Results
In this phase, the answer of the research question can be provided. Several steps were applied to find the results. Step 1 belonged to quantitative analysis. In this step, histograms and regular calculations of raw data illustrated whether the poor delivery performance was caused by average lateness or variance of lateness, or inaccurate due date quotations. However, this step was also a pilot study preparing for further interview researches. Step 2 detected the problems following two directions, which were based on each department of pre-production stage and special too late orders. Step 3 took a deep insight into the roots of problems following the guide of independent variables mentioned in my conceptual model.
Variables Methods Relevant department
Delivery performances Quantitative data ERP System
Planning department
Processing time Quantitative data ERP System
Planning department
Waiting time Quantitative data ERP System
Planning department Accuracy of the due date
quotation
Quantitative data ERP System
Planning department
Complexity of the orders Interview Engineering department
Capacity of engineering department
Interview Engineering department
Planning department Uncertainty of order
specifications
Interview Sales department
Uncertainty of suppliers Interview Planning department
Information exchange Interview All
Sharing plans Interview All
Uncertainty of capacities Interview All
18
Delivery performance problems
In order to analyze the data more accurately, the raw data were classified into 3 classifications based on the sorts of order in RaRo. Furthermore, during the process of data handling, some incorrect data was deleted.
In general, figure 4.1 illustrated that delivery reliability problem with Jobbing belonged to the average lateness of delivery reliability to the customers (instead of the production stage). On the other hand, the problems of Rollepaal can be classified into the variance of lateness. Due to the restriction of raw data (no record of due date to the production stage), whether these problems were caused by the pre - production department or machine failure in the shop floor should be paid more attention. During the interview phase, if the some poor delivery reliability problems were the consequence of machine failure, all these causes would not be taken into consideration. Since the numbers of Moulds are relatively few, histogram cannot show where the problems are. Raw data of Moulds only show a majority of orders were late. Therefore, special Moulds orders were focused to be detected by open question interviews. The horizontal axis indentified the lateness of orders (the negative values mean the products were delivered earlier than the promised due date, while the positive values mean how many days later than the promised due date). The vertical axis show how many orders have been delivered. Therefore, these 2 paragraphs also show that not too many orders have been delivered late. However, when we take a deep insight into the raw data, some unexpected findings were discovered. RaRo calculated their delivery performance based on each part of the project instead of each whole project.
Figure 4.1: histogram (Jobbing) Figure 4.2: histogram (Rollepaal)
For example, there are 10 parts of a huge project. If they can delivery 9 parts on time and the last one 60 days later, the delivery performance is 90%. We assumed that customers were not satisfied with even one single part late, since they cannot start the
whole project without any single part. Hence, we calculated the delivery performance of recent 3 months. Based on parts of the project, the delivery performance of Jobbing was 65.1%. However, based on each project, the result was 69.6%. On the other hand, the delivery performance of Rollepaal turned out 67.2% (based on parts) and 54.8% (based on projects). Even though 3 months data could be not enough, these values can also imply that the delivery performance of RaRo is not as high as they expected. Therefore, we concluded this factor as improper calculation of delivery performance, which is a new discovery besides all factors mentioned in the conceptual model. Table 4.1 illustrated the problems of each kind of products. Next step further analysis was applied by an open question to detect the causes of all problems.
Table 4.1: delivery problems
Causes of the problems
Table 4.2 illustrated the reasons why an order delivered too late. All relevant information was collected from the open questions interview. People from the work preparation department (also in charge of procurement), the sales department, and the engineering department were involved in this interview. Tracking back to my research question and conceptual model (in the middle part), this phase provides an answer to the initial causes of problems. Through comparing and classifying between different person’s answers as well as my own observation, table 4.2 shows the reasons of poor delivery reliability. Based on these answers, all factors were graded. If one factor is mentioned once, it will get 1 point. More than two points were defined as real causes, which provided a direction that in the next phase we should pay more attention to the blocks linking to this one. For example, waiting time linked to the complexity of orders and capacity and experience of the engineering department. All details information about interviews can be found in the Appendix.
Work preparation department
During the open question interview, the interviewee is Ronald Vinke who is in charge of the planning and preparation about all types of orders provided answers.
“In pre production stage I would say; waiting time. To process an order (from making a quote to print the routing cards) is not that difficult or time consuming. Sometimes the routing cards have already been printed while the materials arrive more than a week later.”
The answer shows that for Jobbing and Rollepaal they can prepare everything quite
Jobbing Average lateness/improper calculation
Rollepaal Variance of lateness/improper calculation
20
fast, however, they have to wait for a very long time due to lots of unforeseen. Since the work preparation department is in charge of planning the whole stage, they must clearly know the capability of each stage to make an accurate plan. Therefore, this enhanced the reliability of their answers. They claimed that all their departments can react fast to orders, but they have to wait for the correct data or specification from customers. This lead to a long waiting time
However, they claimed the percentage of waiting time and processing time is hard to estimate. We made a rough estimation based on data collected from the interviews. For Rollepaal and Jobbing, average leading time is 4 weeks. As they claimed, sometimes raw materials arrive more than 1 week later, which is included in the leading time. 25% could be the rough estimation of waiting time. However, this only including waiting for raw materials. If we take other waiting times into consideration, at least 35% waiting time and 65% processing time.
“For moulds: engineering, this always takes more time than calculated. The products are too complex (70-120 production parts defined by RaRo), the customer needs to confirm details, and drawings of all details take more time.”
Since the engineering department has to make the drawing, this process takes the longest time (4-5 weeks on average) in the pre – production stage. They also mentioned every time an order comes, the engineering department has to make a new drawing. Furthermore, the drawings have to be approved by the customers; otherwise, they have to modify them (10%-15% need to be modified, and orders need to be modified 1-2 times on average). The negotiation implied a long waiting time, on the other hand, the complexity and modification of drawing implied a long processing time.
In terms of “how to make the due date quotation?”, the mentioned the following decription.
“Let the system calculate the due date based on budget hours / contact with planning about the capacity figures on the requested machines.”
The system checks the available capacity on work centres, fills in the calculated capacity in the free space available and takes into account the external operations to some extent. It seems to be no problem with the due date quotatopm. However, a few orders were still delivered late.
“Because the customer normally asks for fast delivery, and we don’t like to say “no” to order.”
This implied they promised too early for the customers. In conclusion, the waiting time and promising too early were regarded as two main causes of problems. In general for simple projects, processing time was not a problem. However, for some special complex projects, long processing time can also be the cause of poor delivery reliability.
Sales department
“Average waiting time for answers from customers results in longer processing time. Waiting time makes the pre-production stage long.”
As Evert Zweert, the project leader for the sales department mentioned, waiting time was the main cause of the high average lateness. Since the sales department was in charge of negotiating with customers, they always tend to wait for the answers (e.g.: specifications) from the customers. Therefore, waiting time can be a factor to result in average lateness. However, which kinds of actions really leading to this waiting time still cannot be discovered in this phase. Conclusions were drawn after comparing and classifying all finding from the answers.
The following statement aimed answering “how do you make due date quotation in general?”.
“Mostly in the quote for a mould it is stated that RaRo will handle a delivery termin of 8 weeks for example after receiving the right information. This can result in nice discussions with customers.”
From this answer, we noticed that the quota will not be made until they receive the right information from customers. All due dates had been calculated based on recent workloads and capacities, which implied a quite accurate due date quotation can be made.
From the perspective of sales department, waiting time was pretty sure the causes of problems. On the other hand, they did not mention processing time. From observations and calculations of raw data, we find the processing time of the sales department is not a problem. They always do their job quickly. Finally, wrong due date quotation could not be the cause, since they had nice discussions with customers to ensure getting the right information.
Engineering department
22
As Niek Kodden engineer at RaRo mentioned, average waiting time and the average processing time could both be the causes. On the contrary to sales department explained, they claimed that the customers did not exactly know what they really want when they sign a contract. Therefore, this lead to more time spent on gathering the right information and parameters. Finally, modifications occurred, which resulted in long waiting time and processing time, which was also mentioned by the work preparation department.
“If all the details were known a complete mould can be built from 0 till all drawings finished in approximately 2,5 weeks. Now this takes around 4 weeks.”
“I cannot really estimate a % of waiting time or processing time, but around 25 % of the time can be saved if all details and requirements were known and an engineer could just start with his/her work.”
This is another evidence to prove there is long waiting time and processing time, due to normally 2.5 weeks could be enough to finish a drawing, while 4 weeks in practice. On the other hands, the answers also implied there was some problems in due date quotation. Due to the sales persons claimed quotation cannot be made until right information has been gathered. However, engineers said after the contracts have signed, customers still did not know exactly what they want.
Departments Causes
Work preparation department
Waiting time
Processing time for complex project Due date quotation for complex project Sales department Waiting time
Engineering department
Waiting time
Processing time consequence of waiting
Due date quotation before gathering right information
Waiting time Processing time Due date quotation
3 2 2
Table 4.2: original causes of problems
Deep insight into the causes
Table 4.3 shows the roots of the problems based on the original conceptual model and interview answers. Tracking back to the conceptual model (the beginning part), all open questions for this phase were based on this part. The goal of this phase was to take insight into the causes of problems, therefore, the findings of first interviews were the basis of open questions in this part. Since the waiting time was the main cause of problem, interview questions should be mainly related to this factor. On the other hand, the other two factors should also be taken into consideration. Furthermore, with regards of the conceptual model, all questions were specific to all blocks (e.g. Complexity of orders) related to the results of table 4.2. Finally, following the same rule as table 4.2, comparison and classification between the answers of every department illustrated the roots of problems. All relevant information about interviews can be found in the Appendix.
Work preparation department
Taking into a deep insight into the late orders, the majority reasons for orders delivered late were complexity of projects.
“Products are too complex, customer needs to confirm details, drawings of all details take more time, “invent the wheel again”, unforeseen things or extra things that have been added are all common factors.”
24
more complex an order is, the more percentage modification is required (Jobbing and Rollepaal: 0–5%, Moulds: 10-15%). Too many uncertainties implied wrong due date quotation.
“Lacks of capability? No. Our departments have overall quite skilled employees.. However we are
not quite capable to think about doing things similar to previous jobs sometimes. (Like modular design tricks or standardization)”
“By weekly meetings about status updates with respect to projects, and also “open door policy”, planning and work preparation is in the 1 area, Communication with the rest of the company on daily contact (phone, mail, and drop by in the office).”
“Sometimes we do task double or we lack to do it because of bad internal communication. (This has been improved by weekly meeting)”
“Because sometimes people start to forget some things because they are too busy. Or just not communicate important things.”
Capacity and internal integration were not the causes of poor delivery performance, as they mentioned they do a really good job in these points. However, from the observation as well as interviewees mentioned, RaRo always redo some similar projects, and they never think about whether it is a waste without standardization. This could be a potential factor can be improved.
Furthermore, because a good internal communication (weekly meetings), there are almost no possibilities that one department waiting for a long time due the capacity of next department is full. This ensured two key points. One is that better internal communication ensures more reliable information. The other is enough time for the engineering department to finish drawing on time, if there is no modification.
Weekly meetings can improve internal communication, however, RaRo also mentioned that some problems due to bad internal communication. As they explained there could be some incidental problems such like key person absent the weekly meeting due to illness or changes forgot to be informed to the key person. These problems cannot be avoided.
“External communication can be improved; this can also affect the customer because sometimes both planning, work preparation and production does not contact the customer that their product will be late.”
“External communications are not poor, sometimes it happens. I have to deal with all the transportation and that means that for all the external operations I have to call suppliers if they deliver on time, and sometimes you just have to trust your supplier that they will deliver on time.”
their products will be late. They can only get information when the products were not delivered on time. On the other hand, raw materials late or changes of specification can also not be informed to RaRo on time. For example, suppliers made mistakes such like delivering wrong parts to RaRo. The majority of orders from Siemens cannot be delivered on time due to the lateness of raw materials. The raw materials were delivered by their customer (Siemens). Considering the differences between their customers (Rollepaal and Siemens), Rollepaal, as their mother company, who is really close to RaRo (located next to RaRo). On the other hand, Siemens located in Germany. Transportation time is one of the key factors of uncertainties, as well as other suppliers who are not close to RaRo. Improving the ERP system could be a good solution to improve external communication (Woodside 2007). For example, building connection between RaRo and their major suppliers, which implied they can share information to ensure all raw materials delivering on time.
Sales department
“Complex projects are for example moulds. See questions 1 and 3. First a “loose” delivery date is set in the quote. After signing the contract details are often not known, which results in a shift of the pre-set plan.”
Since complex project was one potential cause of waiting time as the conceptual model mentioned, one important question during the interview tried to gain insight into the relationship between complexity of orders and performance. The answer indicated that the sales department provides “loose” delivery date to prevent long waiting time, processing time, and error of quotation. However, we noticed that one point was different with the statement in the previous interview. For complex projects details are often not known, which further proved that details unclear are the causes of problems. “Cannot say no to the customers” could be the reason, which implied even the customers did not accept the “loose” delivery date; they still had to accept the orders.
“For each project there is a project-map which is available for everybody. Next to that the information system can show a lot of information to the persons involved about the progress and costs made. And there are also weekly meetings were key-points are mentioned.”
26
“Communication is okay, however sometimes not effective because employees are too busy or plans with pre-phase / production / schedules / (could be anywhere in the company) need to be changed (that is also because we have to deal with 3 product groups). Besides that the new information system can provide us with some information; however some reports are still not working properly.”
Moreover, as the sales person mentioned, communication is still not effective enough, due to their employees are too busy. From the observation, there could be two reasons. They accepted too many orders, since they never say no to customers and deal with 3 product groups. On the other hand, they installed new ERP system; persons still need time to proficient use the system. Both assumptions are proved by the answers.
Complexity should be a key factor, since it influenced the waiting time, which further influence the processing time. Furthermore, even there is “loose” delivery date for complex projects, misunderstanding about what is the right information still affects the accuracy of the due date quotation. In other words, they tried to provide tolerances for the uncertainty of customers and capacity. However, there still have spaces to be improved. Familiar with the finding in the work preparation stage, they also realize the importance of reliable information, which can be discovered by weekly meetings and new ERP system. From the perspective of this point, only some incidental problems and degree of proficiency can be improved.
Engineering department
“Simplify a special mould is not that easy, but we try to find examples from our moulds history over the past decades to find similarities.”
Considering the long waiting time, even though they did not mention complexity, they themselves claimed to simplify a special mould is not easy. Therefore, complexity is proved also a cause of the waiting time in this phase. However, they tried to find examples from previous moulds, which had a similar statement as work preparation department claimed. Learning from the previous experience could simplify the project, which reduced the waiting time and processing time.
“But on the other hand, if we make a quote for a mould which the chance that we will get the order is around 10 %, we are not going to make a very detailed plan on this certain mould, (it will cost a lot of time for special moulds and if you won’t get the order, all the time spent on it is for nothing).”
providing an assignment or internship for the student, who can classify the historical drawings. In this way, they easily retrieve drawings to ensure their new similar projects easier.
“On forehand the engineer / person that calculated the mould based on the quote. After making drawings the plans can still be changed due to the work preparation, if they have some other ideas that might work more efficiently in the factory.”
They also pointed that reliable information and sharing planning are important, but plans can be changed to make inconvenience. If all relevant reliable information can update on time from the system, too much modifications can be avoided.
“We sit close to each other, short lines and weekly meetings.”
In terms of how to realize internal integration, they also mentioned weekly meetings. Due to negotiate with the customers is the works for sales and with suppliers is the works for work preparation; external integration was not taken into consideration in this phase.
There is no doubt that complexity is the main causes of problems in this phase. All mentioned problems are related to complexity or how to simplify, which is also related to make full use of capability of the engineering department. Furthermore, internal communication was also mentioned. More reliable information can surely reduce the frequency or percentage of modifications. Both two factors above are realized by weekly meetings. Since there is 10%-15% percentage of modification for complex projects, customers unclear what they really want enhanced the uncertainties. Therefore, the uncertainties of customer implied long waiting time. Even though sales persons claimed that they received the right information before signing contracts, this implied in sales phase they did not clearly know the right information for the engineering department.
Special late orders
“Too complex products; few parts arrived too late (and one of them turned out not to be correct)
and there are some external steps in this product which took longer than expected.”
28
evidence that complexity of projects is the cause of poor delivery performance.
Table 4.3: roots of causes
Table 4.3 illustrated the roots of causes, which is concluded from the results of 3 departments’ interview and special order discussions. We marked the causes in the following rules. If one factor is mentioned in an interview, this factor will get 1 point.
Discovered from where? Scores
Simplifying projects by previous drawings (Complexity of the orders)
Work preparation, sales, engineering, special orders
-4
Simplifying projects by previous drawings (Making full use of capability and experience )
Work preparation, engineering -2
Weekly meetings
(Reliable information and Sharing planning between departments)
Work preparation, sales, engineering +3
Project-map (reliable information)
Sales +1
Incidental problems (too busy to forget informing changes)
Work preparation, sales -2
Incidental problem (suppliers mistake for providing wrong parts)
Work preparation -1
Incidental problem (Absence due to illness)
Work preparation -1
Enhance the ERP system (uncertainties of both suppliers and customers)
Work preparation, sales -2
Gathering right information or not before orders accepted (uncertainties of customers)
Different claims by sales department and engineering department
However, plus grades or minus grades depend on whether RaRo has already done a good job or needs to be improved with regards to this point. Reducing the complexity is the first priority to take actions. Meanwhile, making full use of capability and experience can also be improved at the same time, due to the relationship mentioned before in this chapter. Enhancing the ERP system is an unexpected finding, which was not mentioned in my conceptual model. As time goes by, this can be surely improved. Another unexpected finding is the misunderstanding between the sales department and the engineering department. This due to the reason mentioned by work preparation department, the sales department cannot say no to customers even they did not gather enough information. The last unexpected finding is wrong calculating rule of delivery reliability, which reduced customer satification. However, the incidental problems are inevitable, because nobody can control illness or forgetting. On the other hand, weekly meetings are one positive factor to ensure reliable information and better communication. Therefore, all these positive factors should keep moving.
30 Finger 4.1: revised conceptual model
- + - + - + - - + + + - + + + - + - -
Discussion
The main purpose of this research is to discover the main causes of poor delivery performance problem through diagnosing the pre-production stage of the ETO Companies. Therefore, a gap can be filled, due to previous research only focus on the production stage of the ETO companies. The results of a typical single case study illustrated that all relevant variables have more or less impacts on the delivery performance. Specifically: more complex projects implied more uncertainties in pre-production stage, therefore, poor internal and external integration turned out long waiting time and wrong due date quotation. In order to ensure a high delivery performance, companies should enjoy a better communication with both suppliers and customers. Furthermore, ETO companies should also spend more effort on the internal communications, since selected case applied weekly meeting to ensure the reliability of the information.
Based on the causes have been found, some suggestions were also provided. Suggestion 1 is aimed to solve the problems caused by the complexity of orders and capability of the engineering department. As mentioned before, engineering department always redo some complex previous projects, to some extent, parts of the projects are similar. Since redo previous projects means less innovativeness but efficient, now they are trying to find a solution to improve efficiency. Standardization can easily lead to a focus on efficiency, which implies saving lots of waiting time and processing time (Konijnendijk 1994). The engineering department can make a list of different types of project based on the previous orders, therefore, customers can select from the list and engineering department can only modify the drawing based on the customization. If RaRo considers this wasting too much time of the engineers, internship trainees or university students could be hired to finish this job.
32
the buffer is relatively lower than the cost of contacting the supplier, since some simple parts are really cheap.
Considering the new ERP system, they can build connections with their main suppliers. Otherwise, providing authorities to suppliers and customers (Woodside 2007). On one hand, suppliers can update information about delivery time of raw material. On the other hand, customers can get information with on time update of the status of projects. In this situation, internal and external communications can be improved, simultaneously; RaRo can reduce the workloads of staffs such as connecting with phone or email.
The last point is drawn from the result is that calculate the delivery performance based on customer satisfaction. A customer would like to get the whole project on time rather than some parts quite early while few parts relatively late. Calculating the delivery performance based on part may show a better delivery performance than based on the whole project. However, companies will lose customer satisfaction, which means lose market share in the competitive environment in the long term.
Conclusion and future research
Tracking back to my original research question of this thesis: how to organize the pre-production stage to realize the high delivery performance, several advises were provided to reduce the complexity and uncertainty in the pre-production stage. As the interview results mentioned, weekly meetings and open door policy improved the internal communication. Keeping all relevant departments communicating with customers and suppliers ensure a better external communication. Furthermore, building a large ERP system network to ensure efficient connections with main suppliers and customers can improve the reliability of information as well as remove uncertainties. Finally, even an ETO company has enough capabilities to handle quite complex project, standardization of moulds (which includes lots of complex but similar orders) can be a potential solution to make full use of the ability of the engineering department.
method suitable to the pre-production stage Finally, this research confirmed the previous findings by Land & Gaalman (2009).
The main managerial contribution of this study is that this research provides several reasons of why typical ETO companies took longer time to prepare in the pre-production stage. Managers should try their best to enable a better internal and external communication to keep the customer satisfaction in first priority. Furthermore, they should also make a tradeoff between the costs of keeping inventories or rush communicating with the suppliers. Finally, they should standardize the existing products to maximize the capability. Combining of all suggestions above should result in an improvement of the delivery reliability.
Several limitations of this research can be noticed. On the limitation is the scale of raw data is relatively small, which is only 3 months finished orders. Therefore, the raw data can only support pilot study. This limitation could lead to some error. Furthermore, due to the limitation of raw data (there is no record mentioned the waiting times and the processing times of each department in pre-production stage), some more accurate analysis tools cannot be applied in this research. For example, the combination of throughput diagram and order progress diagram is another potential solution. Measured by Throughput diagram, which is helpful to draw conclusions from the input and output control decisions through the whole time period. Measured by Order Progress Diagram, which indicates the progress of the individual compared to the average order progress pattern, we can observe which orders are delayed and which are speeded up for each stage of order in pre-production phase. Finally with regard to the generalizability, single case study is not enough to draw a suitable conclusion for all ETO companies, due to different characteristics of all ETO companies.
Comparing with the previous research, this one takes a deep insight in an underdeveloped area (the pre-production stage). Similar diagnostic tools (histogram) were applied in this research as the diagnosis of the production stage. A practical revised conceptual model was built to discover the roots of poor delivery performance. However, there are still some drawbacks in this research. For example, as mentioned before, the research data are relatively small, which also result in more accurate diagnosis tools cannot be applied in this research. This could be interesting to research in the future.
34
(SMEs ETO companies who are struggling with poor delivery performance) are advised. Therefore, the conclusions can be generalized to find the roots of delivery performance problems of all ETO companies.
Acknowledgement
First of all I want to thank my two supervisors who provided me significant support through this research, namely Erik Soepenberg as the supervisor from the university and Ronald Vinke as the supervisor from Ramix/Romit Company. Erik always answered my question and reviewed my thesis on time. He also provided great supports to keep me in the right direction of research and how to get rid of my restriction of my bachelor education background. Ronald offered me great helps in every aspect. Every time we had a meeting, he picked me up to the company. He told me how he thought during the master thesis phase to help me build my research. He spared all his effort to help me collect all information, even though he was really busy with his own works.
I also want to thank the co-accessor Stuart Zhu. He provided me useful feedback during the final draft presentation. Furthermore, I would really want to thank Niek Kodden (Engineer at Ramix/Romit) and Evert Zweers (project leader at Ramix). I really appreciate the patience you spend time on my questions. Last but not least, I want to say thank you to my two friends, David Han and Chongjing Jiang, who helped me a lot in checking grammar and spelling mistakes.
Thank you so much to everyone who helped me.
Reference
Anon, 2003. mature supply chain planning. Logistics & Transport Focus, 5(10), pp.9–9.
Bertrand, J. & Muntslag, D., 1993. Production control in engineer-to-order firms. International Journal of Production Economics, 30, pp.3–22.
Carroll, G. & Hannan, M.T., 2000. The Demography of Corporations and Industries., Princeton University Press.
Droge, C., Jayaram, J. & Vickery, S.K., 2004. The effects of internal versus external integration practices on time-based performance and overall firm performance. Journal of Operations Management, 22(6), pp.557–573.
Ebadian, M. et al., 2009. Hierarchical production planning and scheduling in make-to-order environments: reaching short and reliable delivery dates. International Journal of Production Research, 47(20), pp.5761–5789.
Greve, H., Pozner, J. & Rao, H., 2006. Vox Populi: Resource Partitioning, Organizational Proliferation, and the Cultural Impact of the Insurgent Microradio Movement1. American Journal of Sociology, 112(3), pp.802–837.
Gupta, D. & Wang, L., 2007. Capacity Management for Contract Manufacturing. Operations Research, 55(2), pp.367–377.
Hicks, C, McGovern, T. & Earl, C., 2000. Supply chain management: A strategic issue in engineer to order manufacturing. International Journal of Production Economics, 65(2), pp.179–190. Hicks, Chris, McGOVERN, T. & Earl, C., 2001. A typology of UK engineer-to-order companies.
International Journal of Logistics Research and Applications, (March 2013), pp.37–41. Jovanovic, B., 2004. The pre-producers. , (September).
Karlsson, C., 2009. Researching Operations Management,
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. Kingsman, BG, Tatsiopoulos, I. & Hendry, LC, 1989. A structural methodology for managing
manufacturing lead times in make-to-order companies. European Journal of Operational Research, 40, pp.196–209.
36 Konijnendijk, P., 1994. Coordinating marketing and manufacturing in ETO companies. International
Journal of Production Economics, 37, pp.19–26.
Land, M.J. & Gaalman, G.J.C., 2009. Production planning and control in SMEs: time for change. Production Planning & Control, 20(7), pp.548–558.
Land, MJ, 2004. Workload control in job shops, grasping the tap. University of Groningen.
Leonard-Barton, D., 1990. a dual methodology for case studies: synergistic use of a longitudinal single site with replicated multiple sites. organization science, 1(1), pp.248–266.
Little, D. et al., 2000. Integrated planning and scheduling in the engineer-to-order sector. International Journal of Computer Integrated Manufacturing, (April 2013), pp.37–41. Milgate, M., 2001. Supply chain complexity and delivery performance: an international exploratory
study. Supply Chain Management: An International Journal, 6(3), pp.106–118. Raaymakers, W., 1999. Order acceptance and capacity loading in batch process industries. Rahim, A.R.A. & Baksh, M.S.N., 2003. The need for a new product development framework for
engineer-to-order products. European Journal of Innovation Management, 6(3), pp.182–196. Soepenberg, G., 2012. A framework for diagnosing the delivery reliability performance of
make-to-order companies. International Journal of Production Research, (June 2013), pp.37–41. Soepenberg, G.D., Land, Martin & 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.
Steinar, K., 1996. Interviews An Introduction to Qualitative Research Interviewing,
Sun, D., Shi, H. & Liu, C., 2012. An order planning and scheduling framework in MTO environment. International conference on system of Systems Engineering, 7(2), pp.682–687.
Appendix:
Appendix A
Open questions
Work preparation department
First interview:
Since the average lateness is too high, whether is the average waiting time too long or average processing time too long?
- In pre production stage I would say; waiting time. To process an order (from making a quote to print the routing cards) is not that difficult or time consuming. Average waiting time, waiting for the correct data, answers from customers, purchased materials etc. takes the on average the most time. So processing time is not the problem here. Our work preparation can do things quite fast. It is similar to Rollepaal, sometimes the routing cards have already been printed while the materials arrive more than a week later.
As you mentioned, the waiting time could be the problem. Could you make an estimation of the total leading time % is waiting time? And % is processing time?
- No, Hard to say
On average, which department of pre-production stage takes the longest waiting (processing) time?
Why this department takes that much time to prepare?
- For moulds: engineering, this always takes more time than calculated. The products are too complex, customer needs to confirm details, drawings of all details take more time, “invent the wheel again”, unforeseen things or extra things that have been added are all common factors.
Why in general lots of due date quotation are too early?
Because the customer normally asks for fast delivery. And we don’t like to say “no” to order. And next to that we sometimes think we have the resources, capability and time to finish in the requested time. (Can we negotiate with customers about the due date, to make a reasonable due date)
How do you estimate the due date?
Let the system calculate the due date based on budget hours / contact with planning about the capacity figures on the requested machines.
Can you general explain what does the system do exactly to calculate the due date?
38 Deep insight interview:
Why this department always waits so long?
Too complex? Yes to some extent as I mentioned previously. How do RaRo define a project as complex project? How many parts?
That is a mould; which contain approximately between 70 – 120 production orders / lines (those lines assembled together are the complete mould)
More complex than the abilities of the engineering department? No
Lacks of capability? No. Our departments have overall quite skilled employees.. However we
are not quite capable to think about doing things similar to previous jobs sometimes. (Like modular design tricks or standardization)
Based on the complex drawings of engineering department for Moulds, some questions came up with.
How long average did engineering department take to make a drawing? Depends on the
difficulty of the drawing. For a complete mould a couple of weeks.
For complex moulds projects, how long did they take on average? 4-5 weeks.
How many percent of orders require modification? For Rollepaal and jobbing only a few.
For moulds it is a bit more, because some sizes are not correct or missing or things are unclear that were on forehand unforeseen. Rollepaal and jobbing is 0-5 % and moulds maybe 10-15 %. (For a complete mould with 100 parts, thus 15 parts where the drawing is not correct) those problems are most of the times solved quickly.
How many times an order requires modification? On average. Should not happen, but
if it happens then only once, occasionally two times. Sometimes a customer sent us a drawing revision.
Why orders require modification a lot? You should have agreed on a specification during negotiation. Yes we have; if the customer says that the mould is correct as we designed
it and later on things are changing the delivery date might also be shifted and the extra hours need to be paid. It is the same for jobbing and Rollepaal.
If an order is too complex, do you have to outsource it? Which kind of order has to be outsourced? For example? Normally we only outsource if we have too many works that
we can’t handle it ourselves any more. Work preparation and sales know which sizes our machines can handle etc. The only thing that is outsourced is some external operations like welding or surface treatments what we can’t do or are not certified for.
Why this supplier (customer/department) is not reliable? Can we find other more reliable suppliers?
Some departments are under high work pressure; If there is a lot of work to do (or too much) you start to forget things. We are trying to do something about that, which started with a new ERP system.. This eventually should gain more insight in the processes and should make work easier.
Can we build the buffers?
