Delays at the Pre-production Stage in Engineer-to-Order Companies

Delays at the Pre-production Stage in Engineer-to-Order Companies

Author: J. He

Student number: 2074966 Date: 21-06-2013

MSc Technology Operations & Management

Faculty of Economics and Business, University of Groningen Supervisor: dr. N.D. van Foreest

Table of contents

4.5.1 UNCERTAINTY,COMPLEXITY AND DYNAMICS 19

4.5.2 OTHERS 20 5 DISCUSSION 21 5.1 DELAYS 21 5.2 CAUSES OF DELAYS 21 5.2.1 COMBINATIONS OF CAUSES 21 5.2.2 MEASUREMENT CHANGES 22 5.2.3 PROJECT SIZE 23 5.2.4 CONCRETE CAUSES 24 5.3 GENERALISATION 25 6 CONCLUSION 26

7 LIMITATIONS & FUTURE RESEARCH 27

ACKNOWLEDGEMENTS 28

Abstract

On-time delivery is very important for engineer-to-order companies. Customers consider time as an important resource and delays upon delivery may harm the reputation of the company and endanger future prospective. Uncontrolled delays seem to emerge at the pre-production stage. By means of a single case research, in-depth insights on causes of delays at the pre-production stage are obtained. Results show that the typical characteristics (uncertainty, complexity, dynamics) of engineer-to-order companies, cause the delays at the pre-production stage. Further analysis shows that the most influential characteristic is uncertainty, and in particular specification uncertainty, which is caused by both internal and external factors of engineer-to-order companies. These findings are relevant for managerial purposes, because it reveals the concrete causes of uncontrolled delays. For scientific purposes, the findings provide narrowed down causes of delays for future research at engineer-to-order companies.

1 Introduction

Previously, extensive research covered many operation management issues in the high-volume manufacturing sector. As a result, many theories, control models and systems have been developed for this sector (e.g. MRP, lean, EOQ, CONWIP, etc.). Over the past decades, many researches have highlighted the divergent requirements of engineer-to-order (ETO) companies regarding requirements in general, information systems, supply chain management and production control (Wortmann, 1995; Gosling & Naim, 2009; Bertrand & Muntslag, 1993; Hicks et al., 2000; Aslan et al., 2012). These researches have indicated that proven systems and theories for high-volume sectors cannot be simply applied to ETO companies due to different characteristics of processes and lower volumes.

Next to identifying the causes of delays at the pre-production stage, this paper also aims to address possible solutions suitable to tackle the delay problems. To investigate whether the findings are representative for other ETO companies, this research’s results will be compared to results of other researches with the same topic, conducted at the same time, at different ETO companies in the Netherlands. In order to reach these goals, the following research questions have to be adressed:

1. Is there a delay problem at the pre-production stage? 2. What are the causes of these delays?

3. Are the causes applicable for every ETO company? 4. What are relevant solutions for this problem?

The aim of answering these questions is to have both managerial and scientific contributions. Managerial contribution in the sense that companies gain insights on how to reduce delays, and scientific contribution by identifying the causes of delays at the pre-production stage for further research.

2 Theoretical background

This section describes the theoretical foundation of the study. First there will be a short review on what the pre-production stage in an ETO company actually is and what activities are involved in this particular stage of production. The identified activities will be reviewed in further detail, to get valuable insights on the possible relationships between the activities and delays. The results from the review will form a conceptual model, which summarizes the proposed relations upon delays at the pre-production stage.

2.1

The Pre-production Stage

Attention is needed to define the pre-production stage. Searching for the terms “pre-production” and “preproduction” did not provide any scientific results that are related to ETO companies. The term is commonly used in the entertainment industry, where it stands for all preparation activities before the actual production of a movie or recording of a song. To get a better understanding of the pre-production stage at ETO companies, the focus has to be set on previous work of other authors in this field of research.

2.2

Sales

The process of sales in ETO environments is requires much more time, compared to other sales processes like the ones in high-volume standardized products businesses. Standardization possibilities are limited and requests are often one-of-a-kind in ETO environments (Konijnendijk, 1994). It requires high skills and collaboration between persons and departments to produce suitable, qualitative and feasible solutions that fits the customers’ needs, against a competitive price and at a short lead time (Zorzini et al., 2008; Elgh, 2012). Besides setting and estimating the due date, sales also have to estimate the required time for engineering, purchasing, manufacturing and testing the design (Hvam et al., 2006). Profit and customer satisfaction is partially based on the accuracy of estimations by sales. Tight due date setting might increase sales opportunities, yet it also increases the chance of not achieving it. Causing dissatisfaction or costly overtime for the company itself. Underestimating the required lead time has great impact on the profitability of the project and the performance of the entire company (Zijm, 2000).

Delays only occur when a target has been set that cannot be achieved. This brings us to the main relation between sales and delays. The due date is most likely to be set in the quotation phase, which falls under the responsibility of sales activities. As earlier mentioned the solution and design are already generally specified during the time consuming bid preparation. Availability of capacity and exact customer requirements remain highly uncertain during this phase (Wullink et al., 2004). Other bids may be accepted influencing the capacity, and customers tend to change requirements during the bidding process, in some cases without the understanding that the ETO company requires time to adjust the design and recalculate time and costs.

Quotations can either be accepted or rejected, causing fluctuations in available capacity. This decision-making process is not completely in control of the ETO company. The ETO company however does decide whether to prepare a bid or not. This decision has major impact on the availability of capacity and work in process of the entire company. Decisions upon preparing a bid or not may cause delays at the pre-production stage, because of availability of capacity and skill level of the company.

2.3

Engineering

deliver a working design according to specification. Other measures for productivity are: time to delivery, product costs and quality (Abdalla, 1999).

The amount of time required to engineer the pre-specified design has already been set during the sales activities. Tight setting of the required time for engineering may cause delays. Yet, this is not the only pitfall. In this phase rework has been identified as one of the root causes of delays (Caron & Fiore, 1995; Rahman et al., 2003). During the in-detail engineering of the design, issues may arise, which have been overseen by sales, making the initial proposed solution non-feasible. Another potential issue causing rework is inappropriate handovers between engineers or between sales and engineering. Humans tend to work in their own manner with their own systematics. Having insufficient or unclear protocols and guidelines for organizational activities (e.g. set-up of drawing or documenting quotations, orders, e-mails etc.) may result in differentiation in the way of organising business. Having handovers in that case may cause delays since the other person is not familiar with the organization of that certain project.

2.4

Procurement

Every organization requires supply of materials and/or services in order to keep the business running. Procurement is hold responsible for the supply of these materials and services. Having incoming goods incorrectly, not in time, of insufficient quality or overpriced influence the performance of the entire company (Monczka et al., 2009). The growing importance of procurement activities has been emphasised in many papers (Hicks et al., 2000; Tsinopoulos & Bell, 2009; Ellram & Carr, 1994). Still literature regarding delays during the procurement activities is rather scarce. Most papers in the field of procurement, purchasing and supply management focusses on the changing role of procurement within the supply chain and the value contributing function of it.

purchasers, which potentially cause delays at the pre-production stage. Ellram and Carr (1994) stated that procurement should be internally integrated with other functions like sales, engineering and manufacturing in order to prevent frequent rework situations.

2.5

Process Planning

Projects that ETO companies daily encounter after acceptance of the quotation require careful planning. Engineering and procurement planning, order release, logistic control and supervision of manufacturing flows are main concerns during the development and production process. Process planning i.e. project management, is responsible for planning and controlling these activities. Dynamics, uncertainty and complexity in the physical as well as non-physical goods flow, makes planning and control difficult (Land & Gaalman, 2009; Caron & Fiore, 1995; Bertrand & Muntslag, 1993; Little et al. 2000). Extensive literature is available regarding implementation of different (computer aided) planning systems like ERP, MRPII, dynamic scheduling etc. (Little et al., 2000; Aslan et al., 2012; Hicks et al., 2007). These planning systems unfortunately are mainly aimed at planning and controlling the production stage. This conclusion is supported by Land and Gaalman (2009) who also indicates most of the literature focuses on planning and control of the production stage while the delay problem lies at the pre-production stage.

Process planning is an iterative process, requiring planning throughout the entire development of the product (Hetem, 2000). Process specification may still be unclear during the sales and engineering activities and when it does so, still uncertainty remains regarding the final design. Rework and changes in design at latter stages potentially causes delays during process planning activities. Rush orders are likely to be present at ETO companies e.g. regular customers in a hurry to receive their order or spare parts. These orders have to be planned at a short notice, with the potential to delay other projects.

2.6

The Pre-production Process

cannot be analysed as separate departments, but rather as a complex stage where information exchange is highly intensive, uncertainty remains upon final specifications and changes during the development process are common.

2.7

Conceptual Model

The exploratory nature of this research requires more in general causes than the ones presented in the previous paragraphs. Limitation of scope is unfavourable for this type of research and could bias the outcomes. The potential causes therefore are categorised into more general terms namely changes, estimations and difficulty. The generalized terms correspond with the characteristics described by Bertrand and Muntslag (1993), Land and Gaalman (2009), Caron and Fiore (1995) and Pandit and Zhu (2007). They have indicated dynamics, uncertainty and complexity as typical charateristics of ETO environments. These established terms will be used in the remainder of this research (table 1).

* Initial category

Category Causes

Dynamics (changes*)

Customer requirement changes Non-feasible solutions by sales Incorrect specifications Change in specification Variability in capacity Uncertainty (estimations*)

Unclear specification Tight due date setting Lead time estimations

Uncertainty whether quotation is accepted or not Complexity (difficulty*) Handovers between persons and departments

Skill level of organization

Table 1. Categorisation of potential causes

3 Methodology

The main objective of this research is to identify the causes of delay problems during the pre-production stage at ETO companies. By exposing the concrete causes, solutions can be proposed to tackle the delays and decrease lead time of ETO projects. To reach these objectives an exploratory case research will be conducted. This chapter discusses the chosen methods to obtain the required data.

3.1

Method Selection

To gain insights on relations between the independent variables and delays, an in-depth exploratory case research will be conducted. Corrêa (1992) states that case research is suitable even if variables are difficult to quantify. This is also the case in this research as complexity, dynamics and uncertainty are deliberately chosen broad terms in order not to limit the scope of the research. Eisenhardt (1989) in addition claims that case studies can uncover causal relations in specific situations and help to get an understanding of the relationships identified. Both these findings are confirmed by Meredith (1998) who showed that case study: (1) can study a phenomenon in its natural setting (specific situation), (2) can answer why, what and how questions and (3) is very suitable for exploratory research where variables remain unknown and the phenomenon is not fully understood. All three points stated by Meredith (1998) applies to this research. The first point is of interest because the pre-production stage apparently consist multiple factors that influences delays. Therefore performing the study in its natural setting is essential to find relevant information and draw valid and reliable conclusions. According to point two, case research is suitable for investigating what causes the delays and how to tackle them. The third point specified by Meredith (1998) shows that despite the limited empirical research in literature conducted in this field, case research remains suitable for this research.

3.2

Case Selection

3.3

Data Collection

Exploring the literature did not result in practical measurements for the variables at ETO companies. Due to the complexity of the pre-production stage and the different management approaches of companies it is required to observe and investigate the possibilities of obtaining reliable data regarding complexity, uncertainty, dynamics and delays. To have a general guideline and to aid the reliability and validity of the research, a research protocol is designed (Appendix A). This also includes procedures for the observations and reporting (Yin, 1994). Data can be stored physically or digitally but also qualitative data stored within the knowledge and experience of employees is required to a great extent. From the observations it can be concluded that different channels have to be addressed in order to obtain valuable data.

Bertrand and Muntslag (1993) identified that uncertainty of ETO companies lies in the planning, specification and processes requirements of projects. What production processes are needed is mainly based on the specification of the product. Therefore in this research, uncertainty will be measured by planning and specification uncertainty. Uncertainty can be described as the difference between the level of information required to perform a task and the level of information available (Galbraith, 1973). The company however uses a dynamic planning method, which means that the planning is adjusted daily and there is no structural backup of versions to track back what the initial state of the planning was. Due to this lack of documentation, planning uncertainty and delays cannot be obtained quantitatively. This also applies to specification uncertainty and dynamics (changes); changes are not documented after acceptance of quotation and uncertainty is a rather subjective measure as indicated before. The main project participants are most aware of these discussed issues. Therefore the delays, planning uncertainty, specification uncertainty and dynamics have to be obtained qualitatively, where individual participants are the main data source.

Since dozens of projects passes the pre-production stage monthly, predefined questionnaires will be used to investigate the effects of uncertainty, complexity and dynamics on pre-production delays (Appendix D & E). Also data regarding concrete causes of delays will be obtained through questionnaires (Appendix F). The reliability of data obtained through questionnaires, depends mainly on the quality of the questionnaires and the memory and honesty of the participants. In order to increase the level of reliability several steps have to be taken. First, collection of data will be limited to data of the previous 12 months1_{. The purpose of only using recent data is reducing}

memory bias of participants and to have data reflecting the current state of the pre-production stage. In addition to this, questionnaires will contain a limit of 25 projects for each participant if possible, as participants may become tired or bored of the long list of questions influencing the quality (Helgeson & Ursic, 1994; Herzog & Bachman, 1981). Secondly, participants have to be well aware that the questionnaires are not for personal evaluation purposes and confidentiality will be maintained throughout the entire research. Those who are not competent to answer questions confidently are asked to leave the field blank. This will be classified as missing data and cannot be estimated or recovered. These measures are taken to increase the honesty of participants. Thirdly, questionnaires will initially be pilot tested with one participant before release. The participant is asked to provide feedback regarding the clearness of questions and the questionnaire in general in order to improve the quality of the questionnaire.

3.4

Questionnaire

Questionnaires will be addressed to every individual who had a significant role in the pre-production stage of the selected projects. This will be based on the time an individual spends on the selected projects. As a result certain projects are rated by multiple persons and others by one. Two separate questionnaires will be used in order to not bias the answers and not to fatigue the individuals by exposing them with very long questionnaires. Since no former research was undertaken at this company and through observation projects were mostly rated as difficult or easy, a three-point scale will be used in the first questionnaires – 1, easy/low and 3, high/hard, aimed at investigating the following issues:

- Planning uncertainty: level of uncertainty at the initial start of the project;

- Specification uncertainty: the degree of uncertainty at the initial start regarding specification of the project;

- Complexity: difficulty of the project at that specific moment in time.

- Dynamics: level of changes during the pre-production in reflection to the initial specified design.

4 Results

This chapter summarizes the main findings of the research, obtained through the methods described in the previous chapter. Several topics and relationships will be presented, which will be further discussed in chapter 5.

4.1

General

Two groups can be formed out of the 77 selected projects: (1) the delayed and (2) the not delayed group. 11 projects are delayed during the pre-production stage and the remaining 66 are not. The delays seemed to occur randomly; no patterns can be found regarding season, project type, customer or participants of the delayed projects. This supports the former findings of Land and Gaalman (2009) who indicate uncontrolled delays in this stage of production. Table 3 represents a summary of the findings.

4.2

The Variables

The variables are measured on an ordinal 3-point scale. Due to the small number of delayed projects, it is highly expected that the count will be less than five. Therefore a Fisher-Freeman-Haltot Exact test1 _{will be used to test the variables. Outcome results indicate a significant}

increase of uncertainty, complexity and dynamics in the delayed group, compared to the not delayed group (ρ=0,002, ρ=0,000, ρ=0,019 respectively). From these results it can be concluded that the delays are not coincidently related to the level of uncertainty, complexity and dynamics of the projects. Unfortunately statistical significance does not provide any insight on the size of the effect (Coe, 2002). What is the magnitude of the unit increase? Can it be considered as a small or a large effect? Two commonly used measures to determine the effect size (ES) of ordinal data are Cohen’s d and Hedges’s g. According to Grissom and Kim (2005) the second measure provides more accurate estimates, especially when sample sizes are small. Therefore the Hedges’s g will be used to determine the ES of the variables (table 2). The magnitude will be assessed by the well-known rule of thumb by Cohen (1988). From this it can be concluded that that all variables have a large effect. The most influencing variable however is uncertainty.

* Rule of thumb: 0,2 = small, 0,5 = medium, 0,8 = large effect (Cohen, 1988)

1_{Freeman, G.H., & Halton, T.R. (1951). Note on exact treatment of contingency, goodness-of-fit and other 
 problems of}

significance. Biometrika, 38, 141-149.

Mean (delayed) Mean (not delayed) S (pooled std. dev.) ES

Uncertainty 1,91 1,35 0,37 1,5*

Complexity 1,91 1,42 0,46 1,1*

Dynamics 1,82 1,36 0,51 0,9*

Figure 2. Scatter plots of the characteristics

* The estimation error is calculated by the difference between the actual pre-production time and the planned pre-production time, which is not displayed in this table. A negative value indicates that the actual pre-production time exceeds the planned.

Total Delayed Not delayed

N Min. Max. Mean Std. dev. N Min. Max. Mean Std. dev. N Min. Max. Mean Std. dev. Total costs (€) 77 371 237.181 20.297 42.301 11 5.614 151.442 35.534 44.956 66 371 237.181 17.758 41.659 Pre-production time (hrs.) 77 1 811 40 106 11 4 811 132 234 66 1 392 25 55 Pre-production time estimation error* (hrs.) 76 63,5 -177 -8,6 35,5 11 21 -143 -40,6 54,2 65 63,5 -177 -3,2 28,4 Uncertainty score

(on 3-point scale) 77 1,00 2,75 1,43 0,42 11 1,25 2,75 1,91 0,54 66 1,00 2,50 1,35 0,34

Complexity score

(on 3-point scale) 77 1,00 2,5 1,49 0,49 11 1,50 2,50 1,91 0,49 66 1,00 2,50 1,42 0,45

Dynamics score

(on 3-point scale) 77 1,00 3,00 1,42 0,53 11 1,00 2,50 1,82 0,56 66 1,00 3,00 1,36 0,50

** Correlation is significant at the 0,01 level (2-tailed).

A question that now arises is whether uncertainty, complexity and dynamics separately or that a combination of these variables causes the delays. The scatter plots in figure 2 shows monotonic relationships between all combinations of variables. This is required for further analysis and it also shows that correlation between variables is plausible. Table 4 presents the results of running a Spearman’s Rank Order correlation to determine the relationship between 77 projects’ level of uncertainty, complexity and dynamics. The results show strong correlation between complexity and uncertainty, which implies that if a project is rated as complex, it will be likely that the uncertainty level is also high and vice versa. The remaining combinations show moderate correlations.

4.3

Changing Conditions

Now that the influence of the variables, and between variables is identified, it also becomes interesting to know whether the variables change over time. It is plausible that projects are poorly specified in the beginning, however as the project progresses more details become clear i.e. after acceptance of the client, project specification as well as the planning of the project should be clearer than during the sales phase. Out of the 77 selected project, 65 were handled by another employee after acceptance of quotation (sales phase). Table 5 presents the average change of the variables after sales activities. As expected the results show an overall decreases of uncertainty as projects progresses through the pre-production stage. More interesting is the increase of complexity, specification uncertainty and dynamics.

Uncertainty Complexity Dynamics

Uncertainty 1

Complexity 0,740** 1

Dynamics 0,608** 0,586** 1

Total (65) Delayed (11) Not delayed (54) Average total uncertainty difference

(on 3-point scale) - 0,25 - 0,73 - 0,15

Average planning uncertainty difference - 0,45 - 0,91 - 0,35 Average specification uncertainty difference + 0,20 + 0,18 + 0,20 Complexity difference on average

(on 3-point scale) + 0,20 + 0,36 + 0,17

Dynamics difference on average

(on 3-point scale) + 0,17 + 0,55 + 0,09

4.4

Project Size

Project size can be measured by the total incurred costs and the total time spent on pre-production (table 3). Testing the project size, shows significant difference between the delayed and not delayed group for both: total costs (U(76)=171, Z=2,795, ρ=0,005) and total pre-production time (U(76)=154, Z=3,046, ρ=0,002). From these results it can be concluded that the delays are not coincidently related to the size of projects. As the project size increases, the likelihood of pre-production delay also increases.

4.5

Concrete Causes

In the second questionnaire individuals are asked to indicate concrete causes of delays. These causes are classified according to the following coding: uncertainty, complexity, dynamics, others and not valid. The others-coding will receive a secondary coding to investigate whether patterns exists. The results in table 6 and 7, supports the former findings in table 3, where the delayed projects are higher rated on uncertainty, complexity and dynamics, than the not delayed ones. Here, classification in the others and not valid-coding is larger in the not delayed group. This indicates that uncertainty, complexity and dynamics are less present. The concrete causes appear to be very varied (Appendix D). Even within the same coding it is not likely that a similar cause is mentioned multiple times. Nevertheless the results do reveal some interesting aspects of the variables.

4.5.1 Uncertainty, Complexity and Dynamics

The uncertainty-coding shows interesting insights on this influential variable. Two distinguished causes of uncertainty can be identified through the results: the internal and the external cause. Internally, uncertainty is most likely to be caused by inappropriate documentation of past projects. Externally, the clients themselves are not even certain what the exact specification should be. From the questionnaires it can also be concluded that use of advanced and new techniques, is the main concrete cause of the high level of complexity. Especially researching these new techniques enhances the complexity. Unfortunately the concrete causes of dynamics occur to be that varied, that no pattern could be identified.

Coding

Uncertainty (%) Complexity (%) Dynamics (%) Others (%) Not valid (%)

Total 11,5 19,2 13,5 51,9 3,8

Delayed 18,2 27,3 13,6 36,4 4,5

Not delayed 6,7 13,3 13,3 63,3 3,3

4.5.2 Others

The extensive presence of the others-coding category makes it necessary to explore the secondary coding as well (table 7). As the table shows, no obvious pattern can be found that could indicate a clear concrete cause for the delayed projects. Therefore, highlight should be set on priority setting and choice assessment. These are the two most present secondary coding categories (both 11% of the total), which by all means occur at the delayed projects. From the data it can be concluded that interference of higher prioritised projects seems to be the major cause of delay regarding priority setting. Reviewing and assessing different solutions and possibilities to solve a problem is another delaying activity.

Two other secondary coding categories (due date setting and negotiation) appear to be very present as well. However, both of them are only indicated at the not delayed projects. In addition to this, the high amount of due date setting-coding is caused by the fact that one regular client always demands for fast delivery. Furthermore high level of negotiations with clients does not seem to cause any delays. This can be explained by the fact that due dates are mostly set after or during the negotiations. Therefore these two prominent secondary coding categories are not meaningful, with regard to concrete causes of delays.

Secondary coding Total (%) Delayed (%) Not delayed (%)

Awaiting client 4 0 5

Capacity issue 4 13 0

Choice assessment 11 13 11

Communication client 4 13 0

Communication internal 7 13 5

Due date setting 30 0 42

Modification 7 13 5

Negotiation 19 0 26

Priority setting 11 25 5

Time lack 4 13 0

5 Discussion

To answer the research questions in the first chapter, two propositions are presented in the conceptual model and the methodology. One of the propositions is that the variables (uncertainty, complexity and dynamics) are directly related to the delays that occur. In other words, having a high level of uncertainty, complexity and dynamics would affect the pre-production stage, causing delays. The second proposition is that the concrete causes of delays should support the findings of the first proposition. This chapter issues the research questions and discusses whether the propositions are confirmed by the results.

5.1

Delays

Out of the 77 investigated projects, 11 are delayed (14,3%). According to Land and Gaalman (2009) more than 25% and outliers of 70% of delayed orders are not uncommon at small ETO companies. This finding is supported by another master’s thesis research1 _{conducted at the same}

time as this one. Their results show delays of 35, 30, 33 and 45% at the different product types. Therefore, it can be concluded that indeed ETO companies faces delays. The amount of the delays however, seems to be very varied. In this case the delay problem seems to be limited. An explanation for this could be the continuous effort of the company to standardise the concepts in order to decrease time-consuming design activities.

5.2

Causes of delays

Three broadly specified variables have been tested on causing delays during pre-production activities. Table 3 shows a tremendous difference between the average score of the delayed and not delayed projects regarding uncertainty, complexity and dynamics. In addition to this a Fisher-Freeman-Haltot Exact test confirms a significant difference between the two groups. These results show that indeed uncertainty, complexity and dynamics have influence on the delays. Hedges’s g test in addition shows that uncertainty has the largest effect on delays followed by complexity and then the dynamics.

5.2.1 Combinations of Causes

engineer, purchase and plan the project. Splitting up the uncertainty variable, shows that indeed specification uncertainty correlates much more with complexity and dynamics than planning uncertainty (table 8). This finding is not surprisingly also supported by the concrete causes of delays, where all causes mentioned in the uncertainty coding are related to specification uncertainty (Appendix D). From these results it can be concluded that specification uncertainty is not only a major cause of delays, but also a cause for high level of complexity and dynamics. Therefore decreasing the specification uncertainty is definitely one of the most effective methods to decrease delays, as both complexity and dynamics have a rather high correlation with it.

** Correlation is significant at the 0,01 level (2-tailed). 5.2.2 Measurement Changes

Uncertainty is, especially in an ETO environment, not a static variable as mentioned in §4.3. Expectation here is that over time, details will become more clarified, which should result in a decrease of uncertainty. This statement is partially confirmed by the overall decrease of uncertainty in table 5. However, the separated specification uncertainty and planning uncertainty provides another view on this issue. The overall decrease of uncertainty is caused by the large reduction of planning uncertainty. This can be explained by the fact that due dates are usually already determined at the end of the sales phase. A high level of specification uncertainty correlates with the level of complexity and dynamics. This is also reflected in the measurement changes, where increase of specification uncertainty is associated with increase of both complexity and dynamics. So the increase of specification uncertainty is a troublesome and unexpected phenomenon.

The almost equal increase of specification uncertainty at both the delayed and not delayed group, indicates a structural lack of specification clearness after the sales phase. Another remarkable fact is that all the delayed projects are handled by more than one person. Therefore two possible causes can be identified for the increase of specification uncertainty: (1) handovers between sales and the other pre-production departments are inappropriate; resulting in lost of information, and (2) the subsequent department (mostly engineering) prefer to receive more detailed information.

Specification

uncertainty uncertainty Planning Complexity Dynamics

Specification uncertainty 1

Planning uncertainty 0,343** 1

Complexity 0,712** 0,484** 1

Dynamics 0,615** 0,393** 0,586** 1

Unfortunately no valid conclusions can be drawn with the current data, since there is no data collected regarding the amount of effort taken to clarify project specifications. However, from former findings it is already clear that reducing uncertainty, and in particular specification uncertainty is crucial. Therefore it is very important to specify projects thoroughly during the sales phase and hand it over properly between departments to minimise the uncertainty.

5.2.3 Project Size

As §4.4 shows, project size seems to be related to delays. The interest here goes to the underlying cause of this. Table 9 presents different project sizes traded off against the obtained data. The results once again show a tremendous correlation between specification uncertainty and complexity. Both variables also occur to be related to the project size, as only these two variables and the average total costs increase with the project size.

A relation between project size and the estimation error is unlikely as the table and a Pearson product-moment correlation analysis show (r=0,126, n=76, ρ=0,278). As a result, structural underestimation of required pre-production time by sales is not likely to be present. This finding is supported by the fact that due date setting and time lack-coding were not prominent present in the concrete causes. Still, the large difference of 95% is rather alarming especially for large projects. Sales should be well aware of these numbers to prevent costly delays, that mostly occur at large projects, by not underestimating the time required and provide clients due dates which are achievable.

Project size (based on actual pre-production hours)

1-8 8-20 20-100 100 - …

N 37 15 17 7

Average total costs (€) 3.627 6.686 32.013 111.470

Average uncertainty (on 3-point scale) 1,26 1,45 1,56 1,96

Average planning uncertainty 1,22 1,37 1,24 1,29

Average specification uncertainty 1,31 1,53 1,59 2,14

Average complexity (on 3-point scale) 1,24 1,57 1,71 2,14

Average dynamics (on 3-point scale) 1,19 1,70 1,53 1,86

Average estimation error -11% 89% 48% 95%

5.2.4 Concrete Causes

The answers given in the second questionnaire were very varied. However, it did provide interesting findings, especially for the influential variable uncertainty. It appears that uncertainty is caused externally as well as internally. Clients themselves are not sure what they actually want, causing uncertainty for the ETO company (externally). This also explains the high level of dynamics for these projects (Appendix C). One of the primary competition parameters for ETO companies is meeting customer requirements to functionality as well as quality (Hvam et al., 2006). Therefore every alteration in specification that clients prescribe has to be considered, and clear communication is therefore essential. Uncertainty is also caused internally by inadequate documentation of former projects or concepts. As already indicated the case company puts continuous effort in standardisation of concepts. A pitfall here is overestimation of the standardisation level or repeatability of solutions i.e. assuming a well functioning and documented solution while it is not. Retain and maintain up-to-date data regarding solutions is therefore crucial to decrease uncertainty.

Complexity is caused by new and unfamiliar solutions/technologies, requiring extensive research. Since ETO companies can be regarded as innovative firms, exploration of new techniques and solutions is definitely an element of the pre-production stage. However is possible to decrease complexity through standardisation, as the pre-production members become more familiar with the solution. Another method to decrease complexity, as it is a subjective term, is to increase the knowledge and skills of employees. This can be achieved by providing employees specific courses.

The results of the others-coding category show that priority setting could be a cause of delay. This is interesting since priority setting was included in the initial state of this research. Due to time and data restrictions it was not possible to include this in the research. Therefore it remains unclear whether priority was set correctly or not. This finding however does highlight that priority setting is an interesting variable to research in the future.

5.3

Generalisation

Yin (1994) indicates that generalisation is limited for single case researches. Though, this research contains certain aspects that provide some room for generalisation of results. The independent variables (uncertainty, complexity and dynamics) used in this research, are proven to be the typical characteristics of ETO companies (Bertrand & Muntslag, 1993; Land & Gaalman, 2009; Caron & Fiore, 1995; Pandit & Zhu, 2007). Therefore it is unlikely to find very contradictory results regarding these variables in other ETO companies (§4.1 & §4.2).

However, it is arguable whether the more in-depth analysis like changing conditions (§4.3) and the concrete causes of delays will represent other ETO companies in general as well. Sure, other companies will also deal with clients, changing their specifications, and researching new developments. But, do they for example also put effort in standardisation of concepts? Bertrand and Muntslag (1993) relate ETO companies to one-of-a-kind products, where others say ETO products are customised make-to-order products (Chen, 2006). Hicks et al. (2001) identified four types of ideal ETO companies, which indicate that there are definitely differences between ETO companies. This and other factors like, the number of participants of a project (more handovers), capacity of pre-production (possible capacity issue), etc. might influence the results and limit generalisation for these parts of this research.

This case research is part of the “diagnosing the pre-production” theme, where four students investigate the pre-production stage of different ETO companies. Comparing the different results could provide room for generalisation as stated in the methodology. Despite the low response rate (33%), results do show some similarity. The number one of the top-three causes of delays is also identified in this research, namely complexity of orders. The other causes were: not making full use of capability and poor external communication (with customer and supplier)1_.

The last cause shows some similarity with the external cause of specification uncertainty, where appropriate communication with clients is essential.

6 Conclusion

The starting point of this research is the unidentified cause of uncontrolled delays at the pre-production stage of ETO companies. This study has taken an in-depth case research approach to identify the concrete causes of delays, measuring both qualitative and quantitative data. The findings support former researches indicating uncontrolled and randomly occurring delays at the pre-production stage. All typical ETO characteristics occur to have an effect on the delays. The most influential characteristic is uncertainty. Further analysis shows that uncertainty strongly correlates with the level complexity, which can be supported by the fact that tasks will become more complex if information supply is limited.

In ETO environments uncertainty is measured by both planning and specification uncertainty, where specification uncertainty occur to be far more influential. This is also reflected in the concrete cause of delays where internal and external uncertainty, causes by inappropriate documenting and indecision of customer respectively, is identified. Another concrete cause is the continuous innovation, requiring research to obtain the skills and knowledge to apply them (complexity).

Uncertainty, complexity and dynamics are all typical characteristics of ETO companies. These characteristics cannot be fully eliminated, as they are a part of this type of business. However, from the results it can be concluded that reducing complexity and especially specification uncertainty will surely reduce delays at the pre-production stage. This can be achieved through (sub)standardisation of concepts and maintain and obtain high level of knowledge and skills to decrease complexity and by sound communication with clients and appropriate documentation of data to reduce specification and planning uncertainty.

7 Limitations & Future Research

As with all researches several limitations can be noted. One of the limitations is the lack of quantitative data. Despite of all the effort taken to maintain reliability and validity, results of this research still mainly depend on memory, honesty and integrity of the participants. Especially the second phase questionnaire seems to be lacking completeness. Looking back, semi-structured interviews based on the delayed projects would probably result in more complete and in-depth data regarding the concrete causes.

Another limitation is widely discussed in §5.3, which issues the generalisation of this research. Despite the fact that the used variables are present at all ETO companies, this case company’s continuous aim for standardisation, could limit the generalizability. This, and the fact that literature acknowledges differences between ETO companies reduce the generalizability of in-depth results of this research.

This research shows some interesting research areas for future research. One of them is priority setting of employees at the pre-production stage. As this research advanced it became clear that employees have great amount of different responsibilities and tasks to perform, which are not obviously documented. Therefore it is recommended to perform a long-term case research to investigate the pre-production stage by tracking, tracing, following and analysing projects for a longer period. This will help to acquire in-depth data on other not documented factors, that might influence the pre-production stage like: priority setting or communication issues. This will expand our understanding of delays at the pre-production stage.

Acknowledgements

Bibliography

Abdalla, H. S. (1999). Concurrent engineering for global manufacturing. International Journal of Production Economics , 60, 251-260.

Aslan, B., Stevenson, M., & Hendry, L. C. (2012). Enterprise Resource Planning systems: An assessment of applicability to Make-To-Order companies. Computers in Industry , 63, 692-705. Bertrand, J. W., & Muntslag, D. R. (1993). Production control in engineer-to-order firms. International Journal of Production Economics , 30, 3-22.

Burt, D. N., & Doyle, M. F. (1993). The American Keiretsu. Irwin, Homewood: Business One. Caron, F., & Fiore, A. (1995). 'Engineer to order' companies: how to integrate manufacturing and innovative processes. International Journal of Project Management , 13, 313-319.

Chen, C. (2006). Concurrent engineer-to-order operation in the manufacturing engineering contracting industries. International Journal of Industrial and Systems Engineering , 1, 37-58.

Coe, R. (2002). It’s the effect size, stupid. What effect size is and why it is important. Annual Conference of the British Educational Research Association. England: University of Exeter.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale: NJ: Erlbaum. Corrêa, H. L. (1992). The Links Between Uncertainty, Variablility of Outputs and Flexibility in Manufacturing Systems. PhD Thesis, University of Warwick, UK.

Eisenhardt, K. (1989). Building Theories From Case Research. Academy of Management Review , 14, 532-550.

Elgh, F. (2012). Decision support in the quotation process of engineered-to-order products. Advanced Engineering Informatics , 26, 66-79.

Ellram, L. M., & Carr, A. S. (1994). Strategic purchasing: a history and review of literature. International Journal of Purchasing and Materials Management , 30, 9-11.

Galbraith, J. R. (1973). Designing Complex Coordination. Reading: Addison-Wesley. Gell-Mann, M. (1995). What is complexity? Complexity , 1 (1).

Gosling, J., & Naim, M. M. (2009). Engineer-to-order supply chain management: A literature review and research agenda. International Journal of Production economics , 122, 741-754.

Grissom, R. J., & Kim, J. J. (2005). Effect Sizes for Research: A Broad Practical Approach. Mahwah: NJ: Lawrance Erlbaum.

Helgeson, J. G., & Ursic, M. L. (1994). The role of affective and cognitive decision-making processes during questionnaire completion. Public Opinion Quarterly , 11, 493-510.

Hetem, V. (2000). Communication: computer aided engineering in the next millennium. Computer-Aided Design , 32, 389-394.

Hicks, C., McGovern, T., & Earl, C. F. (2001). A Typology of UK Engineer-to-Order Companies. International Journal of Logistics: Research and Applications , 4 (1).

Hicks, C., McGovern, T., & Earl, C. F. (2000). Supply chain management: A strategic issue in engineer to order manufacturing . International Journal of Production Economics , 65, 179-190. Hicks, C., Song, D., & Earl, C. (2007). Dynamic scheduling for complex engineer-to-order products. International Journal of Production Research , 45, 3477-3503.

Hvam, L., Pape, S., & Nielsen, M. K. (2006). Improving the quotation process with product configuration. Computers in Industry , 57, 607-621.

Karlsson, I. (2009). Research Operations Management. New York: Routlegde.

Konijnendijk, P. A. (1994). Coordinating marketing and menufacturing in ETO companies. International Journal of Production Economics , 37, 19-26.

Land, M. J., & Gaalman, G. J. (2009). Production planning and control in SMEs: time for change. Production Planning & Control , 20, 548-558.

Little, D., Rollins, R., Peck, M., & Porter, J. K. (2000). Integrated planning and scheduling in the engineer-to-order sector. International Journal of Computer Integrated Manufacturing , 13, 545-554. Mahmoud-Jouini, S. B., Midler, C., & Garel, G. (2004). Time-to-market vs. time-to-delivery Managing speed in Engineering, Procurement and Construction projects. International Journal of Project Management , 22, 359-367.

Meredith, J. R. (1998). Building Operations Management theory through case and field research. Journal of Operations Management , 16, 441-454.

Monczka, R.M., Handfield, R.B., Guinipero, L.C., Patterson, J.L. (2009). Purchasing & Supply Chain Management. Andover: Thomas Rennie.

Pandit, A., & Zhu, Y. (2007). An ontology-based approach to support decision-making for the design of ETO (Engineer-To-Order) products. Automation in Construction , 16, 759-770.

Rahman, A., Rahim, A., Shariff, M., & Baksh, N. (2003). The need for a new product development framework for engineer-to-order products. European Journal of Innovation Management , 3, 182-196.

Sahin, F., & Robinson Jr, E. P. (2005). Information sharing and coordination in make-to-order supply chains. Journal of Operations Management , 23, 579-598.

Tsinopoulos, C., & Bell, K. (2009). Supply chain integration systems by small engineer to order companies: The challenge of implemantation. Journal of Manufacturing Technology , 21, 50-62. Willaert, S. S., De Graaf, R., & Minderhoud, S. (1998). Collaborative engineering: A case study of Concurrent Engineering in a wider context. Journal of Engineering and Technology Management , 15, 87-109.

Wullink, G., Gademann, A. J., Hans, E. W., & Harten, A. (2004). Scenario-based approach for flexible resource loading under uncertainty. International Journal of Production Research , 42, 5079-5098.

Yin, R. (1994). Case study research. Beverly Hills: Sage Publications.

Zijm, W. H. (2000). Towards intelligent manufacturing planning and control systems. OR Spectrum , 22, 313-345.

Delays at the Pre-production Stage in Engineer-to-Order Companies

APPENDICES

Author: J. He

Student number: 2074966 Date: 21-06-2013

MSc Technology Operations & Management

Faculty of Economics and Business, University of Groningen Supervisor: dr. N.D. van Foreest

Contents

Appendix A. Research protocol Appendix B. Observation analysis Appendix C. Datasheet phase 1 Appendix D. Datasheet phase 2 Appendix E. Phase 1. Questionnaire

Change Log

Date Version Change

17-04-13 1 - Determine level of uncertainty - Determine level of complexity

20-04-13 2 - Added qualitative measures for uncertainty - Added qualitative measures for complexity

01-05-13 3 - Remove priority setting, due to complexity of this issue 08-05-13 4

- Added questionnaire evaluation

- Added additional measure for uncertainty to make it more specific for participants - Changed moment of measuring complexity from during to before start of the project - Added validation activity during coding process

17-05-13 5 - Added additional question to determine if the pre-production stage is delayed or not. The initial _{measure did not seemed to be reliable} 31-05-13 6 - Change the measurement of dynamics to a 3-point scale closed question

18-06-13 7 - Used different statistical tests _{- Added secondary category for coding process}

1 Introduction

The aim of this research is to identify the causes of delay problems at the pre-production stage of ETO companies. By identifying the causes solutions can be proposed, to tackle the delays and decrease lead-time of ETO projects. This will be achieved by analyzing both quantitative and qualitative data obtained from inquiries and projects that pass through the pre-production stage. This document provides a general guideline that should assist in the formulation and application of the case research.

Throughout the entire research confidentiality will be maintained with the organization and individuals. No gathered data will be used in any report or publication that might incriminate or identify the organization or individual.

2 Design

2.1 Case Selection

Given the exploratory nature of this research and the very limited empirical research conducted in this field, we will perform a single case research in order to gain in-depth insights. Hence, data is obtained from one ETO company.

2.2 Observation

Observations of the researcher are required in order to obtain a better understanding of the production stage of the selected case company. The main objective of observing the pre-production stage is to determine how planning uncertainty, specification uncertainty, complexity, dynamics, delays and causes of delays can be identified. One small project will be tracked during its pre-production by the researcher. Observations will mainly be based on identifying all possible manners to measure the previous described terms. During these observations, continuous validation and communication with the professionals is required in order to derive valid measures. In the end the observations should result in a scheme where valid measurements are set off against the measure as indicated in the following table.

The determined measurements will be crosschecked with head of sales and the head project manager who is responsible for scheduling. From this, several data collection methods are introduced in the next paragraph.

3 Data Collection

Data regarding inquiries and projects passing through the pre-production stage are well documented and contains valuable information. In order to draw relevant conclusions for the current state of practice, only data of the past year will be analyzed. Qualitative data will be obtained through questionnaires with participants of the different projects. The exact data to be obtained in the first phase are presented in the table below.

Table 2. Data collection phase 1

In the second phase questionnaires will be used to determine if there is a delay and the concrete causes of delays of it.

Table 3. Data collection phase 2 !

3.1 ERP System

Data acquired from the ERP system require careful analysis and filtering. Outliers and special cases have to be traced down in order to retrieve reliable results. The cause of the outlier has to be identified in order to conclude whether the data is reliable or not. If additional information is sufficient, data might be recalculated or re-estimated otherwise elimination of an obvious outlier is required.

Data Collection method Storage

Project number ERP System Excel sheet

Project name ERP System Excel sheet

Start date ERP System Excel sheet

End date ERP System Excel sheet

Pre-production delay Questionnaires Excel sheet

Total costs of project ERP System Excel sheet

Predicted pre-production time ERP system/stored documentation Excel sheet

Actual pre-production time ERP System Excel sheet

Level of dynamics Questionnaires Excel sheet

Level of complexity Questionnaires Excel sheet

Level of uncertainty Questionnaires Excel sheet

Data Collection method Storage

Pre-production delay Questionnaires Excel sheet

3.2 Stored documentation

Stored documentation can be e-mails, notes, CAD models or other digitally stored documents regarding revisions of the projects. This information is not accessible through the ERP system and has to be retrieved through the server.

3.3 Questionnaires

Simple questionnaires will be used in order to receive relevant information in a short period of time. Questions will be addressed to the different participants of the projects in order to achieve a certain level of reliability and validity. Two separate questionnaires will be used during the two phases. Each questionnaire starts with an introduction to indicate the importance of the participants’ honesty and a glossary to make the participant familiar with the terms used. The questionnaire will be pilot tested by one respondent to provide feedback on behalf of anything that affects answering and the answer of the targeted participant. Before receiving the questionnaires, participants will receive a verbal explanation regarding the subjects and they also have the opportunity to ask questions.

The reliability of the research is mainly based on the quality of the answers given. To reduce memory bias of participants, data selection will be limited to projects not older than one year. In addition to this each questionnaire will contain a maximum of 25 projects if possible since participants may get tired or bored of the questions, affecting the quality of answers given.

3.3.1 Phase 1

the data cannot be estimated and therefore is missing. An example of the questionnaire is displayed below.

Table 4. Questionnaire phase 1

Uncertainty

According to Galbraith (1989) uncertainty is the difference between the level of information required to perform a task and the level of information available in the organization. Therefore the questions regarding uncertainty will be based on the clearness and completeness of information regarding projects. Bertrand and Muntslag (1993) indicated that uncertainty of ETO companies lies in the planning and specification of projects. Hence, uncertainty will be measured as an average of both planning and specification uncertainty.

Planning uncertainty

Indicate the degree of uncertainty regarding the planning at the start of the project. - Low (well specified, clear and complete information)

- Average (certain amount of unclear/incomplete information) - High (a lot of unclear/incomplete information)

Specification uncertainty

Indicate the degree of uncertainty regarding the project specification at the start of the project. - Low (well specified, clear and complete information)

- Average (certain amount of unclear/incomplete information) - High (a lot of unclear/incomplete information)

Level of complexity

Indicate how difficult (solution/engineering complexity) the project was after confirmation of the client (start of the project). Do not base it on a review, but on your judgement at that specific moment.

- Easy (daily working routine, known concepts and designs) - Medium (known standardized concepts with small adjustments)

- Hard (difficult/challenging project with new techniques and unfamiliar design) Changes during pre-production (design/specifications/solutions etc.)

Indicate the level of changes, which are made during the pre-production of the project in reflection to the initial specified design.

3.3.2 Phase 2

In the second phase questionnaires will be used to identify if delays occurred at the pre-production stage and also information regarding concrete causes of delays will be obtained. At most 25 of the last projects of the participant will be presented as shown in table 4. The individuals are asked to indicate if the pre-production of a project is delayed or not in reflection to the initial planning. They are also asked to identify concrete causes of delays for each project regardless if the pre-production was delayed or not.

Table 5. Questionnaire phase 2

Delayed?

Indicate if the pre-production stage (sales, engineering, process planning and procurement) of the project was delayed in reflection to the initial planning. Remark that the project itself (including production) might be finished in time as planned yet, the pre-production stage could have been delayed.

- Yes (The pre-production stage of this project faced delays)

- No (The pre-production stage of this project did not faced any delays) Causes of delays

Describe briefly what you think that delayed the pre-production stage of that particular project. Multiple causes are possible. Remark that a project might have been delayed due to a certain event but eventually did not result in a delayed pre-production stage. Please also indicate these causes.

4 Data Analysis

The obtained data through the ERP system, stored documentation and questionnaires will in the first phase be used to analyze to what extent uncertainty, complexity and dynamics influences the delays at the pre-production stage. The objective of the second phase of the research is to examine what the main concrete causes are.

4.1 Phase 1

This section describes the analysis of the first phase. 4.1.1 Dynamics

The amount of major changes during the pre-production stage will measure the dynamics of the pre-production stage. This will be determined through the questionnaires as no quantitative data is available concerning major changes.

4.1.2 Uncertainty

Uncertainty will be determined by questionnaires since completeness and clearness of information is not documented and participants are more aware of this particular qualitative issue. Two measures of uncertainty have been identified for ETO companies; uncertainty in planning and uncertainty in specifications.

4.1.3 Complexity

Complexity will be determined by questionnaires since this is rather subjective term. Every individual has his or her own specific experience with a certain project.

4.1.4 Priority setting

Due to the diversity of activities of all members it is not possible to include priority setting in this case research. Documented activities of employees are not reliable due to rush activities at ETO companies and divisions are not clearly separated. Employees’ functions are flexible and consultation with colleagues is daily business in order to share knowledge, opinions and expertise. In addition client and supplier visits interfere employees’ activities. For convenience purpose employees do not register their activities in such detail and in actual sequence. In order to draw valid and reliable conclusions regarding priority setting all activities of employees have to be monitored and stored for a certain period of time including multiple projects from the beginning till the end and all side activities involved. Due to time restrictions we are not capable of including this in our current research.

4.1.5 Relationships

The next step after determining the level of uncertainty, complexity and dynamics of the projects, is to test to what extent these three characteristics influences the delays at the pre-production stage. Therefore the three characteristics will be tested collectively as well as separately. Dependent of the type of variable, different compatible tests will be used. An overview is presented below.

Test # Dependent variable Independent variable Test type

1 Delays Uncertainty Fisher-Freeman-Haltot Exact test, Hedges’s g 2 Delays Complexity Fisher-Freeman-Haltot Exact test, Hedges’s g 3 Delays Dynamics Fisher-Freeman-Haltot Exact test, Hedges’s g

4 Uncertainty

Complexity Dynamics

4.2 Phase 2

This section describes the second phase of the case research. 4.2.1 Transcription and Coding

The concrete causes obtained from the second questionnaire will be transcribed and coded into the following categories:

Table 6. Coding categories

One researcher will perform the coding process unless the cause is difficult to assign to one specific category. In that case a second person (most ideally the participant himself) will be consulted in order to assign the cause to a mutual agreed category. A secondary coding category is added for the others-coding to reveal patterns of this coding.

4.2.2 Analysis

The results of the coding process should reflect former findings in phase 1. With this we can validate the findings of the second questionnaire and draw valid conclusions regarding the concrete causes of delays.

5 Reliability and Validity

Having all participants of a certain project to fill in the questionnaire result in more reliable data. As already mentioned the questionnaires are separated into two parts in order not to bias the answers of the participants. In addition to that, all projects of the participants are presented in the second questionnaire, delayed or not, in order to receive highly reliable answers. In an attempt to achieve triangulation, both qualitative and quantitative measures for complexity, uncertainty and dynamics are used.

6 Reporting and Results

The report should deliver the following documents: - Questionnaire records

- Excel data sheet

- Case study report consisting, data analysis, results and conclusions - Improvements for future research

Appendix B. Observation analysis

Measure Measurement

Planning uncertainty

During the sales phase received inquiries are prioritized according to its due date. Customers mostly already indicate when they require the quotation due to managerial issues and lead time. In this phase planning mostly remains uncertain. How uncertain, depends on the project. This is not documented during the sales phase. After acceptance of quotation, planning will mostly become more specified. Due dates are set in accordance with the customer. Uncertainty regarding planning should become lower than during the sales phase. It is however not documented how well the planning is specified. Best option is to ask the project engineer since he experience the most of planning uncertainty.

Specification uncertainty

In most cases certain amount of information is already specified otherwise sales cannot produce a suitable quotation. However one project is better (more static) specified than another. The level of clearness is not documented.

After the sales phase specifications should be clear and certain however, this is not always the case. Clients sometimes change layouts affecting the design of

transportation systems or changes in machine dimensions affecting enclosure dimensions. How well projects are specified is not documented. Project engineer is most aware of this issue.

Complexity

Complexity of the pre-production depends mainly on the different projects. One project is more difficult than another, resulting in a more difficult pre-production stage. The size of a project does not really reflect the difficulty. Perhaps the ratio between total costs and production hours does. Relatively high cost and low

pre-production hours would indicate an easy project where lower costs and more pre-pre-production hours would indicate a more difficult project since it requires relatively more pre-production time in comparison with the costs. Of course this also depends on the participants of a project. Inexperienced staff or trainees require more time. Therefore a qualitative measure would be perhaps be more accurate.

Dynamics Major changes during sales are well documented as revisions. During the rest of the pre-production stage changes are not documented. Documentation is always as-build. _{Only way to track changes during the rest of the pre-production stage is to ask the project engineers what changed in comparison with the initial design.}

Delays

An objective quantification of delays through the ERP system or planning would be the best measurement. However planning is complex and very dynamic. Due dates/ due date periods change often and changes are not documented structurally. Therefore it is very difficult to track changes in the planning as it changes daily. There is a high level of shifting projects in consultation with clients making it even more difficult to track and trace. The project engineer is most aware of the planning of his projects and delays of activities.