Complexity in a Life Cycle Perspective

Available online at www.sciencedirect.com

ScienceDirect Procedia CIRP 61 (2017) 104 – 109

The 24th CIRP Conference on Life Cycle Engineering

Complexity in a Life Cycle Perspective Christoph J. Veltea,*, Anja Wilfahrtb, Robert Müllera, Rolf Steinhilpera,b b

a Fraunhofer IPA - Project Group Regenerative Production, Universitaetsstr. 9, Bayreuth D-95447, Germany Chair Manufacturing and Remanufacturing Technology, University of Bayreuth, Universitaetsstr. 30, Bayreuth D-95447, Germany

* Corresponding author. Tel.: +49-921-78516-422; fax: +49-921-78516-105. E-mail address: [email protected]

Abstract Transforming linear businesses to circular economies is anticipated by industry and policy as a way to conciliate economic, societal and environmental interests in a life cycle perspective. This integration of aspects however comes at the price of complexity in manifold facets. In this paper, we suggest a conclusive categorization of complexity through a literature review and collect drivers of complexity. Through coding the literature for weighted interdependencies, we are able to show how drivers and categories interrelate in a contingency matrix. The results help companies to better leverage existing means of complexity planning and management for developing circular economies. © 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license © 2017 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/4.0/). Conference on Life Cycle Engineering. Peer-review under responsibility of thecscientific of the 24th CIRP under responsibility of the scientifi committeecommittee of the 24th CIRP Conference on Life Cycle Engineering Peer-review

Keywords: Complexity; Production Systems; System Theory; Qualitative Data Analysis; Complexity Management; Interdependency

1. Introduction Complexity is a very complex thing. It is often talked about, while the very meaning can differ from one source to another, from one field of application to the next [1]. In fact, complexity can be seen as an umbrella term that subsumes different phenomena. Those aspects of complexity will be presented in section 2 of this paper. A problem that arises from the ambiguous meaning of complexity on the one hand, and the limited meaningfulness of talking about complexity without detailing the real problem on the other hand, is the fact that researchers as well as practitioners struggle to find the right measures to effectively deal with the complexityinduced problems that are described. This is especially important when taking into account that complexity is not a singular event, but, when thinking about complexity as a system property as done in various disciplines such as “philosophy, the physical sciences, engineering and management” [2, p.13], complexity is present in a system’s elements and their connections [3]. This results in nonobvious, dynamic effects throughout systems, where local differ from global, and short-run differ from long-run phenomena [3]. In the light of a discrepancy of how

complexity is dealt with today and the role of complexity towards a circular economy [4], analyzing drivers is an important start to the control of interdependencies in a life cycle perspective. To help foresee causes and effects better, this paper presents a number of complexity drivers as they are described in the literature and analyzes them for their interrelations through an extensive data analysis. Thus it can be shown how, through these interrelations, changes in one area of a company (e.g. the process complexity) trigger changes in other, connected areas of the same company (e.g. order fulfillment). To reach this goal, this paper will introduce the state of the art concerning systems and models, complexity and qualitative data analysis in engineering in section 2. Section 3 will present the analyzed complexity driver data and result in a list of complexity drivers, categorized for their area of origin and nature. Section 4 will present how those drivers are analyzed for their interdependencies and result in a contingency matrix that presents the weighted mutual influences of the drivers. The results will show how the drivers and categories interrelate. Finally, section 5 will summarize the findings and give an outlook on forthcoming research based on the results presented in this paper.

2212-8271 © 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the scientific committee of the 24th CIRP Conference on Life Cycle Engineering doi:10.1016/j.procir.2016.11.253

105

Christoph J. Velte et al. / Procedia CIRP 61 (2017) 104 – 109

2. Complexity and systems

results that can be further applied with and deepened through e.g. system dynamics models.

In this section, we introduce the state of the art in systems thinking and models, the definition of complexity and the use of qualitative data analysis techniques to clarify the framework used in this paper.

Feedback

E1

2.1. Systems and models To think of production as a system is a good way to make something that is intangible understandable and describable. In this vein, to familiarize with the term complexity and dealing with complexity in companies, the two schools of systems theory and cybernetics have proven to be capable [6]. For an ease of understanding, systems can be pictured as networks that, in their most basic appearance, have at least two elements that are connected by one relation [7]. A system’s model is a simplification of the very system analyzed, limited to the elements and relations that are to be investigated through this model [8], the subjective view of the modeler and the point in time when it is created [9,10]. Systems theory aims at describing such systems using consistent definitions and well described tools in order to understand the basic interrelations and principles of the system on an abstract level [11]. To frame the necessary understanding of a system as it is relevant in the context of this paper, it is furthermore necessary to state that systems and their elements have some distinct properties that define them. Firstly, a system is always limited by a system boundary, separating the elements that are system-internal from those, that are external to the system. Here we talk about system endogenous and exogenous areas [5] or, meaning the same differentiation, the system area versus a system’s environment [11]. Secondly, elements of the system may influence each other in various constellations. The simplest of those is a cause and effect relationship, where an action or change in one element directly causes a change in another element [5]. In reaction to this stimulus, the affected element may affect another element again. If this reaction from the affected element directly feeds back to the element that sent out the initial stimulus, the relationship between those elements is called a (direct) feedback loop [12]. The third class of properties that define a system, besides the fact that it has boundaries and the described elements and relations within these, are quantity and kind of inputs to the system and outputs of the system that pass through these boundaries. The system is influenced by its environment and might influence back surpassing its own boundaries [5]. This is also the case if the system imagined is a subsystem that communicates with the rest of the bigger system. An overview based on a systems understanding as it is taken up in this work based on [5] can be found in Fig. 1. Though due to the scope of this paper, we will not apply the results presented in system dynamics already, it is important to have this understanding of systems and relations within systems to understand the main motivation for the analyses presented. Thinking in systems makes it necessary to analyze sources of complexity in this spirit in order to create

System boundary

System input

E2

System elements E3

E4

System

System output

System environment

Fig. 1. Main terms and elements of a system based on [5]

2.2. Complexity Having gained an understanding of a system and a system’s model, it is easy to understand what is meant when talked about complexity. Building on complexity as defined e.g. by Wildemann, a first delineation can be drawn between structural complexity and dynamics [13,14]. While structural complexity origins from quantity of elements and connections together with the degree of difference between elements as well as connections (compare also [15]), the dynamic aspect of complexity covers the fact that this structure changes over interdependently and is unpredictable and time indeterminable [16]. Different opinions prevail on whether insecurity and opacity belong to the dynamic component of complexity or are a separate category (compare [16–19]), while in this paper we subsume those elements under dynamics and describe complexity as consisting of the static elements quantity and variety and the dynamic elements dynamics and interdependency (compare [18]). Depending on the degree of structural and dynamic complexity, and transferring this idea of complexity to systems thinking, systems can be classified to be simple systems if both structural and dynamic complexity are low, complicated systems if only the structural complexity is high, dynamic systems if only the dynamic complexity is high, and complex systems if they are determined both by a high structural and dynamic complexity [20]. 2.3. Qualitative data analysis, inter-rater-reliability and contingency matrix To be able to analyze qualitative data, such as written text, social sciences know a technique called qualitative data analysis. This technique enables the structured analysis of written text as to its content in order to interpret the meaning of a text and make sense out of it rather than repeating it [21]. Coming from e.g. interview interpretation and bottom-upcoding, this technique is nowadays applied in different fields of research with all kinds of documents [22] and has also been applied to analyze product development data in the engineering sciences [23]. According to Yin, qualitative research is structured into five phases called compilation of data, disassembly of data, reassembly of data, interpretation

106


of data and conclusion [24]. Following this approach, we will code the collected drivers of complexity in this paper according to their area of origin and the dimension of complexity they represent. Compilation and disassembly of data represents the literature research and collection of single drivers, reassembly of data is the categorization in a structure developed in chapter 3.1, interpretation of data and conclusion will follow in chapters 3.2 and 3.3. To assure that this categorization of complexity drivers is objective and not biased by the opinion of one researcher, it is an accepted and approved way to do the coding by more than one expert independently [23,24] and to calculate the inter-raterreliability using Cohen’s Kappa (Κ) [25]. The interdependency of drivers in this paper is analyzed using a contingency matrix as described e.g. by Meyer [26], where a pairwise comparison of drivers is indicated in the specific row and column of a matrix enlisting the drivers [27]. The intensity of the interdependence can be assigned with a degree of 0 for no interdependence, +1 to +3 for limited, strong and very strong intensity of a positive correlation and -1 to -3 for a limited, strong and very strong intensity of a negative correlation [26,28]. The active sum of a driver in the contingency matrix is calculated for each row and is an indicator for how influencing this row is throughout all interdependencies, the passive sum for each column is an indicator for how influenced the driver in this column is, i.e. how sensitive it reacts to changes in the system [29,30]. 3. Analysis of complexity drivers To get a hold of the complexity drivers as they are described in several streams of literature (such as logistics, systems theory or business administration), a literature review especially in German literature on complexity has been conducted. This limitation was chosen in order to come to an understanding of how complexity is perceived and described as within a regionally limited sub-system. With this limitation in mind, we assume that for this subsystem, what is described in the literature covers the actual main aspects of complexity as they occur in practice. To analyze the drivers and their interrelations, we first present the research on complexity driver fields relevant for a company (section 3.1), before going into detail with an actual list of drivers as they are described and categorizing them according to these fields and the type of complexity they represent (section 3.2). Section 3.3 discusses the results of this analysis. 3.1. Fields of complexity drivers in a company To acquire an understanding of what drives complexity in companies, in this paper we subdivide where complexity origins in a company into different fields. Thus, we can categorize the drivers of complexity as to their origin in these fields and consequently better understand how a change in one field influences other fields of the company and how complexity spreads. These fields are called complexity cluster. In order to do so, we first determine the different areas that complexity in production plays a role in,

subdividing the basic differentiation into internal complexity, external complexity and interface complexity as proposed by several authors (e.g. [31, 32]). In a second step, a literature review ([13,18,31,33–36]) concerning clusters of complexity and the frequency of their indication was conducted. Different authors from different streams of literature use different categories to group complexity drivers they identify. Through an analysis of these categories, a list of most-often used terms was generated. This list is depicted in Table 1. Table 1. Complexity clusters used in the literature Cluster Product Customer Process Technology Competitors Goals Organization Product program Purchase

Indications 9 5 4 4 3 3 3 3 3

Cluster continued Sales Order fulfillment Market Autonomy Coordination Development Disposal Individual complexity Information

Indications 3 2 2 1 1 1 1 1 1

Table 1 shows that, without paying respect to internal, interface or external complexity, product complexity is the most often mentioned cluster of complexity drivers, followed by customer complexity and process complexity. Other categories, such as purchase, the organization, the product program and sales follow. Based on the categories in Table 1, and reintroducing the differentiation between internal, interface and external complexity, we use the following differentiation of clusters in this paper: Internal complexity comprises the categories product (joining product and product program from Table 1), process, order fulfillment and organization. External complexity comprises competition and customer. In addition, to get a grasp on how legislative impacts from outside the system boundaries influence the system within, legislation is included in the external category 1 . Finally, interface complexity comprises complexity drivers arising from (subsuming both information and communication coordination from Table 1), purchase and sales. This leads to the differentiation depicted in Table 2. Table 2. Internal, interface and external clusters of complexity Internal complexity

Interface complexity

External complexity

Product Organization Process Order fulfillment

Purchase Communication Sales

Customer Competition Legislation

In terms of the dynamics of life-cycle-planning, it will be important to highlight interdependencies of those clusters in order to determine what a decider should work on first and how the system will react during the product life cycle.

1

This is also based on the concrete drivers that could be identified during the analyses.

107


3.2. Complexity drivers and categorization The analysis of complexity drivers as they are described in the aforementioned literature resulted in an initial list of 243 individual complexity drivers. In a second step, these drivers were refined, ambiguous and similar drivers were consolidated and drivers that are not sufficiently described within the source as to their origin and meaning were discarded. This process resulted in a refined list of 195 complexity drivers that are analyzed further 2. In the next step, the drivers were categorized as to the complexity clusters presented in section 3.1, mostly oriented at the original categories from the sources, while for each driver, this categorization was challenged and reassured. During this same process, the drivers were categorized into quantity, variety, dynamics or interdependency (compare [18])3. For the results of the coding process, done by two coders individually, Cohens Κ [25], an indicator for the fineness of the codes assigned, reaches a value of 83 %, which represents an “Almost Perfect” agreement referring to Landis and Koch [37, p.165], thus fostering the validity of the codes assigned.

about one quarter of all drivers falling into this category (24 %). Measures that help to control quantity and variety of items thus could already help cure three quarters of all drivers of complexity as experienced in companies. Nevertheless, dynamics with 15 % and interdependency with 11 % make up a share of drivers both well above 10 % and together responsible for more than one quarter of drivers of complexity. Again without going into detail towards the originating field of this driver, methods that help cure dynamics and interdependency above the controllable level can be applied towards every fourth source of complexity. 30%

11% 15% 50%

Quantity Variety Dynamics Interdependency

24%

Sales Interface

21%

Legislation External

18%

20% 15%

13%

10% 4%

5%

6%

5%

4%

3%

2%

0%

3.3. Results of the complexity driver analysis The analyses in chapter 3.1 and 3.2 made it possible to collect drivers, reassure the correct categorization for the complexity clusters as proposed in this paper and analyze them for their type of complexity. Thus we get an impression of what kinds of complexity constellations are described mostly in the literature and, most important, have a look behind drivers that are described e.g. as “product complexity” to see what they origin from. The results of this categorization can be seen in Fig. 2 and Fig. 3.

Internal Process

24%

25%

Fig. 3. Share of complexity drivers per cluster

Fig. 3 depicts the share of complexity drivers that fall into each cluster. Here, internal complexity clearly dominates with about 67 % of all drivers, while the customer cluster from the external drivers of complexity is the fourth-strongest source of complexity after product, organization and process, and more influencing than order fulfillment from the internal cluster. The least described drivers of complexity are interfacedrivers (12 % in total), while purchase complexity with 6 % still exceeds competition (5 %) and legislation (3 %) from the external cluster. Number of indications

20 15

Quantity Variety

10 5

Dynamics

0

Interdependency

Fig. 2. Share of drivers per type of complexity of the 195 drivers identified

The analysis of Fig. 2 clearly shows that according to the drivers described in the literature, quantity is the dominating type of complexity in companies with about 50 % of all drivers of complexity described falling into this category. This does not say anything about where this quantity complexity is experienced yet, but already gives a good hint as to what kinds of complexity management strategies should be further developed and cultivated within companies in order to control a large share of the drivers experienced. The second most often categorized type of complexity is variety with

2

A complete list of those drivers would exceed the extent of this article, but can be requested from the corresponding author if needed. 3 The driver “Amount of orders to be coordinated” e.g. was coded with the complexity type quantity and the complexity cluster order fulfilment.

0-5

5-10

10-15

15-20

Fig. 4. Number of indications per cluster and type of complexity

The area diagram in Fig. 4 shows how often each combination of cluster and type of complexity resulted from coding the drivers identified in the literature, combining the insights from Fig. 2 and Fig. 3. On first sight, there are three main peaks, the first and highest marked by the clusters product organization and process together with the types quantity and variety, the second marked by customer with types quantity and variety, the third marked by a combination of organization and interdependency. In total, the highest peaks are coming from the individual combinations quantity and product (20), followed by variety and product (19),


quantity and organization (16) and quantity and process (12). In this “mountain range of complexity”, the area of interface and external drivers of complexity (cluster) together with dynamics and interdependency is relatively flat, indicating that it is not that much identified as driving complexity in producing companies. On the contrary, those areas that are driving complexity are most often dealing with quantity or variety in terms of the type of complexity, and the internal drivers of complexity in terms of cluster. 4. Interdependency of drivers and clusters After analyzing the cluster and type of complexity of each driver, the next step of analyses focuses on how those drivers interrelate. To ultimately be able to determine feedback loops within the system of complexity drivers, the literature was searched concerning the concrete mention of influence. Certain text bits, as they could be identified in the literature, were given a weight of influence they represent between two drivers described. An overview of these references is shown in Table 3. Table 3. Influencing matrix Value Intensity Text in the literature 0 1

No interrelation Weak interrelation

2

Strong interrelation

3

Very strong interrelation

No source Has impact, complexity increases, is influenced in complexity, complexity grows, additional complexity is generated, can induce complexity, contributes to an increase in complexity, is the cause of growing complexity Impacts strongly, has strong impact, contributes strongly to an increase in complexity, strong correlation Influences significantly, can increase complexity massively, contributes to a massive increase, has an enourmous influence, determines decisively, very strong correlation

In this way, both positive as well as negative influences were collected. In the following analysis, only those drivers are taken into account that could be assigned an influence on another driver or by another driver through the literature research. This limitation resulted in 113 drivers that were analyzed towards their interdependency. Those 113 drivers were pairwise compared to each other, resulting in a contingency matrix that gives the values for each interrelation in the intersection of row and column of the drivers. A calculation of the active sum (ΣA) [26] and passive sum (ΣP) [29] for each driver in the matrix calculates how influencing or sensitive a driver is in total. To calculate an active and passive sum for both negative and positive influences, one would add up the modulus of each influence for the specific driver [38]. In contrast to the modulus calculation, we used this matrix to calculate a contingency matrix for the clusters of complexity to evaluate the total influence between fields rather than the activity. To do so, positive values of influence for each driver within one cluster were added up and negative values of influence were subtracted from this sum for each intersection of clusters. This gave us a good overview of how e.g. an increase of complexity in the cluster products influences other clusters, e.g. the order fulfillment. The contingency table that resulted from this analysis of complexity clusters is shown in Table 4.

Table 4. Contingency matrix of complexity clusters (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) ΣA (1) Order fulfillment (2) Purchase (3) Legislation (4) Communication (5) Customer (6) Organization (7) Process (8) Product (9) Sales (10) Competition ΣP

5 2 1 8 14 16 26 21 2 7 102

2 0 1 0 0 0 -4 0 -1 0 4 0 2 0 10 0 0 2 2 10 26 12

8 8 0 8 1 22 8 1 0 4 60

2 17 0 7 11 10 2 14 16 15 3 68 0 20 8 46 5 1 27 19 74 217

-3 8 11 5 21 24 32 62 0 29 195

19 0 9 2 22 7 1 -2 68 6 11 0 22 2 107 11 7 0 71 4 337 34

1 0 6 0 11 0 0 -5 1 59 83

57 37 68 44 153 148 112 271 18 232

Table 4 highlights the fact that the highest correlation of drivers exists within the cluster product, meaning that if complexity in the cluster product rises through an increasing driver of complexity, it mostly influences other drivers in the cluster product. The second highest active influence is directed from competition to product, the third is from customer to product with the same total number as the influence of drivers within organization on other drivers in organization. Additional to this, Fig. 5 gives a good impression of what clusters are most sensitive to changes in the system on the one hand (reactive, upper left corner), and what clusters are the ones that, once confronted with a rising complexity, quickly spread this increase throughout the system on the other hand. To do so, the clusters are divided into four fields: clusters with rather high active sum and low passive sum are called “active”, rather passive clusters are called “reactive”, clusters where both sums are low are called “inertial”, clusters where both sums are high are “critical” [26]. As active clusters, both customer and competition are highly influential but not very reactive, which makes it worthwhile to work on the complexity in those clusters first, with the dynamic repercussion during a products life-cycle in mind. Critical is the cluster product. 400

reactive

critical

350

Passive Sum

108

1 2

8

300

3 4

250 200

7

inertial

150

5

6

6

active

7

1

100 4 99 2 2

50

0 0

50

8

10 10

5

9 10

3

100

150

200

250

Order fulfillment Purchase Legislation Communication Customer Organization Process Product Sales Competition

300

Active Sum Fig. 5. Active and passive sum of complexity clusters

5. Conclusion and outlook The purpose of this paper was to get an impression of what drivers of complexity are present and described in the literature, but going beyond the state of the art, having a look behind those drivers and decoding them for their actual origin in terms of what we introduced as complexity cluster and type of complexity. This makes it possible to locate the problems within a company. An analysis of the interrelation of those drivers showed that indeed complexity seems to be highly


sensitive to changes and contagious within a system. Those areas that are driving complexity are most often dealing with quantity or variety in terms of the type of complexity, and the internal drivers of complexity in terms of cluster. An analysis of the interrelations gives an impression of how complexity procreates within a producing system during a product life cycle. Fig. 5 shows that customer, competition and product are most beneficial to work on in a long-term, life cycle perspective as they are actively influencing other clusters, but are rather inertial themselves. As a next step within our research, the results generated will be modelled dynamically with the use of system dynamics to numerically show how the interrelations uncovered have a bearing on systems during product life cycles. This will show what equilibria result and enable us to better leverage methods of dealing with complexity and modify or develop adequate new methods where necessary. Acknowledgements This paper is a result of the research project KonPrO at the Fraunhofer Project Group Regenerative Production. The project is funded by the Bavarian Ministry of Economic Affairs and Media, Energy and Technology. The authors would like to express their gratitude for enabling the research. References [1] Lechner, A., 2012. Modellbasierter Ansatz zur Bewertung vielfaltsinduzierter Logistikkomplexität in der variantenreichen Serienfertigung der Automobilindustrie. Verl. Praxiswissen, Dortmund. [2] Bozarth, C.C., Warsing, D.P., Flynn, B.B., Flynn, E.J., 2009. The impact of supply chain complexity on manufacturing plant performance 27, p. 78. [3] Senge, P., 1994. The Fifth Discipline. Currency, New York. [4] Velte, C.J., Steinhilper, R., 2016. Complexity in a Circular Economy: A Need for Rethinking Complexity Management Strategies, in Proceedings of the World Congress on Engineering 2016: WCE 2016, June 29 - July 1, 2016, London, U.K., Newswood Limited, Hong Kong, p. 763. [5] Bossel, H., 1989. Simulation dynamischer Systeme: Grundwissen, Methoden, Programme. Vieweg+Teubner Verlag, Wiesbaden. [6] Malik, F., 2008. Strategie des Managements komplexer Systeme: Ein Beitrag zur Management-Kybernetik evolutionärer Systeme, 10th edn. Haupt-Verl., Bern. [7] Bertalanffy, L.v., 1972. Vorläufer und Begründer der Systhemtehorie, in Systemtheorie, Colloquium Verl., Berlin, p. 17. [8] Tabeling, P., 2006. Softwaresysteme und ihre Modellierung: Grundlagen, Methoden und Techniken ; mit 45 Tabellen. Springer, Berlin. [9] Müller-Steinfahrt, U., 2006. Diffusion logistischen Wissens, Denkens und Verhaltens in Grossunternehmen, 1st edn. Kölner Wissenschaftsverlag, Köln. [10] Kosiol, E., 1968. Einführung in die Betriebswirtschaftslehre: Die Unternehmung als wirtschaftliches Aktionszentrum. Gabler Verlag; Imprint, Wiesbaden. [11] Westkämper, E., 2009. Wandlungsfähige Produktionsunternehmen: Das Stuttgarter Unternehmensmodell. Springer, Berlin. [12] Sterman, J.D., 2000. Business dynamics: Systems thinking and modeling for a complex world. McGraw-Hill, Boston. [13] Wildemann, H., Voigt, K.-I., 2011. Komplexitätsindex-Tool: Entscheidungsgrundlagen für die Produktprogrammgestaltung bei KMU. TCW-Verl., München. [14] Schuh, G., 2005. Produktkomplexität managen. Hanser, München. [15] Luhmann, N., 1973. Komplexität, in Handwörterbuch der Organisation, Poeschel, Stuttgart, p. 1064. [16] Schoeneberg, K.-P., 2014. Komplexitätsmanagement in Unternehmen: Herausforderungen im Umgang mit Dynamik, Unsicherheit und

109

Komplexität meistern. Springer Fachmedien Wiesbaden; Imprint: Springer Gabler, Wiesbaden, s.l. [17] Zielowski, C., 2006. Managementkonzepte aus Sicht der Organisationskultur: Auswahl, Ausgestaltung und Einführung, 1st edn. Dt. Univ.-Verl., Wiesbaden. [18] Blockus, M.-O., 2010. Komplexität in Dienstleistungsunternehmen: Komplexitätsformen, Kosten- und Nutzenwirkungen, empirische Befunde und Managementimplikationen, 1st edn. Gabler, Wiesbaden. [19] Westphal, J.R., 2001. Komplexitätsmanagement in der Produktionslogistik: Ein Ansatz zur flussorientierten Gestaltung und Lenkung heterogener Produktionssysteme, 1st edn. Dt. Univ.-Verl., Wiesbaden. [20] Ulrich, H., Probst, G., 1991. Anleitung zum ganzheitlichen Denken und Handeln: Ein Brevier für Führungskräfte. Haupt, Bern. [21] Wholey, J.S., Hatry, H.P., Newcomer, K.E., 2004. Handbook of practical program evaluation, 2nd edn. Jossey-Bass, San Francisco, CA. [22] Flick, U., 2014. The SAGE handbook of qualitative data analysis. Sage, London. [23] Velte, C.J., Butzer, S., Steinhilper, R., 2015. Qualitative Data Analysis in Product Development: An Exploration of Closed Loop Thinking in Product Requirements, in ASME 2015 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, American Society of Mechanical Engineers, New York, NY, V004T05A054. [24] Yin, R., 2011. Qualitative Research from Start to Finish. The Guilford Press, New York. [25] Cohen, J., 1960. A Coefficient of Agreement for Nominal Scales 20, p. 37. [26] Meyer, C.M., 2007. Integration des Komplexitätsmanagements in den strategischen Führungsprozess der Logistik, 1st edn. Haupt, Bern [u.a.]. [27] Hanenkamp, N., 2004. Entwicklung des Geschäftsprozesses Komplexitätsmanagement in der kundenindividuellen Serienfertigung: Ein Beitrag zum Informationsmanagement in mehrdimensional modellierten Produktionssystemen. Shaker, Aachen. [28] Puhl, H., 1999. Komplexitätsmanagement: ein Konzept zur ganzheitlichen Erfassung, Planung und Regelung der Komplexität in Unternehmensprozessen. Lehrstuhl für Fertigungstechnik und Betriebsorganisation, Univ, Kaiserslautern. [29] Lasch, R., Gießmann, M., 2010. Die Logistik unter dem Einfluss der zunehmenden Komplexität: Ergebnisse einer empirischen Untersuchung zum Komplexitätsmanagement in der Praxis, in Dimensionen der Logistik: Funktionen, Institutionen und Handlungsebenen, Gabler, Wiesbaden, p. 846. [30] Drews, G., Hillebrand, N., 2010. Lexikon der ProjektmanagementMethoden: [die besten Methoden für jede Situation Werkzeugkasten für effizientes Projektmanagement], 2nd edn. Haufe, Freiburg, München [u.a.]. [31] Lübke, E., 2007. Lebenszyklusorientiertes Produktstrukturmanagement: Eine theoretische und empirische Untersuchung, 1st edn. TCW-TransferCentrum, München. [32] Sivadasan, S., Efstathiou, J., Calinescu, A., Huaccho Huatuco, L., 2004. Supply Chain Complexity, in Understanding supply chains: Concepts, critiques, and futures, Oxford University Press, Oxford, New York, p. 133. [33] Klepsch, B., 2004. Komplementäre Produkt- und Fabrikmodularisierung am Beispiel der Automobilindustrie. VDi-Verlag GmbH, Düsseldorf. [34] Bliss, C.A., 1998. Komplexitätsreduktion und Komplexitätsbeherrschung bei der Schmitz-Anhänger FahrzeugbauGesellschaft mbH, in Komplexitätsmanagement, Gabler, Wiesbaden, p. 145. [35] Mayer, A., 2007. Modularisierung der Logistik: Ein Gestaltungsmodell zum Management von Komplexität in der industriellen Logistik, 1st edn. Univ.-Verl. der TU, Univ.-Bibliothek, Berlin. [36] Marti, M., 2007. Complexity management: Optimizing product architecture of industrial products, 1st edn. Dt. Univ.-Verl., Wiesbaden. [37] Landis, J.R., Koch, G.G., 1977. The Measurement of Observer Agreement for Categorical Data 33, p. 159. [38] Giebel, M., 2011. Wertsteigerung durch Qualitätsmanagement: Entwicklung eines Modells zur Beschreibung der Wirkmechanismen und eines Vorgehenskonzepts zu dessen Einführung. Kassel University Press, Kassel.