Design for complexity: a global perspective through ...

Int. J. Industrial and Systems Engineering, Vol. 11, No. 3, 2012

279

Design for complexity: a global perspective through industrial enterprises analyst and designer Ibrahim H. Garbie Department of Mechanical and Industrial Engineering, Sultan Qaboos University, P.O. Box 33, Al-Khoud 123, Muscat, Sultanate of Oman Fax: +968 2441 1316 E-mail: [email protected] Abstract: Industrial enterprises analysts and/or designers should be aware of the impact of complexity in their organisations, although they are often defined as being complex. Nowadays, the researchers focused their attention on design for manufacturing, design for assembly, design for cost or design for quality, design for X, etc. they did not mention design for complexity as an important issue especially during the existing global financial crisis. Design for complexity is a systemic approach that simultaneously considers optimising design objectives (i.e. minimise complexity level), variables (parameters) and constraints. This paper includes how to present the concepts of complexity to guide industrial enterprises analysts and designers with the most effective issues and perspective strategies for analysing, planning and eliminating complexity to satisfy design of industrial enterprises. Based on these aspects, the complexity levels will be analysed and evaluated through identifying four major issues: design for vision complexity, design for system structure, design for operating complexity and design for evaluating complexity. The ultimate goal of this paper is to provide the industrial enterprises designers with such complexity information. This analysis shows that the design for complexity is a huge task and should be optimised and taken into considerations when designing an industrial enterprise. Keywords: analysis of industrial enterprises; design for complexity; complexity analysis. Reference to this paper should be made as follows: Garbie, I.H. (2012) ‘Design for complexity: a global perspective through industrial enterprises analyst and designer’, Int. J. Industrial and Systems Engineering, Vol. 11, No. 3, pp.279–307. Biographical notes: Ibrahim H. Garbie is currently an Assistant Professor in the Department of Mechanical and Industrial Engineering at the Sultan Qaboos University, Sultanate of Oman. He received PhD in the Department of Industrial Engineering from the University of Houston, Texas, USA in 2003. His current research focuses on manufacturing/production systems design, complexity analysis in industrial firms, manufacturing leanness, agility measures and reconfiguration of manufacturing systems. He is a Senior Member of IIE.

Copyright © 2012 Inderscience Enterprises Ltd.

280

1

I.H. Garbie

Introduction

Nowadays, industrial enterprises complexity exists as an immense international interest and knowledge for the scientific basis. Studying the importance of reducing or eliminating complexity into industrial organisations was recommended as one of the several solutions to recovery the existing financial recession (Garbie, 2009, 2010a,b; Garbie and Shikdar, 2009, 2010a,b, 2011). Complexity was defined as systemic characteristics which integrate several key dimensions of the industrial environment including size, variety, information, uncertainty, control, cost and value (Kamrani and Adat, 2008). Flexibility and agility are considered as the most desirability of certain system properties for the industrial enterprises with respect to structural and operational complexity measures. These properties will give industrial enterprises more ability to cope with increased environmental uncertainty and adapting to the faster pace of change of today’s markets (Giachetti et al., 2003). The role of strategic decision-makers was presented in determining which properties the enterprises must exhibit, to what level the enterprise requires those properties and how best to incorporate those properties into the enterprise (Giachetti et al., 2003). The industrial enterprises are sometimes viewed as an intrinsic structural property of the system. The structural property consists of individual system components relate to each other and how the relationship determines overall system behaviour (Arteta and Giachetti, 2004). The complexity arises not only from the size of the enterprise but also from the interrelationships of the enterprise components and the emergent behaviour that cannot be predicted from the individual enterprise components. The structural complexity provides a good description of the inherent complexity of its components, the relationship among them and their influence (Kuzgunkaya and ElMaraghy, 2006). But dynamic complexity requires data normally obtained during actual operations or simulation of the shop floor. Effecting structural complexity on job shop manufacturing system is studied, in which processing time and scheduling rules are taken into consideration (Jenab and Liu, 2010). Relative complexity of business process was recommended as the basic of design of enterprises including business rules, activities, organisation, process, process and resources (Johnson, 2008). There are several concepts, such as product, process and operation complexity, which can be used as a component in industrial enterprises. The product complexity focuses on product features and specifications. Product variety sometimes is considered as one of industrial enterprises complexity. This means increasing in the product variety increases the complexity in the industrial enterprises (Kuzgunkaya and ElMaraghy, 2006). While the process complexity analysis focuses on the tools, equipment and operations are used to manufacture it (Hu et al., 2008). However, operational complexity was considered as the cognitive and physical effort associated with the tasks related to a product/process combination. Supply chain management (SCM) complexities, such as upstream complexity, internal manufacturing complexity and downstream complexity, are considered as complexity issues regarding manufacturing systems (Bozarth et al., 2009). Manufacturing strategy (MS) also plays an important role in complexity in industrial enterprise, such as just-in-time manufacturing (lean manufacturing), flexible manufacturing, cellular manufacturing, agile manufacturing, concurrent engineering, etc. The system complexity for lean manufacturing, e.g. is affected by increased product variety when comparing with mass production system (Kuzgunkaya and ElMaraghy, 2006). New aspects of complexity in industrial enterprises are introduced as the

Design for complexity

281

determinants of complexity levels, such as system vision, system design (SD), system operating and system evaluation (Garbie and Shikdar, 2009). Measuring complexity levels are defined as not only intrinsic to the system being studied but also depend on extrinsic properties of the observer. Although most measurements were concentrated on operational measures, both structural and operational characteristics are important to the performance of the system as a whole. Determining the industrial enterprise complexity still has different concepts and views. Structural (static) complexity measure was introduced by Arteta and Giachetti (2004) as the probabilities associated with the state (uncertainty) of each resources, such as machines, people, parts, routings, etc. although this complexity measure does not include the relationship between resources. Operational complexity measures the uncertainty associated with the material and information flows of the system (Arteta and Giachetti, 2004). Measuring the manufacturing complexity in assembly lines based on assembly activities are presented with different configurations and MSs (Wang and Hu, 2010). Also effecting of scheduling rules with processing times on hybrid flow a system is investigated (Yang, 2010). The complexity levels in industrial firms are estimated through several case studies based on general framework that includes a questionnaire focusing on each issue in a firm (Garbie and Shikdar, 2011). The main purpose of this paper is analysing the major issues of complexity in industrial enterprises. This will lead to suggest a framework to minimise or optimise the complexity levels. This paper is organised into several sections. Section 1 presents the importance of complexity concepts. Section 2 reviews previous research regarding complexity issues. Design requirements and a procedure for estimating complexity level will be provided and a framework regarding these requirements will also be analysed and explained in Section 3. Section 4 presents a hypothetical example of how to implement the proposed procedure. Conclusions and recommendations for future work will be introduced in Section 5.

2

Literature review

Several research works have been published in this area but none of the work mentioned the concept of design for complexity. Kamrani and Adat (2008) used simulation model to analyse the complexity in mixed-model assembly production systems. A framework to analyse the structural properties of the enterprise system was presented by Giachetti et al. (2003). They tried to compare between structural properties and operational measures. A complexity measure for the business process level inside the organisation on one product was developed by Arteta and Giachetti (2004) (e.g. the prepaid phone card). They used business process reengineering to reduce the complexity of the process regarding business. A new metrics for assessing the structural complexity of system configurations was suggested through machine complexity, buffer type complexity and material handling system (MHS) complexity (Kuzgunkaya and ElMaraghy, 2006). Jenab and Liu (2010) studied the impact of structural complexity in job shop manufacturing system including and considering direct and indirect relationship between the processes and the required resources. They used products with their processing time, types of skills and resources. Kristianto and Helo (2011) used manufacturing process based on product development (PD) to design agile supply chain taken into consideration strategic safety stock allocation. Effecting of product variety in assembly systems on manufacturing

282

I.H. Garbie

complexity, incorporating assembly activities with different configurations of these systems and assembly strategies was suggested. Product, process and operational complexity are introduced in modelling and assessment of manufacturing complexity (ElMaraghy and Urbanic, 2003, 2004). Each measurement was evaluated independently. Three different types of complexity to represent and model supply chain complexity are suggested by Bozarth et al. (2009), such as upstream complexity, internal manufacturing complexity and downstream complexity. Garbie and Shikdar (2009) identified complexity in industrial organisations into four main issues, such as system vision, system structure, system operation and system evaluation. Impact of product structure complexity on managerial operating decisions in remanufacturing environment was studied by Guide et al. (1997). Johnson (2008) used a concept of relative complexity of business process based on the design of enterprises architecture issues. These issues are business rules, activities, organisation, processes and resources. A coding system of machines, buffers and material handling equipment (MHE) to measure complexity based on time-independent complexity (static or structural complexity) of those major components was classified and designed by ElMaraghy et al. (2005). They based on different configurations of the system and compared between them according to which one had the lowest complexity level to be selected. Hu et al. (2008) used the same structural (static) complexity measure to evaluate the complexity in mixed assembly lines concentrating on providing parts to first station (feed complexity) and between stations (transfer complexity). Two different types of complexity, which were proposed by Tomiyama et al. (2007) are complexity by design and intrinsic complexity of multi-disciplinary. They considered a product design and PD as an inevitably associated with a multi-disciplinary design. Conducting a comprehensive case study involving 14 Italian companies to investigate how complexity can affect a manufacturing company’s performance and those of its supply chain was conducted by Perona and Miragliotta (2004). An analytical model for measuring system complexity based on information entropy and probability distribution of resource allocations was presented by Cho et al. (2009). They used static complexity measure to evaluate the system complexity based on the number of machines in the system (resources). Operational complexity was measured as a function of cost through SCM systems (Wu et al., 2007). They indicated that inventory costs are associated with operational complexity. Huatuco et al. (2009) used entropic related to complexity measures to quantify the complexity associated with information content of schedules and variations between schedules. Yang (2010) presented a computational approach to investigate effecting of scheduling with processing time on two-stage hybrid flow shop systems. A mathematical model was suggested to identify functional requirements and the associated design parameters regarding complexity level (Tomiyama et al., 2007). Kashyap and Sinha (2011) studied the complexity related to the mental fatigue required by a person for doing a specific job. While they used how to manage stress, they developed a general model to estimate overall complexity of a profession, although they tried to evaluate the job complexity of an engineer. Mazur and Chen (2011) presented organising work (OW), communication and managing conflict as the most important issues among team members to complete the project. They considered multi-functional knowledge, teamwork capabilities and working relationships with OW as a complex problem. A mathematical equation to measure the static complexity was derived by Frizelle and Woodcock (1995). Their equation reflects the complexity of structure of operation

Design for complexity

283

and demonstrates that such complexity only has meaning when considered in terms of the demand placed on it. Sivadasan et al. (2002) developed the operational complexity for measuring and analysing supplier–customer system. Deshmukh et al. (1998) presented a measure for quantifying static complexity in manufacturing systems and observed the behaviour of system performance with respect to the static complexity measure. They also proposed variation in static complexity representing in part similarity in processing requirements, system size and product design changes. Assessment of the structural, dynamic and decision-making (DM) complexity associated with the processing and movement of material and information around a manufacturing system was suggested by Efstathiou et al. (2002). Sharif and Irani (2006) presented the codification of the managerial DM process involved within the case company through the application of a standard fuzzy cognitive mapping and morphological analysis. They presented based on the applicability of fuzzy and morphological approaches to modelling complexity within DM. Tani and Cimatti (2008) used technological complexity as a tool to define the PD requirements and develop the feasibility studies to check the capability of manufacturing system to produce the product. Garbie (2010a) suggested complexity and agility levels as a new evaluation of manufacturing enterprises and it cannot separate between them. A mathematical expression to address machine status was presented as a complexity level for structure complex (Isik, 2010). In addition, relationships between customer and suppliers were modified by adding deviation of change of machine changes. It can be noticed from the previous review that designing for industrial enterprises complexity needs more attention from academicians. Based on the previous review, an in-depth analysis of industrial enterprise is used to introduce a global analysis of all components in an industrial enterprise and suggests a framework to analyse these components through the system vision complexity, system structure complexity and design complexity, system operating complexity and system evaluation complexity.

3

Design for complexity requirements

To identify the main requirements of design for industrial enterprises complexity (DFIEC), there are five important questions to be asked such as the following to describe how the complexity of industrial enterprises can be studied. Some of these questions were presented by Garbie and Shikdar (2011) in analysing and estimating the complexity levels in industrial organisations. 1

How the complexity issues of an industrial enterprise is identified and analysed?

2

How the complexity level of an industrial enterprise is estimated?

3

How can an industrial enterprise reduce its complexity?

4

Which issues are more important than others?

5

How can industrial enterprises identify the adverse factors for reducing complexity?

Based on these questions, full requirements for designing for complexity regarding these issues will be analysed and explained through the recommended issues from perspective of industrial enterprises analysts and designers. These can be represented into four main phases as follows (Figure 1). Figure 1 shows a briefly DFIEC procedure consisting of four phases. Each phase will be discussed with its associated issues. It can be noticed

284

I.H. Garbie

from Figure 1 that the procedure of designing for industrial enterprises complexity follows the four phases parallel and some phases can be revised or modified. The model of industrial enterprise components and the corresponding complexity relationships between them to emphasise on particular vision, design, operation and evaluation is presented in Equations (1) and (2). As each component or element in these systems is a potential source of uncertainty (due to its state), the measuring of complexity for each one is highly valuable. Also, due to human psychology, most people have an intuitive understanding of complexity and they will generally improve what is needed to be measured. Based on these concepts and issues, it can be noticed that the industrial enterprises complexity level (IECL) is a function of these major issues. The issues are industrial enterprise vision complexity (IEVC), industrial enterprise design complexity (IEDC), industrial enterprise operating complexity (IEOC) and industrial enterprise evaluation complexity (IEEC) (as shown in Figure 1). Then, IECL is clearly modelled as Equation (1) as function of previous sub-complexities.

IECL

f  IEVC, IEDC, IEOC, IEEC

(1)

Equation (1) can be rewritten as Equation (2). Each term of Equation (2) represents subcomplexity measure of IECL. Adding these terms with relative weights is considered. These weights can be used as a reason existing to differentiate between major issues of complexity.

IECL

wIEVC [IEVC]  wIEDC [IEDC]  wIEOC [IEOC]  wIEEC [IEEC]

(2)

where IECL is the industrial enterprise complexity level, IEVC is the industrial enterprise vision complexity, IEDC is the industrial enterprise design complexity, IEOC is the industrial enterprise operating complexity and IEEC is the industrial enterprise evaluation complexity. Figure 1

Four phases for design for industrial enterprises complexity (see online version for colours)

Design for complexity

285

The wIEVC, wIEDC, wIEOC and wIEEC are relative weights of enterprise vision, enterprise structure, enterprise operating and enterprise evaluation, respectively. Because the tradeoffs frequently exist between these objectives, a comprehensive analysis for each individual measure is needed. The value of these relative weights may reflect the system analyst’s subjective preferences based on his/her experience or can be estimated using tools, such as analytical hierarchy process (AHP). In this paper, the relative weights using the AHP are estimated and changed frequently according to the new circumstances by decision-maker or a group of decision-makers (Abdi, 2006). These groups are represented in senior management level, manufacturing and/or production engineers, plant managers, operators and suppliers. These relative weights can be estimated using AHP according to the next matrix. For example, suppose wIEDC / wIEOC 4 , then this means that design complexity (IEDC) is four times more important than operating complexity (IEOC).

AIECL

ª wIEVC «w « IEVC « wIEDC « « wIEVC « wIEOC « « wIEVC «w « IEEC «¬ wIEVC

wIEVC wIEDC

wIEVC wIEOC

wIEDC w IEDC

wIEDC wIEOC

wIEOC wIEDC

wIEOC wIEOC

wIEEC wIEDC

wIEEC wIEOC

wIEVC º wIEEC » » wIEDC » » wIEEC » wIEOC » » wIEEC » wIEEC » » wIEEC »¼

For each phase (major issue), there are some sequences can be performed to determine the level of complexity based on determining the main issues of each phase, preparing data collection and data file structure, conducting multiple regression analysis, optimising complexity level and searching for optimal items (components) and sensitivity analysis. This procedure is proposed as summarised in Figure 2.

3.1 Phase 1: industrial enterprises vision complexity The IEVC is the first step in the DFIEC. The major issue of this phase is how to collect the main components (elements) of industrial enterprises vision. A design for enterprise vision usually specifies what SCM representing in number of suppliers (NOS), demand variability (DV) representing in number of customers (NOC), introducing a new product (NP), product life cycle (PLC) representing also in PD and time to market (TTM) (Figure 3) require and how they are affected and effecting on the complexity of industrial enterprises. The IEVC will be represented mathematically as a function of these issues as shown in Equation (3).

IEVC

f (SCM(NOS), DV(NOC), NP, PLC(PD), TTM)

(3)

The data file structure for enterprise vision complexity will be used to determine the optimal level of complexity regarding vision and the main issues of these items, such as SCM (NOS), DV, introducing a NP, PLC (PD) and TTM, as shown in Table 1. The multiple regression analysis procedures for each issue will be conducted relating to each industrial enterprise individually to identify the relationship between each issue and

I.H. Garbie

286

complexity level regarding vision. The multiple regression model of IEVC is shown in Equation (4). K

IEVC

B0 

¦B X j

j

i 1

Figure 2

Procedure chart of estimating complexity level in each phase (see online version for colours)

Figure 3

Elements of industrial enterprises vision complexity (see online version for colours)

(4)

Design for complexity Table 1

287

Data file structure for industrial enterprises vision complexity IEVC

SCM (NOS)

DV (NOC)

NP

PLC (PD)

TTM

1

–

–

–

–

–

–

2

–

–

–

–

–

–

3

–

–

–

–

–

–

Enterprise no.

4

–

–

–

–

–

–

#

–

–

–

–

–

–

n

–

–

–

–

–

–

where B0 , B j are the regression coefficients of the IEVC model and K is the number of factors (independent variables) in the regression model. The uncertainty associated with each of these issues is assumed to follow a uniformly distributed random variable with known parameters (other probability distribution can be also considered, but for simplicity, the continuous uniform distribution is chosen). These regression coefficients can be determined via multiple regressions using several available statistical software packages, such as Minitab and/or SPSS. To determine optimal complexity level regarding vision, this model is basically used to calculate the optimal level of vision complexity taking into account SCM representing in the NOS, DV representing in NOC, introducing a NP, PLC (PD) and TTM. The model’s aim is to minimise the sum of these issues with constrained. So, the IEVC can be rewritten as shown in Equations (5) and (6) to optimise (e.g. minimise) the degree of vision complexity. K

Minimise, IEVC

¦B X j

j

(5)

i 1

s.t. ª NOS t L º ½ « »° « NOS  U » ° « NOC t L » ° « »° « NOC  U » ° « NP t L » °° « »¾ « NP  U » ° « PD t L » ° « »° « PD  U » ° « »° « TTM t L » ° ¬« TTM  U ¼» °¿

(6)

3.2 Phase 2: industrial enterprises design complexity The second phase in DFIEC procedure is that of design for system complexity itself. It is mainly concerned with different elements to represent the complexity of it. The elements

288

I.H. Garbie

are product structure and design (PSD), SD and MSs (Figure 4). For each main element, there are several sub-main elements which play an important role in the value of complexity. For example, the PSD has four different types to represent the complexity in the product design, such as number of parts per product (NPP), number of operations per part (NOP), processing or manufacturing time per operation (PT) and product size and weight (PSW). All of them have a significant effect on the complexity of manufacturing/production process. SD is playing a major role in complexity in industrial organisations. It can be observed from Figure 4 that how complexity is the analysis of the SD. The SD divides the complexity analysis into three major issues: production system size (PSS), MHS and plant layout system (PLS). For the PSS, there are three different classifications of production system: small-sized production system (SSPS), mediumsized production system (MSPS) and large-sized production system (LSPS). Each classification type represents or introduces a significant effect on complexity. For example, the complexity in SSPS is less complex than MSPS and than LSPS. This is logical, but it is not enough evidence. We need to look for other issues like resources capacity, capability, utilisation, etc. Figure 4

Elements of industrial enterprises design complexity (see online version for colours)

Design for complexity

289

Also, MHS and PLS play an important role in identifying the complexity in industrial organisations. The MHS consists of MHEs with different types of equipments (e.g. conveyor, trucks, forklifts, crane, etc.), material handling storage system (MHSS) (e.g. manual or automated storage and retrieval) and identification systems (ISs). How it can be thought about all previous components and degree of complexity related to each one. Facility planning or PLS has a significant effect on complexity by different types of configuration. It can be seem that functional layout (FL) or process layout is more complex than product layout (PL) and/or cellular layout (CL). Complexity with respect to MSs is totally different from SD and PSD because it was looking for which strategy can be applied. Complexity in lean manufacturing system (LMS), e.g. is affected by increased product variety if it is compared with PL in mass production system. In general, complexity in LMS is low when compared with agile manufacturing system (AMS), flexible manufacturing system (FMS) and reconfigurable manufacturing system (RMS). Regarding complexity in AMS, FMS and RMS, they can deal easily with any changes in product design (modifications), unpredictable demand, etc. but infrastructure of these systems itself is more sophisticated and complicated to lead these systems to be more and more complex. This means that if an organisation had the higher technologies, there will be the higher complexity. The mathematical expression of IEDC can be modelled as Equations (7) and (8) in different facets.

IEDC

f (PSD,SD, MS)

(7)

IEDC

f (NPP, NOP, PT, PSW, MHS, PSS, PLS, LMS, FMS, AMS, RMS)

(8)

Equation (8) can be rewritten as a very board general representing the lowest level of information in IEDC as Equation (9). IEDC

f (NNP, NOP, PT, PSW, MHE, MHSS, IS, SSPS,

½ ¾ MSPS, LSPS, FL, CL, PL, LMS, FMS, AMS, RMS)¿

(9)

Equation (7) can be rewritten as Equation (10). Each term represents sub-IEDC measure of total IEDC. Adding these terms with relative weights will be considered. These weights can be used as a reason existing to differentiate between sub-complexity measures.

IEDC

wPSDC [PSDC]  wMSDC [SDC]  wMSC [MSC]

(10)

where IEDC is the industrial enterprises design complexity; PSDC is the product structure and design complexity; SDC is the system design complexity and MSC is the manufacturing strategies complexity. The relative weights wPSDC, wSDC and wMSC are for PSD, SD and MSs, respectively, which is three times more important than MSC.

AIEDC

ª wPSDC « « wPSDC « wSDC «w « PSDC « wMSC «¬ wPSDC

wPSDC wPSDC º » wSDC wMSC » wSDC wSDC » wSDC wMSC » » wMSC wMSC » wSDC wMSC »¼

I.H. Garbie

290

For each sub-complexity level, it should follow the same procedure as mentioned in determining the main issues; collecting data and preparing file structure using multiple regression analysis and optimising the complexity level (see Figure 2). Tables 2–4 show the data file structure for PSDC, SDC and MSC, respectively. In Table 2, the PSDC regression model is based on NPP, NOP, PT, product size (PS) and product weight (PW). Then, the PSDC will be written as Equation (11) and the optimal level of PSDC will be determined based on the constraints as shown in Equations (12) and (13). Table 2

Data file structure for product structure and design

Enterprise no.

PSDC

NPP

NOP

PT

PS

PW

1

–

–

–

–

–

–

2

–

–

–

–

–

–

3

–

–

–

–

–

–

4

–

–

–

–

–

–

#

–

–

–

–

–

–

n

–

–

–

–

–

–

Table 3

Data file structure for SD

Enterprise no.

SDC

SSPS

MSPS

LSPS

FL

CL

1

–

–

–

–

–

–

2

–

–

–

–

–

–

3

–

–

–

–

–

–

4

–

–

–

–

–

–

#

–

–

–

–

–

–

n

–

–

–

–

–

–

Table 4

Data file structure for manufacturing strategies

Enterprise no.

MSC

LMS

AMS

FMS

RMS

1

–

–

–

–

–

2

–

–

–

–

–

3

–

–

–

–

–

4

–

–

–

–

–

#

–

–

–

–

–

n

–

–

–

–

–

PSDC

\

B0 

K\

¦B

j

\

X j\

(11)

i 1

where B0 \ , B j \ are the regression coefficients of the PSDC model; K \ is the number of factors (independent variables) in the regression model.

Design for complexity K\

¦B

Minimise PSDC

j

291 \

X j\

(12)

i 1

s.t.

ª NPP t L º ½ « NPP  U » ° « »° « NOP t L » ° « »° « NOP  U » ° « PT t L » °° « »¾ « PT  U » ° « PS t L » ° « »° « PS  U » ° « »° « PW t L » ° ¬« PW  U ¼» °¿

(13)

With respect to SDC, there are a lot of variables (parameters) playing a major role in complexity in this level. Basically, there are three major terms: production systems, PLS and MHS. Production systems can be representing in size PSS of production systems (SSPS, MSPS and LSPS). PLSs design PLS type is classified as FL (process layout), cellular systems (CL) type and product (flow or mass) systems (PL) type. Regarding the FL, the variety of system will affect on the complexity (low, medium and high), but the CL complexity will be determined based on the number of manufacturing cells. For the PL, the complexity can be estimated according to the variety model (single, batch and mixed models). Also, the MHSs represent in MHSS, MHE and IS. In the MHSS, there are two different storage systems: manual and automated. In the MHE, there are three different types of equipments: trucks, vehicles and cranes and hoists. In the ISs, bar code and radio frequency IS are found. If more technologies are there, the more complexity will be found. In Table 3, the SDC regression model will be based on the previous issues that are included in previous paragraph. Then, the SDC will be written as Equation (14). In addition, the optimal level of SDC will be determined based on the constraints of parameters as shown in Equations (15) and (16), respectively. SDC

\\

B0 

K \\

¦B

j

\\

X j \\

(14)

i 1

where B0 \\ , B j \\ are the regression coefficients of the SDC model and K \\ is the number of factors (independent variables) in the regression model. K \\

minimise SDC

¦B i 1

j

\\

X j \\

(15)

292

I.H. Garbie

s.t. SSPS t L ½ SSPS  U °° MSPS t L ° ° MSPS  U ° LSPS t L ° ° LSPS  U ° ° FL t L ° FL  U ° ° CL t L ° ¾ CL  U ° ° PL t L ° PL  U ° MHSS t L ° ° MHSS  U ° ° MHE t L ° MHE  U ° ° IS t L ° °¿ IS  U

(16)

Regarding MSC, there are four suggested strategies or philosophies in industrial enterprises which are necessary. Implementing these strategies will increase or decrease the complexity level of an industrial enterprise. For example, using lean manufacturing philosophy will reduce the level of complexity but using the agility philosophy may increase this level. FMSs are totally different because it is based on more high technologies and automation in machines (e.g. computerized numerical control), robots, etc. This may lead to increase in the complexity level. Regarding reconfigurable philosophy, the situation is different because the degree of reconfigurable will depend on the complexity level of existing systems. In Table 4, the MSC regression model will be based on the previous four strategies based on the levels of measures of manufacturing leanness, flexibility, agility and reconfigurable. Then, the MSC will be written as Equation (17). In addition, the optimal level of MSC is determined based on the constraints as shown in Equations (18) and (19), respectively. MSC

B0

\\\

K \\\



¦B

j

\\\

X j \\\

(17)

i 1

where B0 \\ \ , B j \\ \ are the regression coefficients of the MSC model and K \\\ is the number of factors (independent variables) in the regression model. K \\\

Minimise MSC

¦B i 1

j

\\\

X j \\\

(18)

Design for complexity

293

s.t. LMS t L ½ LMS  U °° AMS t L ° ° AMS  U ° ¾ FMS t L ° FMS  U ° ° RMS t L ° RMS  U °¿

(19)

3.3 Phase 3: industrial enterprises operating complexity Once the industrial enterprise system vision and design complexities become available to the perspective of industrial systems analysts and designers, the design for complexity related to the system operating becomes urgent to be analysed and evaluated. This phase may involve further activities in data collection and processing. Design for IEOC or IEDC is different than previous ones (system vision and SD). In this analysis, it can be noticed that there are three major items of complexity: resource status of operating complexity (RSOC), work in progress complexity (WIPC) and business operations complexity (BOC) (Figure 5). Resources include both equipment (e.g. machining equipment, forming equipment, MHE, etc.) and human. In this analysis, it will be concentrated on the resource reliability (RR), resource capability or flexibility (RC), resource utilisation (RU), resource scheduling/rescheduling (RS/R) and human scheduling/rescheduling (HS/R). For example, maintenance level plays a vital role in RR. This means that the lower the maintenance level is the lower in machine capacity (reliability). In addition, WIPC representing in buffer between workstations or departments is considered as one of the measuring degree of complexity inside the production plant (factory). There are several important issues that can be used to evaluate the BOC (see Figure 6). The issues are organisation plans (OP), OW, structure of management levels (SML), staffing, developing and motivation (FDM), DM, communication between and within management levels (CML), managing conflict, change, culture and stress (MCS) and finally leadership roles in management (LRs). Then, design for IEOC can be modelled to measure or evaluate the complexity level as given in Equations (20) and (21).

IEOC

f (RSOC, WIPC, BOC)

IEOC

f (RR, RC, RU, RS / R, HS / R, WIP(BS), OW, SML, DM, CML, LR, SDM, MCS)

(20)

(21)

294

I.H. Garbie

Figure 5

Elements of industrial enterprises operating complexity (status of operating) (see online version for colours)

Figure 6

Elements of industrial enterprises operating complexity (business operations) (see online version for colours)

Due to many issues (factors) in Equation (21) and it will be more complicated and difficult to implement in regression models. Equation (20) can be rewritten as Equation (22).

IEOC = wRSO [RSOC]  wWIP [WIPC]  wBOC [BOC]

(22)

where IEOC is the industrial enterprises operating complexity, RSOC is the resource status of operating complexity, WIPC is the work in progress complexity and BOC is the business operating complexity. The wRSOC, wWIPC and wBOC are the relative weights for RSOC, WIPC and BOC, respectively. These relative weights can be estimated using AHP according to the following matrix. Each term in Equation (22) represents sub-IEOC. Adding these terms with relative to weights is highly considered. These weights are used to differentiate between these terms according to the industrial enterprise analysts and designers.

Design for complexity ª wRSOC « « wRSOC « wWIPC « « wRSOC «w « BOC «¬ wRSOC

AIEOC

295

wRSOC wRSOC º » wWIPC wBOC » wWIPC wWIPC » » wWIPC wBOC » wBOC wBOC » » wWIPC wBOC »¼

For resource status of operating (RSO), the same procedure should be followed as given in Figure 2. In Table 5, the RSO is based on RR, RC, RU, RS/R and HS/R. The RSOC will be written as Equation (23).

RSOC

B0*

K



¦B

j

*

X j*

(23)

j 1

where B0* , B j * are the regression coefficients of the RSO and K is the number of factors (independent variables) in the regression model. The optimal level of RSOC can be estimated based on constraints of those parameters which are suggested in Equation (23). This relation is shown in Equations (24) and (25). K

Minimise RSOC

¦B

j

*

X j*

(24)

j 1

s.t. RR t L

½ ° ° ° ° RC  U ° RU t L °° ¾ RU  U ° RS/R t L ° ° RS/R  U ° HS/R t L ° ° HS/R  U °¿ RR  U RC t L

Table 5

(25)

Data file structure for resource status of operating

Enterprise no.

RSOC

RR

RC

RU

RS/R

HS/R

#

– – – – –

– – – – –

– – – – –

– – – – –

– – – – –

– – – – –

n

–

–

–

–

–

–

1 2 3 4

296 Table 6

I.H. Garbie Data file structure of business operations complexity

Enterprise no. 1 2 3 4

# n

BOC – – – – – –

OP – – – – – –

OW – – – – – –

SML – – – – – –

S – – – – – –

D – – – – – –

M – – – – – –

DM – – – – – –

CML – – – – – –

MCS – – – – – –

LR – – – – – –

With respect to BOC, it will follow the same procedure and steps as shown in Figure 2 and data structure in Table 6. In Figure 6, it can be noticed that OP is based on the three plans: strategic plan, operational plan and contingency plan. OW in this analysis means the span of management referring to the number of sub-ordinates a manager can effectively manage with no more than six sub-ordinates reporting to a manager. In view of recent evidence, the span of management concept has been revised to state that the number of people who should report directly to any person should be based on the complexity, variety and proximity of the jobs, the quality of the people filling the jobs and the ability of the manager (Rue and Byars, 2009). There are several basic structures that organisations (SML) may use, such as line (vertical), line and staff, horizontal and matrix structure. These organisations used this structure to easily communicating and logistical technology available. The number of structure levels is based on the span of management. So, from three to seven structure levels will be recommended (author judgement). Staffing (F), developing (D) and motivation (M) representing a very big issue in complexity of industrial firms. The staffing (F) mainly based on two major concepts: recruitment with seeking, attracting, selection and testing new employees and transfers, promotions and separations of the existing ones. Developing (D) employees also depend on training employees and management development. With respect to motivation (M), money as a reward, professional career and social life represent the main issues in motivation complexity in the industrial organisations. MCS is not easily work to analyse but at least they need to be determined into main parameters. Managing conflict can be represented into intrapersonal, interpersonal, intergroup, organisational and political. Stress can be divided into mental stress and physical stress. But managing change can be classified into three major categories: technological changes, environmental changes and internal changes. Complexity related to managing change can be taken into consideration as a behaviour style not technical approach. Complexity regarding managing culture is very difficult but it will be necessary especially during the next period for globalisation and international companies. While the culture is the set of important understandings (often unstated) that members of a community share, the companies must identify some rules or characteristics taken together to reduce the level of complexity regarding the different cultures. This can be represented into individual autonomy, structure, support, identification, performance reward, conflict, tolerance and risk tolerance. DM consists of three main rules: with certainty, with risk and with uncertainty. CML can be represented in four ways: e-mail (intranet or internet), oral, writing and fax. Finally with the LR, there are three different types of leadership that can be used, such as autocratic, laissez-faire and democratic. Equation (26) is used to model the BOC as a multiple regression model. Optimising the level of complexity is shown in Equations (27) and (28).

Design for complexity

BOC

B0**

K



¦B

j

**

297

X j **

(26)

j 1

where B0** , B j ** are the regression coefficients of business operations and K

is the number of factors (independent variables) in the regression model. K

Minimise BOC

¦B

j

**

X j **

(27)

j 1

s.t. OP t L

½ ° ° OW t L ° ° OW  U ° SML t L ° ° SML  U ° ° St L ° SU ° ° Dt L ° D U ° ¾ MtL ° M U ° ° DM t L ° DM  U ° ° CML t L ° CML  U ° ° MCS t L ° MCS  U ° ° LR d L ° ° LR ! U ¿ OP  U

(28)

3.4 Phase 4: industrial enterprises evaluation complexity The fourth phase in the DFIEC (IEEC) procedure is the complexity regarding the system evaluation. As industrial enterprise has a great impact on the performance measurements, they still have a problem in measuring these complexities especially regarding selection of the objectives. In this paper, there are five different objectives that can be used to evaluate the complexity (Figure 7). They are product cost (PC), response (R) representing in manufacturing lead time, system productivity (SP) representing in system utilisation, product quality (PQ) representing in number of scrap (defect rate) and appraising and rewarding performance (ARP). They can also be modelled mathematically as given in Equation (29).

IEEC

f (PC, R,SP, PQ, ARP)

(29)

298

I.H. Garbie

Equation (29) can be estimated using a multiple regression model as Equation (28) incorporates PC, response, SP, PQ and ARP using the data file structure (Table 7). Optimisation of the IEOC is suggested through minimising Equation (30) subject to several constraints as Equations (31) and (32), respectively.

B0***

BOC

K

¦B



j

***

X j ***

(30)

j 1

where B0*** , B j*** are the regression coefficients of system evaluation and K

is the number of factors (independent variables) in the regression model. Figure 7

Table 7

Elements of industrial enterprises evaluation complexity (see online version for colours)

Data file structure of system evaluation complexity

Enterprise no. 1 2 3 4

#

SEC – – – – –

PC – – – – –

R – – – – –

SP – – – – –

PQ – – – – –

ARP – – – – –

n

–

–

–

–

–

–

K

Minimise SEC

¦B

j

***

X j ***

(31)

j 1

s.t.

PC t L ½ PC  U °° RtL ° ° R U ° SP t L °° ¾ SP  U ° PQ t L ° ° PQ  U ° ARP t L ° ° ARP  U °¿

(32)

Design for complexity

4

299

A hypothetical example

This numerical example is used to illustrate the proposed procedure of design for complexity regarding industrial enterprises. This can be implemented through three sequential steps. The first step is used to formulate the multiple regression models for the major and sub-major complexity levels. The values were assumed to follow a uniformly distributed random variable with known parameters as shown in Table 8 for 50 generating values for each issue. The multiple regression models for industrial enterprises vision, PSD, SD, MSs, RSO, business operating and system evaluation are presented in Equations (33)–(39), respectively. This was done by using MINITAB statistical software package to generate a random variable and formulating a multiple regression model.

IEVC

0.0356  0.0374(NOS)  0.000013(NOC)

(33)

 0.0065(PD)  0.0842(TTM) PSDC

0.359  0.00436(NPP)  0.0353(NOP)  0.0123(PT)  0.0759(ind1) – 0.0420(ind2)  0.0241(ind3) – 0.0039(ind4)

(34)

where ind 1, ind 2, ind 3 and ind 4 are used for the PSW.

SDC

0.2440  0.00479(SSPS)  0.00795(MSPS)  0.00018(LSPS)  0.0848(FL)  0.0681(CL)  0.0609(PL)  0.092(MHSS)  0.0290(MHE)  0.189(IS)

(35)

MSC

0.441  0.059(LMS)  0.245(FMS) – 0.028(AMS) – 0.024(RMS)

ROSC

0.598  0.379(RR)  0.509(RC)  0.113(RU) – 0.50(RS / R)  0.233(HS / R)` (37)

BOC

Table 8

0.980 – 0.0029(OP) – 0.0076(OW)  0.0015(SML)½ °  0.171( S )  0.234( D)  0.0050( M ) – 0.0266(DM) ¾ ° – 0.0642(CML)  0.134(MCS) – 0.289(LR) ¿

(36)

(38)

Data for complexity issues based on uniform distribution, U [a,b]

Major issues

Sub-majors Values of majors complexity

IEVC

U [0–1]

IEDC

½ ° ¾ ° ¿

Estimated from Equation (10)

PSDC

Values of Sub- Sub-sub majors Values of sub-sub majors

[0–1]

SCM (NOS)

U [3–7] suppliers

DV(NOC)

U [1,000–8,000] units/0

NP

U [3–5] weeks

PLC (PD)

U [2–5] years

TTM

U [5–8] weeks

NPP

U [20–100]

NOP

U [5–15]

PT

U [0.5–5] min/operation

PS

U [1–5]*

PW

U [1–5]*c

300 Table 8

I.H. Garbie Data for complexity issues based on uniform distribution, U [a,b] (continued)

Major Sub-majors issues Values of majors complexity Values of sub- Sub-sub majors Values of sub-sub majors SDC

[0–1]

SSPS MSPS LSPS FL CL PL

IEOC Estimated from Equation (22)

IEEC U [0–1]

MSC

U [0–1]

RSOC

U [0–1]

WIPC BOC

U [0–1] U [0–1]

U [1–20] machines/enterprise U [12–50] machines/enterprise U [51–200] machines/enterprise U [1–10]** U [2–6]**c

MHSS MHE IS LMS FMS AMS RMS RR RC RU RS/R HS/R

U [1–3]**s U [1–2] U [1–3] U [1–2] U [0–1 U [0–1] U [0–1] U [0–1] U [0.5–1] U [0.5–1] U [0.5–1] U [0.7–1] U [0.7–1]

OP OW SML S D M DM CML MCS LR PC R SP PQ ARP

U [1–3] U [1–6] U [3–7] U [1–2] U [1–2] U [1–3] U [1–3] U [1–4] U [2–4] U [1–3] U [5–50] $/unit U [5–20] days U [50–100]% U [0.02–0.1] defect rate U [0.2–0.8]***

Note: U represents the uniform distribution with [a] the lower limit and [b] is the upper limit, *Represents the PS (small, small–medium, medium, medium–large and large), *crepresents the PW (light, light–medium, medium, medium–heavy and heavy), **represents the degree of variety in FL, **c represents the number of manufacturing cells, **srepresents the variety of models (single, batch and mixed), ***represents the ARP as a percentage.

301

Design for complexity IEEC

0.457  0.00210(PC)  0.0088( R ) – 0.172(SP) ½ ¾ – 0.40(PQ) 0.117(ARP) ¿

(39)

The second step is used to determine the complexity level of IEVC, PSDC, SDC, ROSC and IEEC, respectively, by optimising (e.g. minimising) Equations (33)–(39) with constraints assuming the target values of complexity levels no more than 50% (0.50). The constraints are identified based on the range values of each sub-issue that are listed in Table 8. To determine the relative weights between PSDC, SDC and MSC and between RSOC, WIPC and BOC, the following matrices are used to estimate these values and the pairwise comparison between PSD, SD and MSs is illustrated in these matrices.

AIEDC

1 2º ª1 «1 1 2 »» AIEOC « «¬ 0.50 0.50 1»¼

4 2 º ª1 «0.25 1 0.25» « » «¬0.50 4 1 »¼

With respect to IEDC, it can be noticed that a PSDC is estimated to be equivalent to the SDC and twice as important as a MSC. The same estimation is done relate to the SDC with PSDC and MSC. Regarding the IEOC, it can be noticed that RSOC is estimated to be four times more important than the WIPC and two times as important as the BOC. The BOC is also estimated to be four times more important than WIPC. As a result, the relative weights for IEDC between PSDC, SDC and MSC are estimated at 0.40, 0.40 and 0.20, respectively. In addition, the relative weights for IEOC between RSOC, WIPC and BOC are estimated at 0.544, 0.110 and 0.345, respectively. The results of complexity level in each sub-issue and major issue are illustrated in Table 9. The third step is also used to estimate the level of complexity in industrial enterprises based on the four major issues: IEVC, IEDC, IEOC and IEEC. The relative weights between these issues are estimated by using the AHP as the following matrix based on the pairwise comparisons of the four major issues. It can be noticed that the relative weights of IEVC is estimated to be equivalent to the IEDC, twice as important as the IEOC and four times more important than the IEEC. The IEDC is estimated to be three times more important than the IEOC and four times more important than the IEEC. The IEOC is also estimated to be twice as important as the IEEC. As a result, the values of relative weights are estimated at 0.35, 0.40, 0.16 and 0.09 for IEVC, IEDC, IEOC and IEEC, respectively. Equation (2) is used to calculate the IECL incorporating the four major issues as Equation (40). The results of data are illustrated in Table 10.

AIECL

1 2 ª 1 « 1 1 3 « « 0.50 0.33 1 « ¬ 0.25 0.25 0.50

4º 4 »» 2» » 1¼

IECL

0.35[IEVC]  0.40[IEDC]  0.16[IEOC]  0.09[IEEC]

IECL

0.35[0.4582]  0.40[0.3099]  0.16[0.3023]  0.09[0.4570] 0.3740

(40)

302 Table 9

I.H. Garbie Complexity levels in the four phases of industrial enterprises

Major complexity issue IEVC IEDC

Complexity value 0.4582 0.3099

IEOC

0.3023

IEEC

0.4570

Table 10

Sub-major issue PSDC SDC MSC RSOC WIPC BOC

Complexity of sub-major – 0.3828 0.3364 0.1111 0.4049 0.650 0.030 –

Values of major issues regarding the complexity level

Industrial enterprise complexity level (IECL) 0.3740

Major issues IEVC IEDC IEOC IEEC

Value of major issues Percentage of major issues (%) 0.160 42.78 0.124 33.15 0.050 13.36 0.041 10.70

It can be noticed from Equation (40) and Table 10 that the level of complexity in this enterprise equals to 37.40% and this value seems ranked in a medium range. It seems that vision complexity represents more important (0.16/0.3740 = 42.78%) than design complexity (33.15%). The percentage values of operating and evaluating complexity are 13.36% and 10.70%, respectively. These values can be totally different from an industrial enterprise to another one based on the sub-major issues and the relative weights between sub-major and major issues. Regarding the complexity reduction, it can be noticed that if the values of sub-major issues reduced, the complexity level will be reduced too. This can be deeply illustrated through Equations (33)–(39). It also seems that the vision complexity and design complexity represent the adverse factors for reducing complexity. With respect to IEVC, it is not easy to reduce the NOS, and this number should be increased. Also, for IEDC, the NPP, NOP, PT, PSW and PSS are representing adverse factors for reducing complexity because they cannot be changed. It can be observed from Table 11 that changing the values of relative weights regarding major issues with different values from one perspective to another, the IECLs will also be changed. This seems that the IECL will be affected by these changes within a limited range of complexity level (42.05–33.90%). In this hypothetical example, when increasing the value of relative weight of vision complexity and evaluation complexity vs. design complexity and operating complexity, the value of IECL will be increased. Also, when increasing the design and operating complexities with respect to vision and evaluation complexities, the IECL will be decreased, etc. This means that there are different perspectives of designer and/or analyst. Generally, it can be noticed from this hypothetical example that the complexity issues are identified and analysed deeply based on the perspective of industrial enterprises designers and/or analysts. These analyses are used to identify which issue is more important than others. Also, the relative weights between these issues are subjectively and independently considered. The perspective view of the analysts and/or designer based on environment (industry field), situation (economical level) and finally objectives (e.g. minimise complexity level) is also presented.

Design for complexity Table 11

303

Changing relative weights of major issues

Different perspective 1 2 3 4 5 6 7 8

wIEVC

wIEDC

wIEOC

wIEEC

55 50 45 40 35 30 25 20

20 25 30 35 40 45 50 55

5 8 10 12 16 20 22 24

20 17 15 13 9 5 3 1

IECL (%) 42.05 40.80 39.80 38.74 37.40 36.00 34.90 33.90

The estimation of complexity level is a very useful methodology or performance indicator when design for complexity is used to assign a target or a certain value of complexity before following design procedures. As this paper’s central focus is on complexity design, it should not be eliminated but optimised. The medium value of complexity level, i.e. 50% is the most recommended value (author judgement). Maybe this value changes from one perspective to another. Also, as the industrial enterprises need to reduce their complexity, they cannot do it due to another important concept. This concept must be taken into consideration and it is so-called ‘agility’ (Garbie, 2010a). So, complexity level can be reduced or minimised as it can be used but this is not enough for an industrial enterprise to be agile system. Regarding again with which issues are more important than others, it is depending on the perspective of the analyst and/or design. In the beginning of design for complexity, it should be focused on system vision issue and the associated sub-issues. This includes SCM, forecasting demand, introducing a NP, development of the existing ones and TTM. How can be identified which one of these sub-issues is more important than others, this is not a simple task, although they can be considered as some of the globalisation issues? So, managerial implications should be affected by the design for complexity starting from this point (vision complexity) representing into how to select NOS, estimating customer’s requirements, NPs, development time and cost and TTM. With respect to design complexity, it is also very difficult to determine which subissues are more important than others. PSD, SD and MSs are considered overlapping directions and it cannot be distinguished between them. Based on product design, the SD (e.g. cellular system) must be selected with which MSs (e.g. lean strategy) can be used. Industrial operating complexity is a dynamic of the industrial enterprise. How operating complexity can be minimised or optimised, this is another important question? By the way, this complexity is considered as a consequence of the previous major issues: vision and design complexities. Each sub-issue plays an important role in operating/dynamic complexity, especially BOC. It represents a big issue regarding design for complexity. OP, OW, organisation structures, FDM, DM, CML, MCS and LRs represent a big problem in any industrial enterprises complexity regarding organisation management. How to optimise these issues is not a simple task and design for complexity is used to solve this issue. Regarding the last type of complexity (evaluation complexity), it can be noticed that the evaluation complexity is just used to evaluate all previous phases and it represents not only a big value (quantitative) but also an important value (qualitative) regarding the managerial implications. The evaluation complexity suggests PC/price, response (manufacturing lead time), SP, quality and human resources appraisal. These sub-issues are presenting any industrial enterprises performance measurements.

304

5

I.H. Garbie

Conclusions, contribution and recommendations for future work

It can be noticed from this analysis that understanding the concepts and issues of design for complexity issues is not simple. It required emphasise on each of the main issues and the sub-main. Hence, the DFIEC will involve the four major issues: design for vision complexity, design for SDC, design for system operating complexity and design system evaluation complexity. Designing for complexity in industrial enterprises is based on the four main issues and it can be mathematically expressed as a total global function as Equation (41). DFIEC

f (SCM(NOS), DV(NOC), NP, PLC(PD), TTM, NNP, NOP, PT, PSW,½ MHE, MHSS, IS, SSPS, MSPS, LSPS, FL, CL, PL, LMS, FMS, AMS, °° ¾ RMS, RR(ML), RC, RU, RS/R, HS/R, WIP(BS), OP, OW, SML, DM, ° °¿ CML, LR, SDM, MCS, PC, R, SR, PQ, ARP)

(41)

The DFIEC in the global Equation (41) can be thought about the complexity of the industrial enterprises determination. The DFIEC determination should be dynamic and they should evolve with and adapt to the changing internal and external environment. Until now, the DFIEC remains a research topic of immense international interest. As the complexity requirements are classified according to the four major issues of design and analysis of industrial enterprises, a guideline of estimating complexity level of sub-major issues is presented and illustrated. The new measuring methodology is proposed for determining the level of complexity in each major issue based on design requirements and the relative weights of each issue. The design requirements are analysed and optimised in terms of issues count due to the designer/analyst view. This will represent the degree of freedom of industrial enterprises designers to identify which major and submajor issue is more significant than others. This research needs a huge work and time consuming to collect actual data from different industrial firms. These data need to be classified into different categories: products companies (e.g. metal bars), assembly companies (e.g. cars assembly), food industry (e.g. biscuits), petrochemical companies (e.g. shell), etc. For each sector of these companies, a regression model needed to be built and the complexity level in these companies should be estimated too based on a real data. The main contribution in this paper is how to identify and model the components of industrial enterprises complexity in any industrial firms (organisations) at any time considering these components. This means that complexity level can be estimated since beginning of design of industrial enterprise with all levels with a target value of these complexities. Also, the level of complexity of existing industrial firms can be evaluated based on the existing components using the estimated equations (regression model) for major and sub-major issues with identifying the relative weights of these issues. The author intends to extend the following future research to accomplish the benefits from this research: 1

analysis and formulation of the suggested model towards full validation regarding different sectors in industry (discrete product manufacturing, continuous parts manufacturing, food industry, assembly lines, petrochemical industry, etc.)

Design for complexity

305

2

analysis and estimating the agility level of these companies based on major and submajor issues of components and investigating the relationship between complexity and agility levels inside the industrial enterprise

3

further investigation regarding reconfiguring these industrial companies based on the estimated complexity and agility levels and where the reconfiguration process will be done.

Acknowledgements The author would like to thank the anonymous reviewers for valuable and useful feedback comments that contributed to an improvement of this paper.

References Abdi, M.R. (2006) ‘Performance evaluation of reconfigurable manufacturing’, Proceedings of the 36th CIE Conference on Computer and Industrial Engineering, Taipei, Taiwan, pp.2479–2491. Arteta, B.M. and Giachetti, R.E. (2004) ‘A measure of agility as the complexity of the enterprise system’, Robotics and Computer-Integrated Manufacturing, Vol. 20, pp.495–503. Bozarth, C.C., Warsing, D.P., Flynm, B.B. and Flynn, E.J. (2009) ‘The impact of supply chain complexity on manufacturing plant performance’, Journal of Operations Management, Vol. 27, pp.78–93. Cho, S., Alamoudi, R. and Asfour, S. (2009) ‘Interaction-based complexity measures of manufacturing systems using information entropy’, Int. J. Computer Integrated Manufacturing, Vol. 22, No. 10, pp.909–922. Deshmukh, A.V., Talavage, J.J. and Barash, M.M. (1998) ‘Complexity in manufacturing systems, part 1: analysis of static complexity’, IIE Transactions, Vol. 30, pp.645–65. Efstathiou, J., Calinesscu, A. and Blackburn, G. (2002) ‘A web-based experts system to assess the complexity level of manufacturing organizations’, Robotics and Computers Integrated Manufacturing, Vol. 18, pp.305–311. ElMaraghy, H.A., Kuzgunkaga, O. and Urbanic, R.J. (2005) ‘Manufacturing systems configuration complexity’, Annals of the CIRP, Vol. 54, pp.445–448. ElMaraghy, W.H. and Urbanic, R.J. (2003) ‘Modeling of manufacturing systems complexity’, Annals of the CIRP, Vol. 53, No. 1, pp.363–366. ElMaraghy, W.H. and Urbanic, R.J. (2004) ‘Assessment of manufacturing complexity’, Annals of the CIRP, Vol. 53, No. 1, pp.401–406. Frizelle, G. and Woodcock, E. (1995) ’Measuring complexity as an aid to developing operational strategy’, Int. J. Operations and Production Management, Vol. 15, No. 5, pp.26–39. Garbie, I.H. (2009) ‘A vision for reconfiguring industrial organization due to the global recession’, Proceedings of the 39th of International Conference on Computers and Industrial Engineering, France, 6–8 July 2009, pp.658–663. Garbie, I.H. (2010a) ‘A new evaluation of industrial enterprises’, Proceedings of the ICAMMM 2010 of International Conference on Applied Mechanics, Materials and Manufacturing, Muscat, Oman, 13–15 December 2010, pp.253–257. Garbie, I.H. (2010b) ‘A roadmap for reconfiguring industrial enterprises as a consequence of global economic crisis (GEC)’, Journal of Service Science and Management, Vol. 3, No. 4, pp.419–428.

306

I.H. Garbie

Garbie, I.H. and Shikdar, A. (2009) ‘Determination of complexity levels in industrial firms’, Proceedings of the International Conference of Industrial Engineering Research (IERC), Miami, Florida, USA, 30 May–3 June, pp.1423–1428. Garbie, I.H. and Shikdar, A. (2010a) ‘Complexity level in industrial firms: case studies and implementation’, Proceedings of the 2010 International Conference on Industrial Engineering and Operations Management (IEOM 2010), Dhaka, Bangladesh, 9–10 January 2010. Garbie, I.H. and Shikdar, A. (2010b) ‘Analysis of industrial organizations complexity: a conceptual model’, Proceedings of the ICAMMM 2010 of International Conference on Applied Mechanics, Materials and Manufacturing, Muscat, Oman, 13–15 December 2010, pp.247–252. Garbie, I.H. and Shikdar, A. (2011) ‘Analysis and estimation of complexity level in industrial firms’, Int. J. Industrial and Systems Engineering, Vol. 8, No. 3, pp.175–197. Giachetti, R.E., Martinez, L.D., Saenz, O.A. and Chen, C-S. (2003) ‘Analysis of the structural measures of flexibility and agility using a measurement theoretical framework’, Int. J. Production Economics, Vol. 86, pp.47–62. Guide, V.D.R., Srivastava, R. and Kraus, M.E. (1997) ‘Product structure complexity and scheduling of operations in recoverable manufacturing’, Int. J. Production Research, Vol. 35, No. 1, pp.3179–3199. Hu, S.J., Zhu, X. and Koren, W.Y. (2008) ‘Product variety and manufacturing complexity in assembly systems and supply chains’, Annals of the CIRP, Vol. 57, pp.45–48. Huatuco, L.H., Efstathiou, J., Calinescu, A., Sivadasan, S. and Kariuki, S. (2009) ‘Comparing the impact of different rescheduling strategies on the entropic-related complexity of manufacturing systems’, Int. J. Production Research, Vol. 47, No. 1, pp.4305–4325. Isik, F. (2010) ‘An entropy-based approach for measuring complexity in supply chains’, Int. J. Production Research, Vol. 48, No. 12, pp.3681–3696. Jenab, K. and Liu, D. (2010) ’A graph-based model for manufacturing complexity’, Int. J. Production Research, Vol. 48, No. 11, pp.3383–3392. Johnson, M.E. (2008) ‘Relative complexity of enterprise designs’, Proceedings of IIE, 2008 (IERC), Vancouver, Canada, 17–22 May 2008, pp.17–21. Kamrani, A.K. and Adat, A. (2008) ‘Manufacturing complexity analysis: a simulation-based methodology’, in A.K. Karmrani and E.A. Nasr (Eds.), Collaboration Engineering-Theory and Practice. New York, NY, USA: Springer, pp.227–248. Kashyap, N. and Sinha, S. (2011) ‘Estimation of job complexity in an industrial organization’, Int. J. Industrial and Systems Engineering, Vol. 7, No. 1, pp.26–44. Kristianto, Y. and Helo, P. (2011) ‘Designing supply chain by coordinating manufacturing process and product development process’, Int. J. Industrial and Systems Engineering, Vol. 6, No. 3, pp.360–380. Kuzgunkaya, O. and ElMaraghy, H.A. (2006) ‘Assessing the structural complexity of manufacturing systems configurations’, Int. J. Flexible Manufacturing Systems, Vol. 18, pp.145–171. Mazur, L.M and Chen, S-J. (2011) ‘A task-member assignment model for complex engineering’, Int. J. Industrial and Systems Engineering, Vol. 7, No. 1, pp.1–25. Perona, M. and Miragliotta, G. (2004) ‘Complexity management and supply chain performance assessment’, Int. J. Production Economics, Vol. 90, pp.103–115. Rue, L.W. and Byars, L.L. (2007) Management-Skills and Application (12th ed.). New York: McGraw-Hill/Irwin. Sharif, A.M. and Irani, Z. (2006) ‘Applying a fuzzy morphological approach to complexity within management decision making’, Management Decision, Vol. 44, No. 4, pp.930–961. Sivadasan, S., Efstathiou, J., Frizelle, G., Shirazi, R. and Calinescu, A. (2002) ‘An informationtheoretic methodology for measuring the operational complexity of supplier–customer system’, Int. J. Operations and Production Management, Vol. 22, No. 1, pp.80–102.

Design for complexity

307

Tani, G. and Cimatti, B. (2008) ‘Technological complexity: a support to management decisions for product engineering and manufacturing’, 2008 IIIE International Conference on Industrial Engineering and Engineering Management IEEM 2008, Singapore, pp.6–11. Tomiyama, T., Amelio, V.D., Urbanic, J. and ElMaraghy, H.W. (2007) ‘Complexity of multidisciplinary design’, Annals of the CIRP, Vol. 56, No. 1, pp.185–188. Wang, H. and Hu, S.J. (2010) ‘Manufacturing complexity in assembly systems with hybrid configurations and its impact on throughput’, Annals of CIRP-Manufacturing Technology, Vol. 59, pp.53–56. Wu, Y., Frizelle, G. and Efstathiou, J. (2007) ‘A study on the cost of operational complexity in customer–supplier systems’, Int. J. Production Economics, Vol. 106, pp.217–229. Yang, J. (2010) ‘A new complexity proof for the two-stage hybrid flow shop scheduling problem with dedicated machines’, Int. J. Production Research, Vol. 48, No. 5, pp.1531–1538.