Construction management process reengineering performance ...

6 downloads 16 Views 916KB Size Report
This study develops a construction management process reengineering performance measurement. (CMPRPM) model .... Effectiveness is the degree to which a.

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright

Author's personal copy Automation in Construction 18 (2009) 183–193

Contents lists available at ScienceDirect

Automation in Construction j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / a u t c o n

Construction management process reengineering performance measurements Min-Yuan Cheng, Hsing-Chih Tsai ⁎, Yun-Yan Lai a

Department of Construction Engineering, National Taiwan University of Science and Technology, Taiwan

a r t i c l e

i n f o

Article history: Accepted 12 July 2008 Keywords: Construction management process reengineering Performance evaluation Queuing theory Process value BPR

a b s t r a c t This study develops a construction management process reengineering performance measurement (CMPRPM) model based on an application of business process reengineering philosophy. Process operation time and customer satisfaction are used as efficiency and effectiveness evaluation indices. The CMPRPM model applies queuing theory to calculate process operation time in order to strike an optimal balance between process execution demand and manpower service capacity. In order to achieve customer satisfaction, customer demands are identified and a target attainability index is used to calculate process effectiveness. After integrating efficiency and effectiveness evaluation results, indices of process value (PV) and value improvement (VI) are proposed to allow performance prior to and after reengineering to be measured and compared. The proposed CMPRPM model addresses the performance of initial (“As-Is”) and significantly reengineered (“To-Be”) processes to facilitate successful BPR design. Results show that the construction industry stands to benefit significantly in terms of a successful BPR design by adopting the model proposed in this paper. Crown Copyright © 2008 Published by Elsevier B.V. All rights reserved.

1. Introduction As efforts typically target existing processes that are not ideally suited to address evolving operational and market needs, resources are often wasted. While increasing organizational costs, anticipated performance gains are not achieved. Businesses often pursue process reorganization to overcome this problem. “Business process reengineering” (BPR) was first proposed by Hammer [1]. Hammer defined BPR to address basic issue related to the reengineering process in terms of costs, quality, services and speeds. The three primary applications of BPR were defined by Hammer as process reorganization, information technology implementation and organization redesign [2]. In these primary applications, information technology (IT) represents a fundamental element of reorganization. When IT is utilized, its relationship with BPR in construction projects should be developed and potential IT factors must be addressed [3]. Consequently, considerations of process innovation and organizational changes are essential for either BPR or business process implementation (BPI) success. Engaging in BPR efforts can provide high rewards

⁎ Corresponding author. Tel.: +886 2 27376663; fax: +886 2 27301074. E-mail addresses: [email protected] (M.Y. Cheng), [email protected] (H.C. Tsai).

and increase the likelihood of success; although risks of loss remain high if not properly implemented. The potential for BPR in the construction industry should take into consideration core issues of concern to participants, who include clients, designers, suppliers, specialists and contractors, in order to help formulate objectives within an integrated strategy [4,5]. The effective scope of BPR should also be identified [6]. Brown and Marston [7] described how successful BPR lead to successful improvement efforts at the Tennessee Department of Transportation by focusing on the project-development process for new construction. Abdul-Hadi et al. [8] described 29 prioritizing barriers in their investigations of the Saudi Arabian construction industry, which were ranked according to order of difficulty and importance to BPR implementation success. An evaluation method for business improvement should be developed to reduce risks associated with the BPR process and increase the likelihood of BPR success. Predicting quantitatively the effect of an improvement approach facilitates model construction better [9]. The balanced scorecard philosophy also provides an alternative to develop business process evaluation methods [10,11]. This paper discusses the application of BPR in the construction industry. The queuing theory is employed to quantify the process operation time, which is important to the process efficiency of the proposed Construction Management Process Reengineering Performance Measurement Model (CMPRPM). Finally, target achievement matrices are defined to evaluate BPR performance improvement. A construction company was selected as a case study to test and verify the CMPRPM model.

0926-5805/$ – see front matter. Crown Copyright © 2008 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.autcon.2008.07.005

Author's personal copy 184

M.-Y. Cheng et al. / Automation in Construction 18 (2009) 183–193

2. Model knowledge 2.1. Business process reengineering BPR, a strategy-driven organizational initiative designed to improve and redesign business processes, includes four major steps. These are process representation, process transformation, process valuation, and process redesign. 2.1.1. Process representation In process reengineering, one of the most difficult and important tasks is to identify and describe a company's current process. Accurately describing the categorized operational process is an essential first step in the reengineering program. Process representation develops a systematic definition for processes to assist companies to clarify and define current management processes. Two major sub-steps in this stage are clarification and process selection for reengineering. 2.1.2. Process transformation The transformation process mainly represents the application of the conducted operational analysis and process modeling. The primary purpose of operational analysis is to define a processes operational category and hierarchical structure. Process modeling is used to provide a comprehensive explanation of the relationship between operations. Many different methods and techniques, including IDEF, eEPC, Petri Nets, System Dynamics, Knowledge-based Techniques and Discrete-Event Simulation, can be used for modeling business processes in order to provide an understanding of potential improvement scenarios. The eEPC (extended event-process chain) technique employed in this study is composed of events and processes. Process modeling tools should be able to develop As-Is and To-Be models of business processes, which represent, respectively, existing and alternative processes. BPR seeks first to define and understand the current As-Is business process and then, after modeling and analysis, formulates the future To-Be business process. 2.1.3. Process evaluation As reengineering activities focus on outdated and inefficient processes in order to make changes that achieve the greatest impact, prior to execution, the present process must be reviewed to locate process barriers in order to ensure their targeting in process redesign. Process value (PV), used to evaluate process performance, can be viewed from either of the following two perspectives: (1) efficiency per unit of cost or (2) efficiency per unit of time. Time is an important factor that impacts upon cost as, the longer a process takes to accomplish, the higher the financial price demanded. Evaluating performance to identify problematic areas related to these perspectives provides essential references that can be used to develop and implement a successful process reengineering strategy.

present the main objects of Queuing theory interest [12]. The theory was first developed in “The Theory of Probability and Telephone Conversation” by a Danish engineer, A. K. Erlang, in 1909, when it was employed to study telephone system traffic loads. After World War II, queuing theory was widely applied to various real-world challenges (e.g., computer networks, telephony systems, the Internet, industrial production lines) in various fields (e.g., hospitals, banks, airports, gasoline service stations). Inasmuch as abilities of the queuing theory, ALOHANET and ARPANET, which constitute the basis of Internet, were designed and analyzed with queuing theory. Queuing theory mainly focuses on the steady state of an inspected system with distribution functions of processes. After a sufficient time, the queuing system will stabilize and become independent of its initial conditions. Mathematical probabilistic models for the analyzed queuing system can be formed with assumptions of process distribution functions [12]. The common link between the various applications and fields that apply queuing theory is that they all deal with customers who join a queue to wait for some desired service (see Fig. 1). Identifying the point at which offered services and waiting customers achieve some specific balance and optimizing the benefits of such represent the core objectives of queuing theory. The four components of queuing theory include: arrival pattern, service pattern, queue discipline, and system capacity. Arrival pattern typically describes the arrival time of two contiguous customers in the queue and service pattern describes one or more servers. Service time is defined as the length of time required for a customer to receive a particular service. Both arrival and service times are characterized by distribution functions. Queue length represents the length of the queue in which customers wait. Queue discipline represents the order of who should be served first and may be defined by several discipline regimens, including first come/first served (FCFS), last come/first served (LCFS), and service in random order (SIRO). System capacity represents queuing area/service facility capacity limitations. For example, while queuing at a gasoline service station, both the queuing area and service capacity are limited. However, system capacity limitations are usually ignored in queuing theory application. The symbol for queuing theory is usually represented by Kendall's Notation, which is composed of the five letters A/B/X/Y/Z, in which A means arrival time, B represents service time, X is the number of servers, Y is the allowable capacity for customers, and Z represents the adopted queuing discipline. Distributions for A and B are usually defined as Markovian distributions (M), Deterministic (D), Erlang-k distributions (Ek), and General Independent (in order to discriminate between A and B, GI is used for A, and G for B). When Y is set at “infinite” and Z at FCFS, the queuing model can be denoted by A/B/X. Buzacott [13] proposed using queuing theory as a facility for BPR performance evaluation. Queuing theory can explore the practical impact on system structures that derive from the radical changes effected by BPR.

2.1.4. Process redesign The process redesign effort must also include a review of current business operations. Analysis results derived from the process evaluation model can be used to identify major process defects. Satisfaction of customer demands provided by the process and the requirement of adding new activities can then be identified and determined based on operation target attainability. Because, in general, higher effectiveness requires higher costs, decreasing costs may potentially decrease process effectiveness. Therefore, when evaluating process value, managers must trade off between effectiveness and efficiency to pursuance suitable strategies for companies. 2.2. Queuing theory Queuing theory is a theme explored in Operations Research. As its name implies, queues, or work loads, awaiting server processing re-

Fig. 1. One's queue for desired services.

Author's personal copy M.-Y. Cheng et al. / Automation in Construction 18 (2009) 183–193

185

3. Construction management process reengineering performance measurement (CMPRPM) model This paper employs BPR in the construction industry and evaluates the performance of such in implementation. The CMPRPM model, developed to quantify BPR performance, will be introduced in the following sections.

Fig. 3. Project process queue.

3.1. Factors of CMPRPM Efficiency and effectiveness are the two major factors associated with CMPRPM. Efficiency is evaluated by the amount of resources input in relation to the output result. Effectiveness is the degree to which a target has been achieved with resources applied. After interviewing managers, reductions in process operation times that make projects more efficient were identified as the principal expectation of this group with regard to reengineering. Process reengineering is highly customer-oriented to satisfy internal and external customer needs. Internal customers comprise employees, whose needs include reduced work loads, improved work efficiency and higher engineering quality. External customers include employers (e.g., owners), whose needs include efficiency, productivity and quality. In this study, efficiency is evaluated by the process operation time of As-Is and To-Be processes, and effectiveness is evaluated by customer satisfaction after reengineering. Therefore, process operation time and customer satisfaction represent two major factors of concern in this paper which will be discussed in the following sections. 3.2. The process operation time factor For a business, the operation process is something like a queuing sequence, in which work and documents queue for execution. A reasonable operation process and suitable resource allocation strategy greatly impact upon efficiency. This study employed queuing theory to evaluate the process operation times of As-Is and To-Be processes for reengineering performance evaluation (see Fig. 2). In data analysis, estimates of expected values and variances were calculated for intervals

associated with, respectively, biding projects, successful bids, subcontracts, and process service rates. 3.2.1. The selected queuing model: the GI/G/1 model To evaluate the process operation time a queuing GI/G/1 model was employed [14]. GI/G/1 was selected for reasons including: 1) the arrival rate distribution of each biding projects, successful bids and subcontracts satisfied independent and identically distributed with a general distribution, 2) each service rate was independent and identically distributed with a general distribution, and 3) assuming one server in the queuing system. Therefore, authors chose the GI/G/1 model for this study. Besides, the allowable capacity of customers, Y, was set as infinite and FCFS was adopted for Z. 3.2.2. Process operation time evaluation The process operation time evaluation model includes two parts. The first task evaluates the average process operation time required for each project, which, when summed, can be used to evaluate total process operation time. During project execution, each project departure impacts greatly upon the arrival of the next contiguous project in the queue. The coefficient of variation (COV) of a project's departure is the other value that must be determined. This parameter was termed the departure rate, where the ith project's departure rate equals the (i + 1)th project's arrival rate (see Fig. 3). The ith project is executed when the (i − 1)th project has departed (Di − 1) and ith project has arrived (Ai). While the ith project execution time is Si, the time of the ith project's departure (Di) is: Di ¼ maxðDi−1 ; Ai Þ þ Si

ð1Þ

The server is idle during periods when the (i − 1)th project has departed but the ith project has not yet arrived. Conversely, when the next project arrives and the current project is still executing, a queue will be formed (see Fig. 4). The total time of ith project Ti can be expressed as: Ti ¼ ½Di−1 −Ai þ þ Si ¼ ½Di−1 −Ai−1 −ðAi −Ai−1 Þþ þ Si ¼ ½Ti−1 −τi þ þ Si

ð2Þ

In the equation above, τi denotes the arrival interval of the ith and (i − 1)th projects and []+ means that only plus quantities will be accepted

Fig. 2. Procedure of process operation time evaluation.

Fig. 4. Servers idle verses projects queue.

Author's personal copy 186

M.-Y. Cheng et al. / Automation in Construction 18 (2009) 183–193

and negative values will be treated as zero. τi is equal to (Ai −Ai − 1). When Di − 1 is less than Ai, or Ti − 1 is less than τi, the server is idle. Server idle time Ii can be represented as:

where Ca is the coefficient of variance (COV) of the project arrival interval τ, ρ is the time usage rate, Cs is the COV of the project execution time S, and λ is the project arrival ratio.

Ii ¼ ½τ i −Ti−1 þ

Ca2 ¼

ð3Þ

When project queuing takes place, servers will handle the next project immediately as soon as the service needs of the preceding project have been fulfilled. Therefore, the value of Ii is zero, i.e. the server is not idle. Eqs. (2) and (3) yield the following:

ρ¼

Ti −Ii ¼ Ti−1 −τi þ Si

Cs2 ¼

ð4Þ

Assuming that i approximates an infinite value, the system will be under steady state, i.e. E[Ti] =E[T] (in which E[] denotes expected value). Substituting this assumption into Eq. (4): E½T −E½I ¼ E½T −E½τ þ E½S

ð5Þ

E½I  ¼ E½τ−E½S

ð6Þ

An alternative to Eq. (4) can be expressed as: Ti −Si −Ii ¼ Ti−1 −τ i

ð7Þ

And squaring both sides of the above equation: 2 þ τ2i −2Ti−1 τi Ti2 þ S2i þ Ii2 −2Ti Si −2ðTi −Si ÞIi ¼ Ti−1

ð8Þ

To take the E[] of Eq. (8). Because Ti − Si and Si are independent, E[(Ti − Si)Si] = E[TiSi − S2i ] = (E[Ti] − E[Si])E[Si]. Therefore, E[TiSi] =E[Ti]E[Si]+ var[Si]. While Ti − Si N 0, Ii = 0. Because Ti − Si is a non-negative quantity, (Ti − Si)Ii is equal to zero.        2    þ E τ2i −2E½Ti−1 E½τ i  E Ti2 þ E S2 þ E Ii2 −2E½Ti E½Si −2var½S ¼ E Ti−1 ð9Þ

λ¼

var½τ 

ð15Þ

E ½τ 2

E½S E ½τ 

ð16Þ

var½S

ð17Þ

E½S2

1 E ½τ 

ð18Þ

A parameter Cd is also defined to denote the COV of project process departure. A parameter Δi is used to describe the duration between the (i − 1)th project's departure and the ith project's departure, which can be expressed as: Δi ¼ Di −Di−1 ¼ ½Ai −Di−1 þ þ Si ¼ ½τi −Ti−1 þ þ Si ¼ Ii þ Si

ð19Þ

E ½Δ  ¼ E ½τ 

ð20Þ

var½Δ ¼ var½I  þ var½S

ð21Þ

Cd2 ¼

var½Δ E½Δ2

¼

   E I2  − 1−ρ2 þ ρ2 Cs2 2 E ½τ 

ð22Þ

From Eq. (12):       E τ2 E½I 2 ¼ E½τ2 Ca2 þ 1 ð1−ρÞ2 E I2 z E ½τ 2

ð23Þ

Cd2 zð1−ρÞ2 Ca2 þ ρ2 Cs2

ð24Þ

The steady state of Eq. (9) is:             E τ 2 þ 2var½S−E S2 −E I 2 E τ2 −2E½τ E½S þ E S2 −E I2 ¼ þ E½S E ½T  ¼ 2ðE½τ−E½SÞ 2ðE½τ −E½SÞ

ð10Þ In Eq. (10), E[I2] represents the key to calculating E[T]. However, the exact solution of E[I2] remains unsolved. The approximate solution is an alternative that can be chosen. In the equation below, R is a nonnegative random variable and X is greater or equal than zero: E

  h 2 i E R2  2 ðR−X Þþ ¼ E ðR−X Þþ E ½R2

ð11Þ

According to Eq. (11), E[I2] can be approximated as: "   #   h h   2 ii E τ 2i  E τ2i þ 2 ð Þ zE E τ −T E½Ii 2 ð12Þ E Ii2 ¼ E E ðτ i −Ti−1 Þþ z i i−1 2 E ½τ i  E ½τ i 2 The steady state of above equation can be expressed as:     E I 2 z 1 þ Ca2 E½I2

ð13Þ

Substituting Eq. (13) into Eq. (10), E[T] can be expressed as: E½T V

ρð2−ρÞCa2 þ ρ2 Cs2 þ E ½S  2λð1−ρÞ

ð14Þ

Through the equations above, the process operation time of a project can be calculated in forms of expected values and variances. 3.3. The customer satisfaction factor 3.3.1. Evaluation process for customer satisfaction The satisfaction of customer needs is the primary objective of process reengineering. To this end, the functional target of the process should be customer-oriented. In addition, the level to which the existing process already attains this objective should be assessed to identify existing problems and improvement goals. Using the quality function deployment method, this study transforms company policy and customer concerns into process targets. The target attainability matrix is proposed to examine the attainability of customer satisfaction for measuring process effectiveness. Main evaluation process steps are described as follows (see Fig. 5): 3.3.1.1. Definition of operational strategy and company policy. A company's operations can be viewed as a serial composition of processes, with each process required to achieve certain targets. In this framework, it is essential to consider company policy in tandem with the targets of each process in order to accomplish overall company policy objectives. Before analyzing the process, a company's operation policy must first be defined. Inclusion of policy demands when setting process targets is also essential to the realization of a company's operational strategy and customer needs.

Author's personal copy M.-Y. Cheng et al. / Automation in Construction 18 (2009) 183–193

187

Table 1 Relative importance weight matrix Target components Component 1 Component 2 Component 3 … Component n pj Customer Demand 1 demands Demand 2 Demand 3 ⋮ Demand m Wi

Fig. 5. Customer satisfaction evaluation procedure.

3.3.1.2. Identification of internal and external customers. Direct customers are mainly parties that participate in the process, with the final customers (i.e., acceptors of the final products of a process) generally referred to as consumers. Applied to construction companies, owners are typically the final customer. Internal customers are those who actually participate in a process, while external customers are the consumers who accept the final products of a process. As the targets of customer satisfaction, customers must first be identified in order to determine their needs. 3.3.1.3. Surveying customer requirements. Customers' requirements must be considered when setting process targets. Based on the internal and external customers identified above, customer requirements may be established by interviewing managers and experts. 3.3.1.4. Determination of process targets. To meet various customer demands, a process must be able to allocate appropriate resources to where they are wanted. The process targets deployment (PTD) method developed in this study is used to transform customer demands into process targets. Process target components may be determined following PTD analysis. 3.3.1.5. Analysis of the relative importance of process targets. The relative importance weight evaluation matrix is used to identify the relative importance of process targets. Index j in the matrix relates to customer demand and index i relates to target components. The corresponding number rij represents the relationship between customer demands and target components (rij: 1, 3, 5). The value of rij correlates positively with the degree to which target component i is able to meet customer demand j. Questionnaire and interview results are quantitated into the emphasis degree pi (pi: 1 ∼ 5). The relative importance Wi is calculated by Eq. (24), related to m customer demands and n target components. A higher Wi denotes that the ith component has a more significant effect on customer satisfaction. m

∑ rij  pj Wi ¼

n

j¼1 m

∑ ∑ rij  pj i¼1 j¼1

ð25Þ

1 3 3 ⋮ 3 0.12

⋮ 0.13

… … … rij … …

1 3 5 ⋮ 1 0.09

3 ⋮ 5 0.17

3 5 4 ⋮ 3

3.3.1.6. Target achievement analysis process. A quantitative method was used to calculate the achievement of each process target and a process target achievement matrix was proposed to evaluate overall target achievement. Firstly, the relative importance Wi of target components in Table 1 was used in this step. Based on each process target, the attainability of the kth operation for the ith process target Aik (Aik: 0 ∼ 5) was evaluated by the senior managers, with the higher the Aik, the more specific contribution that target component i makes on operation k. The ith operation target attainability OAi achieved by the process activities could be calculated following Aik evaluation. The target attainability TA and the degree of contribution Ck endowed by kth operation were also identified. A higher OAi denotes higher operation target achievement; a higher Ck means higher contribution of the operation; and higher degree of target attainability with higher value of TA (see Table 2). The equations for calculating OAi, TA, and Ck were demonstrated as follows, as they relate to n target components and g process operations. g

OAi ¼ ∑ Wi  Aik

ð26Þ

k¼1 n

ð27Þ

TA ¼ ∑ OAi i¼1

n

Ck ¼ ∑ Wi  Aik

ð28Þ

i¼1

3.3.1.7. Process comparisons. The As-Is and To-Be processes were all evaluated using this procedure to identify improvements in customer satisfaction. With the process operation time evaluation, the proposed CMPRPM model was determined a practical method by which to evaluate efficiency and effectiveness. 4. A CMPRPM case study A real-world BPR case executed by a construction company (Company “A”) is presented in this paper to confirm the feasibility of the CMPRPM model. The reengineering of biding/contracting, budgets,

Table 2 Total process attainability matrix Target components Component Component Component … 1 2 3 Wi Operation 1 Operation 2 Operation 3 ⋮ Operation g OAi

0.12 2 4 3 ⋮ 0 1.25

0.13 2 0 1 ⋮ 1 0.66

0.09 4 3 1 ⋮ 2 0.96

… … … … Aik … …

Component Contribution n of operation Ck 0.17 1 3 4 ⋮ 1 0.13

0.96 0.85 0.63 ⋮ 0.42 TA

Author's personal copy 188

M.-Y. Cheng et al. / Automation in Construction 18 (2009) 183–193

Table 3 Expected values and variances of arrival intervals

Table 5 Service rate for the As-Is budget process

Arrival interval

Biding projects

Successful bids

Subcontracts

Procedure items

Expected value (h)

Variance(h2)

Number of samples Expected value (h) Variance (h2)

21 160.13 22,440.55

8 457.50 77,320.83

127 20.11 702.73

Input items and quantities of contracts Check quantities Unit price analysis Check unit price Calculate preliminary budget Check preliminary budget Survey budget (Manager of construction department) Review budget (Vice president) Determine budget (CEO) Check contract quantities Check contract unit price Check contract total price Survey contract budget (Manager of construction department) Determine contract budget (Vice president)

1.63 220.76 100.75 124.13 0.01 0.01 45.75

0.34 1449.08 353.07 171.84 0.00 0.00 119.36

25.25 24.38 18.75 103.75 0.01 39.38

15.36 29.98 56.21 341.07 0.00 54.27

23.00

25.43

and subcontracting processes of company “A” were completed preliminarily, so that the To-Be process models were created as alternatives for advanced evaluation and even implementation. The efficiency and the effectiveness were estimated with the queuing theory (GI/G/1 model) and the target achievement analysis respectively proposed in the CMPRPM. Once the performance of To-Be processes satisfied the decision makers of reengineering project team, the qualified processes could be considered for advanced implementation. Accordingly, the case with three main steps including (1) preliminary process reengineering, (2) process efficiency evaluation and (3) process effectiveness evaluation, were described. 4.1. Preliminary process reengineering The construction management process reengineering model (CMPR) addressed by Cheng et al. [4,5] were adopted by company “A” for creating the initial To-Be process models of the target reengineered processes. Based on previous process reengineering model, three main steps, namely, (1) process representation, (2) As-Is process analysis, and (3) new process design were completed subsequently. 4.1.1. Process representation The As-Is models of target reengineered processes were created in this step. The process data collection and process modeling were the two main tasks. For the study case, the As-Is process models were created with eEPC diagram of ARIS (Architecture of Integrated Information System) according to the ISO documentations and organization structure. Moreover, for the process efficiency evaluation with queuing theory, the detailed activity information, such as duration, service rate, etc., need to be surveyed additionally. Accordingly, for company “A”, the historical data, reviewed for a period spanning May 1999 to February 2001, provided statistical information on 21 biding projects, 8 successful bids, and 127 subcontracts (which provided the expected values and variance of arrival interval for this case) (see Table 3). Service rates for biding/contracting, budgets, and subcontracting processes were included. The activities information was summarized as shown in Tables 4–6; the original organization structure was illustrated as Fig. 6.

4.1.2. As-Is process analysis Based on Cheng's CMPR model [4], the process rationality and the validity of data processing were evaluated for the As-Is process models. Taking the budget process of company “A” as an example, the problems derived from the analysis results were illustrated as followings. 1. Most business development processes are executed by manpower, the historical data is hard to be used for budget compilation directly. To re-establish usable data will increase process time of budget compilation. 2. Budget compilation and implementation are handled by different departments of cost control and procurement/subcontracting. The information transmission between two departments increases lots of time and interfaces. 3. Documents of budget compilation and implementation are examined through many departments, so the processes are delayed greatly. 4.1.3. New process design Aiming at the determined problems, the To-Be process can be generated consequently. Fig. 7 illustrates the To-Be process model of budget process of the case study. Based on the newly design processes of company “A”, the operation information of activities of newly design processes need to estimate again [5] for process efficiency evaluation. The To-Be bidding/contracting process procedure is listed in Table 7; budget process in Table 8; and subcontracting process in Table 9 with the expected values and variances. 4.2. Efficiency evaluation The efficiency of construction management process reengineering is evaluated through a comparison of process operation times represented

Table 4 Service rate for the As-Is biding/contracting process Procedure items Reports for biding evaluation Evaluation report auditing (Manager of planning dept.) Preliminary bid examination (Vice president) Establishing cost items Quantities survey Check unit price Total cost estimate Fill quotation table Survey quotation table (Manager of construction dept.) Review quotation table (Vice president) Determine biding price (CEO) Adjust item unit prices Make bid proposal

Expected value (h)

Variance(h2)

8.38 8.62

13.15 8.15

5.48 1.10 106.48 87.38 17.43 2.81 6.33

4.86 0.22 1987.04 266.45 38.76 1.46 8.53

12.29 12.10 2.19 1.67

7.31 4.29 1.16 0.43

Table 6 Service rate for the As-Is subcontracting process Procedure items

Expected value(h) Variance(h2)

Plan construction items for subcontracting 3.44 Select subcontractor candidates 2.56 Identify materials requested for quotation 4.68 Arrange quotations 1.71 Evaluate quotations through price competition or 14.85 negotiation Audit subcontractors 5.15 Re-certification audit for subcontractors 2.50 Determine selected subcontractors and subcontract 1.74 costs

5.83 0.66 1.57 0.68 16.42 8.13 0.37 0.31

Author's personal copy M.-Y. Cheng et al. / Automation in Construction 18 (2009) 183–193

189

Fig. 6. Company A organization prior to BPR.

by As-Is and To-Be processes. The total process operation times of biding/contracting, budget and subcontracting processes are calculated in this section, with As-Is and To-Be execution times used to evaluate BPR implementation. In the following, the first two procedures of As-Is biding/contract process are calculated as an example for process operation time calculation using queuing theory (GI/G/1 model).

The first procedure of the As-Is biding/contracting process is “Reports for biding evaluation” (the first procedure in Table 4). In Table 3, the arrival interval between two biding projects is expected to be 160.13 h, with a variance of 22,440.55 h2. From Table 4, the service rate associated with the first procedure is 8.38 with a variance of 13.15 h2. The COV of the first procedure arrival interval Ca

Fig. 7. Company A organization after BPR.

Author's personal copy 190

M.-Y. Cheng et al. / Automation in Construction 18 (2009) 183–193

Table 7 Service rate for the To-Be biding/contracting process Expected value (h) Variance(h2)

Procedure items

Access biding documents 9.00 Call Meeting for biding or not and discuss division 1.38 of labor Access cost estimate system 1.13 Calculate quantities and input into system 128.38 Unit price analysis 9.19 Check unit price 124.13 Estimate total cost 0.01 Call meeting for total cost 1.38 Adjust cost in system 0.20 Make bid proposal 1.88

18.86 0.13 0.34 3195.13 2.57 171.84 0.00 0.13 0.00 0.41

The second procedure taken from Table 4, “Auditing the evaluation reports (Manager of planning dept.)” has a process operation time E[T] and Cd calculated as: ρð2−ρÞCa2 þ ρ2 Cs2 þ E½S 2λð1−ρÞ   8:62 8:62 8:62 2 8:15 2− 0:786 þ 160:13 160:13 160:13 8:622  þ 8:62 V 1 8:62 1− 2 160:13 160:13 V15:62

E½T V

ð31Þ

Cd2 zð1−ρÞ2 Ca2 þ ρ2 Cs2    8:62 2 8:62 2 8:15 z 1− 0:7862 þ 160:13 160:13 8:622 z0:553

Table 8 Service rate of the To-Be budget process Procedure items

Expected value (h)

Variance(h2)

Input items and quantities of contracts Check quantities Check unit price Calculate preliminary budget Output budget, group by items Check execution and contract budgets Determine execution and contract budgets

0.88 202.75 2.38 0.01 0.01 18.75 13.75

0.13 1449.07 0.13 0.00 0.00 12.50 12.50

Table 9 Service rate for the To-Be subcontracting process Procedure items

Expected value (h)

Output subcontract lists from system Identify materials of requested for quotation / e-mail to selected subcontractors Study and arrange quotations (MS Excel format) Conduct price competition/negotiation and input results into system Determine selected subcontractors and subcontract costs

Variance(h2)

0.10 5.15

0.01 0.97

0.08 14.69

0.00 12.23

0.88

0.17

The total process operation time may be calculated after evaluating all As-Is bidding/contracting process procedures (see Table 10). All of these As-Is and To-Be processes can be evaluated, with the evaluations of total process operation time treated as an index of efficiency evaluation (see Table 11). 4.3. Effectiveness evaluation The effectiveness of construction management process reengineering can be evaluated by the customer satisfaction using target attainability TA in Eq. (26). During this evaluation process, internal customers and external customers must first be identified, where owners and subcontractors are defined as external customers and CEOs, vice presidents, managers and team leaders are defined as internal customers. After interviews and questionnaire surveys, customer demands used to form the relative importance matrixes Wi of the bidding/contracting (see

Table 10 Process operation time for the As-Is biding/contracting process Procedure items

Queuing time Execution time Process time (h)

can be calculated as: C2a = 22,440.55/160.132; the COV of the first procedure execution time equals Cs = 13.15/8.382; and time usage rate equals ρ = 8.38/160.13. Using Eq. (14), the process operation time for the first procedure can be calculated as: ρð2−ρÞCa2 þ ρ2 Cs2 þ E½S 2λð1−ρÞ  8:38 8:38 22440:55 8:382 13:15 2− þ 160:13 160:13 160:132 160:132 8:382 þ 8:38  V 1 8:38 1− 2 160:13 160:13 V15:96

ð32Þ

E½T V

ð29Þ

Reports for biding evaluation 7.58 Audit evaluation reports 7 (Manager of planning dept.) Preliminary bid examination 3.94 (Vice president) Set up cost items 0.73 Survey quantities 45.38 Check unit prices 44.97 Estimate total cost 1.51 Fill quotation table 0.18 Survey quotation table (Manager of 0.4 construction dept.) Review quotation table (Vice president) 0.71 Determine biding price (CEO) 0.6 Adjust item unit prices 0.09 Make bid proposal 0.07 Total process operation time

(h)

(h)

8.38 8.62

15.96 15.62

5.48

9.42

1.10 53.19 87.38 17.43 2.81 6.33

1.83 98.57 132.35 18.94 2.99 6.73

12.29 12.10 2.19 1.67

13 12.70 2.28 1.74 332.13

The departure of the first procedure is denoted in the form of Cd as:

Cd2 zð1−ρÞ2 Ca2

þ    8:38 2 22; 440:55 8:38 2 13:15 z 1− þ 2 2 160:13 160:13 160:13 8:38

z0:786

Table 11 Process operation time for As-Is and To-Be processes

ρ2 Cs2

ð30Þ As-Is process time (h) To-Be process time (h)

Bidding/contracting process

Budget process

Subcontracting process

332.13 214.10

724.45 201.38

71.74 33.65

Author's personal copy

Table 12 Relative importance of bidding/contracting process Target components Customer demands

Internal

Owner

Planning section

Construction section

Vice president

CEO

Lower cost Contract execution No extra cost Accurate quantity Suitable work load Detailed unit price information Information for biding Date of biding expiry Market condition information Resources of subcontractors Simplified process Reasonable biding budget Information for biding Accurate cost and profit Date of biding expiry Immediate biding information Information for biding Simplify certification process Acceptable profit Immediate biding information Information for biding Simplify certification process Decision-making authorization Acceptable profit Immediate biding information Wi

Ability for contract execution

5 3

5 3

Accurate calculation

Integrated process information

Detailed construction information

5

1

5 5

3 3

5

3 5 3 5

3

5

5 3

3

5 5

1 5

5

5

IT implementation

3 3 3 1 3

Efficiency

5

5 3 1 5 1

5 3

5

0.11

5 0.03

0.17

3 5 3 3 5 3 3 3

5

0.10

5 0.12

Allowable profit

1

5 3 5 5

5

5

3

1 5 5

Delimitation decision-making

1

3

5

1 1

5

1

5

5

1

5

5

0.07

0.16

0.08

5 1

3

5

3 5

1 0.08

Accurate budget

3

1 5

3

3 1 3

Information of subcontractors

0.05

5 5

Emphasis degree, pi 5 5 5 5 1 4 3 4 3 3 3 5 3 5 4 2 4 2 5 2 4 2 3 5 2

M.-Y. Cheng et al. / Automation in Construction 18 (2009) 183–193

External

Market condition control

0.04

191

Author's personal copy M.-Y. Cheng et al. / Automation in Construction 18 (2009) 183–193

0.41 0.16 0.16 0.51 1.86 1.32 0.16 0.00 0.09 0.09 0.09 0.00 0.00 TA = 6.8 0.04 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 0.08 0 0 0 3 4 4 1 0 4 4 4 0 0 1.92 0.12 0 0 0 0 3 0 0 0 0 0 0 0 0 0.36 0.10 3 1 1 0 0 0 0 0 0 0 0 0 0 0.50

0.05 0 0 0 0 3 0 0 0 0 0 0 0 0 0.15

0.07 0 0 0 0 0 3 0 0 0 0 0 0 0 0.21

0.16 0 0 0 3 4 4 1 0 0 0 0 0 0 1.92

Delimitation decision-making Accurate budget Information of subcontractors Efficiency IT implementation Detailed construction information

0.08 1 0 0 0 0 0 0 0 0 0 0 0 0 0.08 0.03 1 2 2 1 1 1 0 0 3 3 3 0 0 0.51 Wi Reports for biding evaluation Auditing the evaluation reports Preliminary bid examination Set up the cost items Quantities survey Check unit price Total cost estimate Fill the quotation table Survey the quotation table Review the quotation table Determine the biding price Adjust unit price of items Make the bid proposal OAi

0.11 0 0 0 0 0 4 0 0 0 0 0 0 0 0.44

0.17 0 0 0 0 4 0 0 0 0 0 0 0 0 0.08

Integrated process information Ability for contract execution

Accurate calculation

As-Is TA To-Be TA

Bidding/contracting process

Budget process

Subcontracting process

6.8 11.6

11.2 13.4

7.9 9.4

Table 12), budgets and subcontracting processes are summarized. After Wi is formed, 6 process attainability matrixes can be established according to Table 2. The process attainability matrix of the As-Is biding/ contracting process is shown in Table 13. Finally, 6 target attainability TA values can be obtained to represent an index of customer satisfaction/ effectiveness (see Table 14). 4.4. Summary CMPRPM validity has been established and supported by the results obtained in the two previous sections. To combine results, normalization should be adopted in which all evaluation values are normalized by As-Is quantities. For such, we define two parameters, i.e., process value PV and value improvement VI, as such: PV ¼

Customer Satisfaction TAnormalized ¼ Process Time Total process timenormalized

ð33Þ

VI ¼

PVTo−Be −PV As−Is PVAs−Is

ð34Þ

PV represents an index of BPR implementation performance, with a higher PV correlating with a lower process operation time value, which should deliver higher customer satisfaction. VI denotes a ratio of To-Be process improvements. The results show a dramatic and positive improvement in To-Be processes (see Table 15). This indicates that the proposed approach is essential and applicable for a successful BPR in topic of construction management. 5. Conclusion

Market condition control

Target components

Table 13 Process attainability matrix for As-Is biding/contracting process

Table 14 Target attainability TA of As-Is and To-Be processes

Allowable profit

Contribution of operation, pi

192

In BPR design, the ability to evaluate reengineering performance is a key to construction management process reengineering success. This paper proposed a CMPRPM model that integrates efficiency and effectiveness estimators applicable to construction industry needs and employs queuing theory to estimate process operation time to evaluate efficiency. The queuing theory is a highly suitable and significant tool for process operation time evaluation due to its allowing of dual conditions, i.e., an idle service and customer queuing mechanism. A target attainability matrix was employed to evaluate customer satisfaction achievement in order to evaluate effectiveness based on prior relative importance and operation target attainability. These two estimators quantify crucial factors as impersonal indexes for construction management process reengineering performance evaluation. Concepts of process value and value improvement were proposed for overall evaluation of the CMPRPM model. It is essential that both efficiency and effectiveness should be considered in any

Table 15 Process reengineering evaluation Normalization To-Be processes

Normalized process time (Efficiency)

Normalized target attainability (Effectiveness)

Process value

Value improvement

Bidding/contract process Budget process Subcontracting process

0.64 0.28 0.47

1.71 1.20 1.19

2.67 4.29 2.53

1.67 3.29 1.53

Author's personal copy M.-Y. Cheng et al. / Automation in Construction 18 (2009) 183–193

evaluation of process execution performance or To-Be process optimization in order that business managers can clearly view differences between As-Is and To-Be processes. Therefore, the model proposed in this paper should serve as a valuable tool with which to facilitate a successful BPR design in the construction industry. References [1] M. Hammer, Reengineering work: don't automate, obliterate, Harvard Business Review 68 (4) (1990) 104–112. [2] M. Hammer, J. Champy, Reengineering the Corporation—A Manifesto for Business Revolution, Harper Collins, New York, 1993. [3] C.E.A. Fowler, C. Gray, S.J. Palmer, Searching for success: the relationship between information technology and business process reengineering, International Journal of Computer Applications in Technology 11 (6) (1998) 428–435. [4] M.Y. Cheng, M.H. Tsai, Reengineering of construction management process, Journal of Construction Engineering and Management 129 (1) (2003) 105–114. [5] M.Y. Cheng, M.H. Tsai, Z.W. Xiao, Construction management process reengineering: organizational human resource planning for multiple projects, Automation in Construction 15 (6) (2006) 785–799.

193

[6] S. Mohamed, S. Tucker, Options for applying BPR in the Australian construction industry, International Journal of Project Management 14 (6) (1996) 379–385. [7] B.Z. Brown, J.J. Marston, Tennessee Department of Transportation's vision 2000: reengineering the project-development process, Transportation Research Record 1659 (1999) 129–140. [8] N. Abdul-Hadi, A. Al-Sudairi, S. Alqahtani, Prioritizing barriers to successful business process re-engineering (BPR) efforts in Saudi Arabian construction industry, Construction Management and Economics 23 (3) (2005) 305–315. [9] H. Imanaka, N. Ikeuchi, A simulation system for evaluating operation flow, Electronics & Communications in Japan, Part 1, Communications 80 (5) (1997) 11–22. [10] Y. Hwang, R.A. Leitch, Balanced scorecard: evening the odds of successful BPR, IT Professional 7 (6) (2005) 24–30. [11] X. Liu, Q. Li, Y.L. Chen, N.Y. Ma, H. Shen, Business process evaluation method based on balanced scorecard, Jisuanji Jicheng Zhizao Xitong/Computer Integrated Manufacturing Systems 9 (8) (2003) 661–665 (in Chinese). [12] A. Arazi, E. Ben-Jacob, U. Yechiali, Bridging genetic networks and queueing theory, Physica A: Statistical Mechanics and its Applications 332 (1–4) (2004) 585–616. [13] J.A. Buzacott, Commonalities in reengineering business processes: models and issues, Management Science 42 (5) (1996) 768–782. [14] J.A. Buzacott, J.G. Shanthikumar, Stochastic Models of Manufacturing Systems, Prentice-Hall, New Jersey, 1993.