Professional Documents
Culture Documents
Chou 2010
Chou 2010
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
a r t i c l e i n f o a b s t r a c t
Article history: This work presents a Web-based visualized architecture, design, and implementation for assessing project
Accepted 24 February 2010 performance by integrating earned value (EV) analysis and database management system (DBMS).
Particularly, a novel probabilistic multiple criteria decision making (MCDM) process is applied to identify the
Keywords: optimal software for developing the Web-based DBMS based on interviews with domain experts and
Earned value management professional engineers. EV analysis serves as a control technique that helps project managers (PMs) monitor
MIS
cost overruns and project delays. Leveraging modern technology, Web service enables PMs to access project
Web technology
MCDM
information without constraints of time or place. The developed management information system (MIS)
Project control provides an objective measure of completed work that PMs can use to monitor project progress and assess
Project management project achievements by converting project data into manageable information clusters. Additionally, the
system offers construction personnel a simple and flexible method of tracking, modifying, and updating cost
and time-based project data.
© 2010 Elsevier B.V. All rights reserved.
1. Introduction Data and information systems that offer data access without
constraints of time and place are a global trend driving the popularity
Infrastructure projects frequently experience scheduling problems of Web services. Independence from software applications is the
and cost overruns during the construction phase. Hence, it is strongest advantage of Web-based technology. Such systems do not
necessary to exploit modern technology to boost monitoring require complicated functions, and their relatively low application
capability, scheduling accuracy, and cost estimates in construction costs have further popularized their use [6,8,10]. This study focuses on
engineering. Information technology (IT) alleviates the need for developing project control and monitoring (PC&M) tool during the
project managers to perform the monotonous tasks of documenting, execution phase, and Fig. 1 illustrates the full project data life-cycle
analyzing, and presenting project information. Construction compa- framework and the position of proposed earned value management
nies can improve their business operations and profits by frequently system (EVMS) for assessing project performance.
monitoring key project performance indicators to immediately detect Specifically, Fig. 1 indicates that the effective implementation of
and correct cost and schedule overruns. While contractors can EVMS strongly influences the project execution phase, during which
customize their own automated systems to track project progress, project information must be detailed and the capability to change
developing such systems requires extra expenses besides normal costs and schedule reduces as the project progresses [24]. Work
operations. breakdown structure (WBS) design and other project management
Presenting data visually is a critical business process [8,10,21] techniques are generally recommended for project initiation and
since visual information is clearer and easier to understand than planning [23], and the generated data can be saved into an operational
qualitative and numeric information. Effective visual presentation database for subsequent monitoring and documentation (Fig. 1).
using adequate management technique, such as earned value analysis, Project managers (PMs) can then update and modify the project
eliminates the need for companies to train staff to interpret schedule and cost estimates as necessary. Stakeholders can also use
qualitative and complex numerical information. Automating visual- the proposed EVMS to establish a consensus with their PMs regard-
ization is also essential owing to its ability to reduce the need to ing a scope–time–cost baseline. Project information is subsequently
manually analyze and present information. Therefore, project infor- acquired during construction, fed into the earned value management
mation must be presented visually and automatically for efficient (EVM) process and used to update the EVMS as detailed below.
control process. During the project closeout stage, the EVMS structured data can be
normalized using data warehouse relational modeling for possible
future retrieval. Using Web-based technology, the data warehouse
⁎ Corresponding author. Tel.: +886 2 27376321; fax: +886 2 27376606.
E-mail addresses: jschou@mail.ntust.edu.tw (J.-S. Chou),
(DW) can be made available on the Internet (Fig. 1), enabling PMs to
hungming@mail.ntust.edu.tw (H.-M. Chen), D9605001@mail.ntust.edu.tw (C.-C. Hou), access both project information and the DW whenever necessary. The
m9705504@mail.ntust.edu.tw (C.-W. Lin). DW can also be mined to yield nugget rules or implicit knowledge that
0926-5805/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.autcon.2010.02.006
J.-S. Chou et al. / Automation in Construction 19 (2010) 596–607 597
could be beneficial in future similar project initiation and planning, detect problems early enough to allow corrective action, the
since the DW also stores valuable information regarding project advantages of IT applications have not yet been fully exploited by
functions, features, and characteristics. the construction industry.
A growing trend among business organizations is to expand e- Besides serving as a core PPM technique, EVM helps PMs control
business operations using Web-based systems. Although numerous project progress by increasing prediction accuracy. EVM also provides
corporations have achieved competitive advantages using IT, few a platform for project control by consolidating information and
have developed Web-based management information systems for appropriating risks, enabling unified and accurate project risk controls
PC&M, particularly systems that integrate EVM. This study thus [20,28]. Specifically, EVM is a technique for controlling and managing
designs a visualized database management system (DBMS) with projects by consolidating vital project data and expressing it usefully,
graph drawing package for real-time PC&M using earned value (EV) highlighting risks, and measuring project performance [19]. Conse-
metrics. Additionally, a novel probabilistic multiple criteria decision quently, the analytical results obtained using EVM provide a project
making (MCDM) process, analytical hierarchy process (AHP) with performance measurement baseline (PPMB) for determining and
Monte Carlo simulation (MCS), is applied to identify the optimal controlling project outcomes through appropriate action.
software for developing the Web-based DBMS based on interviews To help PMs reduce the adverse impacts of unexpected situations,
with domain experts and professional engineers. Hemachandra and Ruwanpura (2008) devised a control tool inte-
The rest of this paper is organized as follows. Section 2 reviews grating MCS and EVM [17]. Alternatively, Benjaoran (2009) employed
pertinent literature on project performance measurement and draws a collaborative approach to develop a cost control system for five
links with the proposed research. Section 3 then outlines the system selected small- and medium-sized contractors [4]. The author adopted
framework and methodology adopted for Web-based EVMS devel- a new and efficient cost control system based on EV and developed it
opment. Next, Section 4 describes system requirements with decision using easily-learned ICT (information and communication technolo-
making process, system architecture, design and implementation, and gies). Furthermore, Lipke et al. (2009) improved the capability of
system validation using case study. Finally, Section 5 concludes by PMs to make informed decisions by providing a reliable method of
presenting study contributions, cautions regarding implementation, forecasting final cost and project duration [22].
and areas requiring improvement. Rather than using the EVM essentials, Cheung et al. (2004)
designed a Web-based construction project performance monitoring
2. Pertinent literature system intended to assist PMs with construction project control [8].
The authors measured project performance in terms of people, cost,
Numerous industries and projects have successfully implemented time, quality, safety and health, environment, client satisfaction, and
project performance measurement (PPM) [16,20,22]. Accordingly, communication. Although these studies promoted EVM methods or
information technology is important in effective PPM since it facili- employed project monitoring system, none of them designed a man-
tates data handling [8]. Earned value management is a PPM procedure agement information system to embed the EVM theory into practice
that objectively monitors the physical completion of construction via automatic visual graphing for the construction project control and
work [12,31]. Because EVM integrates schedule and cost performance monitoring.
measurements with overall project scope, it has become an important Successful implementation of EVM requires considerable organi-
tool for construction managers in project control. Although managers zational support, including top-level management support, organiza-
can utilize conventional EV analysis to assess project performance and tional integration, and effective training. Moreover, Kim et al. (2003)
598 J.-S. Chou et al. / Automation in Construction 19 (2010) 596–607
EVMS implementation. Real-time measurement of project perfor- ➢ Work scope data, including
mance using the proposed system requires further analysis and • Detailed activities or work packages in the WBS with required
presentation of the following data: quality.
➢ Project time data, including Notably, experienced managers are responsible for taking appropriate
action if necessary after reviewing the EVMS outputs.
• Actual task start and finish dates.
• Expected task start and finish dates.
• Estimated task completion ratio. 4. Web-based EVMS design and implementation
EVMS can facilitate project control during construction by 4.1. Programming tools and system requirements
providing a platform for users to continuously analyze project
information within an IT architecture. The proposed system also The Web-based system comprises client-side programming
provides a document trail that project stakeholders can follow from language, server-side programming language, a graph drawing
project initiation through to closeout. Consequently, the EVMS can be package, and a DBMS. Exclusively, the proposed system employs
transformed into a data warehouse as project completion, which has a JavaScript, PHP (PHP Hypertext Preprocessor), and Open Flash Chart
potential to become a valuable explicit corporate knowledge asset. for the client-side, server-side language, and flash graph drawing tool,
Fig. 5 shows operating flow of the system framework. On the EVMS respectively. JavaScript is a lightweight interpretation and scripting
main menu, the system provides a Web-based interface that enables language designed to add interactivity to HTML (Hyper Text Markup
the user to input real-world data, as described previously, to facilitate Language) pages [29]. JavaScript, which is generally embedded
a baseline comparison via EV analysis. Baseline information can be directly into HTML pages, is open source. Meanwhile, PHP is a
loaded from preliminary planning or similar projects. server-side scripting language that supports multiple databases [30].
During user analysis, the system fulfills three primary functions to Open Flash Chart can help developers and users to automatically
assess project performance: variance analysis, work efficiency construct miscellaneous charts, including line, bar, pie, and area charts
evaluation, and trend analysis. For variance analysis, the schedule shown in flash animations. Both the PHP and Open Flash Chart are also
and cost variances are automatically calculated and graphically OSS that can be freely downloaded and used for minimal cost [26,32].
presented. Furthermore, the database can store data on tracking and Given these merits, this study uses JavaScript, PHP and Open Flash
documentation. For work efficiency assessment and trend analysis, a Chart as the programming and graphing languages in developing the
project forecasting report can incorporate cost and time outputs. Web-based system.
Regarding the selection of DBMS, which is the core engine for the Table 1
EVMS, current popular software, including MS-Access, MS SQL Server, Verbal assessments of the numerical ratings.
MySQL, and Oracle were targeted as candidates. To clarify the Numerical rating Verbal judgments of preferences
potential criteria in software selection, IT personnel, project man-
9 Extremely preferred
agers, and potential users were surveyed via a questionnaire that 8 Strongly to extremely
included items on the importance of individual criteria from the 7 Strongly preferred
perspective of developers: (1) price, (2) technical support, (3) user 6 Considerably to strongly preferred
5 Considerably preferred
interface, (4) cross-platform ability, (5) database capability, (6) data
4 Moderately to considerably
import/export ability, (7) data limitations, (8) data layout model, (9) 3 Moderately preferred
expansion ability, (10) hardware requirement, (11) security, and (12) 2 Equally to moderately
backup and recovery ability. Since these 12 criteria dimensions result 1 Equally preferred
in burden for decision making and appear equally important, a scree-
plot technique [9] of priority weighting versus criteria was imple-
mented during the second phase to identify the most critical criteria, respondents can use as a reference in making judgments. During this
resulting in database capability, backup and recovery ability, security, phase, a panel of experts and experienced engineers were asked to
user interface, cross-platform ability, and data import/export ability. conduct database rating using a three-point estimation technique
Although the dimensions of the evaluation criteria are reduced, the employing minimum, most likely, and maximum values, since the
priority decision from multiple groups regarding the core engine respondents could assess the constructed triangular distributions
DBMS remains unclear and uncertain. more easily than other distributions [23], and because this technique
is more appropriate for eliciting practical professional experience [7].
4.2. MCS-AHP MCDM process for DBMS selection Table 2 lists the resulting criteria pair-wise comparison matrices for
the databases in a compact format, given that the structure of the AHP
To alleviate concerns regarding DBMS selection by applying critical comparison matrix possesses certain regularity (e.g., a judgment of a
evaluation criteria to multiple groups, a novel AHP with MCS pair-wise comparison (aij) has a reciprocal (1/aij) placed in the
procedure is proposed as a probabilistic MCDM technique to transposed position). The lower triangle of the matrix represents the
determine the optimal software for DBMS implementation. The AHP average three-point assessment (ATPA) for each database in terms of
model, originally developed by Saaty (1980) [25], is designed to how it contributes to the criteria: database capability (DC), backup
quantify relative priorities for a given set of alternatives on a ratio and recovery ability (B), security (S), while the upper triangle
scale [1]. Nevertheless, conventional AHP often causes decision indicates the ATPA of the criteria: user interface (U); cross-platform
dilemma for yielding nearly equivalent priority coefficients. In ability (C); data import/export ability (DI). Fig. 7 shows the above
contrast, simulation modeling performs different decision scenario MCS-AHP flow, which was used repeatedly to synthesize the priority
experiments without disrupting MCDM process. Thus, integration of distributions for the DBMSs. The probabilistic MCDM is repeated until
MCS-AHP can reinforce and facilitate managerial decision making by the judgment consistency reaches the threshold proposed by Saaty.
giving clear attribute weightings under uncertainty and confidence Fig. 8 illustrates the priority cumulative distribution functions
level. Diverse application examples include clearly prioritizing quality (CDFs) for the software candidates after performing required MCS. By
attributes in service management [18], prioritizing contractor pre- intersecting the 95-percent likelihood line with the CDFs, four priority
qualification with criteria determined by owner requirements and coefficients can be extracted and presented with normalization in a
preferences as well as the characteristics of the individual contractors pie chart (Fig. 9) ranking MS SQL Server (41.9%), Oracle (25.3%),
[1], probabilistic EV analysis for predicting project performance under MySQL (21.7%), and MS ACCESS (11.1%) in descending order of
uncertainty [17], and range estimation incorporating desired confi- priority based on the synthesis of the valuation systems of multiple
dence level for construction cost management [11]. groups.
Fig. 6 depicts the AHP hierarchy based on the above critical criteria Based on the above investigations and simulations, including ques-
(Level 2) and candidate software (Level 3). Table 1 lists the relative tionnaire analysis and probabilistic MCDM process, the MS SQL Server
scale measurement with corresponding verbal description [25] that is identified as the ideal DBMS software for EVMS development. This
Table 2
Pair-wise comparison matrices in the form of triangular distributions for the criteria.
basis. Input is performed via two steps. First, the user must specify a users to fill all cells. After step 2, the EVMS displays an “Added
construction project. Second, the user must input daily or weekly successfully” message to confirm the successful loading of the new
construction project data, ensuring that data is entered into every construction project.
cell. If JavaScript detects an empty cell, the user receives a pop-up
reminder to fill that cell. The user then clicks the “Calculate” button 4.4.6. System maintenance
to automatically compute the numerical performance measures with System administrators must periodically maintain user and
embedded mathematical functions. Finally, the user clicks the “Input department profiles on the EVMS. Administrators can use the user/
Project Data” button to save the data to the database. Following step department management function to add, edit, disable/enable, and
2, the EVMS displays an “Input successfully” message, after which the delete system users.
user can click “Generate EVM Graph” to plot a real-time EVM graph
on screen. 4.5. System validation and case study
4.4.5. Add a new construction project This study demonstrates system functionalities using the exam-
The add function allows users to add a new construction project to ple of a typical bridge construction project from the Taiwan Public
the EVMS. First, the user inputs project attributes such as name, Construction Commission database. Specifically, the selected case
location, budget, start date, and completion date. Next, the EVMS study is the JieShou Bridge renovation construction project, located
automatically calculates the number of weeks required for project in Rueifang, Taipei County, Taiwan. This project includes six
completion, and the user inputs a planned weekly budget value for principal work items and has an expected completion time of
the work items. Once again, a JavaScript pop-up dialogue prompts 10 weeks. Fig. 14 illustrates the key steps, as described above,
606 J.-S. Chou et al. / Automation in Construction 19 (2010) 596–607
involved in entering this project into the constructed Web-based tion (Fig. 14(b)). The system then calculates base EVM values,
EVMS. A user must first create a new construction project and input including AC, EV, CV, CPI, BAC, and EAC (Fig. 14(c)). Finally, the
the work items, schedule, and budget (Fig. 14(a)); next, the user system automatically generates a graph of PV, AC, EV, and EAC
must input weekly project data, including AC and percent comple- showing the project status in week six (Fig. 14(d)).
J.-S. Chou et al. / Automation in Construction 19 (2010) 596–607 607
5. Implementation issues and conclusions [4] V. Benjaoran, A cost control system development: a collaborative approach for
small and medium-sized contractors, International Journal of Project Manage-
ment 27 (3) (2009) 270–277.
This study details the preliminary EVMS setup using modern [5] D.C. Bradley, M. Mascaro, S. Santhakumar, A relational database for trial-based
information technology. The developed EVMS streamlines data behavioral experiments, Journal of Neuroscience Methods 141 (2005) 75–82.
[6] O. Cakir, M.S. Canbolat, A web-based decision support system for multi-criteria
storage and seamlessly integrates project scope, time, and cost data. inventory classification using fuzzy AHP methodology, Expert Systems with
The prototype system graphs project progress in terms of EVM Applications 35 (3) (2008) 1367–1378.
metrics and presents it on the Web in real-time for access by project [7] K.W. Chau, The validity of the triangular distribution assumption in Monte Carlo
simulation of construction costs. Empirical evidence from Hong Kong, Construc-
managers. Moreover, the proposed system offers construction tion Management and Economics 13 (1) (1995) 15–21.
personnel a simple and flexible visual means of tracking, modifying, [8] S.O. Cheung, K.K. Cheung, H.C. Suen, CSHM: Web-based safety and health
and updating time and cost -based data that allows the input of monitoring system for construction management, Journal of Safety Research 35
(2004) 159–170.
project scope and quality descriptions during execution.
[9] J.-S. Chou, Web-based CBR system applied to early cost budgeting for pavement
The EVMS is developed using open source programming and flash maintenance project, Expert Systems with Applications 36 (2, Part 2) (2009)
graph drawing software integrated with MS SQL Server and Web 2947–2960.
technologies. Notably, besides the straightforward decision regarding [10] J.-S. Chou, J.T. O'Connor, Internet-based preliminary highway construction cost
estimating database, Automation in Construction 17 (1) (2007) 65–74.
free OSS adoption, the priorities of database development software [11] J.-S. Chou, I.T. Yang, W.K. Chong, Probabilistic simulation for developing likelihood
are determined using the innovative probabilistic MCDM process with distribution of engineering project cost, Automation in Construction 18 (5)
the application of MCS-AHP. The implementation of this EVMS (2009) 570–577.
[12] D.F. Cioffi, Designing project management: a scientific notation and an improved
emphasizes the need for improved project estimation and control formalism for earned value calculations, International Journal of Project
techniques. To maximize system efficiency, controllers and users must Management 24 (2006) 136–144.
standardize their definitions and working methods, since non- [13] S. El-Omari, O. Moselhi, Data acquisition from construction sites for tracking
purposes, Engineering, Construction and Architectural Management 16 (5)
standard definitions and work methods can create conflicts during (2009) 490–503.
data entry if computers do not recognize them. [14] Q.W. Fleming, J.M. Koppelman, Using Earned Value Management, Cost Engineer-
Several project-related issues and limitations must be noted since ing 44 (9) (2002) 32–36.
[15] A.G. Ghanem, Y.A. Abdelrazig, A framework for real-time construction project
they have the potential to affect practical system implementation.
progress tracking, American Society of Civil Engineers, vol. 2006, League City,
Measuring and coordinating information flow can be challenging Houston, TX, United states, 2006, p. 112.
because of the input data generally being appraised by professionals [16] H. Gokhale, M.L. Bhatia, A project planning and monitoring system for research
projects, International Journal of Project Management 15 (3) (1997) 159–163.
based on their personal experience and judgment. Completed work is
[17] P.S. Hemachandra, J.Y. Ruwanpura, Probabilistic earned value forecasting and
not paid for until owners are satisfied, which requires subjective activity performance index using Monte Carlo simulation, Proceedings, Annual
verification if cutting-edge technologies (such as, remote sensing Conference — Canadian Society for Civil Engineering 1 (2008) 220–230.
technology) are employed for automatic data acquisition without [18] T.-H. Hsu, F.F.C. Pan, Application of Monte Carlo AHP in ranking dental quality
attributes, Expert Systems with Applications 36 (2, Part 1) (2009) 2310–2316.
confirmation by on-site engineers. Second, services that are intangible [19] H. Kerzner, Project Management, 9 edJohn Wiley & Sons, Inc, 2005.
and embedded in other work items should be grouped separately, to [20] E. Kim, W.G. Wells Jr., M.R. Duffey, A model for effective implementation of earned
avoid the EVMS artificially inflating subjective measures in situations value management methodology, International Journal of Project Management 21
(2003) 375–382.
where it is unable to differentiate them. [21] J. Li, O. Moselhi, S. Alkass, Internet-based database management system for
Furthermore, frequent changes to project scope or design may project control, Engineering, Construction and Architechtural Management 13 (3)
complicate system automation, making additional IT mechanisms (2006) 242–253.
[22] W. Lipke, O. Zwikael, K. Henderson, F. Anbari, Prediction of project outcome. The
necessary to differentiate approved and unapproved changes. application of statistical methods to earned value management and earned
Communicating approved and unapproved information within the schedule performance indexes, International Journal of Project Management 27
system poses another challenge for real-time system controllers. (4) (2009) 400–407.
[23] PMI, A Guide to the Project Management Body of Knowledge: (Pmbok Guide),
Finally, a construction project bid is frequently itemized, typically in Fourth Edition, Project Management InstituteFourth ed, 2008.
CSI (Construction Specifications Institute) Master format (or in other [24] S.T. Regan, EVMS internal and DCAA audit recommendations, 2006 AACE
formats, depending on client). The consistency of data structure in the International Transactions – 50th Annual Meeting, Jun 19–22 2006, Vol. 2006,
Association for the Advancement of Cost Engineering, Las Vegas, NV, United states,
system requires further attention since not all contractors and clients
2006, pp. EVM.03.01-EVM.03.07.
use the same grouping system. Future investigations can focus on the [25] T.L. Saaty, The Analytical Hierarchy Process (1980) 2–30.
above issues and thus establish a full-scale system to examine ways of [26] teethgrinder.co.uk/open-flash-chart/, Open Flash Chart project
overcoming these limitations as well as applying the EVMS architec- [27] E. Turban, E. Mclean, J.C. Wetherbe, Information Technology for Management:
Transforming Organizations in the Digital Economy, 4th Edition edJohn Wiley &
ture to various industries. Sons, Inc, 2004.
[28] S. Vandevoorde, M. Vanhoucke, A comparison of different project duration
References forecasting methods using earned value metrics, International Journal of Project
Management 24 (2006) 289–302.
[29] w3schools.com, JavaScript Introduction.
[1] K.M.A.-S. Al-Harbi, Application of the AHP in project management, International [30] w3schools.com, PHP Introduction.
Journal of Project Management 19 (1) (2001) 19–27. [31] S.P. Warhoe, The basics of earned value management, 2004 AACE International
[2] A. Alheraish, Design and implementation of home automation system, IEEE Transactions – 48th AACE International Annual Meeting, June 13, 2004–June 16,
Transactions on Consumer Electronics 50 (4) (2004) 1087–1092. 2004, Association for the Advancement of Cost Engineering, Washington, DC,
[3] W. Alsakinj, K. Wikstrom, J. Kiiras, Proactive schedule management of industrial United states, 2004, pp. CSC.07.01-CSC.07.10.
turnkey projects in developing countries, International Journal of Project [32] www.php.net, PHP: Hypertext Preprocessor.
Management 22 (2004) 75–85.