Advanced Engineering Informatics · A three-stage framework for introducing a 4D tool in large consulting firms Meng-Han Tsai, Shih-Chung Kang*, Shang-Hsien Hsieh National Taiwan

Advanced Engineering Informatics 24 (2010) 476–489

Contents lists available at ScienceDirect

Advanced Engineering Informatics

journal homepage: www.elsevier .com/ locate /ae i

A three-stage framework for introducing a 4D tool in large consulting firms

Meng-Han Tsai, Shih-Chung Kang *, Shang-Hsien HsiehNational Taiwan University, Taipei, Taiwan

a r t i c l e i n f o

Article history:Received 20 May 2009Received in revised form 20 April 2010Accepted 25 April 2010Available online 23 May 2010

Keywords:4D toolUsability testWorkflow

1474-0346/$ - see front matter Crown Copyright � 2doi:10.1016/j.aei.2010.04.002

* Corresponding author. Tel.: +886 2 33664346.E-mail address: [email protected] (S.-C. Kang).

1 More information can be found at http://www.ben2 More information can be found at http://ppm.inter3 More information can be found at http://www.BAL

a b s t r a c t

The increase in the use of 4D management tools in recent years within the construction industry has beenphenomenal, partly due to the increasing support available in commercial software packages, and partlyin response to a greater demand for efficient construction management. However, successfully imple-menting a 4D management tool in an engineering firm for use in actual projects remains a challengingtask. This paper presents the authors’ experiences of implementing an in-house 4D management toolat a large engineering, procurement and construction (EPC) firm with a long history of design–build pro-jects. A three-stage consulting framework of system evaluation, usability study, and management plan(SUM) was proposed and implemented for a firm of this size, which included three parts: (1) System eval-uation: requirement analysis and performance evaluation of both hardware and software components ofthe 4D tool; (2) Usability study: usability tests and improvement of the 4D tool; and (3) Managementplan: workflow re-engineering for the firm to be able to successfully implement and apply the 4D man-agement tool to actual projects. We found that the SUM framework was able to effectively identify majorproblems when introducing a 4D tool to a large design–build project and helped to minimize its ownimpact on the firm’s business processes.

Crown Copyright � 2010 Published by Elsevier Ltd. All rights reserved.

4

1.1. Background of 4D technologies

Four-dimensional visualization technologies have been used mostprominently in the construction industry to assist in the 4D planningand scheduling process within the construction and project manage-ment framework. This process combines 3D models with their corre-sponding construction work schedules in its simplest form.Utilization of this type of 4D visualization technology has alsoemerged rapidly [1] in the past decade. This is, primarily becausethe construction industry has been made more aware of the benefitsof using 4D CAD applications to aid project work [2–7]. The benefits ofusing 4D-modeling mechanisms over traditional tools are well de-scribed in several of the cited references, with key benefits includingincreased productivity, improved project coordination capability, andoptimized on-site resource utilization techniques.

The development and the application of 4D technology has alsobecome more popular with the availability of powerful commercial4D management software, such as Bentley’s Navigator1, Inter-graph’s SmartPlant Review2, BALFOUR’s FourDscape3, Common

010 Published by Elsevier Ltd. All r

tley.com/.graph.com/visualization/.4.com/.

Point’s Project 4D, and ConstructSim [8]. Moreover, the volume ofrecent research efforts on investigating the use of this technologyhas transformed the 4D communication technology from carryingout simple 4D animation of construction progressions to interactive4D simulations of alternative construction processes (for examples,see [9]). Four-dimensional construction management tools havebeen experimentally deployed in many projects in recent years withthe purpose of allowing managers to make proactive decisionsthroughout the course of their projects. The San Mateo Health Facil-ity [10] and the Construction Director [1] are two prominentexamples.

Researchers have also discovered that the principal advantageof using such 4D management tools on top of a typical planningprocess is the visual enhancements they provide. During any peri-od of time within the project, construction activities as well as sitespace utilization can be viewed temporally forwards or backwards[11], providing users with a graphical simulation of the work plan[12] to help them understand the construction work sequencesmore effectively. The visualization facilitates team collaborationin removing logical errors in construction operations. Ownersand users of the constructed facilities might have little experiencein construction projects, and are often unable to participate in theconstruction plan development process unless a simple method ofvisualization and communication is made available to them; 4Dmanagement tools allow for this [13–15].

ights reserved.

4 More information can be found at http://www.commonpointinc.com/.

M.-H. Tsai et al. / Advanced Engineering Informatics 24 (2010) 476–489 477

Construction project teams must consider a wide variety of infor-mation when making project decisions. During the constructionprocess, sub-contractors, if limited to the use of traditional tools,may often face problems interpreting intricate and complex designdetails, and spend much time and effort trying to understand the de-sign intent. These problems may cause significant delay particularlyto large projects, especially if members of the on-site team are re-quired to explain every single details [14]. In addition, large-scaledesign–build projects, with vast quantities of data for consideration,incur problems displaying the data in a single chart, especially whenmonitoring the state of each individual activity. Visual media allownon-professionals, especially, in particular to gain a clearer under-standing of such site management processes [11].

Four-dimensional visualization technology seems to be aneffective way of enhancing the many different types of human cog-nition (e.g. [2,16–19]) and can help anticipate potential construc-tion conflicts during the operational stages. Furthermore, it isanticipated that the successful implementation of such systemswill enable the creation of reliable plans and constraint-freeassignments, which will in turn reduce production risks and im-prove on-site productivity [20].

1.2. Concerns of implementing 4D technologies

Construction companies as well as the utility owners are start-ing to become aware of the benefits and capabilities of 4D-model-ing tools. The market has an abundance of commercial 4D softwarepackages, regularly upgraded to cope with advancing technology.Nevertheless, project stakeholders lack confidence in using themwhen it comes to practically adopting these types of tools, espe-cially in large-scale or design–build projects. Concerns towardsimplementation of these technology have arisen have surfacedfrom the existing literature and can be seen from our experiencein working with industrial partners. These concerns fall into threecategories: the design of the computer’s system, the usability ofthe software, and the organizational workflow of company.

Computer system requirements of both software and hardwareare usually the critical concern for most firms, especially large-scaleorganizations. Cost, availability and compatibility with existingarrangements of software are key considerations before implement-ing new systems. Four-dimensional models are also often large andcomplex [21], making computational efficiency and system capacitymajor concerns during the decision-making process [21–23].

Usability of tools has been a topic of research for years. Compa-nies frequently do not welcome new products into their existingpractice perceiving that usability testing is an expense of time andmoney. Research has been done using user-centered design [24–26] principles to manage the usability problems of computer sys-tems [27] and understand user behavior [28]. Usability tests helpusers to find the advantages of products by practically knowinghow quickly products can be learnt by the users, and its productivitywithin the company’s system. Dawood et al. [6] evaluated and im-proved the usefulness and usability of the VIRCON system via a fieldtest. Dawood found that the result of usability testing earn the cus-tomers’ faith in the new product, and convinced the company thatthe new system could provide better service than the existing one.The competitive advantages of VIRCON come into mainstream use.Russell et al. [29] assessed the usability of a dynamic visualizationenvironment and resolved the linking and re-linking limits, a verytime consuming process in 4D CAD. Research has been made in seek-ing improvements to usability of collaborative tools [30]. Changet al. [31] proposed the SEUT procedure to determine color schemesvia the usability tests.

Another major concern of software usability is the cost of edu-cating company engineers who would use the 4D tools in theirworks. Construction firms are particularly interested in the time

required to train their employees, and since engineers have ongo-ing work with tight deadlines, it is difficult to allocate time forthem to attend lengthy training courses [32]. In addition, someengineers, due to their area of specialization, are less familiar withspecial computer operations. They may have difficulties in famil-iarizing themselves with software that have complex interfaces.The literature has found that usability and learnability are two sig-nificant issues which must be addressed when implementing 4Dtechnologies [33–35]. Research has been done on using this tech-nology to change user habits; nevertheless, it is still a challengeto introduce the 4D tool.

The third category of concern is how to introduce the 4D tool tointo the company workflow. The workflows of interest of an orga-nization are the collection of tasks organized to accomplish a spe-cific business process that involves relevant staff (usually companyengineers) and the tool [36]. Ford et al. [37] and Hartmann andFischer [38] proposed that improving existing workflow of a pro-ject needs to be a criterion of 4D application development. Theeffectiveness of using 4D tools within the company’s workflow isof great interest to them. A thorough understanding of a company’sorganization and business processes is essential before re-design-ing its workflow. The company’s management, generally, do notwant frequent modification of workflows standards. Existing work-flow is determined by the existing organizational structure and itswork-sharing approach [39]. Hickey and Davis [40] and Santoroet al. [41] have found a common feature among various businessworkflows based on the improvement approach to capture adescription of existing workflows. Hartmann and Fischer have sug-gested that when a new system is introduced to the organization,the tasks and responsibility of departments and personnel shouldbe redefined [37,42,43]. However, this change may be complexand usually results in the need to modify existing operations.Without detailed and considerable planning and support, the deci-sion-makers might not have sufficient confidence to carry out thechanges to their workflow to implement the 4D tool [7].

Because the end users of these 4D tools are typically not IT pro-fessionals, the establishment of business processes and guides areof considerable worth is required to aid their usage. Koo andFischer [21] suggested the steps between preparing data and build-ing the 4D model. Dawood et al. [6] presents the process to build avirtual environment by using virtual construction site tools. Staub-French et al. [19] proposes a process of integrating CAD and sched-uling. Hu et al. [13] provides a way to create 4D model from 2Ddrawings. Chau [44] describes a 4D model depicting the entire pro-cess of a typical construction activity. Ma et al. [5] explains theworkflow of using 4D tool in a construction site. Although variousstudies propose the process of using 4D tools in several ways, therehas been no investigation into workflow changes re-engineeringwithin organizations for using 4D tools.

1.3. Needs of a consulting framework

To address the aforementioned concerns of firms, a consultingframework should be developed. This framework needs to be easyfor a large consulting firm to follow in order to be able to imple-ment the 4D tool successfully. The framework needs to considerpractical issues, including computer related technical problems,usability problems in software, and the workflow issues withinthe organization. This framework needs to be validated with a suc-cessful implement it in a real context.

2. The SUM framework

The SUM framework is the consulting framework proposed inthis research. It is an acronym for a three-stage framework for

478 M.-H. Tsai et al. / Advanced Engineering Informatics 24 (2010) 476–489

implementing a 4D tool in a large consulting firm: system evalua-tion, usability study, and management plan (Fig. 1).

System evaluation is the first stage, whereby users determinethe minimum system requirements required to operate the 4Dtool. The second stage is a usability study, where it is only whenthe results of the usability study attain a minimum usability per-formance level that users can then consider it suitable for a 4Dmodel. The third stage is the development of a management planfor introducing a 4D tool into existing business workflows. The de-fined workflow can be repeated if the management plan is deter-mined inappropriate for the users’ needs.

3. Background of case study

The case study for our research is on a high-rise constructionproject where the SUM framework is used to develop a practicalplan for introducing the 4D tool.

3.1. The design–build company in the case study

Our study focuses on the introduction of Construction Director,a purpose-built 4D tool, into CTCI Corporation, the largest engi-neering, procurement, and construction (EPC) firm in Taiwan. Theyare a contracting firm, specializing in large-scale design–build pro-jects and have been considering introducing Construction Directorto their standard workflow for considerable time. In our research,we selected a design–build project from the company and haveexperimentally implemented a 4D tool. We observed the resultsof both the existing workflow of the company as well as the rede-fined workflow (implementing the 4D tool) engineered by theauthors. The research was conducted in the context of the con-struction of a 17-story high-rise by CTCI to procure the necessarydata for input into and for building a comprehensive 4D model.

3.2. Four-dimensional tools used in the case study

Construction Director was developed by the Department of CivilEngineering at National Taiwan University (NTU) in collaborationwith CTCI Corporation, and has been chosen as the 4D tool in thiscase study. Currently CTCI mainly uses Intergraph’s Plant DesignSystem (PDS) for 3D plant design and construction, and SmartPlantReview (SPR) for interactive review and analysis of the 3D plantmodels. Although some 4D CAD simulation capabilities are sup-ported by the ScheduleReview function within the additional con-struction module feature of SPR, CTCI still feels the need for a moreuser-friendly, flexible, and intelligent 4D CAD solution on top of theexisting SPR. This is in order to not only facilitate more efficientand effective 4D simulation and management of plant construc-tion, but also to allow for better integration with CTCI’s project re-source management systems.

Performance Test

Capacity Test

System Evaluation Usability S

Field Stu

Lab Tes

Fig. 1. The SUM framework f

A collaborative software development project between NTU andCTCI was initiated in April 2005 to develop a 4D CAD system that bet-ter satisfies CTCI’s needs. Development of Construction Director haspreviously been summarized by Hsieh et al. [1]. The design of Con-struction Director takes advantage of several design patterns [45]that are able to satisfactorily display and convey construction infor-mation of a 3D model as well as the work items on the daily schedule[46]. In addition, each phase of the work plan can be represented by adifferent color [31]. Construction Director also utilizes digital datafor each construction phase to further improve the communicationand management functions, such as tracing the progress of a project.Based closely on the current system (3D model using SPR), the toolsand interfaces in Construction Director were developed to betterfacilitate 4D construction simulations.

3.3. Application of the SUM framework in the case study

In this research, we separated the task of introducing the 4Dtool into a large project into three stages according to the SUMframework: (1) system evaluation, (2) usability study, and (3)management plan. This framework was adopted bearing in mindthe features and implementation methods of a 4D tool. By adheringstrictly to the SUM framework, we were able to maintain the bal-ance between the complex system results of the hardware andsoftware components of the tool and ensure that the frameworkwas easy to work with and pragmatic. The stages of the SUMframework are listed in Table 1 below.

The first stage of the SUM framework was to perform a systemevaluation involving both a capacity and a performance test. In thecapacity test, three different computers each with a different hard-ware setup were tested to find the maximum number of elementseach could successfully compute and display. This allowed identi-fication of the hardware components that hindered computation.In the performance test, the frame refresh time was analyzed whilethe construction simulation was running. Finally, the ideal config-urations for both the hardware equipment (for capacity optimiza-tion) and the model display (for performance requirements) werefound.

The second stage, the usability study, involves two tasks: a fieldstudy and a lab test. In the field study, three sub-experiments wereconducted to help users form a method for creating a 4D model.The respective objectives were to determine (1) the time required,(2) the modeling tasks required, and (3) human resources needed.During the lab test, usability requirements were defined and userperformances evaluated during each of the three sub-experiments.Similarly, improvements on the results of the user-related defi-ciencies in the 4D tool can also be achieved by the usability study.

In the management plan stage, a method of introducing the newtool into existing business workflows was developed. The informa-tion gathering was conducted via face-to-face interviews. Each step

tudy

dy

t

Management Plan

Survey the Existing Workflow

Define the Workflow for Implementing 4D Tool

or introducing a 4D tool.

Table 1Stages of the SUM framework.

Stage Task

System evaluation Determine the minimum systemrequirements for operating the 4D tool

Capacity test Analyze equipment configurationPerformance test Analyze model display time

Usability study Improve the user-related deficiencies in the4D tool

Field study Analyze the 4D model creation and timeconsumption

Lab test Define usability requirements and evaluateuser performances

Management plan Introduce the 4D tool into existing workflowSurvey the existingworkflow

Interview with personnel of differentdepartments of the target company

Define the workflow forimplementing 4D tool

Defined an improved process for introducinga 4D tool

Table 2Quantitative results of computation efficiency.

Name System A System B System C

CPU 7474 4275 5983Memory 4276 3492 5404Graphics 1766 6228 13,278HDD 4864 4769 5454Average 5191 4994 6745

All values are benchmark scores from PCMark2005.

0

500

1,000

1,500

2,000

0 50,000 100,000 150,000 200,000

mem

ory

requ

irem

ent

(MB

)

System A System B System C

M.-H. Tsai et al. / Advanced Engineering Informatics 24 (2010) 476–489 479

in the initial setup was carefully adjusted according to manage-ment’s needs, which were gathered through multiple interviewswith various personnel from different engineering departments.From these interviews, we discovered that the 4D model tool has im-mense potential in reducing communication gaps between techni-cal departments and clients, as well as among the sub-contractors.Therefore, bearing in mind the need to minimize the number ofmajor changes to the existing workflow, we defined an improvedprocess that introduces a 4D tool to replace the existing 3D modelfor CTCI.

4. Stage 1: System evaluation

The first stage of the SUM framework, system evaluation, in-volves determining hardware capacity and performance. Capacitytesting includes the analysis of the efficiency of 4D model quantityanalysis and a comparison of the memory usage of different num-bers of elements. Performance testing focuses on the frame renewtime of the 4D simulation window. The tasks executed for systemevaluation are detailed in this section.

4.1. System performance results

This research used a software program called PCMark20055, ageneral tool for estimating system performance, to determine thecomputing capacity level of hardware, before comparing the equip-ment diversity with different hardware configurations.

This research analyzed equipment of various configurations,each with a different computer interface capacity level. Based onthe test results shown in Table 2, we have suggested improve-ments for hardware capacity and the maximum number of ele-ments that should be used in the model.

4.2. Capacity test

Our capacity test involved determining hardware computationcapacities using a system memory test. The test was able to deter-mine the minimum system requirements for operating the 4D tool.

System memory is used to deal with the information from the3D model, the results of which are used to build the 4D model.We tested the relationship between the number of elements cre-ated and memory consumption by using data from a petrochemicalfactory project, which originally had a total of 286,839 elements.As the Windows XP system used allocates a maximum of only

5 PCMark2005 is a system efficiency testing software published by FuturemarkCorporation.

2 GB for any single program, we chose to consider only 149,086of the elements.

The results of 10 successive memory tests are graphed in Fig. 2.They show that the memory requirements do not vary much acrossthe three system configurations, and that the relationship betweenthe number of elements and the memory requirement is almostlinear. The linear regression equation is represented byy = 0.007x + 520, where x is the number of elements and y is its cor-responding memory requirement. The 2 GB limit on Windows XPallows for handling a maximum of 218,285 elements. Due to thesememory restrictions, this test was only able to show the 3D modelgeneration, rather than completing other operations such as add-ing schedules and modifying the 3D elements. As we increasedthe number of elements, there were more computations to bemade slowing down 3D model generation; therefore, we recom-mended that at least 300 MB of RAM be set aside for preparingthe buffers.

4.3. Performance test

In this step, we test the relationship between the number of ele-ments and frame refresh time during simulation. This experimentwas able to determine a range in the number of elements analyzedthat allows the system to provide a normal level of performancewhen using the simulation function.

The 4D construction simulation (Fig. 3) is the main function ofConstruction Director. This study used 70,000 elements with 70work items at a frequency of 1000 elements per day and recordedthe frame refresh time for each of the three different system con-figurations (Fig. 4).

The three regression curves in Fig. 4 have a close fit with eachother. In comparing this against Table 1, we found that althoughSystem C had a better performance in graphical operation, the per-

number of elements

Fig. 2. Relationship between the number of elements and the memoryrequirement.

Fig. 3. 4D construction simulation.

0

5

10

15

20

25

30

35

0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000

refr

esh

time

(sec

)

number of elements

System A System B System C

Fig. 4. The relationship between the number of elements and frame refresh time foreach system tested.

480 M.-H. Tsai et al. / Advanced Engineering Informatics 24 (2010) 476–489

formance itself did not translate into a shorter refresh time. Thedynamic construction simulation function of System A did notuse much of its CPU resources, having used only 7–8% of totalCPU capacity. However, System B had better memory and hard diskperformances as well as a shorter frame refresh time than SystemA. The main factor impacting frame refresh time in dynamic con-struction simulation seems to be the memory and hard diskperformance.

4.4. Findings from system evaluation

Based on our findings, the following configuration is recom-mended for an ideal computer system setup for performance opti-mization: (1) At least 2 GB of RAM; (2) high-speed RAM such asDDR2 capable of 800 MHz; (3) a hard disk cache buffer of at least1024 KB, (4) a graphics card (512 MB)m and (5) a Intel Core 2Duo E6400, 2.13 GHz CPU.

A suggested memory requirement can be calculated usingx = (y � 500)/0.007 � 520, where x is the maximum number of ele-ments and y is the memory space in megabytes.

Our recommendations for ‘hardware configuration’ and ‘modeldisplay’ provide a baseline for equipment requirements for the4D tool.

The next stage is to consider the usability for introducing the 4Dtool.

5. Stage 2: Usability study

The second stage of the SUM framework, the usability study, in-cludes both a field study and a lab test. The purpose of the usabilitystudy is to improve the overall efficiency of creating a 4D modeland to establish a robust procedure for testing the usability ofthe 4D tool. During the field study, we focused on an actual build-ing project and through experiments introduced ConstructionDirector to the project for use as the 4D tool. Using the field dataof a design–build construction project, the 4D model was then cre-ated, and in particular, we observed the 4D model creation processin detail. During the lab test, we conducted experiments on thetool’s user interface design and functional design, receiving feed-back from the users of the 4D tool. Suggestions from participantsand further improvements of the usability study methods aresummarized in this section.

5.1. Field study

In order to clearly define the progress of implementing the 4Dtool in a real construction platform, this research began by collect-ing data from CTCI’s high-rise office building construction project.The project was a design–build contract that involved the con-struction of a multistoried building of 17 floors above groundand 3 below (Fig. 5). The negotiated contract was worth approxi-mately US$160 million; construction started during March 2006and was scheduled for completion in December 2008. The 3D mod-el of the building, as shown in Fig. 5, has 13,147 elements.

Fig. 5. Four snapshots taken during the building of the 3D model.

M.-H. Tsai et al. / Advanced Engineering Informatics 24 (2010) 476–489 481

5.1.1. Creating the 4D modelWe began the process by collecting data from the construction

site each day, as well as recording any other logs and sources ofinformation to better understand the demand for human and timeresources needed to implement the 4D tool. We then proceeded toestablish a procedure to create the 4D model using ConstructionDirector. The actual work executed on the construction site wasincorporated into the 3D model and then correlated with the cor-responding construction schedule. The following five steps werethen adopted during the 4D model creation process:

5.1.1.1. Step 1. Check the record of construction diary and establishwork items. The 4D model was built using information from theconstruction diary records obtained from the site. Data such aswork item names, times and activities were recorded accordingto the requirements.

5.1.1.2. Step 2. Compare the elements in the construction diary recordwith the elements of 3D model. Every 3D model element has a un-ique linkage number; the 4D model builder can use this informa-tion to find the location of the elements.

5.1.1.3. Step 3. Link the elements of record with work items. Users areable to find the corresponding 3D element number in ConstructionDirector using the same element number as the 3D model. Theycan then link the element of the 3D model and the elements of re-cord with the work items.

5.1.1.4. Step 4. Check the elements of work items against theconstruction diary record. After adding the daily work items, the4D model builder can use the ‘‘highlight” function to display the3D model and verify whether the data is correct.

5.1.1.5. Step 5. Verify model. Finally, the user checks the model forany mistakes such as name errors or schedule conflicts, makingcorrections where necessary.

5.1.2. Analysis of human resourcesOver a total of 192 h within the 14 months allocated for this

project, this research linked 10,394 elements in the 3D model withscheduled work items. A summary of the total time budget isshown below. It is based on analysis of time data as building ofthe 4D model progressed.

� Step 1 (Checking the record of construction diary and establish-ing work items), took 19.5% of the total time.� Step 2 (Comparing the elements in construction diary record

with the elements of the 3D model) took 47.44% of the time.� Step 3 (Linking the elements of record with the work items)

took 23.11% of the time.

In total, the three steps represented 90.05% of the 192 h used.Improving the efficiency of these three steps is the most importantaspect of optimizing the design and reducing the time required toimplement the 4D tool. Table 3 below shows the breakdown oftime consumption during each of the different steps and Fig. 6 de-picts the percentage of total time each step had spent in e the 4Dmodel creation.

Reasons for delays in the model creation discovered upon anal-ysis are tabulated in Table 4 below. The method of inking 3D mod-els with schedule elements, the five-step procedure for 4D modelcreation and time consumption analysis may help future develop-ers of 4D tools to further optimize their building process.

5.2. Lab test

Our lab testing involved recruiting participants who had expe-rience with various software packages. Participants went throughfive video tutorials, conducted the operations testing, participatedin an interview, and answered a questionnaire at the end.

5.2.1. Test participantsEight practicing engineers, each with a different background,

were asked to participate in the test. Their ages ranged from 25–45 years old; seven were male and one was female.

The primary criterion for selecting participants was that theywould have some degree of experience in project managementand in using 3D modeling software. The eight participants involvedin this test can be further categorized by their level of softwareproficiency with project management and 3D modeling. One par-ticipant had used only project management software; one partici-pant had used only 3D modeling software; three participants hadused both types of software but only for between 1 and 3 years;whereas the other three participants had used both types softwarefor at least 5 years.

The participants had between 1 and 10 years of experience inconstruction with backgrounds in construction planning, structuralanalysis and design, project management, and field construction.All participants were either current or potential users of Construc-tion Director. All of the participants had also had a preliminaryunderstanding of 4D simulation and were expected to use 4D toolsfor designing and construction in the near future.

5.2.2. Test planThe eight participants were assigned to carry out the tests on

Construction Director in order to examine how their experiencewith previous software tools affect their results with ConstructionDirector. During the 2-h testing session, the participants were re-quired to go through the video tutorials in succession and subse-quently answer follow-up questions after each tutorial had

Table 3Statistical time for building the 4D model during the different steps.

Year/month Work item number Number of elements Step1 (min) Step2 (min) Step3 (min) Step4 (min) Step5 (min) Total time of a month (min)

2006/03–07 3 287 1.700 39.133 47.067 0.483 24.117 112.5002006/11 22 61 16.517 7.483 4.483 1.834 0.000 30.3172006/12 24 80 15.600 7.200 5.533 2.600 0.000 30.9332007/01 15 198 15.167 39.267 14.816 3.833 1.167 74.2502007/04 1 53 2.133 6.750 2.317 0.117 2.516 13.8332007/05 50 424 26.650 100.700 34.883 10.033 4.317 176.5832007/06 123 956 55.983 176.867 60.733 19.450 8.100 321.1332007/07 176 1179 56.183 142.150 54.233 18.667 5.967 277.2002007/08 79 1062 33.700 108.650 40.167 11.800 6.016 200.3332007/09 85 949 33.417 101.283 37.233 9.650 21.567 203.1502007/10 128 1002 48.800 108.467 49.133 12.000 3.433 221.8332007/11 83 798 39.483 96.950 38.500 11.817 8.950 195.7002007/12 34 224 23.433 23.551 10.900 1.250 2.233 61.3672008/01 38 888 19.270 18.000 7.300 2.730 0.950 48.2502008/02 22 979 10.433 26.633 49.050 0.867 4.900 91.8832008/03 34 1254 21.150 17.717 40.833 4.050 8.600 92.350

Total 917 10,394 419.619 1020.801 497.181 111.181 102.833 2151.615

Step1, 19.50%

Step2, 47.44%

Step3, 23.11%

Step4, 5.17%

Step5, 4.78%

Fig. 6. Time taken for different steps of building the 4D model.

Table 4Account of time delays.

Steps Reasons for delay

Step 1: Checking the record of theconstruction diary andestablishing work items

1. The accuracy of the 4D model2. Work items have long names andfollow no rules in identifyingelements3. Work items may have wrong orincomplete schedule information

Step 2: Comparing the elements inthe construction diary record withthe elements of the 3D model

1. Hard to click and identify desiredelements in the 3D model2. Element linkage number has norule for its identity, and wastherefore difficult to identify

Step 3: Linking the elements ofrecord with the work item

There is no predefined rule forallocating element link numbers

Step 4: Verifying that the elements ofthe work items and theconstruction diary recordreconcile

It is very time consuming to verifythe schedule work item, whichconsist of many elements

Computer and

webcam

Facilitator

Facilitator

Participant

Computer

Observer

(a) The testing room (b) The observation room

Fig. 7. Test environment.

482 M.-H. Tsai et al. / Advanced Engineering Informatics 24 (2010) 476–489

finished. If the majority of the questions were answered correctly,they were allowed to move onto the next tutorial.

The tutorials taught specific functions of Construction Director;The substance of the clips were: (1) an introduction to the Con-struction Director and SmartPlant Review; (2) an introduction tothe search function of Construction Director; (3) the operationalprocedures for 3D elements, and linking them with work items;(4) and introduction to the animation function of ConstructionDirector; and (5) creating a steel building using ConstructionDirector.

5.2.3. Test environmentThe lab test was performed in a testing room and an observa-

tion room as shown in Fig. 7. In the video room, participants wereasked to sit between two facilitators, who controlled the test pro-cedure and helped participants with the 4D tool. A webcam re-corded the participants’ operations and the live feed was sent tothe observation room.

In the observation room, five observers monitored the partici-pants’ computer screens and followed their actions with the 4Dtool. The observers recorded operational mistakes and followedparticipants’ actions on the 4D tool.

5.2.4. Test procedureUser tests were conducted through video, by first transcribing

the basic function in Construction Director and then illustratingthe work creation principle. Participants who viewed the videorecording then carried out the assessment tasks. Feedback fromthe participants was then collected in interviews and question-naires after the video. They were then assigned to perform reorga-nization and analytical studies of the video. Finally, they weresubjected to behavioral testing to verify the success of the 4D tool.

The questions in the questionnaire fell into nine major catego-ries: (1) the operational procedure for establishing the projectschedule, (2) the method of arranging the schedule catalogue, (3)the search function of Construction Director, (4) the operationalprocedure for linking 3D elements with work items, (5) toolbardiagrams, (6) the operational sequences of animation, (7) searchinformation of elements in SmartPlant Review, (8) length of teach-ing video, and (9) time spent on the tests.

Fig. 8. Participants’ responses to linking 3D elements to work items.

M.-H. Tsai et al. / Advanced Engineering Informatics 24 (2010) 476–489 483

5.2.5. Results of the lab testThe test results revealed that the participants felt perplexed by

(1) the complexity of the operational procedure for linking the 3Delements to work items, (2) the toolbar diagram, and (3) the func-tions of SmartPlant Review. Fig. 8 below shows the breakdown ofparticipants’ responses to linking 3D elements with the work items.

Participants in the usability test raised the four problems whichthey encountered while using Construction Director. These arelisted in Table 5. Further improvements to the 4D tool were basedon the participants’ recommendations.

After usability testing, we analyzed the participants’ operationalperformance and their basic conceptual understanding of the 4Dtools. The following six observations were made:

1. All participants were able to learn to use the basic functions ofConstruction Director after watching the videos and could com-plete the integrity examination.

2. The majority of the participants’ performances in general andanimation operations using Construction Director clearly sup-ported our hypothesis that people with more experience incomputer software found basic operations in the 4D tool easier,and vice versa for users with less experience.

3. Participants with experience in using construction cartographysoftware such as AutoCAD had better performance when con-trolling the 3D model.

4. Participants with experience in using project software, such asP3, were able to perform the operating tasks more easily, andwere skilled enough to complete sophisticated functions usingthe 3D model. Therefore, possessing the ability to operate a3D model greatly influenced the participants’ ability to use Con-struction Director.

5. Operation of the animation function was easy to learn.

Table 5Participant feedback on Construction Director.

Problems encountered Possible areas

When a large number of elements are used, the method of linkagebecomes significantly more complex

After selectingof the elementinterface of Co

The interface for linking work items to schedule was not found in themenu toolbars, causing some confusion among users

The ‘locate’ fuassigned for e

The interface to SmartPlant Review requires manual configuration 1. Content synbe quite difficuthe same mac2. When the inminimize, andchanges

Unable to stop the search function during operation The search actoperation. Howmenus and sufor the searchalso be implem

6. The tutorial curriculum allowed participants even with little tono 3D modeling experience to learn the basic functions of Con-struction Director. The questionnaire results revealed that 87%of the participants believed that the videos were of an appropri-ate length.

The results obtained from the usability test allow us to identifyweaknesses in the software’s usability and locate areas forimprovement. The usability test, which includes the observationtest and the users’ feedback appraisal, can be seen to successfullyovercome deficiencies in the Construction Director program.

The field experiment shows us the result of putting the 4D toolin practice. The problems encountered by users, the questions theyhad raised as well as the subsequent solutions and improvementideas will certainly help us in delivering a better 4D tool to clientsin the future.

6. Stage 3: Management plan

After the completion of the system evaluation and usabilitytests of the SUM framework, we integrated the 4D tool withinthe workflow of an actual project. We adopted the following proce-dures in the process.

6.1. Interviews

In this part of the research, we conducted interviews with rele-vant engineering personnel of different departments within CTCIwith a view to evaluating the effectiveness of introducing the 4Dtool to the business workflow of the company. The interviews weredivided into two phases: (1) survey the existing workflow and (2)define the workflow for implementing the 4D tool.

During the first phase, we arranged a half-day interview withthe manager of the design department, the purpose of which wasto obtain an overview of the existing workflow of a typical de-sign–build project. We aimed to define the working items, the per-sonnel involved and the relevant departments affected duringdifferent operational phases of the project. As the manager hadbeen involved in the development of Construction Director, wewere able to obtain very relevant information regarding the newworkflow. We identified that the redistribution of responsibilitiesamong the departments after introducing the 4D tool into theworkflow was one of the most important areas of focus. Based onthis information, we drafted two charts illustrating the existingworkflow and the workflow for implementing the 4D tool. After

of improvement

the 3D model element, the program can be modified to provide the item functionand the schedule. In this case, users do not need to find the item functions in thenstruction Director by themselves

nction could be displayed in the work items toolbar, and a keyboard shortcutase of selectionchronization with SmartPlant Review occurs after every content change, but canlt especially if the two programs (SPR and Construction Director) are running on

hine. Two monitors are recommended to avoid this overlay problemterface completes its update; Construction Director should automaticallySmartPlant Review should be displayed on screen to view the effects of the

ion is operated by SmartPlant Review and is unable to be stopped duringever, this action can be stopped by establishing search results by controlling

b-menus of computer commands. Similarly, an independent window can be usedfunction so users can continue working on the main display. A stop button should

ented to stop the list action

484 M.-H. Tsai et al. / Advanced Engineering Informatics 24 (2010) 476–489

the charts were completed, we continued to the second phase ofthe interview.

The second phase of the interview consisted of two parts. Thefirst part was to interview the manager of the project managementdepartment and to ask for his review of the workflows that wedrafted in the previous phase. We asked him to confirm the rolesand responsibilities of his department in each workflow. The sec-ond interview of this phase was held with three people: the man-ager, a senior engineer, and the human resource engineer of theconstruction department. During the interview, we asked theinterviewees to review the figures of the workflows and to pointout any errors. We also asked them to focus on their relevant roleswithin the workflow, particularly communication with sub-con-tractors, any inter-departmental interactions, and any other possi-ble issues during the construction processes.

After this phase of the interview, we modified the workflowsagain, taking into consideration the feedback we had receivedand carefully examined the possible impact of the 4D tool’s intro-duction into their business processes before developing the finalversion of the defined workflow.

6.2. Symbols used in the workflow

The symbols used in the workflow figures are explained in Table 6below. The arrow represents the flow direction of the data/opera-tion. Two kinds of arrows are used: the solid arrowhead representsa one-to-many relationship, whereas the hollow arrowhead repre-sents a zero relationship between the work items. We also use twodashed arrowheads to represent the data exchange in either the for-ward or backward operations. They represent the interaction,communication, negotiation and understanding among operationalactivities. Generally, the application of data exchange would beaccompanied by cooperation blocks to emphasize the parallelrelationship. Work items are represented by rectangular boxes.Shaded rectangular boxes represent work items associated withthe 4D tool (Construction Director). These shaded boxes will onlybe shown in Table 6 for which the 4D tool is introduced. Diamondbox are used to represent flow judgment. Dashed-line roundedrectangles represent a cooperation block. Work items in thecooperation block represent interaction, communication, negotia-tion, and understanding between the departments. Square-edgeddash line rectangles represent a recursive operation. Activities inthe recursive block need to be performed repeatedly until the goalsare achieved.

Table 6Work item diagrams.

Diagram Representation Diagr

Data flow/operation (one-to-many)

Data flow/operation (zero)

Date exchange/operations both way

Work item

6.3. Mapping the existing workflow

We mapped the existing workflow for CTCI using the symbolsexplained above as shown in Fig. 9. Firstly, CTCI’s client announcesthe initial requirements of the design–build project. These require-ments include the client’s budgetary limit, time allowance, objec-tives, expectations, limitations, and regulatory specifications.CTCI, as the contracting company, confirms the client’s expecta-tions and begins to execute the project. They would do so by ini-tially surveying the project site verifying access points, terrain,site conditions, any environmentally sensitive issues and other rel-evant matters. CTCI then starts developing a project plan and pre-pares deliverables in accordance with the client’s demands andrequirements, before finally submitting it for client approval. Usu-ally the client reviews the project plan proposed by CTCI and thenasks CTCI to explain and redefine it several times until they are sat-isfied with their criteria and expectations. Upon achieving anagreed plan, the client then awards the final contract and CTCIcommences construction of the design–build project.

After obtaining the contract, engineers of the design and archi-tecture departments at CTCI begin to create a 3D model accordingto the agreed plan, and, using this model, they iteratively conductstructural analysis and parameter studies. Fig. 9 shows this processpresented by a recursive block. When any problem occurs duringthe model building process, the engineers are able to modify themodel directly until it achieves the desired goal from a construc-tion point of view and at the same time, complies with all therequirements as set out by the client. Finally, a set of constructiondrawings are prepared to support the work items and activities inthe design–build project.

During the design and construction phases, it is standard forCTCI to pass through three major milestones during their existingworkflow process. Horizontal communication between the clientand the project department is maintained at all times during thisphase. Each milestone includes an inter-departmental meetingand a client review. The first milestone occurs upon the completionof the detailed design, the second is to settle the procurement ofmajor materials and to subcontract out major work items, andthe third is to finalize all the details of the construction phase.All three milestones require involvement of the engineering per-sonnel from different departments (design, planning, procurement,construction) so that the construction management issues can bedealt with effectively and resolved during the construction phase.

Both the meetings and reviews rely heavily upon the 3D modeldeveloped by the design department. The 3D model is usually dis-

am Representation

Work item related to Construction Director (4D tool)

Judgment

Cooperation block

Recursive block

Client Project

department Design & Architecture

department Schedule control

department

Purchase management department

Construction department

Contractor

OK

The first inter-department meeting at CTCI based on 3D model

The third inter-department meeting at CTCI based on 3D model

The second inter-department meeting at CTCI based on 3D model

Revisio

Revisio

Analyze the 3D model

Analyze the

structure and

design of the 3D model

Review the construction

plan

The first model confirmation

Provide the construction plan and the 3D model

OK

NO

Examine the 3D model

Review the construction

plan

The second model fi ti

Provide the construction plan and the 3D model

Control quality and progress according to the process of

constructs

Report the construction progress to

client Report

Comment

Check purchases

conform with laws and

regulations

Purchase according to construction

drawing

Perform construction according to the 3D model

Manufactures plan in the construction

firm

Manufacture plan of

construction

Project completion

Compute construction

quantity

Finalize the construction

drawings

Revisio

The third model confirmation

Provide the construction plan and the 3D model

Review the construction

plan

Propose requirement

Review the plan

Develop a plan

Examine the progress and quality

Plans of disposition in the construction firm

Preliminaryactivities

Contract

Createthe 3D model

Reviewcontractor’s

plan

Propose construction

plan

NG

Fig. 9. The existing workflow of a design–build project.

M.-H. Tsai et al. / Advanced Engineering Informatics 24 (2010) 476–489 485

played on a large screen, enabling engineers from the differentdepartments to review and explain their parts within the 3D mod-el. Elements of the project such as the construction plan, construc-tion schedule, cost, budget, and approximate site planning and

manpower deployment are each reviewed during these meetings.The project department at CTCI usually plays the major role ofgathering relevant information from the different departmentsand resolving any conflicts among them. The client’s review is also

486 M.-H. Tsai et al. / Advanced Engineering Informatics 24 (2010) 476–489

based on the 3D model, and review of the 3D model by the client isrequired before a final decision and further progress can be made.Simultaneously, the construction schedule and any other relevantissues need to be explained to the client. Since most clients donot possess a strong background in construction, they usually hirein-house construction professionals to help them understand theproject plan. These construction activities are represented by acooperation block. The client and the project management teammonitor and examine the quality and progress of the construction,based on the 3D model and the construction plan developed duringthe design phase. The various departments, including schedulecontrol, procurement, construction and the contractors and sub-contractors are then required to perform construction activitiesaccording to the 3D model and the construction plan. The work-flow is depicted below in Fig. 9.

6.4. Defined workflow for implementing the 4D tool

After conducting interviews with all the engineering personnelconcerned, we eventually presented a customized workflow thatincorporated the 4D tool (Construction Director) into the existingworkflow of CTCI. The defined workflow is presented in Fig. 10 be-low. We tried to minimize the number of changes to the existingworkflow and presented the new workflow in a more realisticand rational form. This was to reduce the chance of unforeseen im-pact on the organizational structure and on the processes to whichthe organization’s engineers were already accustomed. Unlikeusing the 3D model for the sole purpose of reviewing the construc-tion plan, the engineers will use Construction Director as the majorpoint of reference in the defined workflow during their inter-departmental meetings. They can also use 4D models to explainand present the plan, design and results, showing the project’s pro-gress to the clients using the 4D tool.

6.5. Comparison of the existing workflow with the defined workflow

We make two comparisons between the existing workflow andthe defined workflow. The first is to compare the use of 3D and 4Dmodels for internal communication and the second is to compareexternal communication between third parties and CTCI. Table 7highlights the major differences between the existing workflow(with 3D models) and the defined workflow (with 4D models).

In the existing workflow, engineers maintain the 3D model intheir inter-departmental conference and project review meetings.Their means of correspondence in the 3D model and the projectschedule were kept in a separate status to devise the project plan.This presented a problem when engineers and the client had tounderstand the optimal productivity of the project prior to con-struction. Moreover, when the different departments present theirindividually assigned project issue solutions at project coordina-tion meetings, they should be able to transfer that information toboth the 3D model and the project schedule. The current systemlacks providing such a means for communication and coordination,especially for a large construction project.

In the defined workflow, we introduced the 4D constructionmanagement tool to resolve the uncertainties and ambiguities ofthe project for the clients. With the introduction of the 4D tool,the site’s physical constraints, as well as the conflicts of concurrentactivities can be displayed visually and identified to avoid conflictsof operations and tasks. The schedule controller or the planningengineer can also synchronize the project progress with the taskcompletions of the construction, giving immediate feedback onthe reorganization of the model. In addition, the 4D model allowsengineers to easily understand the relationships between projecttasks as they have a clear view of the interval relationship betweenthe project progress and planning phases, allowing them to make

arrangements for procurement, equipment and manpower wellin advance.

6.6. Major advantages of the defined workflow

There are four major advantages that the defined workflow hasover the existing workflow. Firstly, the 4D-modeling tool can beutilized effectively during the inter-departmental meetings atCTCI. The models created combine the 3D models with the con-struction schedule through the combined efforts of the Designdepartment and Schedule Control department. As the 4D modelis able to display the construction progress, engineers from differ-ent departments can better understand the temporal relationshipbetween the work items and the construction activities. The useof 4D models can successfully reduce the amount of time spenton explanation and clarification of construction issues during themeetings. This is crucial because earlier research has shown that30% of non-productive time on-site is due to a lack of detailedand space planning [47]. The 4D tool will not only increase meetingproductivity but also improve the quality and success of the deci-sion-making process.

Secondly, the 4D tool can be used for client review. For a de-sign–build project, the contractor is the only person who is ableto communicate directly with the client to facilitate project mat-ters in terms of information and progress. Using the 4D model,the client can understand the design details and construction pro-cesses even if they do not have a strong construction background.This will assist them in locating problems and will help them inconfidently making decisions.

Thirdly, this intelligent tool can be very easily incorporated foruse on construction sites. For a design–build project, one majorconcern is the method of communication amongst sub-contractors,especially for those who work in close proximity to schedulingconflicts. In addition, the main contractor has to communicateand liaise with many specialized sub-contractors, suppliers, andregulating agencies with regard to execution of construction activ-ity, material storage, access points, transport, permits, and safetyissues. The 4D model integrates the information of a 3D model aswell as the construction schedule, providing engineers with a moredirect approach to obtaining the information they require to tracethe predecessors and successors of each construction activity andto avoid conflicts of schedules and tasks.

Finally, the 4D model provides a better method of communica-tion compared to the conventional scheduling tools (such as barchart, CPM, etc.). The 4D model also helps to identify accessibilityissues for movable equipment such as trucks, excavators, concretepumps, mobile cranes, and gantry cranes, as well as the arrange-ment of stationary equipment such as tower cranes, pump stations,and boring machines. With the help of this visualization tool, theengineering manager and site engineers can easily gather informa-tion and make decisions relating to site matters with greater con-fidence. The 4D model can also visually convey the information ofboth 3D models and corresponding schedules in such a way that itis able to improve the communication between the parties in-volved and positively enhance the performance of the entireproject.

7. Validation

The SUM framework’s success in being implemented withinCTCI verifies its validity. This reasons for success lie in the readi-ness of CTCI and other departments to embrace and implementSUM’s defined workflow in place of their own workflow. This hasresulted in CTCI entering into more contracts, one of which beinga design and construct contract of the Dalin Refinery project in

Client Project department

Design & Architecture department

Schedule control department

Purchase management department

Construction department Contractor

Revisio

Revise 4D model

Revise 4D model

Create 4D model

The third inter-department meeting at CTCI based on 4D model

The third model confirmation

Provide the construction plan and the 4D model

Review the construction

plan

The second inter-department meeting at CTCI based on 4D model

OK

Propose requirement

Review the plan

Develop a plan

Examine the progress and quality

Plans of disposition in the construction firm

The first inter-department meeting at CTCI based on 4D model

Preliminary activities

Schedule plan

Contract

Createthe 3D model

Analyze the

structure and

design of the 3D model

Finalize the construction

drawings

Analyze the 3D model

The first model confirmation

Compute construction

quantity

Provide the construction plan and the 4D model

Checkpurchases

conform with laws and

regulations

Manufacture plan of

construction firm

Review the construction

plan

The second model fi ti

Provide the construction plan and the 4D model

Review the construction

plan Purchase

according to construction

drawing OK

NO

OKNO

Manufactures plan in the construction firm and 4D model

Perform construction according to the 4D model

Comment

Report

Project completion

Construction requirement

Contract

Construction plan

Examine the 3D model

Reviewcontractor’s

plan

Revisio

Revisio

Report the construction progress to

client

Control quality and progress according to the process of constructs and 4D model

Fig. 10. The defined workflow for implementing the 4D tool in a design–build project.

M.-H. Tsai et al. / Advanced Engineering Informatics 24 (2010) 476–489 487

Kaohsiung City, Taiwan. The negotiated contract is worth approxi-mately US$650 million; construction had started during October

2008 and is scheduled for completion in October 2011. Anotheris a design and construct contract for Linyuan Petrochemical Plant

Table 7Existing Workflow and Defined Workflow Comparison.

Workflow Internal communication External communication

Existingworkflow(3D)

1. Engineers maintain the 3D model2. Presenting both the 3D model and schedule separately requires longer time to understand3. Communication and coordination for teamwork is difficult4. Unable to use the one 3D model across different departments to assist understanding of the

construction progress5. The construction firm can’t effectively plan for equipment, manpower etc. before construc-

tion begins

1. Client does not understand and has little knowledgeabout the relationship between schedule andprogress

2. Information gaps are created in several phases3. Communication gaps may also be created

Definedworkflow(4D)

1. Engineers maintain the 4D model2. Managing the schedule with the 4D model is a lot less complex3. Because of the clear presentation of the relationship between progress and schedule, users

can make arrangement of space, procurement, equipment and manpower etc. before con-struction begins

4. Presenting an individual department’s issue is easier, thus facilitating better communica-tion and coordination

1. Report the project plan with 4D model2. Clients can understand related issues of the project3. Information conveyed visually is easier to understand

488 M.-H. Tsai et al. / Advanced Engineering Informatics 24 (2010) 476–489

project in Kaohsiung County, Taiwan. The negotiated contract isworth approximately US$710 million; construction started duringAugust 2009 and is scheduled for completion in December 2010.

As 4D models are more capable of displaying the constructionprogress, engineers from different departments can better under-stand the temporal relationship between the work items and theconstruction activities. CTCI found that using the 4D model is ableto aid them in brainstorming suggestions regarding specific use ofmaterials or equipment to specific work from quality controllers orconsultants earlier on.

The 4D model was also used to indicate to the client the currentstate of construction progress and show the quality of the workdone. This allows CTCI to lift client satisfaction. The experiencegained by CTCI in using the 4D tool successfully via the SUM frame-work will allow for easier application in the future. Currently, inApril 2010, CTCI is writing up a proposal to win a contract forthe Petrochemical Plant project in United Arab Emirates.

8. Benefits of this research

The three-stage SUM framework proposed in this research hasbeen validated by introducing it at CTCI. From this experience, fourmajor benefits have surfaced:

(1) Knowing the limitation of 4D tools: In a large project, themodel typically includes a large number of elements. Know-ing the capacity of the software and hardware and their per-formance capabilities in advance is important in allowingengineers to determine the level of detail and maximumnumber of sub-models that can be used.

(2) Understanding the capacity of engineers: Usability problemscan be effectively identified via carefully designed user tests.By recruiting a small number of current and potential userswith different backgrounds, we were able to identify andeliminate usability issues before introducing the 4D tool toactual projects.

(3) Identifying viable workflow: The workflow amongst depart-ments changes upon the introduction of the 4D tool. Recur-sive interviews and discussions with key personnel arenecessary to reengineer a well-rounded workflow, whichcan be effective in reducing the possible difficulties of intro-ducing the 4D tool to a large organization.

(4) A systematical approach to solve problems: The SUM frame-work provides guidance to consulting firms who would liketo introduce 4D tools into their organization. From our expe-riences of applying the SUM framework in a consulting firm,we found that the SUM framework provides an effective andsystematical approach in evaluating the major issues sur-rounding tools, users, and management.

9. Conclusions

This paper presents the SUM framework, a consulting frame-work designed to assist with the introduction of a 4D tool toconsulting firms that have large organizational structures andwell-established workflows. The SUM framework includes threestages: system evaluation, usability test and management plan.To validate the proposed framework, we worked with an industrialpartner to apply the SUM framework in practice. The results showthat this framework can effectively identify the problems encoun-tered during the introduction of a 4D tool, including problems insystem limitation, engineers’ capacities, and workflow changes.In future, the SUM framework will provide guidance as well asan evaluative approach to companies who wish to introduce 4Dtools to their organization.

Acknowledgements

We would like to thank members of the CTCI Corp. for theirgenerous participation and feedback in this research. We also wantto thank Mr. Yuan-Fu Liao of CTCI Corporation, Prof. Chuin-ShanChen and Mr. Cheng-Han Kuo of National Taiwan University fortheir assistance in the development of the 4D construction man-agement tool.

References

[1] S.H. Hsieh, C.S. Chen, Y.F. Liao, C.T. Yang, I.C. Wu, Construction Director: 4Dsimulation system for plant construction, in: Proceedings of the EASEC-10,Bangkok, Thailand, August 3–5, 2006.

[2] K. McKinney, M. Fischer, Generating, evaluating and visualizing constructionschedules with CAD tools, Automation in Construction 7 (1998) 433–447.

[3] C. Kam, M. Fischer, R. Hänninen, A. Karjalainen, J. Laitinen, The product modeland fourth dimension project, ITcon 8 (2003) 137–166.

[4] B. Akinci, K. Tantisevi, E. Ergen, Assessment of the capabilities of a commercial4D CAD system to visualize equipment space requirements on constructionsites, in: Proceedings of the CRC, Honolulu, Hawaii, March 19–21, 2003.

[5] Z. Ma, Q. Shen, J. Zhang, Application of 4D for dynamic site layout andmanagement of construction projects, Automation in Construction 14 (2005)369–381.

[6] N. Dawood, D. Scott, E. Sriprasert, Z. Mallasi, The virtual construction site(VIRCON) tools: an industrial evaluation, ITcon 10 (2005) 43–54.

[7] S. Staub-French, A. Khanzode, 3D and 4D modeling for design and constructioncoordination: issues and lessons learned, ITcon 12 (2007) 381–407.

[8] L.M. Sheppard, Virtual building for construction projects, IEEE ComputerGraphics and Applications 24 (1) (2004) 6–12.

[9] K. McKinney, J. Kim, M. Fischer, C. Howard, Interactive 4D-CAD, in: Proceedingsof the 3rd Congress on Computing in Civil Engineering, Anaheim, CA, USA, June17–19, 1996.

[10] E. Collier, M. Fischer, Visual-based scheduling: 4D modeling on the San Mateocounty health center, in: Proceedings of the 3rd Congress on Computing inCivil Engineering, Anaheim, CA, June 17–19, 1996.

[11] J.P. Zhang, Z.Y. Ma, P.U. Cheng, 4D visualization of construction sitemanagement, in: Proceedings of the 5th International Conference onInformation Visualisation, London, UK, July 25–27, 2001.

M.-H. Tsai et al. / Advanced Engineering Informatics 24 (2010) 476–489 489

[12] Ma Zhaoyang, Q. Shen, J. Zhang, Application of 4D dynamic site layout andmanagement of construction projects, Automation in Construction 14 (2005)369–389.

[13] W. Hu, X. He, J.H. Kang, From 3D to 4D visualization in building construction,in: Proceedings of the Computing in Civil Engineering, Cancun, Mexico, July12–15, 2005.

[14] A.A. Ganah, N.B. Bouchlaghem, C.J. Anumba, VISCON: computer visualizationsupport for constructability, ITcon 10 (2005) 69–83.

[15] J.H. Kang, S.D. Anderson, M.J. Clayton, Empirical study on the merit of web-based 4D visualization in collaborative construction planning and scheduling,Journal of Construction Engineering and Management 133 (6) (2007) 447–461.

[16] K. McKinney, M. Fischer, J. Kunz, Visualization of construction planninginformation, in: Proceedings of the International Conference on IntelligentUser Interface, New York, January 6–9, 1998.

[17] P.N. Johnson-Laird, P. Legrenzi, M. Sonino-Legrenzi, Reasoning and a sense ofreality, British Journal of Psychology 63 (3) (1972) 395–400.

[18] K.W. Chan, M. Anson, J.P. Zhang, Four-dimensional visualization ofconstruction scheduling and site utilization, Journal of ConstructionEngineering and Management 130 (4) (2004) 598–606.

[19] S. Staub-French, A. Russell, N. Tran, Linear scheduling and 4D visualization,Journal of Computing in Civil Engineering 22 (3) (2008) 192–205.

[20] E. Sriprasert, N. Dawood, Requirements identification for 4D constraint basedconstruction planning and control system, in: Proceedings of CIB w78Conference, Aarhus School of Architecture, June 12–14, 2002.

[21] B. Koo, M. Fischer, Feasibility study of 4D CAD in commercial construction,Journal of Construction Engineering and Management 126 (4) (2000)251–260.

[22] L.S. Kang, S.B. Jee, C.H. Kim, 4D system for visualizing schedule progress ofhorizontal construction project including earthwork, in: Proceedings of 6thCONVR, Orlando, Florida, USA, August 3–4, 2006.

[23] G. Boloix, System evaluation and quality improvement, Journal of Systems andSoftware 36 (3) (1997) 297–311.

[24] G. Susman, Action research: a sociotechnical systems perspective, in: G.Morgan (Ed.), Beyond Method: Strategies for Social Research, Sage, NewburyPark, 1983, pp. 95–113.

[25] H.N. Ahuja, S.P. Dozzi, S.M. AbouRizk, Project Management: Techniques inPlanning and Controlling Construction Projects, J. Wiley, New York, 1994.

[26] C.H. Caldas, L. Soibelman, Automating hierarchical document classification forconstruction management information systems, Automation in Construction12 (2003) 395–407.

[27] C.C. Lu, S.C. Kang, S.H. Hseih, R.H. Shiu, Improvement of a computer-basedsurveyor-training tool using a user-centered approach, Advanced EngineeringInformatics 23 (2009) 81–92.

[28] C.H. Chen, K. Ssto, K.P. Lee, Editorial: human-centered product design anddevelopment, Advanced Engineering Informatics 23 (2009) 140–141.

[29] A. Russell, S. Staub-French, N. Tran, W. Wong, Visualizing high-rise buildingconstruction strategies using linear scheduling and 4D CAD, Automation inConstruction 18 (2009) 219–236.

[30] S. Casera, P. Kropf, Collaboration in scientific visualization, AdvancedEngineering Informatics 24 (2010) 188–195.

[31] H.S. Chang, S.C. Kang, P.H. Chen, Systematic procedure of determining an idealcolor scheme on 4D models, Advanced Engineering Informatics 23 (2009) 463–473.

[32] K.W. Chau, M. Anson, D.D. De Saram, 4D dynamic construction managementand visualization software: 2. Site trial, Automation in Construction 14 (4)(2005) 525–536.

[33] D. Borenstein, Towards a practical method to validate decision supportsystems, Decision Support Systems 23 (3) (1998) 227–239.

[34] P.J. Stappers, H.V. Rijn, S.C. Kistemaker, A.E. Hennink, F. Sleeswijk Visser,Designing for other people’s strengths and motivations: three cases usingcontext, visions, and experiential prototypes, Advanced EngineeringInformatics 23 (2009) 174–183.

[35] Y.T. Chong, C.H. Chen, K.F. Leong, Human-centric product conceptualizationusing a design space framework, Advanced Engineering Informatics 23 (2)(2009) 149–156.

[36] P. Senkul, I.H. Toroslu, An architecture for workflow scheduling under resourceallocation constraints, Information Systems 30 (5) (2005) 399–422.

[37] J.D. Ford, L.W. Ford, A. D’Amelio, Resistance to change: the rest of the story,Academy of Management Review 33 (2) (2008) 362.

[38] T. Hartmann, M. Fischer, A process view on end-user resistance duringconstruction IT implementations, ITcon 14 (2009) 353–365.

[39] C.S. Dossick, G. Neff, Organizational divisions in BIM-enabled commercialconstruction, Journal of Construction Engineering and Management 136 (4)(2009) 459–467.

[40] A. Hickey, A. Davis, A unified model of requirements elicitation, Journal ofManagement Information Systems 20 (4) (2004) 65–84.

[41] F.M. Santoro, M.R.S. Borges, J.A. Pino, Acquiring knowledge on businessprocesses from stakeholders’ stories, Advanced Engineering Informatics 24(2010) 138–148.

[42] T. Hartmann, M. Fischer, Supporting the constructability review with 3D/4Dmodels, Building Research and Information 35 (1) (2007) 70–80.

[43] T. Hartmann, J. Gao, M. Fischer, Areas of application for 3D and 4D models onconstruction projects, Journal of Construction Engineering and Management134 (10) (2008) 776–785.

[44] K.W. Chau, M. Anson, J.P. Zhang, Four-dimensional visualization ofconstruction scheduling and site utilization, Journal of ConstructionEngineering and Management 130 (4) (2004) 598–606.

[45] E. Gamma, R. Helm, R. Johnson, J. Vlissides, Design Patterns, Addison Wesley,1998.

[46] M.H. Tsai, S.C. Kang, S.H. Hsieh, C.H. Kuo, T.G. Pei, K.C. Yeh, Experiences onstudy of setting-up a 4D construction management tool in a high-riseconstruction project, in: Proceedings of 12th ICCCBE XII & INCITE 2008,Beijing, China, October 16–18, 2008.

[47] B. Kanagasabapathi, K. Ananthanarayanan, Implementation of 4D visualizationas a planning tool in the Indian AEC industry, Journal of the Institution ofEngineers (India): Architectural Engineering Division 85 (2004) 35–40.