A practical approach to project scheduling: considering the potential quality

loss cost in the timecost tradeoff problem

JongYul Kim
a
, ChangWook Kang
b,

, InKeuk Hwang
c
a
Department of Industrial Engineering, Graduate School, Hanyang University, Seoul 133-791, Republic of Korea
b
Department of Industrial and Management Engineering, Hanyang University, Gyeonggi-do 426-791, Republic of Korea
c
Department of Industrial and Systems Engineering, Kongju National University, Chungcheongnam-do 330-717, Republic of Korea
Received 16 April 2010; received in revised form 12 May 2011; accepted 31 May 2011
Abstract
Crashing project activities is a typical way to shorten their completion times to meet project due dates, and previous research on quality in
timecost tradeoff problems focused on maximizing the individual activity quality of projects. However, implementing project scheduling that
takes into account the potential quality loss cost (PQLC) in timecost tradeoff problems is a practical approach, since individual activity quality is
defined by conformance to project contractor requirements. We propose a mixed integer linear programming model that considers the PQLC for
excessive crashing activities. This model will help project planners develop practical project schedules.
Crown Copyright 2011 Published by Elsevier Ltd. APM and IPMA. All rights reserved.
Keywords: Project scheduling; Quality; Potential quality loss cost; Timecost tradeoff; Critical path method
1. Introduction
Successful projects should be completed before project due
dates and within budget; however, these limits are sometimes
surpassed. There may therefore be significant variance between
the assumptions made regarding a project and actual outcomes.
Sudden unexpected changes in construction technology, tech-
niques, materials, or human resources can create budgetary and
scheduling pressures that in turn may increase the possibility of
failure (Zeng et al., 2007). A survey exploring the completion of
construction projects in Saudi Arabia showed that 76% of project
contractors experienced delays of 1030% of the projected
duration (Assaf and Al-Hejji, 2006).
A typical technique used to mitigate scheduling pressure is to
crash project activities. Crashing activities involves allocating
more resources (such as materials, labor, and equipment) than
planned in order to complete a project more quickly (Kessler and
Chakrabarti, 1999).
In timecost tradeoff problems, projects are not always
completed as scheduled without reworking or modification. A
project is a one-time task constrained by time, cost, and quality,
and its success depends on howwell these constraints are balanced
(Atkinson, 1999). If any one of the constraints is overemphasized,
burdens may fall on the other two. Hence, crashing project
activities should be considered a significant factor in the timecost
tradeoff problem.
Some previous studies have treated quality as an important
factor in tradeoff problems, claiming that overall project quality
attained by project activities should be maximized within a given
deadline and budget. These studies have also promoted using a
continuous scale from zero to one to specify the quality of each
activity (Babu and Suresh, 1996; Tareghian and Taheri, 2006,
2007).
However, a study evaluating the application of the timecost
quality tradeoff model to linear programming for a cement factory
construction project in Thailand revealed two facts: overall
project quality cannot be sacrificed by crashing, and individual
activity quality is primarily determined by subjective judgements,

Corresponding author. Tel.: +82 31 400 5264; fax: +82 31 409 2423.
E-mail addresses: jykimle@gmail.com (J. Kim), cwkang57@hanyang.ac.kr
(C. Kang), ikhwang@kongju.ac.kr (I. Hwang).
0263-7863/$ - see front matter. Crown Copyright 2011 Published by Elsevier Ltd. APM and IPMA. All rights reserved.
doi:10.1016/j.ijproman.2011.05.004
Available online at www.sciencedirect.com
International Journal of Project Management 30 (2012) 264272
www.elsevier.com/locate/ijproman
with the exception of a few measurable activities (Khang and
Myint, 1999). In real situations, even if overall project quality
meets project quality targets, if any single project activities do not
meet the project contractor's requirements, rework or modifica-
tion may be necessary and are associated with time delay and cost
overrun. The possibility of rework or modifications must be
considered when crashing project activities to develop practical
and cost effective project schedules.
This paper proposes a mixed integer linear programming
model and procedure that accounts for potential quality loss cost
(PQLC) associated with rework or modifications that may occur
due to excessive crashing activities. The rest of the paper is
organized as follows. Section 2 summarizes previous research
related to timecost tradeoff problems considering project
quality. Section 3 describes the mixed integer linear programming
model used to compute direct project costs, which was previously
considered to be equal to the nonconformance activity rate of the
project activities. Section 4 validates this model with an example
and discusses the PQLC estimation method of the example
project. Section 5 provides the conclusions of the study.
2. Literature review
The critical path method (CPM) that is used for all types of
projects, such as construction, engineering, facility maintenance,
software development, and research and development is a
mathematical algorithm used to schedule a set of activities in a
project. This method is fundamentally related to the tradeoff
between completion time and the costs of the project (Kelley and
Walker, 1959), and is suitable for application to deterministic
conditions rather than probabilistic conditions. The CPM can be
used to determine the timecost tradeoff for activities that meet
given completion times at minimumcost, and is useful when there
are similar experiences from previous projects (Hillier and
Lieberman, 1995).
Timecost tradeoff problems from the late 1950s mostly
concentrated on shortening overall project duration by crashing
the time required to complete individual activities. Researches
in this area include linear programming models (Elmaghraby
and Salem, 1982; Goyal, 1975; Kelley and Walker, 1959; Kelly,
1961; Perera, 1980; Phillips and Dessouky, 1977; Siemens,
1971) and nonlinear programming models (Deckro et al., 1987,
1995; Fulkerson, 1961; Meyer and Shaffer, 1963; Patterson and
Huber, 1974). Under the assumption that time and cost tradeoffs
for individual activities are linear, the relationship can be
represented as a straight line on a graph depicting the relationship
between activity time and cost (Wiest and Levy, 1997). The cost
of completing the activity varies linearly between the normal time
and the crash time (Fulkerson, 1961).
If there is concern over quality degradation then crashing
project activities is not desirable, and more time should be
allowed to finish the project (Deckro et al., 1995; Vrat and
Kriengkrairut, 1986). In such cases, preventive actions should be
taken to avoid rework or modifications that might occur during
project execution.
Time, cost, and contractor requirements for project manage-
ment are significant elements for judging the successes of
information systems and technology projects (Wateridge, 1998).
Once a project has been completed, the time and cost tradeoff
problem is no longer an issue for the project manager, and quality
or performance becomes key issues (Avots, 1984). The earned
quality method assists project managers in detecting the quality
variance of project activities, and allows them to take early
corrective actions by comparing actual quality with planned
quality (Paquin et al., 2000). Project quality is a consequence of
the accumulated contributions of all individual activities executed
during a project's life cycle.
If the outcome of a project meets or exceeds the project
contractor's expectations, the project is deemed successful (Martin
and Tate, 2001). The project contractor gives priority to the
availability of the outcome in the longer-termperspective, because
the project must be profitable. Simply completing the project by
the given due date and within budget is not sufficient, because the
work must also be of acceptable quality.
Previous research indicates that the quality of project
scheduling is not only more important than other factors such
as time and cost, but also that it is significant for defining project
success. Contractor satisfaction is necessary for success, since the
project outcome is transferred to the contractor (Icmeli-Tukel and
Rom, 1997).
Linear programming models that simultaneously consider
time, cost, and quality were proposed in a previous study (Babu
and Suresh, 1996), who explored five different lower bounded
constraints for project quality. In an binary integer programming
model and the meta-heuristic solution procedure that solves
discrete time, cost, and quality tradeoff problems (Tareghian and
Taheri, 2006, 2007), the quality level of 0.750.99 classified the
quality of each project activity. These quality level classifications
are theoretically significant, but are inapplicable to real problems,
since project contractors do not accept quality degradation.
Hence, project planners require mathematical models that are
applicable to real problems related to project duration crashing.
3. Proposed mathematical model
Project completion time and cost are affected by the crashing of
individual activities. If individual activities are excessively
crashed, rework, modifications, or even project failure may
occur. Quality checks must be performed immediately after the
completion of each individual activity, and corrective actions such
as rework or modification can be taken if the quality is not
acceptable. The PQLC is needed to execute such corrective
actions.
3.1. Problem description
Project costs are generally classified into two categories: the
direct costs related to individual activities and the indirect costs
related to overhead items. The problem we explore in this paper
focuses on individual activities under the assumption that the
timecost tradeoffs for project activities are linear (Swink et al.,
2006). The direct costs, considering the PQLC, are minimized
under the following assumptions:
265 J. Kim et al. / International Journal of Project Management 30 (2012) 264272
Nonconformance activity requires rework or modification,
and the contractor requirements or specifications are
identified immediately after the activity is completed.
The activity can start immediately after all of the preceding
activities have been completed.
The rework or modification time of the activity is bounded
by crash duration.
The rework or modification cost of the activity is bounded by
the direct costs of crashing.
3.2. Mathematical model formation
The timecost tradeoff problem in project management
originates when activity time can be reduced with some extra
direct cost (Schwindt, 2005). Previous approaches to solving
the project duration problem include mixed integer linear
programming formulations (Wiest, 1963) and linear program-
ming with integer variables (Brucker and Knust, 2006). Below,
we describe a mixed integer linear programming model that
incorporates the PQLC. This PQLC is the estimated direct cost
of rework or modifications related to nonconformance activity
that may occur if excessive crashing of project duration is
required. Each project consists of individual activities, and the
completion time for each individual activity must be crashed in
order to meet project deadlines. When activities with additional
direct costs are crashed, the PQLC for these activities are
considered at the same time. The notations used for this model
are indicated below.
Variables
Y
j
Crash time for the completion of activity j
Z
j
1 if activity j is selected as a nonconformance risk
activity, otherwise 0
Parameters
m
j
Direct cost per unit time for activity j
t
j
Normal time required when activity j is performed
under normal conditions
c
j
Normal direct cost when activity j is performed in the
normal time t
j
t
j
Crash time required to complete activity j by assigning
resources beyond those originally allocated
c
j
Crash direct cost when activity j is completed in the
crash time t
j
E Additional direct cost when activity j is completed in
the crash time t
j
R
j
Reduced time for activity j
X
j
Start time for activity j
X
i
Start time for predecessor activity i
K Normal completion time for predecessor activity i
Y
i
Crash time for predecessor activity i
X
n
Start time for activity n
t
n
Normal completion time for activity n
Y
n
Crash time for activity n
D Due date of the project
N Number of activities in the project
q
j
Potential quality loss cost for activity j
Nonconformance risk activity rate predetermined by
the project manager
k An arbitrarily large number
The model is formulated as:
Min
n
j =1
m
j
Y
j
+
n
j =1
q
j
Z
j
Subject to
Y
j
R
j
for all j 1
X
j
X
i
+ KY
i
2
X
n
+ t
n
Y
n
D 3
Z
j
N 4
kY
j
Z
j
5
Z
j
1; 0 f g
Y
j
0
m
j
=
c

j
c
j
t
j
t

j
:
The above formulation includes five constraints. The first
constraint (1) states that the time for each activity cannot be
reduced by more than its maximum time reduction. The start
time of each activity in the second constraint (2) must be at least
as great as the finish time of all of the immediate predecessors,
because the activity finish time is reduced by the amount of time
that each activity is crashed. The third constraint (3) indicates
that the project must be completed by its due date. The fourth
constraint (4) states that the project manager limits the number
of nonconformance risk activities in the project. The arbitrarily
large number k is given in the fifth constraint (5) to prevent the
equation from becoming binding.
Timecost tradeoff problems in project scheduling have
been researched extensively since the introduction of CPM in
the late 1950s. A number of models and/or procedures for
solving the timecost tradeoff problems have been proposed by
linear programming (LP), nonlinear programming (NP), integer
266 J. Kim et al. / International Journal of Project Management 30 (2012) 264272
programming (IP), dynamic programming (DP), mixed integer
linear programming (MILP) and heuristic algorithms (HA), but
most of the models and/or procedures did not consider activity
quality in the problems. Hence, existing models and/or pro-
cedures without consideration of the activity quality are too
optimistic. Although some researchers have attempted to take into
account the activity quality in the timecost tradeoff problems
since the late 1990s, unfortunately their models and/or procedures
are limited to the theoretical approach of a project. Table 1
indicates the notable differences between existing models and the
proposed model.
As indicated in Table 1, the proposed model and procedure
makes it possible to solve the real problem by considering the
PQLCto be occurred during the execution of the project activities.
Lowering the quality of individual activities in real life projects is
generally unacceptable. The quality of each individual activity
can be classified as either conforming or nonconforming with the
project's requirements or specifications (Summers, 2005). An
activity that is classified as nonconforming requires additional
direct costs for rework or modification. Unlike previous
approaches, the proposed model (in which the PQLC is
considered) will be practical in real life project scheduling.
4. Application of the model
4.1. Critical path determination
To performa validitytest of the model described inSection3.2,
a robot type palletizing system installation (RTPSI) project for Y
Company in South Korea was selected as an example. Suppose
this example project aims to increase the packaging productivity
of feminine care products by installing a RTPSI immediately after
the current manufacturing process. Fig. 1 depicts the layout for the
RTPSI.
The business leader of the feminine care products division of
this company appointed an engineering manager to complete the
Table 1
Comparisons between existing models and the proposed model.
Problem
focus
Base
formulation
Researcher Solution
method
Remarks (difference)
Time
cost
LP Kelly, 1961 LP/network
flow
Present a theoretical base that incorporates sequence information, durations, costs, and crashing
concept for project activities; have no consideration for the project activities' quality.
NP Deckro et al., 1995 Quadratic
program
Develop nonlinear time/cost tradeoff models for solving an example; have no consideration for
project quality.
DP Robinson, 1975 DP Produce optimum allocations for networks of project activities with arbitrary costtime functions;
may not be optimal because the project activities' quality is not considered.
DP Demeulemeester
et al., 1996
DP/B&B Propose the optimal procedures based on DP logic with a seriesparallel network and a branch &
bound (B&B) search tree to solve a discrete timecost tradeoff problem; may not be practical for
solving a real-life problem and have no consideration on project activities' quality.
GA Feng et al., 1997 GA/Pareto
front
Develop a new algorithm and computer program for optimizing construction timecost decisions;
have no consideration for project activities' quality.
GA Wuliang and
Chengen, 2009
GA Present an improved genetic algorithm to solve a multi-mode resource-constrained discrete timecost
tradeoff problem; are applicable to special knowledge intensive projects and do not consider project
activities' quality.
MILP Sakellaropoulos and
Chassiakos, 2004
MILP Provide an optimal project timecost curve and a minimum cost schedule with the parameters
(generalized precedence relationships, activity planning constraints, external activity constraints, and
late penalty/early bonus existence); have no consideration for project activities' quality.
HA Siemens, 1971 Approximation Present the algorithm (approximation method) offers a systematic procedure for expediting project
duration but do not consider project activities' quality.
HA Vanhoucke and
Debels, 2007
Heuristic
search
Develop a new meta-heuristic procedure to provide near-optimal heuristic solutions for different
problems but need to improve this procedure for the discrete time/cost tradeoff problem with time-
switch constraints or net present value maximization; have no consideration for project activities'
quality.
Time
cost
quality
LP Babu and
Suresh, 1996
LP Develop linear programming models to study the interrelationship of three functions (time, cost, and
quality) and adopt the continuous scale to determine each activity quality. This continuous scale at the
activity level is inapplicable for solving real problems.
LP Khang and
Myint, 1999
LP Apply the Babu-Suresh model to a real project and evaluate the applicability of their model; find the
limitations of continuous scale use in the process of activity quality inspection.
IP Tareghian and
Taheri, 2006
IP Develop three inter-related (time, cost and quality) binary integer programming models and specify
project quality between 0.75 and 0.99 in increments of 0.05; may be inapplicable for the quality of a
real project.
HA Tareghian and
Taheri, 2007
IP/scatter
search
Propose a meta-heuristic solution procedure for solving the time, cost, and quality tradeoff problem
and complete a project with maximum quality at a given deadline; may be impractical in a real project
because the maximum quality is ambiguous.
MILP the proposed MILP/PQLC Propose a MILP model and procedure that takes into account the potential quality loss cost in the
timecost tradeoff problem. This MILP and procedure makes it possible to solve the real problem of a
project.
267 J. Kim et al. / International Journal of Project Management 30 (2012) 264272
RTPSI project within 19 days, and assigned a project manager
who was responsible for satisfying the business leader's deadline
requirements while staying within the assigned budget. The
project manager wanted to use crashing activities to shorten the
project duration and maintain nonconformance risk activity rate
for all project activities.
Table 2 presents data related to the timecost tradeoff for
each activity in the RTPSI project. The cost (c
j
or c
j
) of each
activity in Table 2 includes expenses related to labor, materials,
and equipment. The project manager was required to prepare a
practical and cost effective project schedule that would result in
the successful completion of the RTPSI project by the due date.
An activity on node (AON) diagram is useful for represent-
ing precedence relationships among project activities.
Fig. 2 shows an AON diagram in which the RTPSI project is
broken down into 18 individual activities. They are A) surface
leveling, B) electric wiring for the conveyor, C) electric wiring
for the wrapping machine, D) electric wiring for the robot,
E) installing the electrical panel for the conveyor, F) installing
the electrical panel for the wrapping machine, G) installing the
electrical panel for the robot, H) assembling the parts for the
conveyor, I) framing the wrapping machine base, J) assembling
the parts for the robot, K) installing the conveyor, L) installing
the palletizer, M) installing the robot, N) setting up the robot
program, O) testing the robot, P) testing for fabricated systems,
Q) inspecting the engineering, and R) inspecting the process.
Activities B, C, and D in Fig. 2 cannot begin until activity A is
completed. Activities K, L, and O must be completed before
activity P is initiated.
CPM is a useful technique for finding the longest path of
planned activities on the AON diagram, which is necessary in
order to determine the minimum time required for project
completion. There are three distinct paths describing the RTPSI
project: path 1) ABEHKPQR, path 2) ACFIL
PQR, and path 3) ADGJMNOPQR.
Each path length indicated in Table 3 can be obtained by
estimating the time t
j
in Table 2 for each respective node in the
RTPSI project network. Of the three paths shown in Table 3,
path 3 is the critical path that requires the longest time for the
completion of the project.
The activities along critical path 3 are used to determine the
completion time for the RTPSI project and have zero flexibility.
The entire project duration may be reduced by crashing the time
required to complete one or more of the activities in path 3.
4.2. PQLC estimation
A systematic process for PQLC estimation that includes
procedures, scale, and definition is divided into three steps:
nonconformance risk identification and coding for project
activities, nonconformance risk analysis for project activities,
and PQLC estimation with nonconformance risk activity rate.
4.2.1. Step 1: Nonconformance risk identification and coding
for project activities
A nonconformance risk is defined as any uncertainty that
would negatively affect project activity cost if it occurs after an
activity is finished. Three nonconformance risks for each activity
Fig. 1. The RTPSI layout.
Table 2
Data related to the timecost tradeoffs for individual activities.
Activity Immediate
predecessor
t
j
,
days
c
j
,
$100
t
j
,
day
(s)
c
j
,
$100
R
j
,
day
(s)
E,
$100
m
j
,
$100/
day
A 3.0 17.88 2.0 20.11 1.0 2.23 2.23
B A 3.0 41.71 1.5 44.32 1.5 2.61 1.74
C A 4.0 69.41 2.0 89.37 2.0 19.96 9.98
D A 4.0 65.96 2.0 84.54 2.0 18.58 9.29
E B 3.0 138.10 1.5 170.06 1.5 31.96 21.31
F C 2.0 327.81 1.5 431.13 0.5 103.32 206.64
G D 3.0 189.99 1.5 210.22 1.5 20.23 13.49
H E 1.5 28.14 1.0 38.29 0.5 10.15 20.30
I F 2.0 16.83 1.0 22.08 1.0 5.25 5.25
J G 1.5 79.86 1.0 92.55 0.5 12.69 25.38
K H 6.0 123.86 3.0 167.37 3.0 43.51 14.50
L I 2.0 62.62 1.5 74.91 0.5 12.29 24.58
M J 4.0 625.93 2.0 655.86 2.0 29.93 14.97
N M 3.0 51.97 1.0 71.85 2.0 19.88 9.94
O N 2.0 14.90 1.5 20.11 0.5 5.21 10.42
P K, L, O 2.0 35.75 1.0 43.94 1.0 8.19 8.19
Q P 2.0 59.89 1.5 76.94 0.5 17.05 34.10
R Q 1.5 40.61 1.0 51.34 0.5 10.73 21.46
Total 1991.22 2364.99 373.77
268 J. Kim et al. / International Journal of Project Management 30 (2012) 264272
were identified through brainstorming among experts and
stakeholders involved in the RTPSI project. A code and
description for each nonconformance risk identified was specified
in the code and description register of nonconformance risks,
respectively. The codes and descriptions are shown in Table 4.
4.2.2. Step 2: Nonconformance risk analysis for project
activities
The probability and impact of each nonconformance risk
identified in Step 1 were assessed through interviews with
knowledgeable and experienced project teammember(s) or expert
(s) from outside the project. The previous experiences of experts
may be helpful for probability and impact assessment (Scott-
Young and Samson, 2008). For qualitative analysis, the
probability and impact of the nonconformance risk were assessed
with numerical scales: 0.10, 0.30, 0.50, 0.70, and 0.90. The
numerical scales are defined in Tables 5 and 6, respectively.
The numerical scales of Tables 5 and 6 are used to develop the
probability and impact matrix of Table 7. This matrix specifies
combinations of probability and impact that lead to scoring each
nonconformance risk identified in step 1, and is used to prioritize
nonconformance risks. Numeric values in Table 7 are derived
from the nonconformance risk score; (NRS) =probability
(P) impact (I).
The probability and impact matrix is useful at the beginning of
nonconformance risk analysis when an assessor has limited
information about the risks associated with an activity. Individual
NRSs assessed by this matrix are shown in Table 8. If there is
more than one assessor, the sumof the NRSs fromeach assessor is
divided by the number of assessors to obtain the average score for
an individual nonconformance risk.
To formulate a priority ranking of the individual codes based
on NRSs in Table 8, the NRSs were rearranged in descending
order in Table 9. If there is more than one code in the same
ranking row of Table 9, priority is given to the nonconformance
risk code of the activity with relatively high crash direct cost of
them.
The nonconformance risk analysis for project activities can
be a rapid and cost-effective means to prioritizing before the
PQLC estimation.
4.2.3. Step 3: PQLC estimation with nonconformance risk
activity rate
The PQLC for nonconformance risk activity was estimated
under the assumption that the rework or modification costs for a
nonconformance risk activity are equivalent to its crash direct cost
(c
j
). The acceptable number of nonconformance risk activities is
decided using Eq. (4). The RTPSI project PQLC is computed by
inserting the nonconformance risk activities of related codes from
Table 9 into the proposed model formula in Section 3.2.
For the RTPSI project, the nonconformance risk activity rates
from5%to 25% were given in increments of 5%, and the PQLCs
for each nonconformance risk activity rate candidate were
estimated at 5%, 10%, 15%, 20%, and 25%, respectively.
4.3. Computational results
The example was solved using LINGOsoftware (version 6.01)
on a personal computer (Hewlett-Packard Compaq Intel
Centrino 2.0 GHz with 2 GB RAM and 80 GB hard disk). The
estimated completion time of the RTPSI project according to the
constraint Eq. (3) was 17.5 days, which conformed with the
desired project due date (D). Activities A, B, D, G, N, O, and P of
the RTPSI project were identified as crashing activities that could
influence the project due date.
The estimation of the nonconformance risk activity rate for
the RTPSI project depends on the project manager's strategic
approach regarding the business priorities of Y Company. The
number of nonconformance risk activities Z
j
was calculated
according to the fourth constraint (4).
Table 10 presents each optimal PQLC for the five noncon-
formance risk activity rate candidates. The nonconformance risk
Fig. 2. Project network for RTPSI.
Table 3
Path length calculations.
Path Time estimation at each node Path length (days)
1 3+3+3+1.5+6+2+2+1.5 22.0
2 3+4+2+2+2+2+2+1.5 18.5
3 3+4+3+1.5+4+3+2+2+2+1.5 26.0
269 J. Kim et al. / International Journal of Project Management 30 (2012) 264272
activities shown in Table 10 had higher NRS than the other
activities. The RTPSI project PQLC was directly related to the
value or to the number of nonconformance risk activities. The
PQLC is visualized in Fig. 3.
If no rework or modification occurs while executing
nonconformance risk activities, the PQLCis saved. If any rework
or modifications occur during the execution of these activities, the
PQLC is payable in the budget. Although the RTPSI project
requires excessive crashing for nonconformance risk activities,
cases in which preventive measures against rework or modifica-
tions can be taken allow the project manager to choose a low
value without concern about the PQLC.
The information presented in Table 10 and Fig. 3 allowed the
project manager to complete the RTPSI project within the
assigned budget and due date in this real-life case.
5. Conclusions
In this paper, we propose a mixed integer linear programming
model that accounts for both the nonconformance risks and the
PQLC of the project activities. In a computational application of
this model we found that the crashing activities were properly
selected, and identified a need for special care regarding
nonconformance risk activities. By identifying these nonconfor-
mance risk activities in the process of project scheduling, the
project manager can take preventive actions that eliminate the
need for rework or modification, which in turn facilitates the
completion of the RTPSI project with minimum direct costs and
before the due date. Although rework or modification of
nonconformance risk activities cannot be completely avoided
during project execution, project cost overruns can be avoided
because the direct cost already includes the PQLC. This is similar
to a worst-case design concept in terms of reliability.
In conclusion, we present a valid and practical model that can
minimize PQLC influence on project cost due to excessive
Table 4
Code and description register of nonconformance risks.
Code Description of nonconformance risk Code Description of nonconformance risk
A1 Uneven floor when doing concrete work J1 Level error of suction disks
A2 Contaminated floor J2 Defectively assembled robot arm
A3 Low hardness of concrete floor J3 Inaccurate position where robot arm takes up box
B1 Reverse wiring between power supply cables K1 Loose bolts of supports for conveyor installation
B2 Poor contact of connectors or terminal blocks K2 Time interval error between conveyor sensors
B3 Interior wiring error of conveyor cables K3 Touching corner at turning point of L-shape
C1 Cables loosened while turning wrapping machine L1 Level error of magazine stand
C2 Poor contact of terminal blocks L2 Trouble related to pallet release from magazine stand
C3 Interior wiring error of wrapping machine cables L3 Incomplete operation of oil pressure, pneumatic brakes, and switches
D1 Touch of flexible cables while moving robot arm M1 Programming error; robot position
D2 Poor contact of robot's terminal blocks M2 Readjustment of pattern; technical location
D3 Interior wiring error of robot cables M3 Malfunction of interlocking system
E1 Broken terminal blocks inside panel N1 Trouble in data processing system
E2 Damage to control unit inside panel N2 Inspection of programs related to each pattern
E3 Leakage of electricity due to accumulated dust N3 Preparing backup files
F1 Broken terminal blocks inside electrical panel O1 Inoperative box counter
F2 Damage to controller or drive inside electrical panel O2 Taking a long time to test number of boxes
F3 Inflow of dust into electrical panel O3 Test run data collection for long periods of time
G1 Broken terminal blocks inside of electrical panel P1 Trouble with oil pressure or pneumatic brakes
G2 Damage to control unit inside of panel P2 Trouble with conveyor transportation
G3 Leakage of electricity due to accumulated dust P3 Malfunction of bar code scanner
H1 Misaligned conveyor chains Q1 Loose cable connection
H2 Collision between boxes at junction of conveyor lines Q2 Rotation in the reverse direction
H3 Trouble with automatic reject system Q3 Trouble with interlocking systems
I1 Malfunction of wrapper-lifting system R1 Faulty connection between processes
I2 Collision with boxes during wrapper-lifting motion R2 Poor contact of safety devices
I3 Uneven space between wrapping arm and boxes R3 Incomplete operation of automatic control system
Table 5
The scale and definition of probability related to nonconformance risks.
Probability scale Definition (based on occurrence of rework or
modification after completion of an activity)
0.10 Rare
0.30 Possible
0.50 Likely
0.70 Highly likely
0.90 Almost certain
Table 6
The scale and definition of impact related to activity cost.
Impact scale Definition
0.10 b20% increase in activity cost
0.30 2040% increase in activity cost
0.50 4060% increase in activity cost
0.70 6080% increase in activity cost
0.90 N80% increase in activity cost
270 J. Kim et al. / International Journal of Project Management 30 (2012) 264272
crashing activities. In an example project, we determined that
quality is a significant factor in timecost tradeoff problems when
project activities are excessively crashed. When compared with
the previous models in Table 1, the proposed model will help
project planners develop practical project schedules, and is a
meaningful approach in view of its possible utilization in real life
projects. Preventive actions for nonconformance risk activities
may incur additional cost. In future research, the proposed model
may be extended to estimate the additional costs of preventative
actions that are related to nonconformance risk activities and to
use this information to formulate newapproaches that incorporate
quality concepts into project scheduling problems.
References
Assaf, S., Al-Hejji, S., 2006. Causes of delay in large construction projects.
International Journal of Project Management 24, 349357.
Atkinson, R., 1999. Project management: cost, time and quality, two best
guesses and a phenomenon, its time to accept other success criteria.
International Journal of Project Management 17 (6), 337342.
Avots, I., 1984. Information systems for matrix organizations in Cleveland. In:
David, I. (Ed.), Matrix Management Systems Handbook. Van Nostrand
Reinhold, New York, pp. 535537.
Babu, A., Suresh, N., 1996. Project management with time, cost, and quality
considerations. Journal of Operational Research 88, 320327.
Brucker, P., Knust, S., 2006. Complex Scheduling. Springer-Verlag.
Deckro, R.F., Hebert, J.E., Verdini, W.A., 1987. Nonlinear multiple criteria
time/cost tradeoffs. Proceedings of the Sixteenth Annual Western DSI
Meeting, pp. 134137.
Deckro, R.F., Hebert, J.E., Verdini, W.A., 1995. Nonlinear time/cost tradeoff
models in project management. Computers &Industrial Engineering 28 (2),
219229.
Demeulemeester, E.L., Herroelen, W.S., Elmaghraby, S.E., 1996. Optimal
procedures for the discrete time/cost tradeoff problem in project networks.
European Journal of Operational Research 88, 5068.
Elmaghraby, S.E., Salem, A., 1982. Optimal project compression under
quadratic cost functions. Applications of Management Science 2, 139.
Feng, C.W., Liu, L., Burns, S.A., 1997. Using genetic algorithms to solve
construction timecost tradeoff problems. Journal of Computing in Civil
Engineering 11 (3), 184189.
Fulkerson, D., 1961. A network flow computation for project cost curves.
Management Science 7 (2), 167178.
Goyal, S.K., 1975. A note on simple CPM timecost tradeoff algorithm.
Management Science 21 (6), 718722.
Hillier, F.S., Lieberman, G.J., 1995. Introduction to Mathematical Program-
ming. McGraw-Hill Inc, New York.
Icmeli-Tukel, O., Rom, W., 1997. Ensuring quality in resource constrained
project scheduling. European Journal of Operational Research 103,
483496.
Kelley, J.T., Walker, M.R., 1959. Critical Path Planning and Scheduling: an
Introduction. Mauchly Associates, Ambler (PA).
Kelly, J.E., 1961. Critical-path planning and scheduling: mathematical basis.
Operations Research 9 (3), 296320.
Kessler, E.H., Chakrabarti, A.K., 1999. Speeding up the pace of new product
development. Journal of Product Innovation Management 16, 231247.
Table 7
Probability and impact matrix.
Probability 0.10 0.30 0.50 0.70 0.90
Impact
0.10 0.01 0.03 0.05 0.07 0.09
0.30 0.03 0.09 0.15 0.21 0.27
0.50 0.05 0.15 0.25 0.35 0.45
0.70 0.07 0.21 0.35 0.49 0.63
0.90 0.09 0.27 0.45 0.63 0.81
Table 8
NRS for each nonconformance risk code.
Code NRS Code NRS Code NRS Code NRS
A1 0.27 E3 0.20 J2 0.35 O1 0.28
A2 0.07 F1 0.09 J3 0.18 O2 0.20
A3 0.45 F2 0.40 K1 0.18 O3 0.14
B1 0.24 F3 0.20 K2 0.21 P1 0.30
B2 0.30 G1 0.15 K3 0.28 P2 0.18
B3 0.21 G2 0.45 L1 0.24 P3 0.32
C1 0.24 G3 0.12 L2 0.08 Q1 0.18
C2 0.25 H1 0.17 L3 0.21 Q2 0.30
C3 0.28 H2 0.31 M1 0.27 Q3 0.30
D1 0.37 H3 0.47 M2 0.27 R1 0.18
D2 0.25 I1 0.21 M3 0.24 R2 0.34
D3 0.21 I2 0.27 N1 0.27 R3 0.21
E1 0.15 I3 0.21 N2 0.14
E2 0.40 J1 0.24 N3 0.18
Table 9
Priority ranking of individual codes.
Ranking Code NRS Ranking Code NRS
1 H3 0.47 13 B1,C1,J1,L1,M3 0.24
2 A3,G2 0.45 14 B3,D3,I1,I3,K2,L3,R3 0.21
3 E2,F2 0.40 15 E3,F3,O2 0.20
4 D1 0.37 16 J3,K1,N3,P2,Q1,R1 0.18
5 J2 0.35 17 H1 0.17
6 R2 0.34 18 E1,G1 0.15
7 P3 0.32 19 N2,O3 0.14
8 H2 0.31 20 G3 0.12
9 B2,P1,Q2,Q3 0.30 21 F1 0.09
10 C3,K3,O1 0.28 22 L2 0.08
11 A1,I2,M1,M2,N1 0.27 23 A2 0.07
12 C2,D2 0.25
Table 10
PQLC for each nonconformance risk activity rate candidate.
,% 5 10 15 20 25
Z
j
1 2 3 4 5
Nonconformance risk activity H H,G H,G,A H,G,A,F H,G,A,F,E
PQLC, $100 38.29 248.51 268.62 699.75 869.81
Fig. 3. PQLC increase related to value.
271 J. Kim et al. / International Journal of Project Management 30 (2012) 264272
Khang, D., Myint, Y., 1999. Time, cost and quality tradeoff in project
management: a case study. International Journal of Project Management 17
(4), 249256.
Martin, P.K., Tate, K., 2001. Getting Started in Project Management. John
Wiley & Sons Inc, New York, pp. 2627.
Meyer, W.L., Shaffer, L.R., 1963. Extensions of the Critical Path Method
Through the Application of Integer Programming. University of Illinois,
Department of Civil Engineering.
Paquin, J.P., Couillard, J., Ferrand, D.J., 2000. Assessing and controlling the
quality of a project end product: the earned quality method. IEEE
Transaction on Engineering Management 47, 8897.
Patterson, J.H., Huber, D., 1974. A horizon-varying, zero-one approach to
project scheduling. Management Science 20 (6), 990998.
Perera, S., 1980. Linear programming solution to network compression. ASCE
Journal of the Construction Division 106, 315327.
Phillips, S., Dessouky, M.I., 1977. Solving the project time/cost tradeoff problem
using the minimal cut concept. Management Science 24 (4), 393400.
Robinson, D.R., 1975. A dynamic programming solution to costtime tradeoff
for CPM. Management Science 22 (2), 158166.
Sakellaropoulos, S., Chassiakos, A.P., 2004. Project timecost analysis under
generalised precedence relations. Advances in Engineering Software 35,
715724.
Schwindt, C., 2005. Resource Allocation in Project Management. Springer-Verlag.
Scott-Young, C., Samson, D., 2008. Project success and project team
management: evidence from capital projects in the process industries.
Journal of Operations Management 26, 749766.
Siemens, N., 1971. A simple CPM timecost tradeoff algorithm. Management
Science 17 (6), B354B363.
Summers, D., 2005. Quality Management: Upper Saddle River. Pearson
Education, Inc., New Jersey.
Swink, M., Talluri, S., Pandejpong, T., 2006. Faster, better, cheaper: a study of
NPD project efficiency and performance tradeoffs. Journal of Operation
Management 24, 542562.
Tareghian, H., Taheri, S., 2006. On the discrete time, cost and quality tradeoff
problem. Applied Mathematics and Computation 181, 13051312.
Tareghian, H., Taheri, S., 2007. A solution procedure for the discrete time, cost
and quality tradeoff problem using electromagnetic scatter search. Applied
Mathematics and Computation 190, 11361145.
Vanhoucke, M., Debels, D., 2007. The discrete time/cost tradeoff problem:
extensions and heuristic procedures. Journal of Scheduling 10, 311326.
Vrat, P., Kriengkrairut, C., 1986. A goal programming model for project
crashing with piecewise linear timecost trade-offs. Engineering costs
production. Economics 10, 161.
Wateridge, J., 1998. How can IS/IT projects be measured for success?
International Journal of Project Management 16, 5963.
Wiest, J.D., 1963. The Scheduling of Large Projects with Limited Resources.
Unpublished PhD Thesis, Carnegie Institute of Technology, Pittsburgh.
Wiest, J.D., Levy, F.K., 1997. A Management Guide to PERT/CPM:
Englewood Cliffs. Prentice-Hall, Inc, New Jersey.
Wuliang, P., Chengen, W., 2009. A multi-mode resource-constrained discrete
timecost tradeoff problem and its genetic algorithm based solution.
International Journal of Project Management 27, 600609.
Zeng, J., An, M., Smith, N., 2007. Application of a fuzzy based decision making
methodology to construction project risk assessment. International Journal
of Project Management 25, 589600.
272 J. Kim et al. / International Journal of Project Management 30 (2012) 264272