Time-cost-risk optimization in construction work by using ant colony algorithm

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 04 | Apr-2018 www.irjet.net p-ISSN: 2395-0072
© 2018, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 2298
Time-Cost-Risk Optimization in Construction Work by using Ant
Colony Algorithm
Vishnu Vijayan1, Achu V2, Riyana MS3, Jayakrishnan R4
1,2,3 Assistant Professor, Department of Civil Engineering, BMCE, Kerala, India
4Assistant Professor, Department of CSE, BMCE, Kerala, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract – Time and cost are the most important
parameters considered in every construction project.
Without the timely completion within the scheduled
duration, the project may incur loss. Hence completing the
project within the scheduled time frame is paramount
importance.
Time, cost, risk of project delivery are among the crucial
aspects of each project. The need to identify and manage
risks in project delivery is crucial to establish ways and
methods to deliver a project to meet end user’s expectation
and satisfaction. Construction plannersface the challengeof
optimizing resource utilization to compromise between
different aspects, especially time and cost. It may be
beneficial to consider the risk in performance of project in
addition to time and cost.
Besides time and cost of activities, every resourceutilization
option will yield a specific performance quality according to
the involved. The optimization of trade-off between these
time, cost and risk parametersisdone using Multi-Objective
Ant Colony Optimisation technique. The optimization
program was run for a number of giving weightage for the
parameters. The total time, cost and risk for the project
when executed in variouscombinations of alternativeswere
taken the trade-off between the parameters are analyzed.
The method could be adopted for the time, cost, and risk
optimization of any project with its quantified values. The
risk values may vary with various factors and the choice of
risk values depends entirely on the project.
Key Words: Time cost, risk optimization, ant colony
algorithm, resource utilization, project management,
Multi-Objective Ant Colony Optimisation technique.
1. INTRODUCTION
The prime objective in construction industry is to complete
the planned activities of project in time. Inordertomaximize
the return, both the client and contractor should strive to
optimize the project duration and cost concurrently. Time
and cost are the most important parameters considered in
every construction project. Without the timely completion
within the scheduled duration, the project may incur losses.
Hence completing the project within the scheduled time
frame is of paramount importance. Time, cost and risk of
project delivery are among the crucial aspects of each
project. Maximizing the quality of projects while minimizing
its time and cost, require the development of innovative
models considering risk. Each and every activity of the
project may get affected by various factors and may
adversely affect the cost and duration of the entire project.
Great care should be taken to reduce or minimize these
impacts. This requires an insight into the probable risks
associated with each of the activities and the alternatives
that can be adopted to reduce such risks. It is needed to
quantify the associated potential impact oftheidentifiedrisk
and implementing measures to manage and mitigate the
potential impact. This involves deciding on the relationship
between the likelihood (frequency or probability) and the
consequences (the impacts) of the risks that have been
identified.
The need to identify and manage risks in project
delivery is crucial to establish ways and methodstodelivera
project to meet end user’s expectation and satisfaction. In
order to do so, the common problem in constructionprojects
such as delays in completing the project, over budget,
unsatisfactory product quality, unsafeworkingenvironment
and so on needs to be eliminated as far as possible. Paying
attention to the risk aspects in construction project will
direct to profitable outcome. Risk arisesout ofuncertainty.It
is measured in terms of the likelihood of it happening and
the consequences if it does happen.
RISK= PROBABILTY X IMPACT
Construction planners face the challenge of
optimizing resource utilization to compromise between
different aspects of projects, especially time and cost. It may
prove to be beneficial to consider the risk in performance of
projects in addition to time and cost. These multi-objective
naturesimpose an increasing pressureondecisionmakersin
the construction industry to search for an optimal/near-
optimal resource utilization plan that minimizes the
construction cost and the time, while minimizing its risk.
When the time interval of the activities is minimized we see
the cost will increase due to more resources allocation in
rapid accomplishment for the completion of work. Although
lower project duration leads to higher direct costs, indirect
costs are also decreased. In such situations it isimportant to
study the trade-off between completion time, risk involved
in each resource option and cost of the project. Crashing of
activities results in increase in resource utilization which
eventually leads to increased construction cost for the
project and vice versa. In addition to time and cost of
activities, each resource utilization option will yield specific
risk. Trade-off between these conflicting aspects of project
results in challenging job and as such plannersarefacedwith
numerous possible combinations for project delivery. If

durations of the activities are crashed, the cost shows an
increasing trend due to more resource being allocatedforits
rapid accomplishment. However byutilizingfewerresources
results in extended duration of activities.
Besides time and cost of activities, every resourceutilization
option will yield a specific performance quality according to
the risk involved. Trading between these conflicting aspects
of project is a challenging job to construction planners.Asan
example, the number of possible combinations in a project
with 18 activities and 4 possible resource utilizationoptions
for each activity will be more than 6 billion. An effective
searching tool would then be worthwhile forcomprehensive
yet efficient time-cost- risk trade-off problem (El-Rayes etal
2005). In this paper, a new met heuristic approachisapplied
for optimization of three objective time-cost-risk problem
based on multi-colony ant algorithm. In this thesis a sample
project of 7 activities with time and duration of various
resource options is taken. The risk for each resource option
is calculated from the results of a questionnaire surveydone
on 35 different companies. The optimization of trade-off
between these time, cost and risk parameters was done
using Multi-Objective Ant Colony Optimisation technique.
The optimization program was run for a number of times
giving different weightage for the parameters.Thetotaltime,
cost and risk for the project when executed in various
combinations of alternatives were taken and the trade-off
between the parameters are analyzed.
2. SCOPE
The report focuses on a 7 activity project with time and
cost details of various activity options. The risk values are
calculated from the data collected through a questionnaire
survey done on 35 companiesin the constructionsector.The
method could be used for any construction project,provided
the quantified values of the three parametersare known. All
the parameters, namely, time, cost and risk considered in
this project are minimization in nature; however this could
be used for objectives that are of maximization in nature by
adopting minor changes. The time and cost of the activities
could be easily estimated, but the quantification of risk
values cannot be done easily. The risk of each activitymaybe
calculated based on surveys or even from experience. The
method could be adopted for the time, cost and risk
Optimisation of any project with their quantifiedvalues.The
risk values may vary with various factors and the choice of
risk values depends entirely on the project.
3. ANT COLONY ALGORITHMS
In AS, solutions are constructed based on the transition
probability,
Where ηij is a local heuristic, α and β are two parameters
that determine the relative influence of the pheromone trail
(τ) and the heuristic, and U denotes the set of candidate
solutions to be chosen. The pheromone trail can be updated
as τij
new =ρ τij
old +τij
k
Where ρ is a parameter that controls the pheromone
persistence, i.e., 1-ρ represents the proportion of the
pheromone evaporated NA is number of ants, i.e.,allantscan
contribute to pheromone trail accumulation in the AS
algorithm. Dorigo et al. (l99Ia, 1991b, 1996) propose three
different approachesto find ∆τijvaluesfor theTSPasfollows:
Where Q denotes a constant related the quantity of
pheromone trail ants laid, i j represents the distance d
between cities i and j, and is the total tour length of the ant.
The ant density and ant quantity L,k ,Kth approachesdeposit
pheromone every time an ant goes from i to j, but the ant
cycle deposits it only after a complete tour. Experiments
indicate that ant cycle outperforms the other two
approaches (Colomi et al. 1997, Dorigo et al. 1991a, 1991b,
1996).
Multi-objective optimization (or multi-objective
programmingor "paretooptimization")alsoknownasmulti-
criteria or multi-attribute optimization, is the process of
simultaneously optimizing two or more conflicting
objectives subject to certain constraints.
A lot of work has been done on the multi- objective
optimization in construction projects. Initial trials were to
model multi-objective optimization toolsfortimecosttrade-
off problems. Later the efforts were extended for increased
number of objectives. The technique has been in use for
optimization of quality, resources, safety etc. in addition to
time and cost.
4. MULTI-OBJECTIVE OPTIMIZATION IN
CONSTRUCTION PROJECTS
Multi objective optimization methods are used in
construction projects for a variety of parameters in addition
to time and cost, such as, risk, quality, safety, environmental
impact, resource leveling etc. Researchers have done the
optimization of these conflicting parameters using various
optimization methods. In the beginning the problem was
solved taking two objectives at a time mainly time and cost
and linear programming was generally used for this. Many
researchers attempted to solve TCT by linear programming

(Babu and Suresh1996, Burns et al. 1996, Khang and Myint
1999, Wei and Wang 2003, Vanhoucke 2005, Bidhandi
2006).
5. THE ANT COLONY OPTIMIZATION (ACO)
ALGORITHM
The Ant colony optimization (ACO) algorithm was
developed by Dorigo et al. ACO algorithms may be
considered to be part of swarm intelligence, that is, the
research field that studies algorithms inspired by the
observation of the behavior of swarms. Swarm intelligence
algorithms are made up of simple individualsthat cooperate
through self-organization. The Ant colony optimization
(ACO) algorithm is a part of the swarm intelligence theory.
The main idea behind the ACO algorithm is that the self-
organizing principles that allow the highly coordinated
behavior among ants in real life can be exploited to
coordinate populations of artificial agentsthatcollaborateto
solve computational problems. Communication between
ants relies on a chemical substance called pheromone. To
search for food, ants follow pheromone pathsmade byother
ants, and at the same time they themselves deposit
pheromones along the path they take. Most ants will
habitually take the path on which more pheromones are
deposited. Consequently, the effect of this positive feedback
will lead ants to take the path that has more pheromones
deposited on it.
Marco Dorigo and colleaguesintroduced thefirstAnyColony
Optimization (ACO) algorithms in the early 1990’s which is
inferred by the social behavior of ant colonies. Ants live in
colonies and their behavior is governed by the goalofcolony
survival rather than being focused on the survival of
individuals. When searching for food, ants randomly search
through the possible paths between the food and theirnests.
Ants leave a chemical called pheromone which they can also
smell. If an ant should make a decision about selectingapath
to follow, it selects the one which containshighestamountof
pheromone. The shortest path would have the highest
pheromone concentration if all paths are initially preferred
by same number of ants. This is because the ants would
travel the shortest path in shorter duration and the path
would be travelled by more ants when compared with the
other paths. As this is the case, the pheromoneconcentration
of the path would be increased by the ants and the
probability of the shortest path would be higher than the
other paths. During the return trip, the quantity of
pheromone that an ant leaves on the groundmay depend on
the quantity and quality of the food. The pheromone trails
will guide other ants to the food source. The indirect
communication between ants via pheromone trails, known
as stigmergy, enables them to find shortest paths between
their nest and food sources.
In an experiment known as the “double bridge
experiment"(Deneubourg et al.), the nest of a colony of
Argentine ants was connected to a food source by two
bridges of equal lengths. In such a setting, ants start to
explore the surroundings of the nest and eventually reach
the food source. Along their path between food source and
nest, Argentine ants deposit pheromone. Initially, each ant
randomly chooses one of the two bridges. However, due to
random fluctuations, after some time one of the two bridges
presents a higher concentration of pheromone than the
other and, therefore, attracts more ants.Thisbringsafurther
amount of pheromone on that bridge making it more
attractive with the result that after some time the whole
colony converges toward the use of the same bridge. in
another experiment conducted by Goss et al. When the nest
is connected to the food source by a shorter path and a
longer path the ants choosesthe shorter path andalltheants
gets converted to this path much earlier.
Fig 5.1 Double bridge experiment
Later various meta-heuristic methods which are more
efficient and useful in solving multi-objective optimization
problems were used ,such as, Genetic Algorithm (GA),
Genetic Algorithm with Simulated Annealing(GASA), ),
Particle Swarm Optimization (PSO),AntColonyOptimization
(ACO) and Electromagnetic Scatter Search (ESS) etc .These
methodsgivemore reliable results when multipleobjectives
are considered.
The multi-objective ant colony algorithm is very effective in
solving multi-objective optimizationinconstructionprojects.
Marco Dorigo and colleaguesintroduced thefirstAnyColony
Optimization (ACO) algorithms in the early 1990’s. Ant
Colony Optimization (ACO) had been implementedbyKuang
and Xiong(2005). They solved a small project with 7
activities and find the global optimumwithACObysearching
only 6.63% of the possible search space.
6. THE THREE MAJOR PARAMETERS IN
CONSTRUCTION
A construction project contains many uncertainties. It
requires a number of resources and a large amount of
investment. Time and cost are main management goals.
Contractors want to get the highest profit so they must plan
to complete the job in early time with a minimum cost. A
project may be regarded as successful if it is completed
within budget, on time, without any accidents, to the
specified quality standards and overall client satisfaction.

6.1 TIME
There are many reasons why time management isimportant
in construction. One of these is the fact that workers are
often paid by the hour, so time managementwillhelpcontrol
salary costs. Another reason time management in
construction is important is that having work delayed or
behind schedule can hinder the overall project, especially
when one group or company must wait foranothercompany
to finish a certain type of work before beginning the next
step in the project. Time management in construction alsois
vital because, if projects are not finished in a timely manner,
or at least as quickly as promised, then this can cause the
construction company to lose paymentforbreachofcontract
or monetary penalty according to the contract.
6.2 COST
The most important factor consideredinmostofthe
construction projects is cost. The costs of a construction
project include both the initial capital cost and the
subsequent operation and maintenance costs. Each of these
major cost categories consists of a number of cost sub
components. The capital cost for any construction project
includes the expenses related to the initial establishment of
the facilities such as land acquisition, planningandfeasibility
studies, architectural and engineering design,inspectionand
testing etc. The operation and maintenance cost in
subsequent years over the project life cycle includes land
rent(if applicable),operating staff, labour and material for
maintenance and repairs etc.
The magnitude of each of these cost components
depends on the nature, size and location of the project as
well as the organization. The owner is interested in
achieving the lowest possible overall project cost that is
possible within its investment .It is important to realize that
while the construction cost may be the single largest
component of the capital cost, other costcomponentsarenot
insignificant. For example, land acquisition costs areamajor
expenditure for building construction in high-density urban
areas.
The total project cost consists of two parts, direct
costs and indirect costs. Direct costs are the activity based
costs which are the labor, material, equipment and
machinery costs of the project. As the activity’s normal
duration is crashed, it is expected that the direct costs are
increased. This is because in order to finish a certainactivity,
larger crews are assigned or the crew is over-timed. Larger
crew size or over-timing decreases the productivity. As a
result of this, unit cost per unit output increases. Indirect
costs are the overhead costs and possible delay penalties of
the projects. Overhead cost includes any costs of the project
which cannot be associated with the activitiesof the project.
Salaries of the cook, security staff and office staff can be
counted as overhead costs. Heating or cooling, illuminating
of the site and other buildings are also included in overhead
costs of the project. Overhead costs fluctuate during the
project as it affected by the climate and the number of the
workers in the site. In order to simplify the cost
computations, overhead cost is usually assumed constant
during the project. Consequently, indirect cost decreases
when the project duration decreases.
6.3 RISK
Risks are unavoidable in almost every construction
project whether if it is building projects, civil works, or any
other type of construction projects. Risk is inherent in all
human endeavors, including construction activities, and the
risk factors involved are diverse and varied. Risks are
uncertainties, liabilities, or vulnerabilities, which may cause
a project to deviate from its defined plan. While risk in a
project environment cannot be totally eliminated or
transferred, it can be monitored and minimizedormitigated
wherever possible. To succeed, organizations must commit
to addressing risk management throughout the project
lifecycle.
Risk is defined as “the chance of something happening
that will have an impact on objectives”. Risks and
uncertainties appear in various stages and types. Risk is an
inevitable aspect of most projects, but even the most
proficient managers have difficulty handling it. To avoid the
projects ending up with an overrun schedule, overflowing
budget and compromised specifications, the common
problem in construction projects such as delays in
completing the project, over budget, unsatisfactory product
quality, unsafe working environment and so on needs to be
eliminated asfar aspossible . Risk analysis and management
in construction depend mainly on intuition, judgment and
experience. In projects the objectives are most often related
to time, cost, quality etc.
The risks in construction projects could be broadly
classified as:
 Cost related risks:
 Time related risks:
 Quality related risks:
 Environment related risks:
 Safety related risks:
This project mainly concentrates on selecting
appropriate resource options for every activity to obtainthe
objective of time, cost and risk of a project. Quantifying the
risk involved in performing different activity as a functionof
different resource utilizations is a challenging task because
of the difficulty in measuring impact of these strategies of
performing activities. Moreover, it is a complicated work to
evaluate the contribution of individual activities’ risk on
performance of the project. The objective of risk may be
evaluated with following function:
Mean Risk =Σ (r x w)/Σw

Where r is the risk of the optimized resource option
for each activity w is the activity weight for each activity. A
common approach is to define risk asthe combination ofthe
probability (or likelihood) and impact of an event (or
outcome or result of exposure). This gives rise to the widely
used concept of risk:
Severity (risk) =Probability x Impact
6.4 TIME COST TRADE-OFF(TCT) PROBLEM
The objectives of the TCT analysis are to compress
the project to the optimum duration which minimizes the
total project cost. TCT analysis is performed by evaluating
the possible crashing alternatives of the activities. Possible
crashing durations can be obtained in three ways: by over
sizing the crew, by over timing or by executing another
construction technique which is faster but more expensive.
Crashing of activities increases the activity’s cost, thus
increasing the direct project cost. However, crashing of
activities reduces the project duration and decreases the
indirect project costs. Summation of direct and indirect
project costs is aimed to be minimized in TCT problem.
TCT is one of the major interests of the construction
management, since the optimum solution of TCT problems
directly increases the productivity and thus the profit of the
project. It is expected that the cost slope of activitiesincrease
asthe activities are further crashed. This ismainlycausedby
the reduced productivity of labour and machinery as the
crew size or amount of overtime is increased. A relationship
exists between the duration of a construction activityandits
cost. Similarly, construction cost and time for undertaking a
construction project are interrelated. Generally, the total
project cost includes both the direct and indirect costs of
performing construction work. The higher curve in Figure 1
illustrates the relationship between cost and time for a
construction project.
Fig 6.1 time cost relationship of construction project
At the normal point, the construction costs are
minimized. Construction costs may increase if any variation
in time occurs from this point. If the construction time from
the point is reduced, the direct costs will increase while the
indirect costs decrease (and vice versa). The TCT analysis is
helpful in providing monetary benefits and such cost
reductions add up to the increase the profits.
6.5 TIME-COST- RISK OPTIMISATION
It is important to study the trade-off between
completion time, risk involved in each resource option and
cost of the project. Crashing of activitiesresultsinincreasein
resource utilization which eventually leads to increased
construction cost for the project and vice versa. In addition
to time and cost of activities, each resourceutilizationoption
will yield specific risk. Trade-off between these conflicting
aspects of project results in challenging job and as such
plannersare faced with numerouspossiblecombinationsfor
project delivery. If durations of the activitiesare crashed,the
cost shows an increasing trend due to more resource being
allocated for itsrapid accomplishment. However byutilizing
fewer resources results in extended duration of activities.
Moreover time and cost of activities, every resource
utilization option will yield a specific performance quality
according to the risk involved.
Time Cost Risk Trade off analysis is the
compression of the project schedule to achieve a more
favorable outcome in terms of project duration, cost, and
risk. The time-cost trade-off problem (TCTP) is the most
important content in the project management.Althoughalot
of research has been done on time cost trade-off problems ,
the TCTP considering the risk elements has not been solved
to such extend effectively. Each activity has certain options
for resource utilization with varying cost, time and risk
factors. The goal is to find the optimal or near optimal
solution with optimal trade-off. Quantifyingtheriskinvolved
in performing different activity as a function of different
resource utilizations is a challenging task because of the
difficulty in measuring impact of these strategies of
performing activities. Moreover, it is a complicated work to
evaluate proportion of individual activities’ risk on
performance of the project. The trade off analysis between
time cost and risk is more advantageous and helps in
selecting a project delivery with less risk. The parameters
used for the quantification may be purely subjective. The
risk factor may be quantified on various grounds such as
safety, quality, cost overrun, schedule overrun,availabilityof
skilled laborers, availability of machineriesetc.Thetimecost
risk trade off analysis helps not only in the comparison
between time and cost, but also help in comparing various
alternatives with respect to the risk associated with the
alternative and helps in better decision making.
7. MODEL DEVELOPMENT
7.1 description of the algorithm
This thesis aims to develop an optimization tool for
the time-cost-risk trade-off based project planning and
optimization problems. The problems in concern have a
common point in which all have the objective function as
minimization of an objective value. Project duration can
often be shortened by accelerating some of its activitiesatan
additional expense. Obtaining optimum or near-optimum

solution for the project planning problemsconsideringtime,
cost and risk is the main objective of this project.
Optimization procedure is performed by softwaregenerated
during the thesis study. Time-Cost-Risk Trade-off analysisis
the compression of the project schedule to achieve a more
favorable outcome in terms of project duration, cost, and
risk.
The ant colony optimization is used commonly for
various multi-objective optimizationproblems.Inthisreport
multi-objective ant colony optimizationmethodologyisused
to find optimal or near optimal solutions for time-cost-risk
trade off for a sample project consisting of 7 activities with
different resource option alternatives. The ant colony
algorithm used in the optimization process is explained in
detail.
In the ant colony optimization adopted in this study, each
activity is considered to be a path for the ants. A colony of
ants is assumed to visit the nodes and an objective is
assigned to the colony of ants. This colony of antschecksthe
solution and updates the pheromone by considering this
objective alone. The start and end of each activity is
considered to be a node (similar to the cities in travelling
sales man problem). But unlike in travelling salesman
problem the ants do not take all the pathsrandomly;theants
can travel from one node to the next node by taking one of
the path connecting those nodes (using a resource
option).The ants takes up the paths randomly in the first
visit as no pheromone trail is present in any of the paths and
all the paths are equally attractive to the ants. Once the ant
takes up a path the ant movesto the next activity.Theending
node of the first activity serves as the starting node for the
next activity.
Fig 7.1 Graphical representation of problem R.
Shrivastava etal(2011)
Once all the ants have visited all the nodes ( all the
activities are done by every ant by taking one resource
option for each activity), an iteration is completedandallthe
paths is deposited with different amounts of pheromone
depending on the goodness of that path. Then a new colony
of antsis assigned a different objective and is allowedtovisit
the nodes. This colony of ants checks the solutions based on
this objective alone. The new colony of ants chooses the
paths based on the pheromone content (deposited by the
ants of the preceding colony) in the paths.
Since the second colony of ants makes their visit
only after all the ants in the first colony completes the
iteration, there is a time lag between the visitsof the two ant
colonies at every node. The pheromone is a volatile
substance and it evaporates with the passage of time. In
order to simulate the pheromone evaporation, the
pheromone evaporation coefficient (ρ) is defined which
enables greater exploration of the search space and
minimizes the chance of premature convergence to
suboptimal solutions. Thepheromoneevaporationalsohelps
in eliminating the localization of search, by evaporating the
pheromone content of all the pathsby a factor,thedifference
in pheromone content of all the paths are reduced. This
makes even the worst solution to be considerably more
favorable for ants to be visited in the subsequent iterations
and hence helps in conducting a wider search and avoids
localization and early convergence of solution. The
probability of ants choosing a path from one node the next
node depends on the pheromone content in the paths. The
path with higher amount of pheromone attracts more
number of ants and vice versa.
The number of colony of antsmade to visit the node
in one cycle is equal to the number of objectives considered.
Once all the colonies visits the nodes an optimal or near
optimal solution is obtained which hasbeenevaluatedonthe
grounds of all the objectives. The quality of the solution
increases as the number of cycles increases. After a number
of cycles the solution converges to a single solution which
gives a near optimal solution with respect to all the
objectivesconsidered. The weightage of eachobjectivecould
be varied and this givesa different solutiondependingonthe
assigned weightages.
At the end of each cycle a number of optimal/near
optimal solutions are obtained. After running the
optimization program for a number of cycles all the optimal
/near optimal solutions obtained are analyzed and the
pareto-optimal solutions are grouped. Then the constraints
are imposed on the set of pareto-optimal solutions by
providing upper limits for the objectives. If the value of at
least one of the objectives exceeds the upper limit of the
corresponding objective, it is eliminated. This provides
refined set pareto-optimal solutions representing the
solutions with all the parameters of objectives within in the
acceptable limits. The solutions thusobtained are compared
to obtain the results.
7.2 THE MAIN PHASES OF ACO ALGORITHM
 Initialisation of parameters
 Selection of paths by the ants.
 the pheromone update and the pheromone
evaporation
 finding the optimal path( solution)

7.2.1 INITIALISATION OF PARAMETERS
Various parameters are used in the ant colony optimization
algorithms. Most of them are used to simulate the real life
ant colony behavior, which they use to find the shortestpath
from nest to the food source. The parameters use in this
algorithm and the values used are.
1. The number of ants in each ant colony, antcol= 50
2. The maximum number of activities that can be
evaluated= 50
3. Pheromone updation amount, ∆τij
4. The coefficient controlling the influence of
pheromone ,α=1
5. Heuristic function is not considered in the
algorithm, therefore the parameter controlling its
influence β=0.
6. The maximum number of resource options that can
be assigned to each activity=15
7. Initial pheromone in each path, τo=1/ number of
resource option for the activity
8. The pheromone evaporation coefficient ,ρ=5%
9. The number of iteration in each cycle=3
10. The number of cycles=1000
11. Total number of iterations=3000
7.2.2 Selection of Paths by Ants:
When the ants come to a node and has to select one among
the ants chooses the paths based on probability which in
turn is governed by the pheromone content present in the
path. The probability of an ant choosing a path is;
Prob (path) = (τpath)α/(∑τpaths)α
Where τpath is the pheromone content in thedesiredpathand
∑τpaths is the sum of pheromone content of all the paths for
that activity.
7.2.3 Pheromone Updation:
After each iteration, the pheromone values are updated by
the m ants that have chosen the same path. The pheromone
τij for path ij is updated as
Where i is the activity number j is the resource option
chosen ρ is pheromone evaparation coefficient ( [0,1] )
∆ τij is the amount of pheromone updated by each ant. m is
the number of ants using the path ij k is the ant taking the
path ij. Once the ants completes the travel it updates the
pheromone on the path depending on the goodness of the
parameter considered. The amount of pheromone updated
by each ant ∆ τij is calculated.
7.2.4 Finding Optimal Solution or Path:
In first iteration the colony of ants evaluates the paths and
updates pheromone considering the time. In the second
iteration the ants evaluates the paths and updates
pheromone considering cost and risk in the third. At the end
of each cycle (3 iterations) the ants provides a solution
considering all the objectives. The paths containing the
highest amount of pheromone at the end of each path
becomes the solution of that cycle. The solutions changes as
the program evaluatesthe alternates, and finallytheytendto
converge to the near optimal solution based on the weights
assigned to the objectives.
7.2 ASSUMPTIONS MADE
The main assumptions made in this project are
1. All the activities have finish start relationship
2. Only one activity is done at a time and all the activities
have the preceding activity as its predecessor.
3. Indirect cost for each activity is Rs 5000/day.
8. CASE STUDY
The time and cost data is taken as that for a 7
activity project with different resource options for each
activity. The cost consists of direct cost and indirect cost.
Indirect cost of Rs 5000/day is assumed. The activities are
assumed to occur one after the other on a finish to star
condition (finish of a preceding activity is a condition forthe
start of the succeeding activity.
Sl
No
Activity
description
Option Duration
(day)
Cost
(Rs)
1
Site
Preparation
1crew+equipment 1 14 990000
2 crew +equipment2 20 820000
3crew+equipment 3 24 600000
2
Forms and
Rebar
method 1 15 195000
method 2 18 186000
method 3 20 172000
method 4 23 175000
method 5 25 165000
3
Excavation
equipment 1 15 255000
equipment 2 22 270000
4 Precast
ConcreteGirder
method 1 12 1875000
method 2 16 1480000
5
Foundations
and Piers
method 1 22 910000
method 2 24 820000
method 3 28 740000

method 4 30 550000
6
Deliver PC
girders
Railroad 14 1670000
Truck 18 1370000
7
Erect girders
1 crew+ crane 1 9 1245000
2 crew+crane 2 15 1035000
3crew+crane3 18 870000
Table 8.1: Initial Data (R. Shrivastava et al.(2011))
9. RESULTS AND DISCUSSIONS
SOLUTIONS USING THE MULTI OBJECTIVE ANT COLONY
OPTIMISATION
The program was used to get the solution for the
time cost risk trade-off for the sample project. Running the
project for a number of times gives a number of solutions
with different combination of resource option for each
activities. From these solutions the pareto-optimalsolutions
are selected to obtain the optimal/near optimal solutions. A
total of 70 solutions were obtained using the ant colony.
9.1 REPRESENTATION FEASIBLE SOLUTIONS
Fig 9.1: Time-Cost-Risk Combination Graph
Fig 9.2: Risk-Time Trade Off
Inference:
In fig 7.8, considering the trade-off between time and risk in
the feasible pareto-optimal solutions, the risk valueofallthe
options lie within a small range of [-0.25, +0.1], ie,
irrespective of the change in duration of the project the risk
values do not show much significant trade-off for the
options. But still when required to compare twooptions the
graph could be used to make a better decision on the basisof
risk and time.
10. CONCLUSIONS
In this report the Multi-Objective Ant colony
Optimisation wasused for the optimizationontime-cost-risk
trade-off, for a sample project of 7 activities with different
resource options. The method provided 70 distinct possible
solutions considering all the three parameters, 48 pareto-
optimal solutions and after providing the constraints a final
set of 22 feasible solutions were obtained.
Analysing the results, showed the relationship
between the parameters. The time and cost are inversely
related and time and risk are inversely related.Furthermore
on analysing the final set of feasible solutions, it was found
that time and cost shows significant changes when various
options are changes, but the risk does not show such
significant changesin values. Hence time and costisfoundto
be the dominant parametersin choosing theoptions,butrisk
also plays a role in choosing a solution among comparable
solutions, ie, if the time and cost of two or more solutionsare
comparable to choose a better option the risk parameter
could be used effectively.
The quantification of risk depends on the
parametersconsidered, availabilityofdata,methodusedand
so on. In smaller projects the risk variation between various
options are not very large and hence the risk is not as
influential astime and cost but could be effectiveinchoosing
among the comparable solutions. Choosing an option with
lesser risk leads to less undesirable impacts during the
construction such as cost overrun, schedule exceedance,
safety during construction, shortage of resources etc. In this
thesis, for the sample project considered, the risk level
associated with each options do not show large variations
for different alternatives, but in large projects or new
innovative projects where the uncertainty is large the risk
variations between various options, when quantified, could
be high. If such projects are considered the risk values for
the alternatives may have larger differenceswhencompared
to each other and in such cases the risk parameter may play
a dominant role in choosing the resource options.
11. LIMITATION
The main limitation is the problem of quantification of risk.
The quantification of risk of risk is done for the operational
stage of the project alone (riskswill be apparent at all stages
of the life of a construction project: at appraisal, sanction,

construction and operation). The risk values may vary with
various factors and the choice of risk valuesdependsentirely
on the project. The value of risk for the same activity with
the same resource option may vary from project to project
depending on circumstances such as ground condition,
weather conditions etc. Hence a general quantification of
risk cannot be made for all the projects.
In this project a general quantification of risk is made froma
questionnaire survey for the calculation of risk for various
activities but the risk may vary from depending on thefirms,
experience, working conditions etc. In reality the risk values
are very much subjective than it can be generalized.
12. FUTURE EXPANSION.
The proposed ant colony Optimisationalgorithmcanbeused
to solve multi-objective problems of any 3 parameters
provided all the parameters are to be minimized. The
program could be modified even to solve minimisation
problems too. The quantification of risk is a major constrain
in the problem, as it is more difficult to quantify parameters
such as risk. Risk again consists of two vague components,
namely, probability and severity of risk. The accuracy and
dependability of the solution could be improved if better
methods for risk quantification are used, such as a more
detailed survey. This project takes up a sample project and
optimises the time, cost and risk for the constructionstageof
the project. It can be extended to other stages such as
planning stage, maintenance stage etc.
13. ACKNOWLEDGEMENT
The author thanksDr.K.R HareendranNair(HCT-AbuDhabi),
my parents and my friends for their valuable time spending
with me for the completion of this project.
14. REFERENCES
1. Afshar A., Kaveh A., and Shoghli O. R. (2007),“Multi-
objective optimization of time-cost-quality using
multi-colony ant algorithm”, Asian Journal of civil
engineering
(Building and Housing), Vol. 8 (2), pp. 113 – 124.
2. Babu, A.J.G., Suresh, N., Project management with
time, cost and quality considerations, European J.
Operations Research, (1996), 88, 320-327.
3. Arun c. (2011) ,”Time-cost-risk trade off using ant
colony optimization”, Journal of Construction in
Developing Countries, 2011
4. Bidhandi H. M. (2006), “A New Approach Based on
the Surrogating Method in the Project Time
Compression Problems”. Ann Oper Res, 143: 237 –
250.
5. Blum C. and Samples M. (2004), “An ant colony
optimization algorithm for shop scheduling
problems.” J Math Modelling Algorithms, Vol. 3(3),
pp. 285 – 308
6. Burns S. A., Liu L., and Feng C. W., (1996), “The
LP/IP hybrid Method for Construction Time-Cost
Trade-off Analysis”. Construction Management and
Economics, 14: 265 – 276.
7. Bullnheimer, B., R.F. Hartl, C. Strauss, 1999. "A New
Rank Based Version of the Ant System – A
Computational Study," CentralEuropeanJournalfor
Operations Research and Economics, 7: 25-38
8. Colorni, A., M. Dorigo, F. Maffioli, V. Maniezzo, G.
Righini and M. Trubian, 1996. "Heuristics from
Nature for Hard Combinatorial Optimization
Problems," International Transactions in
Operational Research,3(1): 1-21.
9. Deneubourg, J.L., S. Aron, S. Goss and J.M. Pasteels,
1990. The self-organizing exploratorypatternofthe
argentine ant, J. Insect Behav., 3: 159–168.
10. Dorigo, M., V. Maniezzo, A. Colorni, 1996. Ant
system: optimization by a colony of cooperating
agents, IEEE Trans. Systems, Man, Cybernet.-Part B
26(1): 29–41
11. Dorigo, M., T. Stutzle, 2004. Ant Colony
Optimization, MIT Press, Cambridge, MA. Fogel, L.J.,
Owens ,A.J., Walsh, M.J .1996. Artificial Intelligence
Through Simulated Evolution, Wiley, New York.

Time-cost-risk optimization in construction work by using ant colony algorithm

More Related Content

What's hot (19)

Similar to Time-cost-risk optimization in construction work by using ant colony algorithm (20)

More from VISHNU VIJAYAN (20)

Recently uploaded (20)

Time-cost-risk optimization in construction work by using ant colony algorithm