Computers and Operations Research: Dirk Briskorn, Michael Dienstknecht

Computers and Operations Research 92 (2018) 194–207
Contents lists available at ScienceDirect
Computers and Operations Research

journal homepage: www.elsevier.com/locate/cor
Review
Survey of quantitative methods in construction

Dirk Briskorn1, Michael Dienstknecht1,∗
Bergische Universität Wuppertal Lehrstuhl für BWL, insbesondere Produktion und Logistik Rainer-Gruenter-Str. 21, Wuppertal, 42119 Germany
a r t i c l e i n f o a b s t r a c t
Article history: Quantitative methods and techniques from operations research (OR) are well-accepted in many industries,
Received 25 March 2017 e.g. the manufacturing industry or the transportation industry. Similarly, numerous applications in the
Revised 1 November 2017
field of construction can be found in the academic literature. This paper gives an extensive overview
Accepted 20 November 2017
on a number of popular fields where OR methods are applied in the construction industry. These fields
Available online 22 November 2017
comprise layout and location planning for construction facilities, scheduling of construction projects and
Keywords: problems related to construction cranes. While the first two topics relate to traditional problems in the
Survey field of OR, the latter one is motivated mainly by the construction industry. In either case, this review
Layout planning presents a survey of papers in the scientific literature. Each paper is categorized and problem setting as
Project management well as techniques applied are briefly described.
Construction cranes
© 2017 Elsevier Ltd. All rights reserved.
1. Introduction reflect uncertainties as well as complex interdependencies is

particularly suitable and allows for analyzing different scenarios.
In 1960, Heiman (1960) wondered whether operations research Regarding the aspects of uncertainties and planning complexity,
(OR) could be applied to construction in order to increase effi- Chan et al. (2009) promote the use of fuzzy techniques and
ciency compared to planning by gut feeling or experience. It has give examples for their employment (e.g. contractor selection,
been stated that OR had proven to be helpful in other disci- scheduling, layout planning, risk assessment, productivity estima-
plines and, thus, could support planning in construction projects. tion) in eight construction-oriented journals from 1996 to 2005.
Since then, a vast amount of OR-related literature tackling planning Liao et al. (2011) focus on the use of metaheuristics such as tabu
problems in construction by different means has been published. search, GAs and simulated annealing for various problems during
Several review papers have appeared in the past. the construction life cycle. Similarly, Sarker et al. (2012) give an
Goh (2008) analyzes the application of quantitative techniques overview on the application of quantitative techniques such as
based on publications in two particular journals – Construction heuristics, mathematical programming or simulation in construc-
Management and Economics and Journal of Construction Engineering tion projects. Finally, Jato-Espino et al. (2014) review more than
and Management – from 1983 to 2006. The survey distinguishes 20 techniques for multi-criteria decision making in construction
conventional techniques (regression models, time series analysis, environments and name cases of application in the period from
probability functions, simulations) and AI techniques (artificial 1992 to 2013.
neural networks, genetic algorithms (GAs), fuzzy techniques, case- In contrast to most of the above mentioned reviews, this paper
based reasoning, expert systems). These are applied to problems in is not structured according to the OR techniques applied, but ac-
construction economics (bid estimation, cost estimation, demand cording to the field of application. It gives an extensive overview
forecasting) and construction management (resource allocation, on literature covering a selection of planning problems that arise
equipment selection, layout planning, scheduling). It is found in construction management. This selection is limited to the three
that there is a positive trend in employing AI techniques to fields that are most comprehensively considered in our opinion.
construction management. AbouRizk et al. (2011) give a review This paper’s contribution is twofold. Naturally, it first summa-
on existing simulation tools such as CYCLONE and COSYE and rizes existing literature with focus on the problem variants being
their application. They argue that simulation with its ability to tackled. Methods applied are briefly stated, as well. We categorize
the problems tackled according to their structure rather than real-
world concepts being represented. This is due to the fact that the
∗
Corresponding author. problems’ structure often is identical although different concepts
E-mail addresses: briskorn@uni-wuppertal.de (D. Briskorn), dienstknecht@uni- are represented. By focussing on the structure we, thus, reveal
wuppertal.de (M. Dienstknecht).
1
similarities or even equivalences between problem variants devel-
http://prodlog.wiwi.uni-wuppertal.de .
https://doi.org/10.1016/j.cor.2017.11.012
0305-0548/© 2017 Elsevier Ltd. All rights reserved.
D. Briskorn, M. Dienstknecht / Computers and Operations Research 92 (2018) 194–207 195
oped independently. This enables researchers to re-apply methods tural commonalities and unify the phrasing we refer to this
to seemingly different problem variants that, in fact, just differ in component of objective functions as total weighted proximity
their verbal descriptions. Secondly, for each type of planning prob- cost (TWPC).
lem we identify generic OR models which can be found at the core • The linear AP employs assignment costs for each object and
of many more involved problem variants. This relates the problem each position. A layout is then evaluated by the total cost
variants discussed to classical OR problems and points to basic ap- of chosen assignments. Again, various interpretations can be
proaches to tackle them. found, e.g. set-up costs, associated risk or utility when installing
The remainder of the paper is organized as follows: the fol- a facility in a certain position. We refer to this component of
lowing two sections are dedicated to two well-known optimiza- objective functions as total assignment cost (TAC). As opposed
tion problems with applications in the field of construction, i.e. to the quadratic AP, the linear AP can be solved in polynomial
Section 2 and Section 3 summarize literature on construction site time and is, therefore, used either as a simplifying problem
layout planning and construction project scheduling, respectively. capturing the main characteristics or as subproblem in order to
In Section 4, research regarding construction cranes – i.e. rather tackle problems in numerous applications.
construction-specific problems – is studied. Finally, Section 5 con-
cludes the paper with a short summary and an outlook on possible
aspects for future research. Most often the objects to be positioned are any kind of con-
struction facilities, but some papers are dedicated to a specific type
2. Construction site layout planning of facility, e.g. storage areas. In the following, the general term fa-
cility is used and it is specified more accurately whenever neces-
Sadeghpour and Andayesh (2015) have reviewed papers on con- sary. Regarding the term position, it can be broadly distinguished
struction site layout planning with a rather application-oriented between discrete approaches where a predefined finite set of avail-
perspective. We, contrastingly, take a rather abstract, methodolog- able locations is given and continuous approaches where any point
ical perspective with a focus on structural problem aspects and on the construction site that is not occupied by any existing struc-
will only briefly discuss common constraints and objectives from ture is available for placing an object. More precisely, we refer to
a practical point of view at the end of this section’s introductory a model or an approach as continuous if there are two different
part. Thus, we start by defining general terms that will be used locations available so that each location in between these two is
while reviewing the literature. Layout planning on a construction available, as well. It should be noted that most researchers dis-
site – as in other fields of application – is concerned with assign- cretize a continuous space by laying a grid over the site. Hence,
ing positions to objects. Usually, a set of objects to be positioned we differentiate with respect to the original problem description
is given and restrictions regarding the positioning have to be con- rather than to the model and categorize papers according to their
sidered, e.g. objects must not overlap and must be placed within problem description rather than the model and solution procedure
a given area. It should be emphasized that this section only cov- developed. If a paper considers a truly continuous solution space,
ers research concerned with placing objects within the boundaries this will be explicitly stated. Another distinction can be made with
of a single construction site. Note that some authors employ con- respect to time. In static approaches, a single layout is planned
cepts for facility location planning in order to tackle such prob- and considered to be valid throughout the planning horizon. A dy-
lems. These approaches, consequently, are summarized in the sec- namic approach, in contrast, respects requirements changing over
tion at hand, as well. Often an objective function is given implying time. For example, a storage place for bricks is needed maybe prior
that not only a feasible positioning but an optimal one or at least to and definitely during building the walls, but afterwards it can
a good one is desired. be removed from the site and its position is free for other equip-
In the literature different objective functions have been pro- ment. Most of these dynamic approaches respect the time dimen-
posed in order to evaluate a given assignment. Two common ones sion by subdividing the whole construction life cycle into periods
are instantiated by the quadratic assignment problem (AP) and the or phases that are planned successively. Andayesh and Sadegh-
linear AP, respectively. pour (2014) correctly point out, that this is more of a phased per-
spective rather than a dynamic one. However, this type of approach
• The quadratic AP employs distances between positions, amount is considered dynamic in this review, since the dynamic nature of
of material to be transported between objects, and – potentially the problem has been recognized and is reflected.
– a cost factor. The effort for transport from one object to an- Most of the reviewed papers are classified in dynamic and
other equals the amount to be transported times the distance static as well as discrete and continuous approaches. According
between the assigned locations times the cost factor. The ob- to this classification, they are listed in Table 1 and are presented
jective of the quadratic AP is to minimize total effort for trans- in more detail in the corresponding Sections 2.1–2.4. Additionally,
port. The quadratic AP is NP-hard, that is it is hard to solve, within these categories, we distinguish single-objective and multi-
and it cannot even be approximated within a constant factor in objective problems. The modelling variety in discrete approaches is
polynomial time (see Burkard, 1984 for details). Nevertheless, much smaller than in continuous approaches. We therefore lay the
since it is one of the most intensively analyzed optimization emphasis on outlining structural commonalities when discussing
problems there are many solution methods available in the lit- these in Sections 2.1 and 2.2. When reviewing continuous ap-
erature, see Loiola et al. (2007). proaches in Sections 2.3 and 2.4 we provide more details about
In construction engineering there is a variety of concepts re- the model and the actual application.
garding the objective above. Most of them rely on the distance There are papers that are rather related to classic facility lo-
between two objects as a first factor. The distance is multiplied cation problems (FLPs); see Klose and Drexl (2005) for a detailed
by a second factor depending on the pair of objects. The inter- review and categorization. These are presented in Section 2.5. Sim-
pretation of this second factor varies among different papers. It ilar to the layout problems from Sections 2.1 to 2.4, FLPs can be
may represent, e.g., amount of material transported, safety fac- categorized into discrete and continuous and static and dynamic
tors, preferences, or simply be an abstract value. Sometimes, a problem variants, as well. While the categorization with respect to
third factor, mostly reflecting variable costs is employed. Note time does not differ from the one for layout problems, we briefly
that all these different interpretations do not influence the ob- outline the difference between discrete and continuous FLPs in the
jective function’s structure. In order to emphasize these struc- following.
196 D. Briskorn, M. Dienstknecht / Computers and Operations Research 92 (2018) 194–207
Table 1
Classification of layout- and location-related research.
discrete continuous
single-obj. multi-obj. single-obj. multi-obj.
layout static Cheng and Lien (2012); Ning and Lam (2013) Easa and Hossain (2008); El-Rayes and Hyari (2005);
Lam et al. (2005); Elbeltagi and Hegazy (2001); El-Rayes and
Lam et al. (2007); Hammad et al. (2017); Khalafallah (2005);
Lam et al. (2009); Li and Hegazy and Elbeltagi (1999); Hammad et al. (2016);
Love (1998); Li and Love (20 0 0); Imam and Mir (1998); Khalafallah and
Liang and Chao (2008); Lien and Mawdesley et al. (2002); El-Rayes (2011); Xu and
Cheng (2012); Mawdesley and Osman et al. (2003); Song (2014)
Al-Jibouri (2003); Sanad et al. (2008);
Warszawski (1972); Zhou et al. (2009);
Wong et al. (2010); Yeh (1995); Zouein et al. (2002)
Yeh (2006); Zhang and
Wang (2008)
dynamic Ning et al. (2010); Ning et al. (2011); Xu and Andayesh and Said and El-Rayes (2010);
Ning et al. (2011) Li (2012) Sadeghpour (2013); Said and El-Rayes (2013a);
El-Rayes and Said (2009); Xu and Song (2015);
Elbeltagi et al. (2001); Yahya and Saka (2014)
Elbeltagi et al. (2004);
Hammad et al. (2015);
Said and El-Rayes (2011);
Said and El-Rayes (2013b);
Zouein and Tommelein (1999)
location static Huang et al. (2010); Warszawski (1972)
Warszawski (1972);
Warszawski (1973);
Warszawski and Peer (1973)
dynamic Chau (2004); Warszawski (1972);
Warszawski and Peer (1973)
In the discrete version, there is a set of customers with given APs traditionally include constraints requiring at most one fa-
demands of a single product and there is a set of potential facility per position and – in the generalized version – constraints
cilities with given supply capacities. Opening a facility is charged allowing only a subset of positions for having assigned a certain fa-
with a facility-dependent fixed cost and gives the opportunity to cility. In construction engineering these constraints are rather com-
supply customers from this facility. Supplying a customer from a mon as well. However, there are additional constraints that are re-
facility yields transportation cost per unit depending on both cus- spected in several papers, e.g., prescribed maximum and/or mini-
tomer and supplier. The objective is to decide which facilities to mum distances between facilities, constraints requiring or prohibit-
open and which quantities to ship so that customer demand is ful- ing the (joint) location of certain facilities in specific site areas or
filled at minimal total cost. In contrast to the assignment problems the need to determine a facility’s orientation on-site.
discussed in Sections 2.1 and 2.2, discrete FLPs are not concerned
with assigning facilities to predefined locations, but to select given 2.1. Static discrete approaches
facilities (or facility locations) on-site. Additionally, transportation
does not occur between facilities to be located, but between fa- In static discrete problems the goal is to identify an assignment
cilities to be located and given customers (i.e. there are no inter- of facilities to positions so that each facility gets a position and
facility flows). no position gets more than one facility. In the basic variant, each
In continuous FLPs, a given number of facilities has to be lo- facility can be assigned to any position. In a generalized problem
cated in the plane. If only a single facility has to be located, the version each facility can be assigned only to a subset of positions.
problem is known as the Weber problem; if more than one facility We first review papers that focus on the basic variant and consider
has to be located (and if it has to be decided which demand point the generalized variant afterwards. With one exception (Ning and
is served completely by which facility), this is referred to as the Lam, 2013) all papers in this section focus on a single objective.
multi-source Weber problem. Again, transportation does not occur The basic variant is equivalent to the quadratic AP if the objec-
between the single facilities, but between facilities and given cus- tive is to minimize TWPC. Lam et al. (2007), Lam et al. (2009),
tomers. Liang and Chao (2008), and Lien and Cheng (2012) employ
Especially for the layout-related research, there is a consider- an ant colony optimization, a hybrid of GA and max-min ant
able variety of practically motivated constraints and goals. When system, tabu search, and a particle bee algorithm, respectively.
defining the basic concepts for our categorization, we already men- Lam et al. (2005) and Li and Love (1998) develop GAs. In ad-
tioned different practical interpretations of the classical OR objec- dition to their GA, Lam et al. (2005) present fuzzy techniques
tives TAC and TWPC like location-dependent set-up costs, risk or and the entropy technique for determining the proximity weights.
utility (for TAC) and transportation effort, safety risk or layout pref- Wong et al. (2010) restrict their scope to an area on a construction
erences (for TWPC). As can be seen in the following subsections, site where pre-cast concrete is produced. They basically consider
most papers rely on these concepts, but there are quite specific ob- the quadratic AP, propose a GA and a mathematical program that
jectives due to special applications and environments that, in turn, is solved by a standard solver.
necessitate specific formulations, as well. These include, e.g., noise Warszawski (1972) presents a mathematical program that ba-
pollution, illuminance (when locating lighting equipment) as well sically models an AP with the objective to minimize TWPC plus
as aviation safety and airport security (at an airport construction TAC. For the same problem, Mawdesley and Al-Jibouri (2003) and
site). Yeh (1995) present a GA and a solution approach that combines
simulated annealing with artificial neural networks, respectively.
Other authors tackle an AP with the objective to maximize developing a decision support system that first identifies the nec-
TWPC plus TAC. A hybrid approach of simulated annealing and essary facilities and their sizes via an expert system, then employs
artificial neural networks (Yeh, 2006), a particle bee algorithm fuzzy techniques to determine the proximity weights between fa-
(Cheng and Lien, 2012) and tabu search (Liang and Chao, 2008) are cilities and finally solves the problem from Hegazy and Elbelt-
employed. agi (1999). Zouein et al. (2002) consider a site on which a num-
The generalized variant of the problem where some objects ber of rectangular facilities with given sizes and known material
cannot be assigned to each position turns out to be less popu- flows has to be placed, i.e. location and orientation have to be de-
lar in the literature. Li and Love (20 0 0) extend their approach in termined in a continuous space. In addition to the above men-
Li and Love (1998) by including unequal-size facilities. Zhang and tioned general constraints, prescribed minimum and maximum
Wang (2008) consider the objective to minimize the sum of TWPC distances between facilities as well as relative positions of facili-
and the TAC and solve the corresponding problem by particle ties have to be respected. A GA is proposed for solving the prob-
swarm optimization. lem. Osman et al. (2003) employ a GA for planning the layout of a
Ning and Lam (2013) are the only ones to perform multi- site of arbitrary shape modelled as a grid. Rectangular facilities of
objective optimization. For the generalized version of the assign- different sizes have to be placed. Easa and Hossain (2008) formu-
ment problem, they propose an ant colony optimization. From a late a mathematical program for the problem to position a number
structural perspective, both objectives – minimization of trans- of facilities on a site that is divided into several rectangular ar-
portation effort and minimization of safety risks – express TWPC. eas. Within these areas, facilities can be positioned arbitrarily with
regard to several constraints comprising minimum and maximum
2.2. Dynamic discrete approaches distances between facilities and prohibited areas for certain facili-
ties. Sanad et al. (2008) tackle a similar problem with a given con-
In dynamic discrete problems the goal is to identify an assign- struction site of arbitrary shape, given facilities of given sizes and
ment of facilities to positions over time. We have a discretized shapes. Minimum and maximum distances are required between
time horizon partitioned into a number of periods. Each facility facilities. The authors discretize the site using a grid and propose
may occupy the assigned position for multiple consecutive periods. a GA to solve the problem. Hammad et al. (2017) consider a rect-
No position can get more than one facility in any period. Again, we angular site with rectangular barriers on it. Rectangular facilities
start with single-objective approaches and consider multi-objective have to be placed and oriented on-site overlapping neither each
ones afterwards. other nor barriers (i.e. barriers render the feasible region for facil-
Ning et al. (2010) propose a sort of ant colony optimization for ity placement non-convex). The goal is to minimize TWPC. As bar-
a problem that is a quadratic AP over T periods with the objective riers are infeasible regions for transportation movements, distances
of minimizing the weighted sum of transportation effort and safety among facilities have to account for these regions. This problem
risk over the whole planning horizon. Both transportation effort setting is formulated as a mathematical program. However, in or-
and safety risk are TWPC from a structural perspective. The prob- der to simplify the setting the continuous region is discretized us-
lem is altered by Ning et al. (2011) in two ways: first, unequal-size ing a grid. For the discretized version, two variants of problem set-
facilities are included and second, besides the weighted objective ting are proposed. In the first one, there may be at most one facil-
function, true multi-objective optimization is employed. For both ity per cell, in the second one, multiple facilities may share a cell.
approaches, ant colony optimization is used and the final decision Both variants are modelled as MIPs. For the first variant, an ex-
is supported by fuzzy techniques. act approach – a cutting plane algorithm – is proposed and tested
Multiple-objective optimization in a construction project that favourably against a standard solver.
spans over T periods is also considered by Xu and Li (2012). They Mawdesley et al. (2002) consider a rectangular construction site
employ a multi-objective particle swarm optimization algorithm that is modelled as a grid and a number of rectangular facilities of
in order to minimize the net present value of costs – compris- different sizes. There are prohibited areas for facilities and distance
ing both deterministic and stochastic elements – on the one hand constraints between facilities. The problem to minimize TAC plus
and safety risks related to the positions of high-risk and high- TWPC is approached using a GA.
protection facilities on the other hand. Zhou et al. (2009) consider a construction site of arbitrary
shape with rectangular or circular zones where no facility can be
2.3. Static continuous approaches placed and rectangular facilities have different sizes. They develop
a GA that respects both hard constraints, e.g. non-overlapping fa-
In continuous layout problems, there are no predefined loca- cilities, and soft constraints, e.g. preferred orientation of a facility,
tions for placing objects, but the whole site area – excluding areas when locating facilities in continuous space.
of fixed objects – is available. Often, this continuous space is dis- A number of papers aim at multi-objective optimization with
cretized by laying a grid over the site. In any case, for each object rather specific applications and, thus, specific components for lay-
the coordinates have to be determined. Some papers additionally out evaluation. El-Rayes and Khalafallah (2005) consider a con-
focus on an object’s orientation, which is usually limited to deter- struction site of a given shape that can be decomposed into rectan-
mining whether an object is placed horizontally or vertically in a gles, where a number of rectangular facilities with different sizes
coordinate system. The objectives are similar to those in discrete has to be located so that safety is maximized and TWPC is min-
approaches and are generally subject to the constraints of non- imized. They propose a GA to solve this problem. El-Rayes and
overlapping of objects and placing objects completely within the Hyari (2005) develop a GA for illuminating a highway construc-
site boundaries. Subsequently, we first present single-objective and tion site. The objective is to determine number, type, configu-
afterwards multi-objective approaches. ration and location of lighting equipment so that on the one
Most papers propose quantitative approaches for minimizing hand average illuminance and lighting uniformity are maximized
TWPC. Imam and Mir (1998) consider a continuous site that can and on the other hand lighting costs and glare are minimized.
be decomposed into rectangles where a number of rectangu- Khalafallah and El-Rayes (2011) propose a GA for multi-objective
lar facilities of given sizes has to be placed. Hegazy and Elbelt- optimization on an airport construction site. A number of rectan-
agi (1999) use a GA for placing facilities of arbitrary shapes on gular facilities has to be placed in order to maximize construction
a given construction site of arbitrary shape that is modelled as a safety, aviation safety and airport security and to minimize lay-
grid. This approach is extended in Elbeltagi and Hegazy (2001) by out costs. Xu and Song (2014) propose multi-objective optimiza-
tion via particle swarm optimization for placing given facilities on areas; their sizes depend on the inventory levels which result from
a given continuous site. The objectives are minimization of TWPC the procurement decisions. The authors develop a GA to minimize
and maximization of both site utilization and logistics relevancy. the sum of relocation cost, TWPC, stock-out cost, capital and order-
In Hammad et al. (2016), the construction site is partitioned into ing cost.
several rectangular areas within which rectangular facilities have Few authors perform multi-objective optimization. Yahya and
to be located continuously. Multiple facilities may be located in Saka (2014) consider a construction project of multiple periods. In
the same area, but facilities must not overlap and must not reach each period, a number of rectangular facilities has to be placed on
out of the respective area. The first objective function is to mini- the site – modelled as a grid – with regard to location and orien-
mize TWPC while the second objective function seeks to minimize tation. The objectives are minimization of TWPC and minimization
noise pollution (measured as the maximum noise level recorded of safety risk – from a structural perspective, a kind of TWPC –
at given on-site receivers). The problem is modelled as a mixed- in each period subject to the fact that some facilities have to be
integer non-linear program. In the model, a location-dependent placed within the radius of a tower crane. The optimization is per-
distance measure rather than the classical Euclidean or Manhattan formed by a multi-objective artificial bee colony algorithm.
distance is used. Trade-off solutions are identified by employing In contrast to Yahya and Saka (2014), the following papers al-
the -constraint method. low relocation of facilities. Said and El-Rayes (2010) consider a spe-
cial type of construction project – a critical infrastructure project.
2.4. Dynamic continuous approaches The problem is similar to that in El-Rayes and Said (2009), but
now security is an additional issue. Therefore, relocation cost and
Just as in Section 2.2, the goal here is to identify an assign- TWPC have to be minimized while security is maximized. With a
ment of facilities to positions over time. Again, we have a dis- GA, non-dominated solutions are found. The paper by Said and El-
cretized time horizon and, first, present single-objective and af- Rayes (2011) is extended by Said and El-Rayes (2013a) in such a
terwards multi-objective approaches. Within these subsections, we way that areas within the building structure can be used as storage
distinguish between the problems that allow relocation of facilities areas as well and the construction schedule can be altered in order
and those that do not. to make areas within the building available at certain times. Two
Elbeltagi et al. (2001) and Elbeltagi et al. (2004)extend the pa- objectives, namely changes in the schedule and the sum of relo-
per by Elbeltagi and Hegazy (2001) discussed in Section 2.3 to a cation cost, TWPC, stock-out cost, capital and ordering cost, are to
multi-period problem. Andayesh and Sadeghpour (2013) consider be minimized. A GA is employed to find non-dominated solutions.
a construction site of arbitrary shape. Circular facilities that re- Xu and Song (2015) consider a multi-period layout problem where
main on-site for given time intervals have to be placed so that rectangular facilities have to be located on a rectangular construc-
TWPC is minimized. The authors discretize neither the construction site. There are two objective functions, namely the minimiza-
tion site nor the planning horizon and propose a concept from tion of distances between the facilities’ centers and the minimiza-
physics called minimum total potential energy for layout optimization of the sum of TWPC and relocation cost over the complete
tion. Minimizing TWPC in a multi-period problem is also con- planning horizon. In the problem, the cost factor for calculating
sidered by Hammad et al. (2015). The construction site is parti- TWPC accounts for uncertainty. A mathematical program is formu-
tioned into several rectangular areas. Rectangular facilities have to lated and a particle swarm algorithm is proposed for solving the
be positioned continuously within these areas. There may be more problem.
than one facility per area, but a facility must not reach out of
an area and must not overlap with another facility. The problem 2.5. Location problems
is formulated as a mathematical program. Additionally, the paper
has two special features regarding the calculation of TWPC (i.e. In Warszawski (1972), a static discrete FLP is modelled
the sum of the products of inter-facility distances and travel fre- with customers representing demand points on the con-
quencies): first, travel frequencies are derived from work schedule struction site and facilities representing feasible supply lo-
and Building Information Modeling (BIM) information. Second, dis- cations. Warszawski and Peer (1973) consider the special
tances are computed depending on the facilities’ locations like in case of locating only a single supply point. Both papers
Hammad et al. (2016). present exact and heuristic solution procedures. An extended
Zouein and Tommelein (1999) propose an approach that com- model that considers multiple supply locations and materi-
bines mathematical programming and a customized heuristic for als is presented in Warszawski (1973) and Warszawski and
the following problem: in each period, a number of rectangular Peer (1973) along with several exact and heuristic solution ap-
facilities has to be placed in terms of location and orientation proaches. Huang et al. (2010) formulate a mathematical program
on a given continuous construction site so that TWPC and to- for the construction of a high rise. Each storey implies demand
tal relocation cost are minimized over all periods. As constraints, for multiple materials. Some storeys which are divided into cells
both minimum and maximum distances between facilities and rel- can be used for storing materials and supplying others. Such cells
ative positions of facilities have to be respected as well as restric- correspond to facilities in the FLP while storeys with demand
tions regarding the assignment of facilities to areas. El-Rayes and correspond to customers.
Said (2009) focus on a rectangular construction site – modelled Some papers consider dynamic discrete FLPs over multiple
as a grid – over a number of planning periods. Rectangular fa- periods. Warszawski (1972) presents an extended mathematical
cilities have to be placed in terms of location and orientation in program that considers multiple periods and materials and re-
order to minimize relocation cost and TWPC over the whole plan- spects period-, material- and location-specific operating costs of
ning horizon. There are constraints regarding minimum and maxi- supply points. Similarly, Warszawski (1973) and Warszawski and
mum distances between facilities and area constraints (i.e. certain Peer (1973) introduce a model considering a single material and
facilities have to be or must not be placed in certain areas of the multiple periods and propose a number of exact and heuristic so-
site). The problem is solved by approximate dynamic programming lution procedures. Chau (2004) formulates a mathematical pro-
(DP). An alternative solution procedure, namely a GA, is proposed gram for a multi-echelon distribution of material. It can be trans-
in Said and El-Rayes (2013b). In Said and El-Rayes (2011) a similar ported directly from a supply to a demand point or via a trans-
problem is considered. A special feature of this paper is the inte- shipment center. There is a given set of available transshipment
gration of procurement planning. Some of the facilities are storage centers. Opening, operating and closing a transshipment center is
charged with costs. For each period it has to be decided which Although the problems presented in this section are close to
transshipment centers are used and which material flows occur. RCPSP and MRCPSP regarding their structure, there is a consid-
The objective is to minimize total costs over the planning horizon. erable variety of objectives and constraints that can be observed
A two-stage solution approach is proposed. While a GA searches in construction project scheduling. Regarding objectives, it can be
for good configurations of transshipment centers, a transshipment broadly distinguished between four categories: time-oriented, cost-
problem is solved in order to evaluate such a configuration. oriented, resource-oriented and, less prominent, quality-oriented
Warszawski (1972) considers a static continuous version of the objectives. Time-oriented objectives mostly pick up the classical
FLP where demand points with given coordinates and demands goal of minimizing a project’s makespan. With respect to cost ori-
have to be supplied by a given number of supply points that have entation, there are different types of costs that are considered.
to be located. These can be costs related to the project’s duration or costs related
to the single activities. The latter may be related to the activities’
3. Construction project scheduling durations if these can be decided, e.g. by selecting a certain mode
to perform an activity. Project duration-related costs may account
In general, project scheduling is concerned with assigning start for penalties for violating given deadlines or rewards for keeping
times and completion times to each activity in a given set. Usually, them. Resource-oriented objectives focus on minimizing peak re-
times cannot be assigned arbitrarily but a set of constraints has source usage or minimizing deviations from a certain usage level.
to be considered. A well-known project scheduling problem in OR Less prominent is the consideration of work quality in the objec-
is the resource-constrained project scheduling problem (RCPSP). It tives, i.e. performing activities in a reduced amount of time dimin-
encompasses a set of activities with precedence relationships and ishes the project’s quality and, e.g., quality loss shall be minimized.
a set of capacitated resources that are needed for executing ac- In addition to the spectrum of objectives, there exists a range
tivities. The objective is to find a schedule that minimizes the of constraints, as well. Besides classical constraints like precedence
makespan, i.e. the completion time of the last activity, while sat- constraints among activities and limited capacities of resources,
isfying resource capacities and precedence relationships. An ex- there are constraints that account for the planning environment of
tended problem is the multiple-mode RCPSP (MRCPSP) where each construction projects. There may be milestones for example, i.e. a
activity can be performed in a number of modes. With each mode, certain construction progress has to be realized at a given point
a certain duration, cost and resource consumption are associated. in time. Furthermore, there may be given budgets for either the
Thus, it has to be decided in which mode an activity is per- whole project or sets of activities. Quality of activities is consid-
formed. There are numerous variations of RCPSP. For an overview, ered, as well, i.e. activities may be performed at different speed or
the reader is referred to Brucker et al. (1999) and Hartmann and in different modes with a certain associated quality, but a mini-
Briskorn (2010). Although even fairly special cases of RCPSP and mum quality must be met. Finally, there are sets of activities that
MRCPSP are NP-hard (see, e.g., Briskorn and Fliedner, 2012 and have to be performed simultaneously, while others must not be
Brucker et al., 1999) researchers can build on an extensive collec- performed simultaneously. Some researchers include uncertainty,
tion of solution methods since both problems and further variants e.g. for activity durations or budgets.
have been tackled in countless papers, see Brucker et al. (1999). For scheduling construction projects, Zhou et al. (2013) present
In the literature, we mainly encounter variants of RCPSP and a survey. They state that construction scheduling basically is an
MRCPSP, respectively, potentially differing in the objective. We application of the above mentioned variations of project schedul-
would like to emphasize two important variants here which of- ing. The reviewed papers are classified into three categories: (i)
ten are not put into relation to RCPSP in the literature. Firstly, mathematical methods like the critical path method, DP, linear and
a repetitive construction project is a special type of construction mixed-integer programming, (ii) problem-specific heuristic meth-
project where a construction phase consisting of a set of activi- ods, and (iii) metaheuristics like GAs, simulated annealing and tabu
ties is repeated multiple times. Applications for this type of project search. This section focuses on an update of Zhou et al. (2013) and,
are, e.g., construction projects of highways, pipelines, and high- thus, picks up the classification scheme. Table 2 summarizes the
rises. Secondly, some construction projects allow a special feature papers gathered and classified accordingly in Sections 3.1–3.3. In
called fast-tracking. Fast-tracking means that succeeding activities addition, it incorporates a categorization with respect to the prob-
may overlap in order to reduce the makespan.
Table 2
Classification of scheduling-related research.
Methodology Problem variant
RCPSP MRCPSP repetitive fast-track other
Mathematical Lorterapong and Menesi and Hegazy (2015); Ammar (2013); Liu and El-Kholy (2015); Wang et al. (2015)
methods Ussavadilokrit (2013) Menesi et al. (2013) Wang (2012); Tang et al. Khoueiry et al. (2013);
(2014a); Tang et al. Kim et al. (2012);
(2014b) Klansek and
Psunder (2012)
Problem-specific Li et al. (2015); Heravi and Faeghi (2014) Bakry et al. (2014); Bakry et al. (2014);
heuristics Ng et al. (2013) Dolabi et al. (2014) Moselhi and
Roofigari-Esfahan (2013)
Metaheuristics Alghazi et al. (2013); Ashuri and Agrama (2014); Cho and Hastak (2013) Sonmez and Uysal (2014)
Bettemir and Tavakolan (2015); Damci et al. (2013);
Sonmez (2015); Fu (2014); Fan et al. (2012)
Bozejko et al. (2013); Ghoddousi et al. (2013);
Cheng et al. (2014); Li and Zhang (2013);
Dong et al. (2012); Monghasemi et al. (2015);
Faghihi et al. (2014) Sonmez and
Bettemir (2012);
Yang et al. (2014)
lem variant – RCPSP, MRCPSP, repetitive project, fast-tracking or a given minimum duration. Additionally, crashing an activity im-
other. poses a given quality loss cost for this activity. It has to be decided
Note that project scheduling is not the only scheduling applica- which activities to crash by which amount of time so that total
tion in the field of construction. An interesting problem that inte- cost is minimized. Klansek and Psunder (2012) consider a simi-
grates scheduling and vehicle routing is the concrete delivery prob- lar problem and propose a non-linear MIP. For a project consist-
lem. However, since Kinable et al. (2014) provide a state-of-the-art ing of precedence-related activities, a budget and a deadline are
review on related research in their original research paper, we re- given. Activities can be crashed, but in turn, costs increase. The
strict ourselves to the more common project scheduling literature objective is to find start and completion times for all activities so
here. that total project cost is minimized. Total project costs include di-
rect costs for each activity (a function of an activity’s duration),
3.1. Mathematical methods indirect costs (a function of the project’s duration) and a penalty
or reward for violating or satisfying the deadline. The research by
We first consider papers concerning variants of the RCPSP and El-Kholy (2015) is dedicated to a time-cost trade-off problem, as
MRCPSP, respectively. Lorterapong and Ussavadilokrit (2013) con- well, but includes stochasticity. A set of precedence-related activi-
sider a project scheduling problem that is captured by the RCPSP. ties has to be performed. Each activity imposes costs depending on
The problem is represented as a constraint satisfaction problem its duration which is uncertain. The available funding of the project
and a solution procedure consisting of multiple stages is proposed. is uncertain, as well, and must not be exceeded with a certain
Menesi et al. (2013) focus on a variant of the MRCPSP instead. probability. Additionally, a project deadline has to be respected.
Violating or satisfying a given deadline results in a penalty or re- The objective is to minimize the costs for performing all activi-
ward. The objective is to minimize the sum of direct costs (of ac- ties. A mathematical program is formulated and solved by chance-
tivities), indirect costs (depending on the project duration), penalty constrained programming.
and reward. In a follow-up paper (Menesi and Hegazy, 2015), A construction program – i.e. a number of interrelated construc-
project deadline and costs are neglected. Instead, two variants of tion projects – is considered by Wang et al. (2015). Projects have
objective functions are considered: the first aims at minimizing the different priorities and there are alternative projects that can be
makespan, the second additionally aims at minimizing the peak re- chosen from. Furthermore, alternative activities within a project
source usage. Each of these problems is solved by constraint pro- are considered. For example, to pass a river, a bridge or alterna-
gramming (CP) techniques. tively a tunnel can be built. A modified critical path method is
Also, repetitive construction projects are tackled using math- used for scheduling the activities.
ematical methods. Liu and Wang (2012) propose CP in order to
schedule a repetitive construction project. In each construction 3.2. Problem-specific heuristics
phase, a set of precedence-related activities has to be performed
by a number of crews. There are single-skilled and multi-skilled Again, we focus on variants of the RCPSP and MRCPSP first. An
crews. The time needed for performing an activity depends on iterative heuristic approach for preemptively scheduling activities
the assigned crews: usually, a single-skilled crew executes an ac- and assigning resources in a project is proposed by Ng et al. (2013).
tivity, but employing an additional multi-skilled crew speeds up In each step, the procedure identifies maximum subsets of activi-
the activity. Starting times of activities and the assignment of ties that can be feasibly scheduled at the moment. Among these
multi-skilled crews to activities have to be determined so that the subsets, the one with the resource demand closest to the free re-
project makespan is minimized. Ammar (2013) develops a schedul- source capacity is started. Processing of these activities stops as
ing approach for repetitive projects that combines the critical path soon as the shortest activity in the group is completed. All re-
method with the graphical line-of-balance method. Tang et al. sources occupied by the group are released and the next itera-
(2014b) investigate a repetitive construction project with a given tion starts, until all activities are completed. Li et al. (2015) in-
duration where precedence relations, time buffers between activi- clude stochastic activity durations in their research. Each activity
ties as well as minimum and maximum resource usage levels for has a stochastic duration and certain resource consumption per
each activity have to be considered. Activities have to be scheduled time unit. A given deadline has to be satisfied with a certain prob-
and resource usage levels have to be assigned to activities so that ability. The objective is to determine a scheduling policy, i.e. a list
the resource usage is balanced. The objective is formulated as min- of activities, that minimizes the expected sum of the weighted co-
imizing the sum of the absolute resource usage deviations on two efficient of variation of the total resource usage. Two customized
succeeding days. The problem is tackled via CP. In another paper, heuristic approaches are presented.
Tang et al. (2014a) extend the objective function, which is now the Heravi and Faeghi (2014) consider a variant of MRCPSP where
weighted sum of resource usage deviations and schedule changes. modes are associated with different durations, costs and qualities.
The latter is measured as the absolute deviation in total resource Three objectives are treated: minimization of makespan, minimiza-
usage of an initial schedule and a revised schedule. Again, CP is tion of costs and maximization of quality. The proposed solution
employed for problem solving. procedure generates a set of solutions that are evaluated with re-
Another part of the literature is concerned with reducing activ- spect to these objectives. Afterwards, the alternatives are ranked
ities’ durations or the makespan by means of overlapping activities based on the evaluation via group decision making. Since activ-
or speeding up activities. Khoueiry et al. (2013) consider a problem ity durations and costs are stochastic, Monte Carlo simulation is
with fast-tracking. Reducing the makespan gains money, but fast- employed during the evaluation. For estimating the quality, fuzzy
tracking might cause costs for rework. The objective is to find the – techniques are used.
in terms of associated net benefit – optimal point in time for start- Only a few customized heuristics tackling repetitive project
ing the activity overlap. The authors propose a mathematical pro- scheduling problems or concerning reduction of activities’ dura-
gram. Another way of reducing the makespan is crashing certain tions can be found in the recent literature. Dolabi et al. (2014) pro-
activities, i.e. reducing their durations. Kim et al. (2012) formu- pose two heuristics for linear construction projects that are based
late a mixed-integer program (MIP) for such a problem. A project on the research by Ammar (2013). Crews and start dates have to be
consisting of multiple activities with precedence relations has to assigned to activities while a deadline has to be kept and prece-
be finished before a certain deadline. A given number of activi- dence relations between activities have to be respected. The ob-
ties might be crashed, but each activity cannot be crashed below jective is to minimize the number of active crews in the project.
Bakry et al. (2014) are concerned with speeding up a repetitive afterwards via Shannon’s entropy technique and evidential rea-
construction project in a cost-optimal way. In a first step, the soning. A variation of the MRCPSP is considered by Ashuri and
project progress is updated and the project is re-scheduled. Then, Tavakolan (2015). Milestones, i.e. certain levels of construction
a heuristic procedure is used to select activities to be sped up, progress at certain dates, are given. Each mode determines the ac-
evaluate the costs of different strategies for reducing activity du- tivity’s duration, cost of processing, whether the activity can be
rations (overtime, double shifts etc.) and assign strategies to ac- preempted and the costs related to interrupting and continuing the
tivities. The procedure is repeated until the necessary reduction in activity. For each activity, a mode, start times, and times when to
project completion time is realized or no further reduction is pos- interrupt have to be chosen. Multiple objectives, namely minimiza-
sible. tion of project duration, project cost, and sum of squared resource
Moselhi and Roofigari-Esfahan (2013) consider a construction consumptions, are considered in a shuffled frog-leaping method
project of multiple precedence-related activities with a given that aims at finding non-dominated solutions.
project target duration. In order to keep the target duration, sev- Fan et al. (2012) consider repetitive projects where the se-
eral activities have to be crashed. An iterative heuristic procedure quence of activities in construction phases may vary and activ-
is proposed to identify critical activities, assign crashing priorities ities might be processed in parallel. A GA is presented for se-
and determine crash durations. quencing activities and assigning resources to activities in order to
minimize total project cost consisting of direct cost and duration-
3.3. Metaheuristics related cost. Damci et al. (2013) focus on a repetitive construction
project where multiple resources have to be leveled, i.e. the daily
In the project studied by Dong et al. (2012), a number of rooms deviations of the resource usage from the average resource usage
has to be built. Within each room a number of activities with have to be minimized. Line-of-balance scheduling and a GA are in-
given durations and precedence relations is to be executed. Re- tegrated in order to tackle this setting. Agrama (2014) employs a
sources – here, i.e. crews – are capacitated. Additionally, zone and GA for scheduling the construction of a multi-storey building. Mul-
blocking constraints have to be respected. Zone constraints force tiple objective values – project duration, number of active crews
some activities to be performed simultaneously by the same crew. and activity interruptions – are addressed in a weighted objective
A blocking constraint means that performing an activity in a cer- function.
tain room blocks other rooms, i.e. no operation can take place A GA for finding an optimal trade-off between project cost and
in these rooms. Activities have to be scheduled in order to mini- duration in a project where fast-tracking is allowed is proposed by
mize project duration or cost. The problem is formulated as math- Cho and Hastak (2013).
ematical program and solved by a GA. Bozejko et al. (2013) inter- Sonmez and Uysal (2014) consider several projects that have to
pret a construction scheduling problem as a total weighted tardi- be scheduled simultaneously. Within the single projects, activities
ness flowshop problem and propose a solution approach that com- are precedence-related and consume capacitated resources. The ac-
bines scatter search, GA, simulated annealing and tabu search. In tivities have to be scheduled in a way that minimizes the latest
Alghazi et al. (2013), a GA is used to incorporate cash constraints finishing time among all projects. For solving the problem, a com-
in project scheduling. Whereas Cheng et al. (2014) propose a hy- bination of critical path scheduling, simulated annealing and a GA
brid approach of differential evolution, fuzzy c-means clustering is presented.
and chaotic technique for solving a standard RCPSP, Bettemir and
Sonmez (2015) combine simulated annealing and a GA. The inte- 4. Construction cranes
gration of a building information model and a GA for scheduling
a project with the goal of maximizing its constructability is pre- Construction cranes give rise to various optimization prob-
sented by Faghihi et al. (2014). lems and have been treated in the literature considerably. For an
Sonmez and Bettemir (2012) consider a variant of the MR- overview on cranes and crane operations the reader is referred to
CPSP where total project cost, consisting of direct cost imposed Hornaday et al. (1993) and Shapira et al. (2007). The most promi-
by modes and indirect cost that depend on the project dura- nent decisions addressed in optimization problems consider crane
tion, are to be minimized. A combination of GA, simulated an- selection, crane location and lift planning. These types of problems
nealing and quantum simulated annealing is employed for solving are covered by Sections 4.1, 4.2 and 4.3, respectively. Finally, ap-
the problem. An ant colony optimization algorithm is employed proaches that do not fit into this classification scheme or that inte-
by Li and Zhang (2013) in order to tackle a standard MRCPSP. In grate different fields of application are presented in Section 4.4. In
Ghoddousi et al. (2013), each activity can be performed in dif- each section, we distinguish between approaches for mobile cranes
ferent modes on alternative resources. For this setting, a GA is and those for tower cranes. The literature and its categorization are
used in order to find non-dominated solutions, i.e. a start time, summarized in Table 3.
a resource and a mode assigned to each activity, with regard to While the modelling variety in crane selection and lift planning
minimization of makespan, project cost, and squared deviation problems is rather small, crane location problems are more diverse.
from the average resource usage for all resources. Fu (2014) in- Regardless of the considered crane type, they differ with respect to
tegrates procurement planning in the MRCPSP. It has to be de- the perspective of crane location which is either discrete or contin-
cided which activity to start when and in which mode and how uous, i.e. there is either a finite set of locations to be chosen from
many batches of a material to order at which time. The goal is or an infinite set. Additionally, it can be distinguished between ap-
to minimize the sum of schedule- and procurement-related cost proaches locating a single crane or several cranes – in the latter
terms. This problem is formulated as an MIP and solved by a hy- case, it can be further differentiated between approaches for ho-
brid approach of harmony search and GA. Yang et al. (2014) in- mogeneous and heterogeneous cranes. With regard to objectives,
tegrate Monte Carlo simulation, support vector regression and there are three broad categories: time-, cost-, and safety-oriented
multi-objective particle swarm optimization in order to find non- objectives. Time-oriented objectives focus on minimizing (average)
dominated solutions with respect to minimization of project costs transportation times by locating cranes. Depending on the cost to
and makespan in an MRCPSP with stochastic activity durations and be considered, cost-oriented objectives are closely related to time-
costs. Monghasemi et al. (2015) consider a similar problem that oriented objectives. This is the case when transportation amounts
is deterministic, but includes maximizing quality as a third ob- are weighted by a cost factor instead of a rate for transportation
jective. A GA identifies non-dominated solutions that are ranked time. Sometimes, costs include – maybe location-specific – charges
Table 3
Classification of crane-related research.
Field of application Literature
Crane selection Al-Hussein et al. (2001); Al-Hussein et al. (2005); Hanna and Lotfallah (1999); Hasan et al. (2013); Sawhney and Mund (2001); Shapira and
Goldenberg (2005); Shapira and Goldenberg (2007); Shapira and Schexnayder (1999); Wu et al. (2011)
Crane location Huang et al. (2011); Irizarry and Karan (2012); Lien and Cheng (2014); Rodriguez-Ramos and Francis (1983); Safouhi et al. (2011); Tam and
Tong (2003); Tam et al. (2001); Tantisevi and Akinci (2008); Zhang et al. (1996); Zhang et al. (1999)
Lift planning AlBahnassi and Hammad (2012); Ali et al. (2005); Chang et al. (2012); Kang and Miranda (2006); Kang and Miranda (2008); Lei et al. (2013a);
Lei et al. (2013b); Leung and Tam (1999); Leung et al. (2001); Lin et al. (2014); Olearczyk et al. (2014); Reddy and Varghese (2002);
Sivakumar et al. (2003); Tam et al. (2002); Tam et al. (2004); Tantisevi and Akinci (2009); Zavichi et al. (2014); Zhang and Hammad (2012a);
Zhang and Hammad (2012b)
Miscellaneous Al-Humaidi and Tan (2009); Briskorn and Dienstknecht (2017); Furusaka and Gray (1984); Kim et al. (2011); Lin and Haas (1996); Shapira and
Lyachin (2009); Shapira and Simcha (2009a); Shapira and Simcha (2009b); Shapira et al. (2012); Tam and Fung (2011)
for setting up and dismantling cranes, as well. Safety-oriented ap- pose algorithms for selecting mobile cranes for a given construc-
proaches usually focus on collision potential among cranes and, tion project. An iterative procedure to filter all technically feasible
thus, have to operationalize this goal, e.g. by the extent to which mobile crane configurations from a database for a given lift opera-
the cranes’ operating areas overlap. tion is introduced by Wu et al. (2011).
Restrictions are often not clearly defined or are implied in some For a given construction project, Hasan et al. (2013) present a
assumptions. Most approaches just require the coverage of certain framework for tower crane selection that respects simulation re-
objects (these can be geometric objects like points or polygons or sults, the crane’s productivity and carbon emission.
even the whole site). Different interpretations of coverage can be
found in the literature: an object may be covered jointly by several
4.2. Crane location problems
cranes or it has to be covered completely by at least one crane.
However, with regard to constraints, there may be some potential
In crane location planning, the location of a single crane or
for future research as the set of considered restrictions appears to
several cranes has to be determined considering various site con-
be quite marginal to date. Such attempts could focus on, e.g. in-
straints. Reference problems from OR depend on the exact prob-
troducing infeasible areas for crane location or enforcing safety as-
lem variant under consideration, i.e. the features of cranes cap-
pects by hard constraints.
tured by the problem setting. In a simple variant, cranes are han-
dled like any other facility on the site. Consequently, this can be
4.1. Crane selection problems regarded as a specific application of a layout or location problem
from Section 2 where cranes (facilities) have to be placed in order
Crane selection is concerned with the selection of the appro- to supply certain areas with material. A less simplifying problem
priate crane types, cranes, and crane configurations for a construc- variant focuses on a two-dimensional perspective where cranes
tion project. Most decision support approaches tackling this type of (discs) have to be placed in some area (the construction site) so
problem have their focus on laying the ground for the application that other areas (buildings, e.g.) are covered. Covering problems,
of OR approaches, e.g. by identifying important drivers, rather than in general, have received considerable attention in the OR litera-
on formulating models or applying algorithms. However, a classic ture (cf. Farahani et al., 2012). We would like to emphasize that
OR problem capturing a flavour of crane selection certainly is the the problem described has a set cover flavour and we even have
set cover problem, see Caprara et al. (20 0 0), as a set of require- a geometric interpretation of subsets to be chosen (buildings be-
ments regarding different types of capabilities has to be met by ing covered when a particular crane is placed in a particular posi-
the selection of cranes. In the first part of this section, we present tion) so that each element of the groundset is covered. Problems
approaches not restricted to a particular type of crane. Afterwards, of this type are referred to as geometric set cover problems, cf.
the focus is on mobile cranes and, finally, tower cranes are consid- Clarkson and Varadarajan (2007). If discs – i.e. working areas of
ered. cranes – are preferred not to overlap, we may have a flavour of
Shapira and Goldenberg (2005) emphasize the importance of circle packing problems, see Castillo et al. (2008).
soft factors in equipment selection in general (not restricted to For a given construction project, Tantisevi and Ak-
cranes) and present an analytical hierarchy process-based approach inci (2008) present a simulation-based procedure to identify
for the quantification of soft factors which then can be com- possible mobile crane locations that avoid spatial conflicts and
bined with hard factors to compute a score for each equipment. minimize the number of crane relocations when performing a
The same authors identify 27 important soft factors in equipment number of lift tasks. Safouhi et al. (2011) propose an algorithm to
selection (not restricted to, but mainly focused on cranes) and determine collision-free areas for positioning a given mobile crane.
critisize their lack of consideration in both practice and research Location of tower cranes in the literature is mostly concerned
(Shapira and Goldenberg, 2007). Furthermore, they find that there with single cranes only. Rodriguez-Ramos and Francis (1983) con-
is no structured equipment selection process in practice and thus sider a problem where a tower crane serves a set of construc-
propose one that respects both hard and soft factors. Hanna and tion facilities with known locations. Servicing a facility with its
Lotfallah (1999) use fuzzy techniques to select the right crane type specific demand causes transportation cost depending on the de-
for a given construction project based on both quantitative and mand and the distance between facility and crane. The authors
qualitative factors. Sawhney and Mund (2001) develop a tool called treat this problem as a single facility location problem with rec-
IntelliCranes that helps the user in selecting the right crane type tilinear distances and solve it by means of graph theory. In
and crane for a given project. A neural network is employed to find Zhang et al. (1996), a set of supply points and a set of demand
the best crane type. Finally, an expert system finds feasible crane points are given. Transportation amounts for each pair of supply
models. and demand point are known. Additionally, a set of feasible crane
Shapira and Schexnayder (1999) conduct interviews with ex- locations is given. A simulation based approach is used to find the
perts in order to identify and rank relevant factors in mobile crane optimal crane location, i.e. the location with the minimal average
selection. Al-Hussein et al. (2001) and Al-Hussein et al. (2005) pro- transportation time. Tam et al. (2001) develop a GA in order to
solve another problem concerning tower cranes. For a given set crane. A special type of mobile crane that is allowed to move dur-
of demand points, a given set of possible supply points, known ing the lift is considered by Lin et al. (2014). They aim at finding
transportation amounts between supply and demand points, given a short feasible lift path by employing a rapidly-exploring random
transportation cost rates and a set of possible crane locations, trees algorithm.
the objective is to select supply points and a crane location so Kang and Miranda (2006) and Kang and Miranda (2008) first
that total transportation cost is minimized. In a follow-up paper, develop several search algorithms for finding and optimizing a lift
Tam and Tong (2003) extend the proposed approach by adding path for a single tower crane that transports a given object from its
the prediction of hoisting times via an artificial neural network. pick point to its destination and then introduce an iterative pro-
Huang et al. (2011) formulate an MIP for a more general tower cedure to coordinate multiple tower cranes to avoid collisions. A
crane problem where more than one type of material is consid- framework that allows for coordination of multiple cranes is also
ered. provided by AlBahnassi and Hammad (2012). Paths are planned
Only a few papers focus on the location of multiple tower initially via rapidly-exploring random trees algorithms and can be
cranes. Zhang et al. (1999) consider a setting similar to the one adapted in short time when new information is available. The co-
in Zhang et al. (1996). Here, however, there is a group of cranes, ordination of cranes is achieved by prioritization of cranes. The
but transportation between two points cannot be conducted jointly higher a crane’s priority, the earlier it is planned, and is then
by more than one crane. Irizarry and Karan (2012) propose a considered as a dynamic obstacle for cranes with lower priority.
graphically supported planning approach to locate tower cranes Zhang and Hammad (2012b) propose a multi-agent system for co-
on a construction site and reduce collision potential. Lien and ordinating multiple cranes.
Cheng (2014) consider a tower crane problem with given supply Some authors study cooperative lifts of multiple cranes.
and demand points where the supply capacity of each supply point Sivakumar et al. (2003) apply an A∗ algorithm and hill climbing
is limited and each demand point has a certain given demand. For for finding a lift path for two identical mobile cranes lifting a
a given set of cranes and a given set of possible crane locations, given object. A GA is used by Ali et al. (2005) to find a colli-
a particle bee algorithm is proposed to determine transportation sion free lift path for an object that is lifted by multiple cranes.
amounts between supply and demand points and to select the Chang et al. (2012) rely on the probabilistic roadmap method for
crane locations that minimize total costs, that comprise transporta- planning the lift path of a given object that is moved either by a
tion and crane costs. single crane or two cranes cooperatively.
The prediction of hoisting times for both mobile and
4.3. Lift planning problems tower cranes is another topic treated in the literature. For
tower cranes, Leung and Tam (1999), Leung et al. (2001) and
Lift planning problems are concerned with the transport of Tam et al. (2002) identify factors impacting hoisting times and ap-
objects by means of cranes. There are two types of typical ply multiple regression and artificial neural network models for
OR problems related to this area. The problem to sequence a prediction. Similarly, Tam et al. (2004) propose different artificial
set of transport requests to be conducted by some mean of neural network models for predicting mobile crane hoisting times.
transport is captured by the pickup and delivery problem, see
Berbeglia et al. (2007). There are even variants where pickup and 4.4. Integrated problems and other construction cranes problems
delivery locations are positioned on a circle which comes very
close to the moving characteristics of many crane types on con- First, we focus on problems that integrate multiple of the deci-
struction sites. The second typical OR problem stems from compu- sions discussed in Sections 4.1–4.3.
tational geometry and is referred to as path or motion planning. In Furusaka and Gray (1984) employ an MIP-based approach and
a basic variant, a robot – i.e. a construct of links and joints – has DP for crane selection and location so that the construction site is
to move from a given starting position to a given destination in fully covered. The objective is to minimize total costs comprised of
two-dimensional space under the presence of polygonal obstacles. hire, assembly and dismantling costs over the length of the con-
A feasible or even a shortest path from the starting position to the struction project. Lin and Haas (1996) tackle crane selection and
destination has to be found (cf. de Berg et al., 2008). location integratedly. They employ an MIP-based approach in or-
Zavichi et al. (2014) formulate an MIP for the crane service se- der to minimize the number of crane relocations while respect-
quencing problem. A single tower crane with a given location is ing safety measures. Finally, Briskorn and Dienstknecht (2017) also
surrounded by a set of supply and demand points with known consider crane selection and crane positioning on a given construc-
coordinates. Transportation amounts between supply and demand tion site to be decided integratedly. Cranes have to be selected
points and hook travel times between locations are known as well. from a given set of crane types so that the chosen cranes can
The goal is to find a time minimal service sequence. be positioned to fulfill cover requirements at minimum cost. The
Lift path planning problems aim at finding a feasible – i.e. colli- cover requirements necessitate to cover pairs of buildings and stor-
sion free – path on which a given crane located at a given position age locations jointly by a single crane each in order to implement
can move a lifting object from a pick position to its destination on the transport relation between them. Cranes cannot be placed in
a given construction site. Most – but not all – approaches for lift buildings, in one of a set of given blocked areas or outside the
path planning are based on concepts from kinematics / robotics. construction site. The authors reduce the infinite set of position
For a mobile crane lift, Reddy and Varghese (2002) propose a candidates in the area to a finite set. This enables them to repre-
heuristic search procedure to find a lift path. Tantisevi and Ak- sent the problem as a special case of the set cover problem and to
inci (2009) use a cyclic coordinate descent-based procedure to gen- employ CPLEX in order to find optimum solutions.
erate a sequence of mobile crane moves to lift an object from its Furthermore, there are several approaches in the literature tack-
pick point to its destination. Zhang and Hammad (2012a) employ ling problems that hardly fit in Sections 4.1, 4.2 or 4.3.
rapidly-exploring random trees algorithms to first plan a lift path Kim et al. (2011) present an iterative simulation-based ap-
of a hydraulic crane and later adapt the plan when new informa- proach for determining a cost-optimal foundation for a given
tion is available. For a single mobile crane, Lei et al. (2013a) and tower crane. Several researchers focus on safety aspects. Tam and
Lei et al. (2013b) develop methods for checking whether there is a Fung (2011) conduct a survey on tower crane safety, whereas
feasible lift path for a mobile crane. Olearczyk et al. (2014) propose Al-Humaidi and Tan (2009) use fuzzy set approaches to link
an algorithm for finding a smooth feasible lift path for a mobile the clearance of a mobile crane to an overhead powerline with
the probability of an electrocution accident. In order to evalu- test sets are used. Therefore, results achieved by different ap-
ate safety related to the operations of tower cranes on a given proaches are hardly comparable.
construction site, Shapira and Lyachin (2009) first interview in- • With regard to techniques applied, Zhou et al. (2013) have al-
dustry experts to identify factors affecting tower crane safety. ready pointed out that most construction planners do not have
Then, weights of the factors are determined (Shapira and Sim- the required methodological knowledge for developing exact
cha, 2009a) and – for two of the identified factors – the measure- approaches or formulating mathematical models that allow for
ment of a factor on a given site and the derivation of the asso- the proper use of standard software. This is reflected in our re-
ciated risk are presented (Shapira and Simcha, 2009b). Based on view, as well: a majority of papers relies on metaheuristics that
that, Shapira et al. (2012) finally calculate a site safety index. can often be applied without exploiting the problem structure.
However, as pointed out by Sörensen (2015), this often leads to
5. Conclusions and outlook wasted potential if it comes to run times or solution quality.
• There is a surge of “new” approaches which introduce seem-
The paper at hand gives an overview of problem settings in the ingly new metaheuristic concepts. Although this is by no means
field of construction where OR methods have been applied. In par- exclusive to optimization techniques applied in construction en-
ticular, we classify the literature in these fields. We see, first, that gineering we nevertheless found quite some examples. As im-
well-known generic problems from OR, namely APs, the facility lo- pressively demonstrated by Weyland (2015), often such new
cation problem (Sections 2 and 4) and the project scheduling prob- concepts do not add anything to the toolbox available for tack-
lem (Section 3), have been applied without major modifications. ling optimization problems.
However, often the problem setting is not referred to using the Finally, we identified several promising streams for future re-
terms established in the OR community and, hence, the connec- search.
tion cannot be made instantly. Still, methods from the huge cat-
alogue of approaches tackling APs, facility location problems, or
• Often multiple objectives have to be respected. Nevertheless,
scheduling problems might be applicable without or with minor only few papers include multiple objectives and if they do
modifications only. Secondly, we observe applications of OR meth- so, the objectives are often weighted and expressed in a sin-
ods where apparently no generic problem setting fully covers the gle objective function. However, these objectives might include
applied problem. For example, for lift planning for tower cranes classical ones like the minimization of project duration and
in Section 4.3 to the best of our knowledge no generic problem other aspects like maximization of quality, e.g. in multi-mode
setting is established. Nevertheless, we identified generic problems project scheduling. These objectives might not be comparable
at the very core which can be used as a building block for more and, thus, a weighted sum of different objectives usually falls
involved problem settings. Again, lift planning may serve as an short of covering the nature of the problem. Thus, an impor-
example since path planning problems surely capture important tant contribution for the future is performing multi-objective
characteristics of the lift planning problem. optimization aiming at Pareto-optimal solutions.
This review aims at bridging the gap between the two fields
• The integration of distinct but interrelated decisions is ad-
of research in order to enable researchers in construction to fall dressed only rarely. For example, although the location of tower
back on established methods (cf. Burkard, 1984 for layout prob- cranes is an important issue that can be put at the heart of
lems, Klose and Drexl, 2005 for location problems, Castillo et al., layout decisions at the construction site it would be interest-
2008 for packing problems, Farahani et al., 2012 for covering to integrate layout decisions as presented in Section 2 and
ing problems, Brucker et al., 1999 for project scheduling prob- crane location decisions as discussed in Section 4.2. Similarly,
lems, Berbeglia et al., 2007 for pickup and delivery problems, scheduling construction activities could be integrated with pro-
de Berg et al., 2008 for motion planning) and to highlight promis- curement planning. Finally, having holistic supply chain integra-
ing applications of OR methods to researchers in the OR commu- tion in mind, even production planning might be included as
nity. well, e.g. the concrete delivery problem that is dedicated to lo-
Hereafter, some general aspects for future research shall be gistics could be extended towards planning the production of
pointed out. A common issue with literature in construction engi- ready-mixed concrete.
neering applying optimization techniques is that the vast majority
• If an efficient solution method is at hand to tackle an opti-
of papers does not meet the reporting standards established, e.g., mization problem decision support can be provided on a higher
in the OR literature with regard to traceability. One reason may be level, as well. For example, managerial implications can be
that construction projects are unique as stated, e.g., in Zhou et al., drawn from solving various instances of the optimization prob-
2013. However, the models applied often are quite generic and, lem. The goal might be to determine parameter settings that al-
therefore, we believe that it is worth to exploit structural com- low for extraordinarily good solutions or solutions being robust
monalities (cf. Section 2) in order to reach reusability and compa- with respect to modifications in the parameter values. Such pa-
rability (at least to some degree). Here, however, we observe four rameter settings, then, are desirable to be established in the
shortcomings in the existing literature. real-world setting. Despite the potential from both, a theoret-
ical and a practical perspective, we did not find any published
• Problem settings and/or solution methods often lack accurate findings regarding these questions.
descriptions which has been stated also in Sadeghpour and An- • Algorithms and models, after being specified accurately, should
dayesh (2015). Usually, key features are mentioned, but details be compared to each other by means of computational stud-
are often left out. This allows for a general conception of prob- ies. In order to support such studies it would be good to have
lems and approaches, but not for a detailed analysis. It is, e.g., test cases available. So, building up a testbed accounting for
hardly possible to tell whether an approach can be used to problem instances that either are challenging from a theoret-
solve a problem introduced in another paper. ical point or are typical for real-world cases would be of great
• Computational tests are often not convincing. Approaches are value.
often evaluated using a single case study. While these cases • Applications of standard OR models in construction are rather
motivate the research in the first place it would surely be straightforward. Here, two directions of future research might
meaningful to investigate the performance on a wide range turn out to be fruitful. Firstly, it would be interesting to de-
of problem instances. Moreover, it seems that no common termine properties of, e.g., the quadratic AP, that can be found
when planning layouts on construction sites. This may speed up Brucker, P., Drexl, A., Möhring, R., Neumann, K., Pesch, E., 1999. Resource-con-
solution methods considerably or may allow for design of more strained project scheduling: notation, classification, models, and methods. Eur.
J. Oper. Res. 112, 3–41.
efficient solution approaches. Secondly, the standard problem Burkard, R., 1984. Quadratic assignment problems. Eur. J. Oper. Res. 15, 283–289.
setting can be enriched by real-world aspects. Caprara, A., Toth, P., Fischetti, M., 20 0 0. Algorithms for the set covering problem.
• Finally, as far as we can see, most solution approaches used Ann. Oper. Res. 98, 353–371.
Castillo, I., Kampas, F., Pinter, J., 2008. Solving circle packing problems by global
to tackle problem variants accounting for real-world constraints optimization: numerical results and industrial applications. Eur. J. Oper. Res. 191,
are monolithic metaheuristic procedures. Approaches hardly ex- 786–802.
ploit subproblems which can be solved rather easily. The iden- Chan, A., Chan, D., Yeung, J., 2009. Overview of the application of fuzzy techniques
in construction management research. J. Constr. Eng. Manag. 135, 1241–1252.
tification of such subproblems might lead to decomposition
Chang, Y.-C., Hung, W., Kang, S.-C., 2012. A fast path planning method for single and
approaches where complex components of the problem and dual crane erections. Autom. Constr. 22, 468–480.
rather easy to solve components are treated separately. While Chau, K., 2004. A two-stage dynamic model on allocation of construction facilities
with genetic algorithm. Autom. Constr. 13, 481–490.
the complex components might still be tackled by metaheuris-
Cheng, M.-Y., Lien, L.-C., 2012. Hybrid artificial intelligence-based PBA for bench-
tics their search might then be guided by (computationally) ef- mark functions and facility layout design optimization. J. Comput. Civil Eng. 26,
ficient modules investigating the remaining components. There- 612–624.
fore, we believe the development of decomposition approaches Cheng, M.-Y., Tran, D.-H., Wu, Y.-W., 2014. Using a fuzzy clustering chaotic-based
differential evolution with serial method to solve resource-constrained project
to be a promising field of future research. scheduling problems. Autom. Constr. 37, 88–97.
Cho, K., Hastak, M., 2013. Time and cost-optimized decision support model for fast–
An interesting trend with respect to the application of OR mod- track projects. J. Constr. Eng. Manag. 139, 90–101.
Clarkson, K., Varadarajan, K., 2007. Improved approximation algorithms for geomet-
els and methods is the increasing deployment of BIM. BIM provides
ric set cover. Discret. Comput. Geom. 37, 43–58.
a detailed database covering all aspects of a building during plan- Damci, A., Arditi, D., Polat, G., 2013. Multiresource leveling in line-of-balance
ning, construction and afterwards. Since reliable data is an absolute scheduling. J. Constr. Eng. Manag. 139, 1108–1116.
requirement for OR techniques their usage may fail if such data is Dolabi, H., Afshar, A., Abbasnia, R., 2014. CPM/LOB scheduling method for project
deadline constraint satisfaction. Autom. Constr. 48, 107–118.
not available. Indeed, this was often the case in the past. BIM be- Dong, N., Ge, D., Fischer, M., Haddad, Z., 2012. A genetic algorithm-based method
coming more common is at least lowering the barrier. Ideally, all for look-ahead scheduling in the finishing phase of construction projects. Adv.
data necessary for instantiating models or methods will be read- Eng. Inf. 26, 737–748.
Easa, S., Hossain, K., 2008. New mathematical optimization model for construction
ily available. Recent projects, for example in Faghihi et al. (2014), site layout. J. Constr. Eng. Manag. 134, 653–662.
Irizarry and Karan (2012), Kim et al. (2013), Moon et al. (2014), El-Kholy, A., 2015. New aspects in time-cost tradeoff analysis. J. Manage. Eng. 31,
Moon et al. (2015), and Song et al. (2012), show that this opportu- 04014051–1–04014051–8.
El-Rayes, K., Hyari, K., 2005. Optimal lighting arrangements for nighttime highway
nity can be taken. Nevertheless, research in this area is in an early construction projects. J. Constr. Eng. Manag. 131, 1292–1300.
stage and, therefore, the full potential is still to be exploited. El-Rayes, K., Khalafallah, A., 2005. Trade-off between safety and cost in planning
construction site layouts. J. Constr. Eng. Manag. 131, 1186–1195.
El-Rayes, K., Said, H., 2009. Dynamic site layout planning using approximate dy-
References namic programming. J. Comput. Civil Eng. 23, 119–127.
Elbeltagi, E., Hegazy, T., 2001. A hybrid AI-based system for site layout planning in
AbouRizk, S., Halpin, D., Mohamed, Y., Hermann, U., 2011. Research in modeling and construction. Comput.-Aided Civ. Infrastruct. Eng. 16, 79–93.
simulation for improving construction engineering operations. J. Constr. Eng. Elbeltagi, E., Hegazy, T., Eldosouky, A., 2004. Dynamic layout of construction tempo-
Manag. 137, 843–852. rary facilities considering safety. J. Constr. Eng. Manag. 130, 534–541.
Agrama, F., 2014. Multi-objective genetic optimization for scheduling a multi-storey Elbeltagi, E., Hegazy, T., Hosny, A., Eldosouky, A., 2001. Schedule-dependent evolu-
building. Autom. Constr. 44, 119–128. tion of site layout planning. Constr. Manag. Econ. 19, 689–697.
Al-Humaidi, H., Tan, F., 2009. Mobile crane safe operation approach to prevent elec- Faghihi, V., Reinschmidt, K., Kang, J., 2014. Construction scheduling using genetic al-
trocution using fuzzy-set logic models. Adv. Eng. Softw. 40, 686–696. gorithm based on building information model. Expert Syst. Appl. 41, 7565–7578.
Al-Hussein, M., Alkass, S., Moselhi, O., 2001. An algorithm for mobile crane selection Fan, S.-L., Sun, K.-S., Wang, Y.-R., 2012. GA optimization model for repetitive projects
and location on construction sites. Constr. Innov. 1, 91–105. with soft logic. Autom. Constr. 21, 253–261.
Al-Hussein, M., Alkass, S., Moselhi, O., 2005. Optimization algorithm for selection Farahani, R., Asgari, N., Heidari, N., Hosseininia, M., Goh, M., 2012. Covering prob-
and on site location of mobile cranes. J. Constr. Eng. Manag. 131, 579–590. lems in facility location: a review. Comput. Ind. Eng. 62, 368–407.
AlBahnassi, H., Hammad, A., 2012. Near real-time motion planning and simulation Fu, F., 2014. Integrated scheduling and batch ordering for construction project. Appl.
of cranes in construction: framework and system architecture. J. Comput. Civil Math. Model. 38, 784–797.
Eng. 26, 54–63. Furusaka, S., Gray, C., 1984. A model for the selection of the optimum crane for
Alghazi, A., Elazouni, A., Selim, S., 2013. Improved genetic algorithm for fi- construction sites. Constr. Manag. Econ. 2, 157–176.
nance-based scheduling. J. Comput. Civil Eng. 27, 379–394. Ghoddousi, P., Eshtehardian, E., Jooybanpour, S., Javanmardi, A., 2013. Multi–
Ali, M., Babu, N., Varghese, K., 2005. Collision free path planning of cooperative mode resource-constrained discrete time-cost-resource optimization in project
crane manipulators using genetic algorithm. J. Comput. Civil Eng. 19, 182–193. scheduling using non-dominated sorting genetic algorithm. Autom. Constr. 30,
Ammar, M., 2013. LOB and CPM integrated method for scheduling repetitive 216–227.
projects. J. Constr. Eng. Manag. 139, 44–50. Goh, B., 2008. The state of applications of quantitative analysis techniques to con-
Andayesh, M., Sadeghpour, F., 2013. Dynamic site layout planning through mini- struction economics and management (1983 to 2006). Constr. Manag. Econ. 26,
mization of total potential energy. Autom. Constr. 31, 92–102. 485–497.
Andayesh, M., Sadeghpour, F., 2014. The time dimension in site layout planning. Au- Hammad, A., Akbarnezhad, A., Rey, D., 2016. A multi-objective mixed integer nonlin-
tom. Constr. 44, 129–139. ear programming model for construction site layout planning to minimise noise
Ashuri, B., Tavakolan, M., 2015. Shuffled frog-leaping model for solving time-cost-re- pollution and transport costs. Autom. Constr. 61, 73–85.
source optimization problems in construction project planning. J. Comput. Civil Hammad, A., Akbarnezhad, A., Rey, D., Waller, S., 2015. A computational method for
Eng. 29, 04014026–1–04014026–11. estimating travel frequencies in site layout planning. J. Constr. Eng. Manag. 142,
Bakry, I., Moselhi, O., Zayed, T., 2014. Optimized acceleration of repetitive construc- 04015102–1–04015102–13.
tion projects. Autom. Constr. 39, 145–151. Hammad, A., Rey, D., Akbarnezhad, A., 2017. A cutting plane algorithm for the site
Berbeglia, G., Cordeau, J.-F., Gribkovskaia, I., Laporte, G., 2007. Static pickup and de- layout planning problem with travel barriers. Comput. Oper. Res. 82, 36–51.
livery problems: a classification scheme and survey. TOP 15 (1), 1–31. Hanna, A., Lotfallah, W., 1999. A fuzzy logic approach to the selection of cranes.
de Berg, M., Cheong, O., van Kreveld, M., Overmars, M., 2008. Computational Geom- Autom. Consrtr. 8, 597–608.
etry: Algorithms and Applications, 3 Springer. Hartmann, S., Briskorn, D., 2010. A survey of variants and extensions of the re-
Bettemir, O., Sonmez, R., 2015. Hybrid genetic algorithm with simulated an- source-constrained project scheduling problem. Eur. J. Oper. Res. 207, 1–14.
nealing for resource-constrained project scheduling. J. Manage. Eng. 31, Hasan, S., Bouferguene, A., Al-Hussein, M., Gillis, P., Telyas, A., 2013. Productivity and
04014082–1–04014082–8. CO2 emission analysis for tower crane utilization on high-rise building projects.
Bozejko, W., Hejducki, Z., Wodecki, M., 2013. Applying metaheuristic strategies in Autom. Constr. 31, 255–264.
construction projects management. J. Civil Eng. Manag. 18, 621–630. Hegazy, T., Elbeltagi, E., 1999. Evosite: evolution-based model for site layout plan-
Briskorn, D., Dienstknecht, M., 2017. Covering polygons with discs: the problem of ning. J. Comput. Civil Eng. 13, 198–206.
crane selection and location on construction sites. Working Paper. Heiman, D., 1960. Operations research as applied to construction. Manag. Technol.
Briskorn, D., Fliedner, M., 2012. Packing chained items in aligned bins with applica- 1, 20–25.
tions to container transshipment and project scheduling. Math. Methods Oper.
Res. 75, 305–326.
Heravi, G., Faeghi, S., 2014. Group decision making for stochastic optimization Mawdesley, M., Al-Jibouri, S., 2003. Proposed genetic algorithms for construction
of time, cost and quality in construction projects. J. Comput. Civil Eng. 28, site layout. Eng. Appl. Artif. Intell. 16, 501–509.
275–283. Mawdesley, M., Al-Jibouri, S., Yang, H., 2002. Genetic algorithms for construction
Hornaday, W., Haas, C., O’Connor, J., Wen, J., 1993. Computer-aided planning for site layout in project planning. J. Constr. Eng. Manag. 128, 418–426.
heavy lifts. J. Constr. Eng. Manag. 119, 498–515. Menesi, W., Golzapoor, B., Hegazy, T., 2013. Fast and near-optimum schedule opti-
Huang, C., Wong, C., Tam, C., 2010. Optimization of material hoisting operations and mization for large-scale projects. J. Constr. Eng. Manag. 139, 1117–1124.
storage locations in multi-storey building construction by mixed-integer pro- Menesi, W., Hegazy, T., 2015. Multimode resource-constrained scheduling and level-
gramming. Autom. Constr. 19, 656–663. ing for practical-size projects. J. Manage. Eng. 31, 04014092–1–04014092–7.
Huang, C., Wong, C., Tam, C., 2011. Optimization of tower crane and material sup- Monghasemi, S., Nikoo, M., Fasaee, M., Adamowski, J., 2015. A novel multi criteria
ply locations in a high-rise building site by mixed-integer linear programming. decision making model for optimizing time-cost-quality trade-off problems in
Autom. Constr. 20, 571–580. construction projects. Expert Syst. Appl. 42, 3089–3104.
Imam, M., Mir, M., 1998. Cluster boundary search algorithm for building block lay- Moon, H., Kim, H., Kamat, V., Kang, L., 2015. BIM-based construction scheduling
out optimization. Adv. Eng. Softw. 29, 165–173. method using optimization theory for reducing activity overlaps. J. Comput.
Irizarry, J., Karan, E., 2012. Optimizing location of tower cranes on construction sites Civil Eng. 29, 04014048–1–04014048–16.
through GIS and BIM integration. J. Inf. Technol. Constr. 17, 351–366. Moon, H., Kim, H., Kim, C., Kang, L., 2014. Development of a schedule-workspace
Jato-Espino, D., Castillo-Lopez, E., Rodriguez-Hernandez, J., Canteras-Jordana, J., interference management system simultaneously considering the overlap level
2014. A review of application of multi-criteria decision making methods in con- of parallel schedules and workspaces. Autom. Constr. 39, 93–105.
struction. Autom. Constr. 45, 151–162. Moselhi, O., Roofigari-Esfahan, N., 2013. Project schedule compression: a multi-ob-
Kang, S., Miranda, E., 2006. Planning and visualization for automated robotic crane jective methodology. Constr. Innov. 13, 374–393.
erection processes in construction. Autom. Constr. 15, 398–414. Ng, S., Zheng, D., Xie, J., 2013. Allocation of construction resources through a pull–
Kang, S.-C., Miranda, E., 2008. Computational methods for coordinating multiple driven approach. Constr. Innov. 13, 77–97.
construction cranes. J. Comput. Civil Eng. 22, 252–263. Ning, X., Lam, K., 2013. Cost–safety trade-off in unequal-area construction site lay-
Khalafallah, A., El-Rayes, K., 2011. Automated multi-objective optimization system out planning. Autom. Constr. 32, 96–103.
for airport site layouts. Autom. Constr. 20, 313–320. Ning, X., Lam, K., Lam, M.-K., 2010. Dynamic construction site layout planning using
Khoueiry, Y., Srour, I., Yassine, A., 2013. An optimization-based model for maxi- max-min ant system. Autom. Constr. 19, 55–65.
mizing the benefits of fast-track construction activities. J. Oper. Res. Soc. 64, Ning, X., Lam, K., Lam, M.-K., 2011. A decision-making system for construction site
1137–1146. layout planning. Autom. Constr. 20, 459–473.
Kim, H., Anderson, K., Lee, S., Hildreth, J., 2013. Generating construction schedules Olearczyk, J., Bouferguene, A., Al-Hussein, M., Hermann, U., 2014. Automating mo-
through automatic data extraction using open BIM (building information mod- tion trajectory of crane-lifted loads. Autom. Constr. 45, 178–186.
eling) technology. Autom. Constr. 35, 285–295. Osman, H., Georgy, M., Ibrahim, M., 2003. A hybrid CAD-based construction site lay-
Kim, J., Kang, C., Hwang, I., 2012. A practical approach to project scheduling: con- out planning system using genetic algorithms. Autom. Constr. 12, 749–764.
sidering the potential quality loss cost in the time cost tradeoff problem. Int. J. Reddy, H., Varghese, K., 2002. Automated path planning for mobile crane lifts. Com-
Project Manage. 30, 264–272. put.-Aided Civ. Infrastruct. Eng. 17, 439–448.
Kim, S.-K., Kim, J.-Y., Lee, D., Ryu, S.-Y., 2011. Automatic optimal design algorithm for Rodriguez-Ramos, W., Francis, R., 1983. Single crane location optimization. J. Constr.
the foundation of tower cranes. Autom. Constr. 20, 56–65. Eng. Manag. 109, 387–397.
Kinable, J., Wauters, T., Vanden Berghe, G., 2014. The concrete delivery problem. Sadeghpour, F., Andayesh, M., 2015. The constructs of site layout modeling: an
Comput. Oper. Res. 48, 53–68. overview. Can. J. Civ. Eng. 42, 199–212.
Klansek, U., Psunder, M., 2012. MINLP optimization model for the nonlinear discrete Safouhi, H., Mouattamid, M., Hermann, U., Hendi, A., 2011. An algorithm for the cal-
time-cost trade-off problem. Adv. Eng. Softw. 48, 6–16. culation of feasible mobile crane position areas. Autom. Constr. 20, 360–367.
Klose, A., Drexl, A., 2005. Facility location models for distribution system design. Said, H., El-Rayes, K., 2010. Optimizing the planning of construction site security for
Eur. J. Oper. Res. 162, 4–29. critical infrastructure projects. Autom. Constr. 19, 221–234.
Lam, K., Tang, C., Lee, W., 2005. Application of the entropy technique and genetic Said, H., El-Rayes, K., 2011. Optimizing material procurement and storage on con-
algorithms to construction site layout planning of medium-size projects. Constr. struction sites. J. Constr. Eng. Manag. 137, 421–431.
Manag. Econ. 23, 127–145. Said, H., El-Rayes, K., 2013. Optimal utilization of interior building spaces for mate-
Lam, K.-C., Ning, X., Lam, M.-K., 2009. Conjoining MMAS to GA to solve construction rial procurement and storage in congested construction sites. Autom. Constr. 31,
site layout planning problem. J. Constr. Eng. Manag. 135, 1049–1057. 292–306.
Lam, K.-C., Ning, X., Ng, T., 2007. The application of the ant colony optimization al- Said, H., El-Rayes, K., 2013. Performance of global optimization models for dynamic
gorithm to the construction site layout planning problem. Constr. Manag. Econ. site layout planning of construction projects. Autom. Constr. 36, 71–78.
25, 359–374. Sanad, H., Ammar, M., Ibrahim, M., 2008. Optimal construction site layout consider-
Lei, Z., Taghaddos, H., Hermann, U., Al-Hussein, M., 2013a. A methodology for mobile ing safety and environmental aspects. J. Constr. Eng. Manag. 134, 536–544.
crane lift path checking in heavy industrial projects. Autom. Constr. 31, 41–53. Sarker, B., Egbelu, P., Liao, T., Yu, J., 2012. Planning and design models for construc-
Lei, Z., Taghaddos, H., Olearczyk, J., Al-Hussein, M., Hermann, U., 2013b. Automated tion industry: a critical survey. Autom. Constr. 22, 123–134.
method for checking crane paths for heavy lifts in industrial projects. J. Constr. Sawhney, A., Mund, A., 2001. Intellicranes: an integrated crane type and model se-
Eng. Manag. 139, 04013011–1–04013011–8. lection system. Constr. Manag. Econ. 19, 227–237.
Leung, A., Tam, C., 1999. Models for assessing hoisting times of tower cranes. J. Con- Shapira, A., Goldenberg, M., 2005. AHP-Based equipment selection model for con-
str. Eng. Manag. 125, 385–391. struction projects. J. Constr. Eng. Manag. 131, 1263–1273.
Leung, A., Tam, C., Liu, D., 2001. Comparative study of artificial neural networks and Shapira, A., Goldenberg, M., 2007. Soft considerations in equipment selection for
multiple regression analysis for predicting hoisting times of tower cranes. Build. building construction projects. J. Constr. Eng. Manag. 133, 749–760.
Environ. 36, 457–467. Shapira, A., Lucko, G., Schexnayder, C., 2007. Cranes for building construction
Li, H., Love, P., 1998. Site-level facilities layout using genetic algorithms. J. Comput. projects. J. Constr. Eng. Manag. 133, 690–700.
Civil Eng. 12, 227–231. Shapira, A., Lyachin, B., 2009. Identification and analysis of factors affecting safety
Li, H., Love, P., 20 0 0. Genetic search for solving construction site-level unequal-area on construction sites with tower cranes. J. Constr. Eng. Manag. 135, 24–33.
facility layout problems. Autom. Constr. 9, 217–226. Shapira, A., Schexnayder, C., 1999. Selection of mobile cranes for building construc-
Li, H., Xu, Z., Demeulemeester, E., 2015. Scheduling policies for the stochastic re- tion projects. Constr. Manag. Econ. 17, 519–527.
source leveling problem. J. Constr. Eng. Manag. 141, 04014072–1–04014072–9. Shapira, A., Simcha, M., 2009. AHP-based weighting of factors affecting safety on
Li, H., Zhang, H., 2013. Ant colony optimization-based multi-mode scheduling under construction sites with tower cranes. J. Constr. Eng. Manag. 135, 307–318.
renewable and nonrenewable resource constraints. Autom. Constr. 35, 431–438. Shapira, A., Simcha, M., 2009. Measurement and risk-scales of crane-related safety
Liang, L., Chao, W., 2008. The strategies of tabu search technique for facility layout factors on construction sites. J. Constr. Eng. Manag. 135, 979–989.
optimization. Autom. Constr. 17, 657–669. Shapira, A., Simcha, M., Goldenberg, M., 2012. Integrative model for quantitative
Liao, T., Egbelu, P., Sarker, B., Leu, S., 2011. Metaheuristics for project and construc- evaluation of safety on construction sites with tower cranes. J. Constr. Eng.
tion management – a state-of-the-art review. Autom. Constr. 20, 491–505. Manag. 138, 1281–1293.
Lien, L.-C., Cheng, M.-Y., 2012. A hybrid swarm intelligence based particle-bee Sivakumar, P., Varghese, K., Babu, N., 2003. Automated path planning of cooperative
algorithm for construction site layout optimization. Expert Syst. Appl. 39, crane lifts using heuristic search. J. Comput. Civil Eng. 17, 197–207.
9642–9650. Song, S., Yang, J., Kim, N., 2012. Development of a BIM-based structural framework
Lien, L.-C., Cheng, M.-Y., 2014. Particle bee algorithm for tower crane layout with optimization and simulation system for building construction. Comput. Ind. 63,
material quantity supply and demand optimization. Autom. Constr. 45, 25–32. 895–912.
Lin, K., Haas, C., 1996. Multiple heavy lifts optimization. J. Constr. Eng. Manag. 122, Sonmez, R., Bettemir, O., 2012. A hybrid genetic algorithm for the discrete time-cost
354–362. trade-off problem. Expert Syst. Appl. 39, 11428–11434.
Lin, Y., Wu, D., Wang, X., Wang, X., Gao, S., 2014. Lift path planning for a nonholo- Sonmez, R., Uysal, F., 2014. Backward-forward hybrid genetic algorithm for re-
nomic crawler crane. Autom. Constr. 44, 12–24. source-constrained multiproject scheduling problem. J. Comput. Civil Eng. 29,
Liu, S.-S., Wang, C.-J., 2012. Optimizing linear project scheduling with multi-skilled 04014072–1–04014072–9.
crews. Autom. Constr. 24, 16–23. Sörensen, K., 2015. Metaheuristics – the metaphor exposed. Int. Trans. Oper. Res. 22,
Loiola, E.M., de Abreu, N.M.M., Boaventura-Netto, P.O., Hahn, P., Querido, T., 2007. A 3–18.
survey for the quadratic assignment problem. Eur. J. Oper. Res. 176, 657–690. Tam, C., Leung, A., Liu, D., 2002. Nonlinear models for predicting hoisting times of
Lorterapong, P., Ussavadilokrit, M., 2013. Construction scheduling using the con- tower cranes. J. Comput. Civil Eng. 16, 76–81.
straint satisfaction problem method. J. Constr. Eng. Manag. 139, 414–422.
Tam, C., Tong, T., 2003. GA-ANN model for optimizing the locations of tower crane Xu, J., Song, X., 2015. Multi-objective dynamic layout problem for temporary con-
and supply points for high-rise public housing projects. Constr. Manag. Econ. 21, struction facilities with unequal-area departments under fuzzy random envi-
257–266. ronment. Knowl. Based Syst. 81, 30–45.
Tam, C., Tong, T., Chan, W., 2001. Genetic algorithm for optimizing supply locations Yahya, M., Saka, M., 2014. Construction site layout planning using multi-objective
around tower crane. J. Constr. Eng. Manag. 127, 315–321. artificial bee colony algorithm with Levy flights. Autom. Constr. 38, 14–29.
Tam, C., Tong, T., Tse, S., 2004. Modelling hook times of mobile cranes using artificial Yang, I.-T., Lin, Y.-C., Lee, H.-Y., 2014. Use of support vector regression to improve
neural networks. Constr. Manag. Econ. 22, 839–849. computational efficiency of stochastic time-cost trade-off. J. Constr. Eng. Manag.
Tam, V., Fung, I., 2011. Tower crane safety in the construction industry: a Hong Kong 140, 04013036–1–04013036–13.
study. Saf. Sci. 49, 208–215. Yeh, I.-C., 1995. Construction-site layout using annealed neural network. J. Comput.
Tang, Y., Liu, R., Sun, Q., 2014. Schedule control model for linear projects based Civil Eng. 9, 201–208.
on linear scheduling method and constraint programming. Autom. Constr. 37, Yeh, I.-C., 2006. Architectural layout optimization using annealed neural network.
22–37. Autom. Constr. 15, 531–539.
Tang, Y., Liu, R., Sun, Q., 2014. Two-stage scheduling model for resource leveling of Zavichi, A., Madani, K., Xanthopoulos, P., Oloufa, A., 2014. Enhanced crane operations
linear projects. J. Constr. Eng. Manag. 140, 04014022–1–04014022–10. in construction using service request optimization. Autom. Constr. 47, 69–77.
Tantisevi, K., Akinci, B., 2008. Simulation-based identification of possible locations Zhang, C., Hammad, A., 2012. Improving lifting motion planning and re-planning of
for mobile cranes on construction sites. J. Comput. Civil Eng. 22, 21–30. cranes with consideration for safety and efficiency. Adv. Eng. Inf. 26, 396–410.
Tantisevi, K., Akinci, B., 2009. Transformation of a 4D product and process model to Zhang, C., Hammad, A., 2012. Multiagent approach for real-time collision avoidance
generate motion of mobile cranes. Autom. Constr. 18, 458–468. and path replanning for cranes. J. Comput. Civil Eng. 26, 782–794.
Wang, Y., Le, Y., Dai, J., 2015. Incorporation of alternatives and importance Zhang, H., Wang, J., 2008. Particle swarm optimization for construction site un-
levels in scheduling complex construction programs. J. Manage. Eng. 31, equal-area layout. J. Constr. Eng. Manag. 134, 739–748.
04014098–1–04014098–10. Zhang, P., Harris, F., Olomolaiye, P., 1996. A computer-based model for optimizing
Warszawski, A., 1972. Layout and location problems – the quantitative approach. the location of a single tower crane. Build. Res. Inf. 24, 113–123.
Build. Sci. 7, 61–67. Zhang, P., Harris, F., Olomolaiye, P., Holt, G., 1999. Location optimzation for a group
Warszawski, A., 1973. Multi-dimensional location problems. J. Oper. Res. Soc. 24, of tower cranes. J. Constr. Eng. Manag. 125, 115–122.
165–179. Zhou, F., AbouRizk, S., Al-Battaineh, H., 2009. Optimisation of construction site lay-
Warszawski, A., Peer, S., 1973. Optimizing the location of facilities on a building site. out using a hybrid simulation-based system. Simul. Modell. Pract. Theory 17,
J. Oper. Res. Soc. 24, 35–44. 348–363.
Weyland, D., 2015. A critical analysis of the harmony search algorithm – how not to Zhou, J., Love, P., Wang, X., Teo, K., Irani, Z., 2013. A review of methods and algo-
solve sudoku. Oper. Res. Perspect. 2, 97–105. rithms for optimizing construction scheduling. J. Oper. Res. Soc. 64, 1091–1105.
Wong, C., Fung, I., Tam, C., 2010. Comparison of using mixed-integer programming Zouein, P., Harmanani, H., Hajar, A., 2002. Genetic algorithm for solving site layout
and genetic algorithms for construction site facility layout planning. J. Constr. problem with unequal-size and constrained facilities. J. Comput. Civil Eng. 16,
Eng. Manag. 136, 1116–1128. 143–151.
Wu, D., Lin, Y., Wang, X., Wang, X., Gao, S., 2011. Algorithm of crane selection for Zouein, P., Tommelein, I., 1999. Dynamic layout planning using a hybrid incremental
heavy lifts. J. Comput. Civil Eng. 25, 57–65. solution method. J. Constr. Eng. Manag. 125, 400–408.
Xu, J., Li, Z., 2012. Multi-objective dynamic construction site layout planning in
fuzzy random environment. Autom. Constr. 27, 155–169.
Xu, J., Song, X., 2014. Suggestions for temporary construction facilities’ layout
problems in large-scale construction projects. J. Constr. Eng. Manag. 140,
06014001–1–0604001–7.

Computers and Operations Research: Dirk Briskorn, Michael Dienstknecht

Uploaded by

Copyright:

Available Formats

Computers and Operations Research: Dirk Briskorn, Michael Dienstknecht

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Computers and Operations Research: Dirk Briskorn, Michael Dienstknecht

Uploaded by

Copyright:

Available Formats

Computers and Operations Research 92 (2018) 194–207

Contents lists available at ScienceDirect

Computers and Operations Research

Survey of quantitative methods in construction

1. Introduction reﬂect uncertainties as well as complex interdependencies is

single-obj. multi-obj. single-obj. multi-obj.

Methodology Problem variant

RCPSP MRCPSP repetitive fast-track other

Field of application Literature

You might also like