Transportation cost and CO2 emissions in location decision models · Transportation Cost and CO 2...
Transcript of Transportation cost and CO2 emissions in location decision models · Transportation Cost and CO 2...
Transportation cost and CO2 emissions in location decisionmodelsCitation for published version (APA):Velazquez Martinez, J. C., Fransoo, J. C., Blanco, E. E., & Mora-Vargas, J. (2014). Transportation cost and CO2emissions in location decision models. (BETA publicatie : working papers; Vol. 451). Technische UniversiteitEindhoven.
Document status and date:Published: 01/01/2014
Document Version:Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)
Please check the document version of this publication:
• A submitted manuscript is the version of the article upon submission and before peer-review. There can beimportant differences between the submitted version and the official published version of record. Peopleinterested in the research are advised to contact the author for the final version of the publication, or visit theDOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and pagenumbers.Link to publication
General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, pleasefollow below link for the End User Agreement:www.tue.nl/taverne
Take down policyIf you believe that this document breaches copyright please contact us at:[email protected] details and we will investigate your claim.
Download date: 07. Feb. 2021
Transportation Cost and CO2 Emissions in Location Decision Models
Josue C. Vélazquez Martínez, Jan C. Fransoo,
Edgar E. Blanco, Jaime Mora-Vargas
Beta Working Paper series 451
BETA publicatie WP 451 (working paper)
ISBN ISSN NUR
804
Eindhoven March 2014
1
Transportation Cost and CO2 Emissions in Location Decision Models
Josue C. Vélazquez-Martínez1, Jan C. Fransoo1*, Edgar E. Blanco2, Jaime Mora-Vargas3
1 Technische Universiteit Eindhoven
School of Industrial Engineering
P.O. Box 513, Pav F4
NL-5600 MB Eindhoven
Netherlands
2 MIT Center for Transportation & Logistics
3 Instituto Tecnológico y de Estudios Superiores de Monterrey, Campus Estado de México
* Corresponding author: [email protected]
2
Transportation Cost and CO2 Emissions in Location Decision Models
Josue C. Vélazquez-Martínez1, Jan C. Fransoo1, Edgar E. Blanco2, Jaime Mora-Vargas3
1 Technische Universiteit Eindhoven, School of Industrial Engineering
2 MIT Center for Transportation & Logistics
3 Instituto Tecnológico y de Estudios Superiores de Monterrey, Campus Estado de México
An increasing number of companies are making their supply chains more sustainable. Because
transportation accounts for a large share of global CO2 emissions, finding logistics alternatives that reduce
carbon emissions while keeping costs low is a priority. In this article, we study the trade-off between cost
and CO2 emissions by using a multiobjective approach for the facility location problem. We propose a
model with new cost and CO2 structures for the p-Median problem to determine a sustainable location. To
solve this problem, we develop the multiobjective combinatorial optimization cross-entropy method, to
address the difficulties of combinatorial models. We test the algorithm against two p-Median problems
from prior literature and show that it can approximate the Pareto frontier efficiently. We also conduct an
experimental study for a consumer packaged goods company in Mexico City to provide insights on the
structure of the Pareto frontier for a practical sustainable facility location case. The study shows that by
selectively changing a subset of locations, companies may achieve a substantial reduction in carbon
emissions under similar costs.
Key words: Sustainable facility location, Multiobjective combinatorial optimization.
1. Introduction Regardless of whether companies’ decision makers believe in green and sustainable practices, a market
shift is occurring, driving firms to measure and disclose their carbon emissions and implement solutions
to reduce those emissions (Hoffman and Woody, 2008). For example, the number of companies around
the world reporting their emissions to the Carbon Disclosure Project (2011), a not-for-profit foundation,
has doubled from 2007 to 2010, and approximately 1,000 of these companies have publicly committed to
a carbon reduction target. Transportation is one of the main contributors of carbon dioxide (CO2)
3
(Intergovernmental Panel on Climate Change, 2007), and predictions indicate increases in the next 20
years (European Commission, 2011); thus, more efficient transport operations would substantially reduce
carbon emissions.
A relevant logistics problem that determines the configuration of a company’s delivery of goods is the
facility location problem. The location of distribution centers is critical to the efficient and effective
operation of a supply chain; poorly placed plants can result in excessive costs and low service level no
matter how well tactical decisions (e.g., vehicle routing, inventory management) are optimized (Daskin et
al., 2005). In addition, this configuration can increase CO2 emissions, in that the main drivers of
transportation carbon emissions are distance, truck load (Greenhouse Gas Protocol Standard, 2011) and
the number of trips required to deliver demand to each customer.
A common facility location model that has multiple real-world applications is the p-Median problem
(ReVelle et al., 2008). The p-Median problem considers p facilities that must be located in a network,
such that the demand in each node must be satisfied while the total weighted demand-distance is
minimized (Hakimi, 1964). Because of the strategic nature of facility location problems, the cost structure
of the p-Median does not take into account capacity or cost of the truck fleet, and it assumes a constant
cost per distance per unit (ReVelle et al, 2008), linear to the distance and demand. However, in reality
these two factors are not linearly related; for example, the amount of items affects the transport cost (and
carbon emissions) when it implies an increase in the number of trucks or trips.
Previous operations research has addressed aggregations. Jalil et al. (2011) study cost aggregation for
real-life spare parts planning, and they find that errors can be significant. Velázquez-Martínez et al.’s
(2013) study addresses the effects of using different aggregation levels to measure transport carbon
emissions in the dynamic lot-sizing model, and they show that errors associated with aggregation tend to
be substantial and systematic. These findings suggest that increasing the level of detail in the p-Median
problem is necessary.
Because decision makers are likely interested in evaluating alternatives to increase carbon efficiency
while keeping low transport costs, we propose a multiobjective approach to analyze the trade-off between
cost and carbon emissions expressions for the p-Median problem. Although literature on multiobjective
facility location problems with environmental concerns has increased in the first decade of the twenty-
first century (e.g., Blanquero and Carrizosa [2002] study a semi-obnoxious location problem with cost
and negative effect; Medaglia et al. [2009] investigate a hospital waste management system; Fonseca et
al. [2009] address stochastic reverse logistics; Du and Evans [2008] study reverse logistics network), to
our knowledge, no studies include sustainability in their location models, because it is difficult to define a
4
structure that measures this attribute (Farahani et al., 2010). In this study, we propose an expression that
takes into consideration CO2 emissions in the p-Median problem.
The p-Median problem with a single objective is classified as NP-hard (Kariv and Hakimi, 1979);
therefore, defining the Pareto frontier for its multiobjective version is not an easy task. Traditional
optimization methods such as gradient- and simplex-based methods are difficult to extend to
multiobjective optimization (Fonseca and Fleming, 1995); however, the use of metaheuristics allows
researchers to manage these problems effectively. These techniques enable researchers to find several
solutions of the set in a single run instead of performing a series of separate runs (Ajith et al., 2005).
Bekker and Aldrich (2011) apply a new metaheuristic to solve multiobjective optimization problems
based on the cross-entropy (CE) method (De Boer et al., 2005), a simulation-based metaheuristic used to
solve single-objective optimization problems. Bekker and Aldrich (2011) adapt the CE method to solve
multiobjective optimization problems and test it against benchmark problems from the literature. They
show that the CE method applied to multiobjective optimization obtains satisfactory results. Although
their test results seem promising for applying the CE method in multiobjective problems, their study does
not address the use of the algorithm on multiobjective combinatorial optimization (MOCO) problems,
which corresponds to the structure of the p-Median model.
In the current research, we study the trade-off between cost and CO2 by using a multiobjective approach
to the p-Median problem. We name this new model the sustainable facility location (SFL) problem. Our
model captures a higher level of detail in the cost and CO2 formulations by taking into account truck
accessibility constraints per demand node and finer-grained expressions for both objective functions. To
approximate the Pareto frontier of the SFL, we develop an adaptation of the CE method applied to
multiobjective optimization problems, called the MOCO CE method (hereinafter, MOCO CEM). We test
our algorithm against problems from prior literature and demonstrate that the MOCO CEM is capable of
approximating the Pareto frontier for the SFL problem efficiently. Furthermore, we conduct an
experimental study of the MOCO CEM by solving the SFL for a consumer packaged goods (CPG)
company that operates in the Mexico City metropolitan area. The experiments provide managerial
insights on the structure of the Pareto frontier for an actual company in a large urban area and show that it
is possible to find location alternatives, such that for small increases in cost, firms can obtain large CO2
emission reductions.
The remainder of this article is organized as follows: Section 2 presents the mathematical formulation of
the sustainable facility location model. Section 3 describes the MOCO CEM. Section 4 presents a case
5
study in which we apply the MOCO CEM to an actual CPG company. In Section 5, we present our
conclusions.
2. The Sustainable Facility Location Problem Our model corresponds to the p-Median problem with cost and CO2 objective functions. The general
assumptions of the p-Median problem are applicable to our model; that is, deterministic demand and the
candidate locations are known in advance. We also assume that the company manages multiple trucks
with different capacities and the trucks are assigned according to demand node constraints. These
assumptions allow the model to include the possibility that certain customers are reachable only by
certain types of trucks, with distinct cost structures.
To formulate the carbon emissions objective function in our model, we use Network for Transport and
Environment’s methodology. For the estimation of carbon emissions, this methodology requires a high
level of detailed parameters: fuel consumption, distance traveled and weight per shipment (Network for
Transport and Environment Road, 2008). Fuel consumption is a function of the type of vehicle, the load
factor and the type of road. The methodology uses the European Assessment and Reliability of Transport
Emission Models and Inventory Systems database, which includes a detailed emissions model for all
transport modes to provide consistent emission estimates at the national, international and regional levels
(Transport Research Laboratory, 2010).
We therefore formulate our model as follows. Let 𝐼 be a set of demand nodes and 𝐽 be a set of candidate
locations. We define the following parameters:
ℎ𝑖 = Demand at node 𝑖 ∈ 𝐼
𝑑𝑖𝑗 = Distance between candidate facility site 𝑗 ∈ 𝐽 and
customer location 𝑖 ∈ 𝐼
𝐴𝑖𝑗 = Fixed cost per trip between candidate facility site
𝑗∈𝐽 and customer location 𝑖∈𝐼
𝑣𝑖𝑗 = Cost per distance traveled between candidate
facility site 𝑗∈𝐽 and customer location 𝑖∈𝐼
𝑊𝑖 = Truck capacity per trip restricted to customer
location 𝑖∈𝐼
𝑛𝑖𝑗 =ℎ𝑖𝑌𝑖𝑗𝑊𝑖
= Number of trips per period required to serve the
customer location 𝑖∈𝐼 from facility site 𝑗∈𝐽
𝑘 = Constant emission factor (2,621 grams of CO2/liter
6
of fuel)
𝑓𝑖𝑒 = Fuel consumption of the empty vehicle used to
serve at customer location 𝑖∈𝐼 (liters/km), and
𝑓𝑖𝑓 = Fuel consumption of the fully loaded vehicle used
to serve at customer location 𝑖∈𝐼 (liters/km)
We use the following decision variables:
𝑋𝑗 = 1 if we locate at site 𝑗 ∈ 𝐽, 0 otherwise.
𝑌𝑖𝑗 = Fraction of demand at customer location 𝑖 ∈ 𝐼 that is
served by facility at site 𝑗 ∈ 𝐽.
We then formulate the multiobjective SFL model as follows:
𝑀𝑖𝑛 → 𝑂𝐹1 = ��𝐴𝑖𝑗𝑛𝑖𝑗𝑖∈𝐼𝑗∈𝐽
+��𝑣𝑖𝑗𝑑𝑖𝑗�𝑛𝑖𝑗�𝑖∈𝐼𝑗∈𝐽
𝑀𝑖𝑛 → 𝑂𝐹2 = 𝑘 ���𝑑𝑖𝑗𝑓𝑖𝑒�𝑛𝑖𝑗�𝑖∈𝐼𝑗∈𝐽
+��𝑑𝑖𝑗𝑛𝑖𝑗�𝑓𝑖𝑓 − 𝑓𝑖𝑒��𝑛𝑖𝑗�
𝑖∈𝐼𝑗∈𝐽
�
subject to
∑ 𝑌𝑖𝑗𝑗∈𝐽 = 1 ∀𝑖 ∈ 𝐼 (1)
∑ 𝑋𝑗𝑗∈𝐽 = 𝑝 (2)
𝑌𝑖𝑗 − 𝑋𝑗 ≤ 0 ∀𝑖 ∈ 𝐼 ∀𝑗 ∈ 𝐽 (3)
𝑋𝑗 ∈ {0,1} ∀𝑗 ∈ 𝐽 (4)
𝑌𝑖𝑗 ≥ 0 ∀𝑖 ∈ 𝐼,∀𝑗 ∈ 𝐽 (5)
The first part of OF1 corresponds to the fixed cost per trip. It includes labor and costs related to the use of
the truck (e.g., depreciation, insurance). In turn, 𝑛𝑖𝑗 is the total number of trips required to deliver the
total demand from location j to demand node i taking into account the truck capacity Wi. Note that we do
not restrict 𝑛𝑖𝑗 to be an integer because of the strategic nature of the model, fixed costs tend to be
allocated across multiple demand nodes and trips. The second part of OF1 corresponds to the variable
cost per distance traveled from candidate facility site 𝑗 ∈ 𝐽 to customer location 𝑖 ∈ 𝐼. We apply a ceiling
7
function to estimate the number of times distance 𝑑𝑖𝑗 is traveled. The first part of OF2 corresponds to the
carbon emissions due to the number of trips required, and the second part corresponds to the carbon
emissions due to the units to be transported. Constraint 1 states that each demand node is covered.
Constraint 2 establishes that p facilities are located. Constraint 3 designates that the facility is opened
when a demand node is assigned. Constraint 4 encompasses the integrality constraints, and constraint 5
encompasses the non-negative constraints.
By reducing the expressions, and because the single sourcing condition holds for both OF1 and OF2, we
formulate the multiobjective SFL problem as follows:
𝑀𝑖𝑛 → 𝑂𝐹1 = ��� 𝐴𝑖𝑗ℎ𝑖𝑊𝑖
+ 𝑣𝑖𝑗𝑑𝑖𝑗 �ℎ𝑖𝑊𝑖��𝑌𝑖𝑗
𝑖∈𝐼𝑗∈𝐽
𝑀𝑖𝑛 → 𝑂𝐹2 = ��𝑑𝑖𝑗 �ℎ𝑖𝑊𝑖� �𝑓𝑖𝑒 + �𝑓𝑖
𝑓 − 𝑓𝑖𝑒�ℎ𝑖𝑊𝑖�𝑌𝑖𝑗
𝑖∈𝐼𝑗∈𝐽
subject to Constraints 1-5.
Note that the objective functions differ from each other: OF1 has fixed and variable costs, but only the
variable part depends on the distance traveled, whereas the entire OF2 expression depends on the distance
traveled. This relationship suggests that for a specific instance, if the highest fixed costs are assigned to
edges with the shortest distance (or the lowest fixed costs are assigned to edges with the largest distance),
SFL may provide multiple solutions, and thus, the Pareto frontier would contain a larger number of
points.
3. The Multiobjective Combinatorial Optimization Cross-Entropy Method 3.1. Multiobjective definitions
In general, the multiobjective optimization (MOO) problem involves having more than one objective
function to be minimized or maximized. Commonly, these objective functions are continuous. As in the
single-objective optimization problem, in the MOO problem, several constraints are clearly defined, and
any feasible solution must satisfy them. However, the main difference compared with single-objective
problems is that with MOO, it is commonly not possible to define an optimal solution for the problem (if
there were an optimal solution, the problem would not require a multiobjective setting, and a single
objective might suffice); instead, the result is a set of “efficient” solutions that are not superior to one
another in all the objective functions. In this sense, single-objective optimization problems are a
degenerate case of the MOO problem, not just an extension of it (Deb, 2001).
8
When a MOO problem involves discrete space, it is considered a MOCO problem. Similar to MOO,
MOCO has several objective functions, but the variables are integers or binary numbers. As in
combinatorial optimization with a single objective, MOCO has a finite and discrete set of feasible
solutions. A general formulation of the MOCO problem is as follows (Ehrgott, 2005):
𝑚𝑖𝑛�𝑓𝑗(𝒙):𝒙 ∈ 𝑋; 𝑗 = 1, … ,𝑚�
𝑋 = {𝒙:𝐴𝒙 = 𝒃,𝒙 ≥ 0 𝑖𝑛𝑡𝑒𝑔𝑒𝑟; 𝑖 = 1, … ,𝑛}
where 𝑓𝑗 is the 𝑗𝑡ℎ objective function and x is the vector of decision variables. The matrix A and vector b
define the set of linear constraints.
In the following subsections, we elaborate on concepts from Deb (2001), Ehrgott (2005) and Coello
(2009) to define relevant terms.
Definition 1. Dominance
A solution x is said to dominate solution y if both the following conditions hold for a minimization
problem:
1. The solution x is not worse than y in all objectives; that is, 𝑓𝑗(𝒙) ≤ 𝑓𝑗(𝒚) for all 𝑗 = 1, … , 𝐽
2. The solution x is strictly better than y in at least one objective; that is, 𝑓𝑗′(𝒙) < 𝑓𝑗′(𝒚) for at least
one 𝑗′ ∈ {1, … ,𝑀}
Definition 2. Efficient or Pareto-optimal solution
A feasible solution is called efficient or Pareto optimal if there is no other feasible solution that
dominates it. If the solution vector x is efficient, the vector 𝑓(𝒙) whose M components are the values of
the objective functions at the point x is called a non-dominated point.
Definition 3. Nondominated set
The nondominated set of solutions P’ is the set of all solutions that are not dominated by any member of
the feasible solution set P.
Definition 4. Pareto-optimal set
The nondominated set of the entire feasible space is the globally Pareto-optimal set. This set is also
called the efficient frontier or Pareto frontier.
9
Thus, the goal in MOO and MOCO is to find members of the Pareto-optimal set. Some solution
approaches obtain all elements of the Pareto set, while other techniques aid in the approximation of this
set. Such an approximation could result in a subset of the exact Pareto set or a set of points within a
specified distance of the Pareto set.
For MOO in continuous problems, a classic approach to obtain the Pareto set involves using the weighted
sum scalarization method, which corresponds to the set of solutions that can be obtained by solving
𝑚𝑖𝑛�∑ 𝜆𝑗𝑓𝑗(𝒙)𝑚𝑗=1 :𝒙 ∈ 𝑋�, where 0 ≤ 𝜆𝑗 ≤ 1 and ∑ 𝜆𝑗𝑚
𝑗=1 = 1. However, this approach does not work
for MOCO problems because of their discrete structure (Ehrgott and Gandibleux, 2000). The solutions for
the MOCO problems that can be obtained by this approach are called supported solutions. Typically,
MOCO problems have more efficient solutions that are not optimal for any weighted sum of the
objectives. These solutions are called nonsupported solutions, and their numbers increase exponentially
with the problem size (for the assignment problem, see Hansen 1980; for the traveling salesperson
problem, see Sergienko and Perepilitsa 1991). Ehrgott (2000) shows that the general MOCO problem is
NP-complete, and thus, it may be computationally prohibitive to use an exact method to determine all
efficient solutions. Therefore, we conclude that approximate methods are appropriate.
In the next section we review the main approximate algorithms for combinatorial multiobjective
problems.
3.2. Approximate algorithms for MOCO problems
Because most MOCO problems are NP-complete, literature mostly focuses on approximate (heuristic)
algorithms that address reasonably sized problems. These algorithms generate an approximated Pareto set
(or a subset of it) with solutions that either belong to the optimal Pareto set or are close to it.
Evolutionary algorithms are metaheuristic procedures that have become increasingly popular in recent
decades (Coello, 2009). Their ability to handle complex problems, involving features such as
discontinuities, multimodality and disjoint feasible spaces, have made them effective techniques to seek
approximately Pareto-optimal solutions for MOO problems (Fonseca and Fleming, 1995; Jaszkiewicz,
2002; Köksalan, 2008; Zitzler et al., 2000). Schaffer (1985) presents the vector evaluated genetic
algorithm (VEGA). Horn et al. (1994) propose the niched Pareto genetic algorithm (NPGA). Zitzler and
Thiele (1998) propose the strength Pareto evolutionary algorithm (SPEA). Deb et al. (2000) suggest an
elitist nondominated sorting genetic algorithm (NSGA-II) that uses an elite preservation strategy together
with an explicit diversity-preserving mechanism.
10
In the current study, we propose using the CE method (Rubinstein & Kroese, 2004) for MOCO problems
by adapting the MOO CE method (hereinafter, MOO CEM) that Bekker and Aldrich (2011) present. The
MOO CEM was tested against benchmarking problems from Coello et al. (2007) and the results were
promising by approximating all the problems to the true Pareto optimal set (Bekker and Aldrich, 2011),
with fewer objective function evaluations than in other experiments reported for the algorithms VEGA,
NPGA and SPEA (Zitzler et al., 2000).
The MOO CEM is a metaheuristic to solve MOO problems using the CE method (De Boer et al., 2005;
Rubinstein and Kroese, 2004). The CE method was motivated by an adaptive algorithm for estimating
probabilities of rare events in complex stochastic networks (Rubinstein, 1997). This algorithm can solve
difficult single-objective combinatorial optimization problems. It works by translating the deterministic
optimization problem into a related stochastic optimization problem and then using rare event simulation
techniques similar to Rubinstein (1997). Several recent applications demonstrate the power of the CE
method as a practical tool for solving NP-hard problems: buffer allocation (Alon et al., 2005), control and
navigation (Helvik and Wittner, 2001), DNA sequence alignment (Keith and Kroese, 2002), vehicle
routing (Chepuri and Homem-de-Mello, 2005), project management (Cohen et al., 2005), heavy-tail
distributions (Rubinstein and Kroese, 2004), network reliability (Hui et al., 2005), maximal cut and
bipartition problems (Rubinstein, 2002) and facility location problems (Caserta and Quiñonez Rico,
2009). De Boer et al. (2005) provide a tutorial on the CE method.
3.3. The MOCO CEM algorithm
3.3.1. The CE Method
The foundation of CE method for optimization is in Importance Sampling and the Kullback–Leibler
distance (Rubinstein and Kroese, 2004). In the following paragraphs, we present a brief description of the
CE method based on some definitions from Rubinstein and Kroese (2004) and include the algorithm to
solve single-objective combinatorial optimization problems (for a more complete review, see De Boer et
al. 2005).
Let 𝚾 = (𝑋1, … ,𝑋𝑛) be a random vector assuming values from some space x, and let f be some real
function on x. To determine the probability that 𝑓(𝚾) is greater than or equal to a real number 𝛾 under a
family of probability density functions ℎ(∙;𝑢) on x, we use 𝑙 = ℙ𝑢(𝑓(𝚾) ≥ 𝛾) = 𝔼𝑢𝐼(𝑓(𝚾)≥𝛾). Then
𝑓(𝚾 ≥ γ) is called a rare event if l is very small, and it can be efficiently estimated using importance
sampling. We thus generate a random sample 𝚾1, … ,𝚾𝑁 from a different density g on x and estimate l
using the following likelihood estimator (Rubinstein and Kroese, 2004): 𝑙 = 1𝑁∑ 𝐼{𝑓(𝚾𝑖)≥𝛾}𝑁𝑖=1
ℎ(𝚾𝑖;𝐮)𝑔(𝚾𝑖)
.
11
Thus, 𝑙 =𝐼�𝑓�𝚾𝑖�≥𝛾�ℎ(𝚾𝑖;𝐮)
𝑔∗(𝚾𝑖). In turn, 𝑔∗ can be approximated within the family of densities {ℎ(∙; 𝐯)} with
reference parameter v such that the distance between 𝑔∗ and ℎ(∙; 𝐯) is minimal. A measure of this
magnitude is called the Kullback–Leibler distance, or the CE between g and h.
To solve single-objective combinatorial optimization problems, the CE method associates an estimation
problem with the original combinatorial optimization problem, characterized by a density function q. The
stochastic problem is then solved by identifying the optimal importance sampling density q*, which is the
sampling density that minimizes the CE with respect to the original density q. The minimization of the
CE leads to the generation of improved feasible vectors, and it terminates when convergence to a point in
the feasible region is achieved. The four steps for the CE method are as follows (De Boer et al., 2005):
1. Start with some 𝒑�0, say 𝒑�0 = (0.5, … ,0.5). Let 𝑡 ∶= 1.
2. Draw a sample 𝚾1 , … ,𝚾𝑁 of Bernoulli vectors with success probability vector 𝒑�𝑡−1. Calculate the
performances of 𝑆(𝚾𝑖) for all i and order them from smallest to largest, 𝑆(1) ≤ ⋯ ≤ 𝑆(𝑁). Let 𝛾�𝑡
be (1− 𝜌) quantile of the performances: 𝛾�𝑡 = 𝑆(⌈1−𝜌⌉𝑁).
3. Use the same sample to calculate 𝒑�𝑡 = ��̂�𝑡,1, … , �̂�𝑡,𝑛� with the following:
𝒑�𝑡,𝑗 =∑ 𝑰�𝑆�𝚾𝑖�≥𝛾�𝑡� 𝑰�𝑋𝑖,𝑗=1�𝑁𝑖=1
∑ 𝑰�𝑆�𝚾𝑖�≥𝛾�𝑡� 𝑁𝑖=1
4. If the stopping criterion (e.g., stop when 𝛾�𝑡 does not change for several subsequent iterations, stop
when vector 𝒑�𝑡 has converged to a binary vector) is met, stop; otherwise set 𝑡 ∶= 𝑡 + 1 and
reiterate from Step 2.
3.3.2. The CE Method for Solving MOCO Problems
The MOO CEM is similar to the CE method for a single objective in terms of deriving the best solution
through the generation of random samples based on a mechanism and then updating the parameters of the
mechanism to obtain better solutions (Bekker and Aldrich, 2011). The algorithm works as follows: We
generate the random sample using the probability density of the truncated normal distribution, based on
the lower and upper bounds of each decision variable. Using these random values, we evaluate all the
solutions in the objective functions and rank them using Goldberg’s (1989) Pareto ranking algorithm. The
best solutions are located in a vector called Elite. We then update this vector by building histograms for
each decision variable, such that the best solutions are more likely to influence the next random sample.
Finally, we use a stopping criterion such as a low standard deviation or a defined number of runs to
determine the Pareto frontier approximation.
12
Although Bekker and Aldrich (2011) demonstrate that the MOO CEM is suitable for solving MOO
problems from the literature, their study does not address issues in solving combinatorial optimization
problems. First, the solution representation is different in terms of the decision variables domain. The
MOO CEM involves continuous variables, which make the application of the truncated normal
distribution effective in terms of updating the parameters to generate new, “better” samples. However, the
decision variables in MOCO problems are either integer or binary; therefore, this approach is not useful.
In addition, the solutions that the MOO CEM obtains are not limited to constraints (just the boundaries of
the decision variables); therefore, an extension to include them in the procedure is not trivial and would
require changing the histogram concept, which is the basis of the MOO CEM.
To address these two issues, we define a new updating rule to provide feasible and better solutions in each
iteration. We propose a new metaheuristic algorithm based on the MOO CEM to solve the SFL problem:
the MOCO CEM.
The first step to define the MOCO CEM is how to represent a solution. Caserta and Quiñonez Rico
(2009) present a CE-based algorithm for solving facility location problems for a single objective function.
In their study, a binary representation was used to model the number of located facilities among the
candidates. Using this approach we propose the following vector representation for a solution: Consider
SFL problem. Let n and m be the number of demand nodes and candidate locations respectively. Table 1
shows the solution representation.
Table 1. Solution representation
Facility location Cost assignment CO2 assignment
𝑋1, … ,𝑋𝑚 𝑌𝑎11, … ,𝑌𝑎𝑛𝑚 𝑌𝑏11, … ,𝑌𝑏𝑛𝑚
Thus, we define a solution k as �𝐗𝑘 ,𝐘𝑎𝑘 ,𝐘𝑏𝑘�, where
𝐗𝑘 = (𝑋1, … ,𝑋𝑚)
= the binary vector of candidate locations for solution k, where
𝑋𝑗 takes value of 1 if a facility is located at site j and 0
otherwise.
𝐘𝑎𝑘 = (𝑌𝑎11, …𝑌𝑎𝑛𝑚) = the binary vector for the optimal cost assignment of solution
𝐗𝑘.
𝐘𝑏𝑘 = �𝑌𝑏11, …𝑌𝑏𝑛𝑚� = the binary vector for the optimal CO2 assignment of solution
𝐗𝑘.
13
Because the assignment subproblem is solvable in polynomial time, it can be solved optimally during the
algorithm procedure. In other words, for a feasible solution 𝐗𝑘, we can calculate the optimal value
𝑆(𝚾𝑘) = �𝑆(𝐘𝑎𝑘), 𝑆�𝐘𝑏𝑘��, based on the optimal assignment subproblem 𝐘𝑎𝑘 and 𝐘𝑏𝑘 of SFL. This
representation allows the algorithm to deal only with 𝐗𝑘 vectors. The MOCO CEM differs from previous
applications of the CE method in several ways:
1. To generate random samples, we propose the Bernoulli distribution as in Caserta and Quiñonez
Rico (2009). Because the decision variables 𝑋𝑗 for all 1 ≤ 𝑗 ≤ 𝑚 are binary, the truncated normal
distribution that Bekker and Aldrich (2011) suggest is no longer adequate for the problem.
2. To generate more feasible solutions 𝐗𝑘, we propose using the success probability vector with
𝐏�0 = �𝑝𝑚
, … , 𝑝𝑚�, such that p is the number of facilities to be located and m is the number of
candidate locations, which enables us to obtain more feasible solutions using smaller population
sizes. Then, to test feasibility, we evaluate each solution for the decision variables with the
expression ∑ 𝑋𝑖𝑚𝑖=1 = 𝑚.
Previous applications in single combinatorial optimization problems suggest initializing the CE
method with 𝐏�0 = (0.5, … ,0.5) (Caserta and Quiñonez Rico, 2009; De Boer et al., 2005), which
implies assuming equal probability for locating a facility in all candidate locations. This approach
might be appropriate when the number of facilities to be located is approximately half the
candidate locations, in which case we could obtain several feasible solutions rapidly in the
random sample. However, if the number of facilities is substantially more or less than half,
obtaining feasible solutions may require a very large random sample.
3. To avoid premature convergence, the MOO CEM adjusts the histogram frequencies during each
iteration t with a preset probability between .1 to .3, by inverting all the histograms for all
decision variables. Because our calculations involve the Bernoulli distribution and the histogram
class vector is reduced to two options, {0,1}, we adapted the histogram inversion operation by
changing the probability of the original vector 𝐏�𝑗 = �𝑃�𝑡1, …𝑃�𝑡𝑚� by 𝐏�𝑗 = ��1− 𝑃�𝑡1�, … �1−
𝑃�𝑡𝑚�� with the same probability range of .1 to .3.
Next we present a narrative description of the MOCO CEM. We create a vector E that will contain the
efficient solutions, and initialize the iteration counter 𝑡 = 1 and the population size N. To generate sample
vectors 𝐗𝑗 for 𝑗 = 1, … ,𝑁, we use 𝐏�𝑡−1 = �𝑝𝑚
, … , 𝑝𝑚� with m elements, as success probability values for
the Bernoulli distribution. Each 𝐗𝑗 represents a feasible location of SFL. Then we calculate 𝐘𝑎𝑗 and 𝐘𝑏𝑗
14
and the values of the fitness functions 𝑆�𝐘𝑎𝑗� and 𝑆�𝐘𝑏𝑗� for the cost and CO2 emissions functions,
respectively (see Table 2).
Table 2. Representation of the random sample 𝚾1, … ,𝚾𝑁
Facility location Cost assignment CO2 assignment 𝑆�𝚾𝑗�
𝐗1 𝐘𝑎1 𝐘𝑏1 𝑆(𝐘𝑎1) 𝑆�𝐘𝑏1�
⋮ ⋮ ⋮ ⋮ ⋮
𝐗𝑁 𝐘𝑎𝑁 𝐘𝑏𝑁 𝑆(𝐘𝑎𝑁) 𝑆�𝐘𝑏𝑁�
We rank the solution using the Pareto-ranking algorithm (Goldberg, 1989), with a threshold (th) value of
2, as in Bekker and Aldrich (2011) (i.e., we keep all solutions dominated by no more than two solutions in
E and delete the rest). Next, we update vector 𝐩�𝑡 = ��̂�𝑡,1, … , �̂�𝑡,𝑚� by using �̂�𝑡,𝑗 =∑ 𝑰�𝑆�𝚾𝑗� ∈ 𝐸� 𝑰�𝑋𝑗=1�𝑟𝑗=1
|𝐸| ,
where 𝐩�𝑡 contains the new success probability values for the Bernoulli distribution. We smooth 𝐩�𝑡 using
�̂�𝑡,𝑗 = 𝛼𝑝�𝑡,𝑗 + (1− 𝛼)�̂�𝑡−1,𝑗, and we generate another random sample of vectors 𝐗𝑗 . After a given
number of iterations, we rank E again with 𝑡ℎ = 1. Finally, using the most updated version of E, we rank
the vector with 𝑡ℎ = 0 to obtain the final version of E, which contains the efficient solutions.
The pseudo-code for the MOCO CEM algorithm to solve the SFL problem is as follows:
1. Set vector 𝐄 = ∅. Set 𝑡 = 1, 𝑘 = 1.
2. Start with 𝐏�0 = �𝑝𝑚
, … , 𝑝𝑚� with m elements.
3. Draw a sample 𝚾1, … ,𝚾𝑁 of Bernoulli vectors with success probability vector 𝐩�𝑡−1.
4. Calculate the performances 𝑆�𝚾𝑗�, 1 ≤ 𝑗 ≤ 𝑁 by solving the assignment problem of SFL.
5. Rank the performance values using the Pareto-ranking algorithm (Goldberg, 1989), with a relaxed
𝑡ℎ = 2 to obtain an updated elite vector E.
6. Use the same sample to calculate 𝐩�𝑡 = ��̂�𝑡,1, … , �̂�𝑡,𝑚� with the following equation:
�̂�𝑡,𝑗 =∑ 𝑰�𝑆�𝚾𝑗� ∈ 𝐸� 𝑰�𝑋𝑗=1�𝑟𝑗=1
|𝐸|
7. Smooth the vectors 𝐩�𝑡 using �̂�𝑡,𝑗 = 𝛼𝑝�𝑡,𝑗 + (1 − 𝛼)�̂�𝑡−1,𝑗 .
8. If all 𝜎𝑡,𝑗 > 𝜖 or less than the allowable number of evaluations has been done, increment t and
reiterate from Step 3.
9. Rank E using the Pareto-ranking algorithm (Goldberg, 1989) with 𝑡ℎ = 1.
15
10. Increment k.
11. If k is less than the allowable number of loops, return to Step 2.
12. Rank E using the Pareto-ranking algorithm (Goldberg, 1989) with 𝑡ℎ = 0 to obtain the final
vector of efficient solutions.
3.4. Validation
This section presents an analysis of the application of the MOCO CEM in solving p-Median problems
from the literature. Our purpose is to show that the algorithm can solve the problems efficiently by
approximating to the Pareto frontier; thus, we do not provide a full experimental analysis of its
performance against other algorithms.
To validate the MOCO CEM applied in p-Median problems, we solve two test instances from Beasley
(1985) (see Table 3). Because there are no extant test instances for multiobjective p-Median problems
with known Pareto frontiers, we selected the first two instances simply to exemplify the application of the
algorithm. The instances were designed to test heuristics for single objective problems; therefore, they
include only the transport cost 𝑐𝑖𝑗 and not the carbon emissions 𝑐′𝑖𝑗. We calculate the carbon emissions as
follows: 𝑐′𝑖𝑗 = �𝑐𝑖𝑗 − 𝑚𝑎𝑥�𝑐𝑖𝑗|𝑗 ∈ 𝐽, 𝑖 ∈ 𝐼��. This operation does not change the complexity of the
problem; we calculate it with the purpose of defining a different carbon emissions objective function so
that the Pareto frontier can contain more points.
Table 3. Instances from Beasley (1985)
Instances nodes edges P possible solutions
pmed1 100 200 5 7.53 x 107
pmed2 100 200 10 1.73 x 1013
To measure the performance of our algorithm, we calculate the supported solutions by using the weighted
sum scalarization method (see Section 3.1) to identify some points of the real Pareto frontier. Note that
this method can be used for pmed1 and pmed2 because neither instance is as complex as the rest of the
instances Beasley (1985) provides. Therefore, we can compare solutions in terms of their quality. We use
∆𝜆 = 0.01 such that 𝜆𝑗 = 𝜆𝑗−1 + ∆𝜆 for 𝑗 = 1, … ,100 and 𝜆0 = 0. We obtain three supported solutions in
each instance. Using the efficient solutions the MOCO CEM obtains, we select the following metrics
from Knowles and Corne (2002) to assess the efficient solutions against the supported solutions: error
ratio (ER), general distance (GD), the maximum Pareto front error (MPFE), the overall nondominated
vector generation (ONVG) and the overall nondominated vector generation ratio (ONVGR). We
16
programmed the MOCO CEM in Mathematica 9 and performed the tests on an EliteBook laptop with two
Intel Core i7 CPUs and 8 GB memory. We used 𝛼 = 0.7 and a preset probability of .1 as Bekker and
Aldrich (2011) suggest. Table 4 presents the results.
Table 4. Test results for the MOCO CEM in solving pmed1 and pmed2
Instance N ER GD MFE ONVG ONVGR
pmed1 100 0% 0% 0% 11 3.67
pmed2 200 0% 0% 0% 24 8
Because the MOCO CEM generates all the supported solutions, the metrics ER, GD and MFE are equal
to zero. Furthermore, the ONVG indicates the cardinality of the nondominated set, which is equivalent to
the number of efficient solutions. The ONVGR shows the ratio between the nondominated solutions and
the real Pareto points—in this case, estimated as efficient solutions divided by supported solutions. It
indicates that the MOCO CEM reveals 3.67 and 8 times more efficient solutions than supported solutions
for pmed1 and pmed2, respectively. Figures 1 and 2 show the efficient and supported solutions (ES and
SS) for both instances. Note that none of the efficient solutions is dominated by any supported solutions.
Figure 1- Pareto approximation for pmed1.
13800
13900
14000
14100
14200
14300
14400
14500
14600
14700
13950 14000 14050 14100 14150 14200 14250 14300
Cost
Kg of CO2
ES
SS
17
Figure 2- Pareto approximation for pmed2.
The MOCO CEM thus is capable of solving the SFL problem. In the following section, we apply the
MOCO CEM to solve the SFL for a CPG company in the Mexico City metropolitan area. We aim to
provide managerial insights regarding the structure of efficient solutions for a practical case.
4. Case Study: CPG Company in the Mexico City Metropolitan Area
The Mexico City metropolitan area is among the largest regions in any developing country; it includes
more than 18 million inhabitants, more than 2.5 million vehicles and 35,000 industries that consume more
than 44 million liters of fuel per day (Molina and Molina, 2000). Therefore, a reduction proposal
conducted in this region could significantly affect global carbon emissions. To study the type of solutions
for the SFL, we ran experiments using data from a CPG company operating in this area. This company
has operations throughout Mexico. It manages approximately 80,000 delivery points in the Mexico City
metropolitan area, with an average demand per year of approximately 2.3 million tons. The company
assigns each truck to a specific zone of approximately 100 square kilometers.
The company decision makers want to evaluate the location of 10 distribution centers and have previously
defined 18 candidate locations. Figure 3 shows a map of the area divided by square zones. Dark colors
indicate high concentration of demand. We estimated the distance between points using the great circle
distance approximation (for the distance matrix, see the Appendix).
10000
10100
10200
10300
10400
10500
10600
10700
10800
10900
9300 9500 9700 9900 10100 10300 10500
Cost
Kg of CO2
ES
SS
18
Figure 3- Map of the Mexico City metropolitan area divided by squared clusters of 100 square
kilometers. Darker areas represent higher demand. Black points represent the candidate locations.
The goal of the experiment is to indicate which efficient solutions of the SFL problem are most beneficial
to the company. Because decision makers likely seek alternatives that reduce emissions but keep costs
low, we analyze the conditions in which the efficient frontier contains solutions, such that marginal
increases in cost imply significant reductions in carbon emissions. We also investigate the possible causes
of the problem that drive the number of efficient solutions.
We define the following factors with their respective levels:
1. Truck assignment:
a) Trucks are assigned based on the area’s conditions (e.g., only small trucks can access
downtown).
b) Trucks are assigned without considering the area’s conditions (e.g., large trucks are assigned
to regions with high demand).
19
2. Fixed cost assignment:
a) We identify high fixed costs for short distance trajectories and vice versa (High 𝐴𝑖𝑗 – Short
𝑑𝑖𝑗).
b) We identify high fixed costs for large distance trajectories and vice versa (High 𝐴𝑖𝑗 – Large
𝑑𝑖𝑗).
Our response variables are the number of nondominated solutions (NDS) and the ratio between the
reductions in carbon emissions divided by the increase in costs (RCC). For this last indicator, we consider
only the largest value. We replicate each setting 10 times. Table 6 shows the results for the NDS response
variable. The numbers suggest that when high fixed costs are present in short distance trajectories, the
number of nondominated solutions is larger than when high fixed costs exist in large distance trajectories.
Table 6. Results for NDS response variable
MCMA conditions Large trucks to high demand nodes Replicates High Aij - Large dij High Aij - Short dij High Aij - Large dij High Aij - Short dij
1 2 16 2 15 2 3 18 3 21 3 3 21 3 22 4 3 21 3 22 5 3 21 3 22 6 3 21 3 22 7 3 21 3 22 8 3 21 3 22 9 3 21 3 22 10 3 21 3 22
Table 7 shows the results for the RCC response variable. We notice similar behavior with the NDS
response variable, such that when high fixed costs are present for short distance trajectories, the
percentage of CO2 reduction divided by the cost increase percentage is larger than when high fixed costs
are present for large distance trajectories. We obtain large numbers (e.g., replicate 3 reports 1.9 million)
because the increase in cost is close to 0% (2 E^10-7%).
20
Table 7. Results for RCC response variable (CO2 reduction divided by the increase in cost)
MCMA conditions Large trucks to high demand nodes Replicates High Aij - Large dij High Aij - Short dij High Aij - Large dij High Aij - Short dij
1 1.09 47,442,889.08 0.12 384,331.72
2 0.05 1,946,988.58 0.12 384,721.67
3 0.05 1,946,988.58 0.12 384,721.67
4 0.05 1,946,988.58 0.12 384,721.67
5 0.05 1,946,988.58 0.12 384,721.67
6 0.05 1,946,988.58 0.12 384,721.67
7 0.05 1,946,988.58 0.12 384,721.67
8 0.05 1,946,988.58 0.12 384,721.67
9 0.05 1,946,988.58 0.12 384,721.67
10 0.05 1,946,988.58 0.12 384,721.67
To determine which factors affect the response variables, we conducted two analyses of variance. Table 8
shows the analysis of variance for the NDS response variable. We conclude that the assignment of fixed
costs to distances explains the number of nondominated solutions for the problem, independent of the
area’s conditions.
Table 8. Analysis of variance for the NDS response variable
Source DF SS MS F P
Truck assignment 1 .1 .1 .6 .444
Fixed cost & distance 1 220.9 220.9 1325.4 .000
Interaction 1 .1 .1 .6 .444
Error 36 6 .167
Table 9 shows the analysis of variance for the RCC response variable. We conclude that the truck
assignment, the assignment of fixed costs to distances and their interaction explain the percentage of CO2
reduction divided by the percentage of increase in cost for the problem.
21
Table 9. Analysis of variance for the RCC response variable (CO2 reduction divided by the increase
in cost)
Source DF SS MS F P
Truck assignment 1 2.82909E+18 2.82909E+18 19561.8 .000
Fixed cost & distance 1 6.14065E+18 6.14065E+18 42459.6 .000
Interaction 1 2.82909E+18 2.82909E+18 19561.8 .000
Error 36 5.20644E+15 1.44623E+14
To demonstrate the structure of Pareto approximation for the case study, we present some examples of the
results from the experiment. Figure 4, Panel A, shows an example of the Pareto approximation for the
area’s conditions with high fixed costs in short distance trajectories. Figure 4, Panel B, shows only the
nondominated solutions with an increase in cost close to zero (less than 1.4 E^10-6%). Note that it is
possible to achieve a significant reduction in carbon emissions (~40%) with almost no cost increase,
which may make the solutions equivalent for practical purposes.
Figure 4 (A): Pareto approximation for the area’s conditions with high fixed costs found in short
distance trajectories; and (B): Sample of points with less than 1% of increase in cost.
Figure 5 shows the locations in the map for three Pareto approximation points from Figure 4. Panel A
shows the cost optimal location, Panel B shows a location with 36% of carbon reduction with an increase
in cost of 3 E^10-5% and Panel C shows the carbon emissions optimal location with 12.5% of increase in
cost.
0.0E+00
2.0E-07
4.0E-07
6.0E-07
8.0E-07
1.0E-06
1.2E-06
1.4E-06
1.6E-06
0% 10% 20% 30% 40% 50%
% o
f Cos
t inc
reas
e
% of CO2 reduction
0%
2%
4%
6%
8%
10%
12%
0% 10% 20% 30% 40% 50%
% o
f Cos
t inc
reas
e
% of CO2 reduction
22
A: Cost optimal location B: Location with 36% of carbon reduction and
cost increase of 3 E^10-5%
C: CO2 optimal location
Figure 5- Examples of three Pareto points for the area’s conditions with high fixed costs in short
distance trajectories.
Note that the solutions shown in Panels A and B of Figure 5 are similar in terms of changing only two
locations (X8 and X14 by X15 and X10). Furthermore, the CO2 optimal location (Panel C) is different than
the other solutions in terms of locating facilities closer to regions with high concentration of demand (and
also in terms of truck constraints).
23
Figure 6 shows an example of the Pareto approximation with high fixed costs in large distance
trajectories. Note that for these conditions, the three solutions result in similar costs and carbon emissions
(the largest difference is .025% of CO2 reduction with a cost increase of .55%).
Figure 6- Pareto approximation with high fixed costs in large distance trajectories
Figure 7, Panel A, shows an example of the Pareto approximation when the company does not need to
accommodate Mexico City metropolitan area conditions (e.g., large trucks may be assigned to regions
with high demand) and when high fixed costs exist in short distance trajectories. Figure 7, Panel B, shows
only the nondominated solutions with an increase in cost close to zero (less than 5 E^10-6%). Note that it
is possible to achieve high reductions in carbon emissions (~25%) with almost no cost increase, similar to
the solutions shown in Figure 4, which may also make the solutions equivalent for practical purposes.
Figure 7 (A): Pareto approximation when the company does not accommodate Mexico City and
metropolitan area conditions and with high fixed costs in short distance trajectories; (B): Sample of
points with less than 1% of increase in cost.
0.0%
0.1%
0.2%
0.3%
0.4%
0.5%
0.6%
0.000% 0.005% 0.010% 0.015% 0.020% 0.025% 0.030%
% o
f Cos
t inc
reas
e
% of CO2 reduction
0.0E+00
5.0E-07
1.0E-06
1.5E-06
2.0E-06
2.5E-06
3.0E-06
3.5E-06
4.0E-06
4.5E-06
5.0E-06
0% 5% 10% 15% 20% 25% 30%
% o
f Cos
t inc
reas
e
% of CO2 reduction
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
10%
0% 5% 10% 15% 20% 25% 30%
% o
f Cos
t inc
reas
e
% of CO2 reduction
24
Figure 8 shows the locations in the map for three Pareto approximation points from Figure 7. Panel A
indicates the cost optimal location, Panel B shows a location with 26% of carbon reduction with an
increase in cost of 4.7 E^10-4% and Panel C shows the carbon emissions optimal location with 9% of
increase in cost.
A: Cost optimal location B: Location with 26% of carbon reduction and
cost increase of 4.7 E^10-4%
C: CO2 optimal location
Figure 8- Examples of three Pareto points when the company does not accommodate Mexico City
and metropolitan area conditions and with high fixed costs in short distance trajectories.
Note that the CO2 optimal location (Panel C) is different than the other solutions in terms of locating
facilities closer to regions with high concentration of demand (and also in terms of truck constraints).
25
Figure 9 shows an example of the Pareto approximation when the company accounts for Mexico City
metropolitan area conditions and experiences high fixed costs with large distance trajectories. Similar to
Figure 6, under these conditions, the three solutions result in similar costs and carbon emissions (the
largest difference is .12% of CO2 reduction with an increase in cost of almost 1.2%).
Figure 9- Pareto approximation for nonconstrained delivery conditions with high fixed costs for
large distance trajectories
5. Conclusions
Transportation is a main contributing factor of carbon emissions, and studies indicate that these emissions
will increase in the future. This situation suggests that firms need more efficient transport logistics
operations to reduce CO2 emissions while keeping costs low. In the current study, we investigate the
trade-off between cost and CO2 using a multiobjective approach for the p-Median problem. We provide
three main contributions: First, we introduce the SFL problem, a multiobjective p-Median model with
new cost and CO2 objective functions that allow us to take into account vehicle access limitations in the
demand nodes. Second, to solve the SFL, we develop an adaptation of the MOO CEM: the MOCO CEM.
We test our algorithm against problems from prior literature and show that the MOCO CEM can solve the
multiobjective p-Median problem efficiently by finding efficient and supported solutions and by
performing better in the different metrics. Finally, we provide managerial insights on the structure of the
Pareto frontier using an application for an actual company: we conduct an experimental study of the
MOCO CEM by solving the SFL for a multinational CPG company that operates in Mexico City
metropolitan area. Our intention is to analyze two scenarios: when high fixed costs exist in short distance
trajectories and when high fixed costs exist in large distance trajectories. The results show that regardless
of Mexico City conditions, when high fixed costs are present in short distance trajectories, CO2 emission
0.0%
0.2%
0.4%
0.6%
0.8%
1.0%
1.2%
1.4%
0.00% 0.02% 0.04% 0.06% 0.08% 0.10% 0.12% 0.14%
% o
f Cos
t inc
reas
e
% of CO2 reduction
26
reductions are much larger (~40% considering area conditions and 25% not considering these conditions)
than the increase in cost (almost 0% in both cases). The results also show that these CO2 savings can be
achieved by only changing a small subset of locations. The study provides insights that can help
companies make better decisions by analyzing the cost and CO2 trade-offs for the facility location
problem.
We leave comparison of the MOCO CEM against other algorithms from the literature as a fruitful future
research avenue. Furthermore, including other type of pollutants—such as noise, particulate matter, CO
and NOx—as possible objective functions in the SFL model is a worthwhile research direction. For this
problem, researchers may need to develop new heuristics strategies to accommodate the complexities.
References
Ajith, A., Jain, L., Goldberg, R. 2005. Evolutionary multiobjective optimization, theoretical advances and
applications. London: Springer-Verlag London Limited.
Alon, G., D.P. Kroese, T. Raviv, R.Y. Rubinstein. 2005. Application of the cross-entropy method to the
buffer allocation problem in a simulation-based environment. Annals of Operations Research
134, 19–67.
Beasley, J.E. 1985. A note on solving large p-median problems. European Journal of Operational
Research 21, 270–273.
Bekker, J., Aldrich, C. 2011. The cross-entropy method in multi-objective optimisation: An assessment.
European Journal of Operational Research 1(211), 12–121.
Blanquero, R., Carrizosa, E. 2002. A DC biobjective location model, Journal of Global Optimization 23,
139–154.
Caserta, M., Quiñonez Rico, E. 2009. A cross entropy-based metaheuristic algorithm for large-scale
capacitated facility location problems. Journal of the Operational Research Society 60(10),
1439–1448.
Carbon Disclosure Project. 2011. Carbon Disclosure Project. Available at www.cdproject.net/, accessed
on March 21, 2011.
Chepuri, K., Homem-de-Mello, T. 2005. Solving the vehicle routing problem with stochastic demands
using the cross-entropy method. Annals of Operations Research 134, 153–181.
Coello, C.A. 2009. Evolutionary multi-objective optimization: Some current research trends and topics
that remain to be explored. Frontiers of Computer Science in China 3(1), 18–30.
Coello, C.A., Lamont, G.B., Van Veldhuizen, D.A., 2007. Evolutionary algorithms for solving multi-
objective problems, 2nd ed. New York: Springer.
27
Cohen, I., Golany, B., Shtub, A. 2005. Managing stochastic finite capacity multi-project systems through
the cross-entropy method. Annals of Operations Research 134, 183–199.
Daskin, M.S., Snyder, L.V., Berger, R.T., 2005. Facility location in supply chain design. In: Logistics
systems design and optimization, ed. Langevin, A. and Riopel, D. New York: Springer.
De Boer, P., Kroese, D.P., Mannor, S., Rubinstein, R.Y. 2005. A tutorial on the cross-entropy method.
Annals of Operations Research 134, 19–67.
Deb, K. 2001. Multi-objective optimization using evolutionary algorithms. Chichester, UK: John Wiley &
Sons, LTD.
Deb, K., Agrawal, S., Pratap, A., Meyarivan, T. 2000. A fast elitist non-dominated sorting genetic
algorithm for multi-objective optimization: NSGA-II. Lecture Notes in Computer Science 1917,
849–858.
Du, F., Evans, G.W. 2008. A bi-objective reverse logistics network analysis for post-sale service.
Computer Operations Research 35, 2617–2634.
Ehrgott, M. 2000. Approximation algorithms for combinatorial multicriteria optimization problems.
International Transactions in Operational Research 7(1) 5-31.
Ehrgott, M. 2005. Multicriteria optimization, 2nd ed. New York: Springer.
Ehrgott, M., Gandibleux, X. 2000. A survey and annotated bibliography of multiobjective combinatorial
optimization. OR-Spektrum 22(4), 425–460.
European Commission 2011. Roadmap to a single European transport area: Towards a competitive and
resource efficient transport system. Available at http://ec.europa.eu/transport/index_en.htm,
accessed on May 27, 2011.
Farahani, R.Z., SteadieSeifi, M., Asgari, N. 2010. Multiple criteria facility location problems: A survey.
Applied Mathematical Modeling 34, 1689–1709.
Fonseca, M.C., Fleming, P.J. 1995. An overview of evolutionary algorithms in multiobjective
optimization. Evolutionary Computation 3(1), 1–16.
Fonseca, M.C., García-Sánchez, A., Ortega-Mier, M., Saldanha-da-Gama, F. 2009. A stochastic bi-
objective location model for strategic reverse logistics. TOP, 18(1), 158–184.
Greenhouse Gas Protocol Standard 2011. The Greenhouse Gas Protocol. Available at
http://www.ghgprotocol.org/standards, accessed on March 11, 2011.
Goldberg, D. 1989. Genetic algorithms in search, optimization, and machine learning. Reading, MA:
Addison-Wesley Professional.
Hakimi, S.L. 1964. Optimum locations of switching centers and the absolute centers and medians of a
graph. Operations Research 12, 450–459.
28
Hansen, P. 1980. Bicriterion path problems. In: Multiple criteria decision making theory and application.
Heidelberg, DE: Springer Berlin Heidelberg, 109–127.
Helvik, B.E., Wittner, O. 2001. Using the cross-entropy method to guide/govern mobile agents’ path
finding in networks. In Mobile Agents for Telecommunication Applications. Spring Berlin
Heidelberg, 255-268.
Hoffman, A.J., Woody, J.G. 2008. Climate change: What’s your business strategy? Cambridge, MA:
Harvard Business Press.
Horn, J., Nafpliotis, N., Goldberg, D.E. 1994. A niched pareto genetic algorithm for multiobjective
optimization. in Evolutionary Computation. In: Proceedings of the First IEEE Conference on
Computational Intelligence. New York: Institute of Electrical and Electronics Engineers.
Hui, K.P., Bean, N., Kraetzl, M., Kroese, D.P. 2005. The cross-entropy method for network reliability
estimation. Annals of Operations Research 134(1), 101–118.
Intergovernmental Panel on Climate Change 2007. Climate change 2007: An assessment of the
Intergovernmental Panel on Climate Change. Available at http://www.ipcc.ch/, accessed on May
27, 2011.
Jalil, M., Zuidwijk, R., Fleischmann, M., Van Nunen, J. 2011. Spare parts logistics and installed base
information. Journal of the Operational Research Society 62(3), 442–457.
Jaszkiewicz, A. 2002. Genetic local search for multi-objective combinatorial optimization. European
Journal of Operational Research 137(1), 50–71.
Kariv, O., Hakimi, S.L. 1979. An algorithmic approach to network location problems, Part II: The p-
median. SIAM Journal of Applied Mathematics 37, 539–560.
Keith, J., Kroese, D.P. 2002. SABRES: Sequence alignment by rare event simulation. In: Proceedings of
the 2002 Winter Simulation Conference. New York: Institute of Electrical and Electronics
Engineers, 320–327.
Knowles, J., Corne, D. 2002. On metrics for comparing nondominated sets. In: Evolutionary
Computation, 2002. Proceedings of the 2002 Congress on, Vol. 1. New York: Institute of
Electrical and Electronics Engineers, 711–716.
Köksalan, M. 2008. Multiobjective combinatorial optimization: Some approaches. Journal of Multi-
Criteria Decision Analysis 15(3–4), 69–78.
Medaglia, A.L., Villegas, J.G., Rodríguez-Coca, D.M. 2009. Hybrid bi-objective evolutionary algorithms
for the design of a hospital waste management network. Journal of Heuristics 15, 153–176.
Molina, M.J., Molina, L.T. 2000. Integrated strategy for air management in Mexico City Metropolitan
Area. Boston: Massachusetts Institute of Technology.
29
Network for Transport and Environment Road 2008. Environmental data for international cargo
transport-road transport. Available at http://www.ntmcalc.se/index.html. Accessed on February
13, 2011.
ReVelle, C.S., Eiselt, H.A., Daskin, M.S., 2008. A bibliography for some fundamental problem
categories in discrete location science. European Journal of Operational Research 184, 817–848.
Rubinstein, R.Y. 1997. Optimization of computer simulation models with rare events. European Journal
of Operations Research 99, 89–112.
Rubinstein, R.Y. 2002. Cross-entropy and rare-events for maximal cut and bipartition problems. In: ACM
transactions on modeling and computer simulation. New York: Association for Computing
Machinery, 27–53.
Rubinstein, R.Y., Kroese, D.P. 2004. The cross-entropy method: A unified approach to combinatorial
optimization, Monte-Carlo simulation, and machine learning. New York: Springer.
Schaffer, J.D. 1985. Multiple objective optimization with vector evaluated genetic algorithms. In:
Proceedings of the First International Conference on Genetic Algorithms. Mahwah, NJ: L.
Erlbaum Associates, 93–100.
Sergienko, I.V., Perepelitsa, V.A. 1991. Finding the set of alternatives in discrete multicriterion problems.
Cybernetics and Systems Analysis 23(5), 673–683.
Transport Research Laboratory. 2010. Information retrieved from ARTEMIS project web site.
http://www.trl.co.uk/artemis/. Accessed on June 5, 2012.
Velázquez-Martínez, J.C., Fransoo, J.C., Blanco, E.E., Mora-Vargas, J. 2013. The impact of carbon
footprinting aggregation on realizing emission reduction targets. Flexible Services and
Manufacturing Journal (January), 1–25.
Zitzler, E., Deb, K., Thiele, L. 2000. Comparison of multiobjective evolutionary algorithms: Empirical
results. Evolutionary Computation 8(2), 173–195.
Zitzler, E., Thiele, L. 1998. An evolutionary algorithm for multiobjective optimization: The strength
pareto approach. Swiss Federal Institute of Technology, TIK-Report 43.
30
APPENDIX: DISTANCE MATRIX (IN KILOMETERS) FOR THE CPG COMPANY
Node/Candidate Location 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 181 57.5 59.3 62.4 78.4 59.1 49.9 59.2 83.5 62.4 62.8 60.6 68.6 81.7 73.8 42.6 83.6 84.1 88.62 63.0 64.6 67.2 82.1 64.1 55.0 65.5 87.5 67.2 67.2 65.7 73.2 85.7 78.7 47.9 88.0 88.6 92.73 45.0 46.8 50.2 66.7 46.7 37.5 46.5 71.6 50.2 50.7 48.3 56.5 69.7 61.4 30.2 71.5 71.9 76.64 50.3 51.9 54.6 70.0 51.5 42.3 52.8 75.2 54.6 54.7 53.0 60.6 73.4 66.1 35.2 75.6 76.1 80.35 56.4 57.9 59.9 74.1 57.2 48.3 59.8 79.6 59.9 59.7 58.7 65.7 77.9 71.5 41.4 80.4 81.0 84.86 28.6 30.8 36.2 54.4 32.2 25.3 25.6 57.6 36.2 38.1 33.4 42.2 55.6 44.7 21.3 55.7 55.9 61.87 29.4 31.5 36.2 54.2 32.3 23.8 28.7 58.2 36.2 37.5 33.6 42.5 56.2 46.2 17.8 57.0 57.3 62.88 32.6 34.5 38.3 55.5 34.6 25.4 33.8 60.1 38.2 39.1 36.1 44.6 58.1 49.2 18.3 59.6 60.0 64.99 37.5 39.2 42.1 58.0 38.9 29.7 40.2 63.1 42.1 42.4 40.4 48.2 61.3 53.5 22.5 63.2 63.7 68.2
10 43.7 45.2 47.2 61.7 44.4 35.6 47.3 67.2 47.2 47.1 46.0 53.1 65.4 58.8 28.8 67.8 68.4 72.311 50.7 51.9 53.3 66.4 50.9 42.5 55.0 72.1 53.3 52.8 52.4 58.8 70.4 64.9 36.1 73.2 73.8 77.312 22.1 24.0 29.7 47.0 26.1 22.6 16.3 49.2 29.8 32.1 26.8 34.8 47.4 35.8 22.5 46.7 46.7 52.913 19.5 21.6 27.0 45.2 23.1 16.9 16.7 48.3 27.1 29.0 24.2 33.0 46.4 35.5 14.7 46.5 46.7 52.614 20.5 22.6 27.1 45.0 23.2 14.6 21.1 49.1 27.1 28.3 24.6 33.4 47.1 37.4 8.9 48.1 48.4 53.715 24.8 26.6 29.8 46.5 26.4 17.1 27.6 51.3 29.7 30.3 27.9 36.0 49.4 41.1 9.8 51.1 51.5 56.316 31.1 32.5 34.5 49.5 31.7 22.9 35.2 54.8 34.5 34.5 33.2 40.4 53.0 46.1 16.4 55.3 55.8 60.017 38.3 39.5 40.7 53.8 38.3 30.2 43.1 59.5 40.6 40.1 39.8 46.1 57.8 52.2 24.3 60.5 61.1 64.618 46.1 47.1 47.6 59.1 45.7 38.1 51.4 64.9 47.6 46.7 47.1 52.6 63.4 58.9 32.7 66.4 67.1 70.119 21.9 23.0 28.0 42.3 25.5 26.6 14.1 43.0 28.0 30.7 25.5 31.1 41.4 29.7 29.7 39.6 39.4 45.720 14.7 16.3 21.8 38.3 18.6 18.1 7.6 40.2 21.9 24.4 19.0 26.3 38.4 26.7 20.9 37.5 37.5 43.721 10.3 12.4 18.0 36.1 14.1 10.1 8.4 39.1 18.0 20.1 15.1 23.8 37.2 26.3 12.2 37.4 37.6 43.422 12.2 14.0 18.0 35.8 14.3 5.4 15.4 40.0 18.0 19.2 15.7 24.4 38.1 28.8 3.5 39.2 39.6 44.823 18.6 19.8 21.9 37.7 19.0 10.4 23.6 42.7 21.8 22.0 20.5 27.9 40.8 33.4 5.4 42.8 43.4 47.824 26.4 27.3 28.0 41.3 25.9 18.5 32.0 46.9 28.0 27.4 27.3 33.4 45.2 39.5 14.1 47.8 48.4 52.125 34.6 35.3 35.3 46.4 33.7 27.0 40.7 52.2 35.2 34.2 35.0 40.0 50.6 46.4 22.9 53.7 54.4 57.426 51.6 52.1 51.2 59.1 50.2 44.3 58.0 65.2 51.2 49.8 51.4 55.1 63.8 61.7 40.4 67.5 68.2 70.127 19.5 19.7 23.4 34.9 22.0 26.3 12.7 34.8 23.4 26.0 21.5 24.8 33.3 22.0 31.4 31.1 30.8 37.128 10.7 11.1 15.5 29.9 13.6 17.7 4.5 31.3 15.5 18.2 13.2 18.4 29.5 17.7 23.3 28.4 28.3 34.629 2.1 3.6 9.3 27.0 5.9 9.4 5.7 29.9 9.4 11.8 6.4 14.7 28.0 17.2 15.9 28.2 28.4 34.230 6.9 7.5 9.3 26.7 6.3 3.9 14.1 31.0 9.3 10.1 7.8 15.6 29.1 20.8 10.7 30.6 31.1 35.931 15.6 15.9 15.5 29.1 14.0 9.8 22.7 34.4 15.5 14.7 15.3 20.7 32.7 26.9 11.5 35.1 35.7 39.632 24.3 24.5 23.4 33.7 22.5 18.1 31.4 39.5 23.4 22.0 23.7 27.6 37.9 34.1 17.4 41.0 41.7 44.733 33.1 33.2 31.7 39.7 31.2 26.7 40.2 45.7 31.7 30.1 32.2 35.4 44.3 41.9 25.0 47.8 48.6 50.834 50.6 50.7 48.9 54.1 48.6 44.1 57.7 60.1 48.8 47.0 49.6 51.8 58.9 58.4 41.6 62.8 63.6 64.835 29.4 28.8 30.6 35.9 30.8 37.2 24.2 33.7 30.7 32.8 29.7 29.4 32.8 24.1 43.0 29.2 28.6 34.236 21.1 20.4 21.9 28.6 22.2 29.2 17.1 27.3 21.9 24.1 21.1 20.9 26.1 16.1 35.5 23.2 22.8 28.937 13.5 12.3 13.1 22.2 13.8 21.7 12.3 22.6 13.2 15.3 12.6 12.6 21.0 9.4 28.5 19.4 19.3 25.638 8.5 6.5 4.5 18.0 6.5 15.6 12.8 20.7 4.5 6.6 4.9 5.9 18.7 8.4 22.9 19.1 19.4 25.039 10.7 9.2 4.5 17.5 6.8 13.1 18.2 22.3 4.5 2.3 6.6 8.0 20.4 14.3 19.7 22.5 23.1 27.340 17.6 16.8 13.2 21.1 14.3 15.9 25.5 26.8 13.1 11.0 14.7 15.7 25.2 22.2 20.1 28.4 29.1 32.041 25.7 25.1 21.9 27.1 22.7 22.0 33.5 33.1 21.9 19.7 23.3 24.1 31.7 30.6 24.0 35.4 36.2 38.142 34.1 33.7 30.7 34.3 31.3 29.5 41.8 40.3 30.6 28.4 32.0 32.7 39.1 39.1 30.0 43.1 43.9 45.143 42.6 42.3 39.4 42.1 39.9 37.6 50.2 48.0 39.4 37.2 40.7 41.4 47.0 47.7 37.1 51.1 51.9 52.644 33.2 32.1 32.3 32.5 33.5 41.3 29.8 29.0 32.3 33.9 32.2 29.2 28.5 22.9 47.9 24.3 23.6 28.245 26.2 24.8 24.1 24.2 25.9 34.3 24.4 21.2 24.2 25.5 24.4 20.6 20.5 14.2 41.2 16.6 16.0 21.546 20.5 18.7 16.6 16.2 19.2 28.2 21.3 14.6 16.6 17.6 17.6 12.1 13.4 5.6 35.5 10.7 10.4 16.747 17.6 15.5 11.1 9.6 14.8 23.9 21.6 11.5 11.1 10.8 13.3 4.9 9.5 3.7 31.1 10.2 10.7 15.848 18.8 16.9 11.1 8.7 14.9 22.3 25.1 14.2 11.1 8.9 14.1 7.3 12.5 12.2 28.8 15.7 16.5 19.349 23.4 21.9 16.6 14.6 19.5 24.0 30.8 20.6 16.6 13.9 19.2 15.3 19.4 20.9 29.1 23.3 24.1 25.550 30.0 28.8 24.1 22.4 26.3 28.5 37.7 28.3 24.1 21.5 26.4 23.9 27.3 29.6 31.9 31.5 32.3 32.851 37.4 36.5 32.3 30.7 34.0 34.6 45.2 36.4 32.2 29.7 34.3 32.5 35.6 38.4 36.6 39.9 40.8 40.752 45.3 44.6 40.7 39.2 42.1 41.7 53.1 44.8 40.6 38.2 42.5 41.2 44.1 47.1 42.7 48.5 49.3 48.953 38.9 37.4 36.2 31.6 38.3 47.0 36.9 26.7 36.3 37.3 36.9 31.8 26.9 25.3 53.9 22.4 21.6 24.454 33.1 31.4 29.2 22.8 31.9 40.9 32.7 18.0 29.3 29.9 30.3 24.2 18.1 17.8 48.1 13.7 12.8 16.155 28.8 26.8 23.4 14.1 26.7 36.0 30.4 9.4 23.5 23.5 25.2 17.5 9.4 12.0 43.3 4.9 4.1 8.856 26.8 24.7 19.9 5.5 23.8 32.7 30.6 2.3 19.9 19.0 22.4 13.5 0.7 11.3 39.8 3.8 4.7 6.957 27.6 25.6 19.9 3.6 23.9 31.5 33.2 8.6 19.9 18.0 22.8 14.6 8.1 16.2 38.0 12.6 13.4 13.158 31.0 29.1 23.5 12.2 27.0 32.8 37.7 17.2 23.4 20.9 26.3 19.8 16.9 23.5 38.2 21.3 22.1 21.159 36.2 34.6 29.3 20.9 32.2 36.2 43.5 25.9 29.2 26.6 31.9 27.0 25.6 31.5 40.4 30.0 30.9 29.660 42.5 41.2 36.3 29.7 38.7 41.2 50.2 34.6 36.2 33.6 38.7 34.9 34.3 39.8 44.2 38.8 39.6 38.261 49.6 48.5 43.9 38.4 46.0 47.3 57.4 43.4 43.9 41.2 46.1 43.1 43.1 48.3 49.3 47.5 48.4 46.862 40.9 39.0 36.0 25.1 39.1 48.4 41.4 19.2 36.0 36.2 37.6 30.2 20.2 24.5 55.6 16.3 15.6 15.163 37.5 35.5 31.5 17.5 35.1 44.3 39.6 11.5 31.5 31.1 33.6 25.2 12.9 20.7 51.5 10.2 9.8 6.764 36.0 33.9 29.0 11.7 32.9 41.7 39.8 7.0 29.0 27.9 31.5 22.6 9.0 20.3 48.6 9.7 10.0 3.965 36.6 34.5 29.0 11.0 33.0 40.8 41.8 10.9 29.0 27.2 31.9 23.3 12.1 23.4 47.2 15.4 16.1 11.866 39.2 37.3 31.5 16.0 35.3 41.7 45.4 18.5 31.5 29.2 34.5 26.9 19.1 28.9 47.4 23.1 23.8 20.367 43.4 41.7 36.0 23.4 39.4 44.4 50.4 26.8 36.0 33.5 38.9 32.5 27.1 35.7 49.2 31.3 32.1 29.068 48.9 47.3 41.9 31.4 44.9 48.6 56.2 35.3 41.9 39.3 44.6 39.3 35.5 43.2 52.3 39.8 40.6 37.769 55.1 53.8 48.7 39.8 51.3 53.9 62.7 43.9 48.6 46.0 51.2 46.8 44.0 51.1 56.7 48.3 49.1 46.470 46.5 44.4 40.1 24.2 43.8 52.9 48.8 18.6 40.1 39.4 42.3 33.7 20.4 29.7 60.1 18.8 18.5 13.571 45.3 43.1 38.1 20.4 42.1 50.7 48.9 16.2 38.1 36.9 40.7 31.8 18.2 29.4 57.6 18.6 18.7 12.472 45.7 43.6 38.1 20.0 42.1 50.0 50.6 18.3 38.1 36.4 41.0 32.3 19.9 31.6 56.4 22.1 22.5 16.673 47.8 45.8 40.1 23.1 44.0 50.8 53.6 23.6 40.1 37.9 43.0 35.0 24.8 35.9 56.6 28.0 28.6 23.574 51.3 49.5 43.7 28.7 47.3 53.0 57.9 30.5 43.7 41.3 46.7 39.5 31.4 41.6 58.1 35.1 35.8 31.375 56.0 54.3 48.7 35.6 52.0 56.6 63.0 38.2 48.7 46.1 51.5 45.2 38.8 48.2 60.8 42.8 43.5 39.576 61.6 60.0 54.6 43.1 57.6 61.1 68.9 46.2 54.6 52.0 57.3 51.8 46.7 55.4 64.6 50.8 51.6 47.977 67.8 66.4 61.2 51.1 63.9 66.5 75.4 54.5 61.2 58.5 63.8 59.0 54.9 63.0 69.3 59.1 59.9 56.478 54.5 52.3 47.3 29.4 51.3 59.8 58.1 25.5 47.3 46.0 49.9 41.0 27.4 38.6 66.7 27.7 27.7 21.579 54.9 52.7 47.3 29.1 51.3 59.2 59.5 26.8 47.3 45.6 50.1 41.4 28.6 40.3 65.6 30.1 30.4 24.280 56.6 54.6 48.9 31.4 52.8 59.9 62.1 30.7 48.9 46.8 51.8 43.5 32.1 43.7 65.8 34.7 35.1 29.381 59.6 57.7 51.9 35.7 55.7 61.8 65.8 36.2 51.9 49.6 54.9 47.2 37.5 48.5 67.0 40.6 41.2 35.982 63.7 61.9 56.2 41.4 59.7 64.8 70.4 42.9 56.1 53.7 59.1 52.1 43.9 54.3 69.4 47.5 48.1 43.283 68.6 66.9 61.4 48.0 64.6 68.9 75.7 50.2 61.3 58.8 64.2 57.9 51.0 60.8 72.8 54.8 55.5 51.084 63.7 61.6 56.5 38.5 60.5 69.0 67.3 34.7 56.5 55.2 59.1 50.2 36.6 47.8 75.7 36.8 36.8 30.685 64.0 61.9 56.5 38.3 60.5 68.4 68.5 35.7 56.5 54.8 59.3 50.5 37.5 49.2 74.8 38.7 38.9 32.686 65.5 63.5 57.8 40.0 61.8 69.0 70.8 38.7 57.8 55.8 60.7 52.3 40.3 52.0 75.0 42.4 42.7 36.6
Working Papers Beta 2009 - 2014 nr. Year Title Author(s) 451 450 449 448 447 446 445 444 443 442 441 440 439
2014 2014 2014 2014 2014 2014 2014 2014 2014 2014 2013 2013 2013
Transportation Cost and CO2 Emissions in Location Decision Models Tracebook: A Dynamic Checklist Support System Intermodal hinterland network design with multiple actors The Share-a-Ride Problem: People and Parcels Sharing Taxis Stochastic inventory models for a single item at a single location Optimal and heuristic repairable stocking and expediting in a fluctuating demand environment Connecting inventory control and repair shop control: a differentiated control structure for repairable spare parts A survey on design and usage of Software Reference Architectures Extending and Adapting the Architecture Tradeoff Analysis Method for the Evaluation of Software Reference Architectures A multimodal network flow problem with product Quality preservation, transshipment, and asset management Integrating passenger and freight transportation: Model formulation and insights The Price of Payment Delay On Characterization of the Core of Lane Covering Games via Dual Solutions
Josue C. Vélazquez-Martínez, Jan C. Fransoo, Edgar E. Blanco, Jaime Mora- Vargas Shan Nan, Pieter Van Gorp, Hendrikus H.M. Korsten, Richard Vdovjak, Uzay Kaymak Yann Bouchery, Jan Fransoo Baoxiang Li, Dmitry Krushinsky, Hajo A. Reijers, Tom Van Woensel K.H. van Donselaar, R.A.C.M. Broekmeulen Joachim Arts, Rob Basten, Geert-Jan van Houtum M.A. Driessen, W.D. Rustenburg, G.J. van Houtum, V.C.S. Wiers Samuil Angelov, Jos Trienekens, Rob Kusters Samuil Angelov, Jos J.M. Trienekens, Paul Grefen Maryam SteadieSeifi, Nico Dellaert, Tom Van Woensel Veaceslav Ghilas, Emrah Demir, Tom Van Woensel K. van der Vliet, M.J. Reindorp, J.C. Fransoo Behzad Hezarkhani, Marco Slikker, Tom van Woensel
438 437 436 435 434 433 432 431 430 429 428
2013 2013 2013 2013 2013 2013 2013 2013 2013 2013 2013
Destocking, the Bullwhip Effect, and the Credit Crisis: Empirical Modeling of Supply Chain Dynamics Methodological support for business process Redesign in healthcare: a systematic literature review Dynamics and equilibria under incremental Horizontal differentiation on the Salop circle Analyzing Conformance to Clinical Protocols Involving Advanced Synchronizations Models for Ambulance Planning on the Strategic and the Tactical Level Mode Allocation and Scheduling of Inland Container Transportation: A Case-Study in the Netherlands Socially responsible transportation and lot sizing: Insights from multiobjective optimization Inventory routing for dynamic waste collection Simulation and Logistics Optimization of an Integrated Emergency Post Last Time Buy and Repair Decisions for Spare Parts A Review of Recent Research on Green Road Freight Transportation
Maximiliano Udenio, Jan C. Fransoo, Robert Peels Rob J.B. Vanwersch, Khurram Shahzad, Irene Vanderfeesten, Kris Vanhaecht, Paul Grefen, Liliane Pintelon, Jan Mendling, Geofridus G. Van Merode, Hajo A. Reijers B. Vermeulen, J.A. La Poutré, A.G. de Kok Hui Yan, Pieter Van Gorp, Uzay Kaymak, Xudong Lu, Richard Vdovjak, Hendriks H.M. Korsten, Huilong Duan J. Theresia van Essen, Johann L. Hurink, Stefan Nickel, Melanie Reuter Stefano Fazi, Tom Van Woensel, Jan C. Fransoo Yann Bouchery, Asma Ghaffari, Zied Jemai, Jan Fransoo Martijn Mes, Marco Schutten, Arturo Pérez Rivera N.J. Borgman, M.R.K. Mes, I.M.H. Vliegen, E.W. Hans S. Behfard, M.C. van der Heijden, A. Al Hanbali, W.H.M. Zijm Emrah Demir, Tolga Bektas, Gilbert Laporte
427 426 425 424 423 422 421 420 419 418 417 416
2013 2013 2013 2013 2013 2013 2013 2013 2013 2013 2013 2013
Typology of Repair Shops for Maintenance Spare Parts A value network development model and Implications for innovation and production network management Single Vehicle Routing with Stochastic Demands: Approximate Dynamic Programming Influence of Spillback Effect on Dynamic Shortest Path Problems with Travel-Time-Dependent Network Disruptions Dynamic Shortest Path Problem with Travel-Time-Dependent Stochastic Disruptions: Hybrid Approximate Dynamic Programming Algorithms with a Clustering Approach System-oriented inventory models for spare parts Lost Sales Inventory Models with Batch Ordering And Handling Costs Response speed and the bullwhip Anticipatory Routing of Police Helicopters Supply Chain Finance: research challenges ahead Improving the Performance of Sorter Systems By Scheduling Inbound Containers Regional logistics land allocation policies: Stimulating spatial concentration of logistics firms
M.A. Driessen, V.C.S. Wiers, G.J. van Houtum, W.D. Rustenburg B. Vermeulen, A.G. de Kok C. Zhang, N.P. Dellaert, L. Zhao, T. Van Woensel, D. Sever Derya Sever, Nico Dellaert, Tom Van Woensel, Ton de Kok Derya Sever, Lei Zhao, Nico Dellaert, Tom Van Woensel, Ton de Kok R.J.I. Basten, G.J. van Houtum T. Van Woensel, N. Erkip, A. Curseu, J.C. Fransoo Maximiliano Udenio, Jan C. Fransoo, Eleni Vatamidou, Nico Dellaert Rick van Urk, Martijn R.K. Mes, Erwin W. Hans Kasper van der Vliet, Matthew J. Reindorp, Jan C. Fransoo S.W.A. Haneyah, J.M.J. Schutten, K. Fikse Frank P. van den Heuvel, Peter W. de Langen, Karel H. van Donselaar, Jan C. Fransoo
415 414 413 412 411 410 409 408 407 406 405 404 403
2013 2013 2013 2013 2013 2013 2013 2013 2013 2013 2013 2013 2013
The development of measures of process harmonization BASE/X. Business Agility through Cross- Organizational Service Engineering The Time-Dependent Vehicle Routing Problem with Soft Time Windows and Stochastic Travel Times Clearing the Sky - Understanding SLA Elements in Cloud Computing Approximations for the waiting time distribution In an M/G/c priority queue To co-locate or not? Location decisions and logistics concentration areas The Time-Dependent Pollution-Routing Problem Scheduling the scheduling task: A time Management perspective on scheduling Clustering Clinical Departments for Wards to Achieve a Prespecified Blocking Probability MyPHRMachines: Personal Health Desktops in the Cloud Maximising the Value of Supply Chain Finance Reaching 50 million nanostores: retail distribution in emerging megacities A Vehicle Routing Problem with Flexible Time Windows
Heidi L. Romero, Remco M. Dijkman, Paul W.P.J. Grefen, Arjan van Weele Paul Grefen, Egon Lüftenegger, Eric van der Linden, Caren Weisleder Duygu Tas, Nico Dellaert, Tom van Woensel, Ton de Kok Marco Comuzzi, Guus Jacobs, Paul Grefen A. Al Hanbali, E.M. Alvarez, M.C. van der van der Heijden Frank P. van den Heuvel, Karel H. van Donselaar, Rob A.C.M. Broekmeulen, Jan C. Fransoo, Peter W. de Langen Anna Franceschetti, Dorothée Honhon,Tom van Woensel, Tolga Bektas, GilbertLaporte. J.A. Larco, V. Wiers, J. Fransoo J. Theresia van Essen, Mark van Houdenhoven, Johann L. Hurink Pieter Van Gorp, Marco Comuzzi Kasper van der Vliet, Matthew J. Reindorp, Jan C. Fransoo Edgar E. Blanco, Jan C. Fransoo Duygu Tas, Ola Jabali, Tom van Woensel
402 401 400 399 398 397 396 395 394 393
2012 2012 2012 2012 2012 2012 2012 2012 2012 2012
The Service Dominant Business Model: A Service Focused Conceptualization Relationship between freight accessibility and Logistics employment in US counties A Condition-Based Maintenance Policy for Multi-Component Systems with a High Maintenance Setup Cost A flexible iterative improvement heuristic to Support creation of feasible shift rosters in Self-rostering Scheduled Service Network Design with Synchronization and Transshipment Constraints For Intermodal Container Transportation Networks Destocking, the bullwhip effect, and the credit Crisis: empirical modeling of supply chain Dynamics Vehicle routing with restricted loading capacities Service differentiation through selective lateral transshipments A Generalized Simulation Model of an Integrated Emergency Post Business Process Technology and the Cloud: Defining a Business Process Cloud Platform
Egon Lüftenegger, Marco Comuzzi, Paul Grefen, Caren Weisleder Frank P. van den Heuvel, Liliana Rivera,Karel H. van Donselaar, Ad de Jong,Yossi Sheffi, Peter W. de Langen, Jan C.Fransoo Qiushi Zhu, Hao Peng, Geert-Jan van Houtum E. van der Veen, J.L. Hurink, J.M.J. Schutten, S.T. Uijland K. Sharypova, T.G. Crainic, T. van Woensel, J.C. Fransoo Maximiliano Udenio, Jan C. Fransoo, Robert Peels J. Gromicho, J.J. van Hoorn, A.L. Kok J.M.J. Schutten E.M. Alvarez, M.C. van der Heijden, I.M.H. Vliegen, W.H.M. Zijm Martijn Mes, Manon Bruens Vasil Stoitsev, Paul Grefen
392 391 390 389 388 387 386 385 384 383 382
2012 2012 2012 2012 2012 2012 2012 2012 2012 2012 2012
Vehicle Routing with Soft Time Windows and Stochastic Travel Times: A Column Generation And Branch-and-Price Solution Approach Improve OR-Schedule to Reduce Number of Required Beds How does development lead time affect performance over the ramp-up lifecycle? Evidence from the consumer electronics industry The Impact of Product Complexity on Ramp- Up Performance Co-location synergies: specialized versus diverse logistics concentration areas Proximity matters: Synergies through co-location of logistics establishments Spatial concentration and location dynamics in logistics:the case of a Dutch province FNet: An Index for Advanced Business Process Querying Defining Various Pathway Terms The Service Dominant Strategy Canvas: Defining and Visualizing a Service Dominant Strategy through the Traditional Strategic Lens
D. Tas, M. Gendreau, N. Dellaert, T. van Woensel, A.G. de Kok J.T. v. Essen, J.M. Bosch, E.W. Hans, M. v. Houdenhoven, J.L. Hurink Andres Pufall, Jan C. Fransoo, Ad de Jong Andreas Pufall, Jan C. Fransoo, Ad de Jong, Ton de Kok Frank P.v.d. Heuvel, Peter W.de Langen, Karel H. v. Donselaar, Jan C. Fransoo Frank P.v.d. Heuvel, Peter W.de Langen, Karel H. v.Donselaar, Jan C. Fransoo Frank P. v.d.Heuvel, Peter W.de Langen, Karel H.v. Donselaar, Jan C. Fransoo Zhiqiang Yan, Remco Dijkman, Paul Grefen W.R. Dalinghaus, P.M.E. Van Gorp Egon Lüftenegger, Paul Grefen, Caren Weisleder Stefano Fazi, Tom van Woensel, Jan C. Fransoo
381 380 379 378 377 375 374 373 372 371 370 369
2012 2012 2012 2012 2012 2012 2012 2012 2012 2012 2012 2011
A Stochastic Variable Size Bin Packing Problem With Time Constraints Coordination and Analysis of Barge Container Hinterland Networks Proximity matters: Synergies through co-location of logistics establishments A literature review in process harmonization: a conceptual framework A Generic Material Flow Control Model for Two Different Industries Improving the performance of sorter systems by scheduling inbound containers Strategies for dynamic appointment making by container terminals MyPHRMachines: Lifelong Personal Health Records in the Cloud Service differentiation in spare parts supply through dedicated stocks Spare parts inventory pooling: how to share the benefits Condition based spare parts supply Using Simulation to Assess the Opportunities of Dynamic Waste Collection Aggregate overhaul and supply chain planning for
K. Sharypova, T. van Woensel, J.C. Fransoo Frank P. van den Heuvel, Peter W. de Langen, Karel H. van Donselaar, Jan C. Fransoo Heidi Romero, Remco Dijkman, Paul Grefen, Arjan van Weele S.W.A. Haneya, J.M.J. Schutten, P.C. Schuur, W.H.M. Zijm H.G.H. Tiemessen, M. Fleischmann, G.J. van Houtum, J.A.E.E. van Nunen, E. Pratsini Albert Douma, Martijn Mes Pieter van Gorp, Marco Comuzzi E.M. Alvarez, M.C. van der Heijden, W.H.M. Zijm Frank Karsten, Rob Basten X.Lin, R.J.I. Basten, A.A. Kranenburg, G.J. van Houtum Martijn Mes J. Arts, S.D. Flapper, K. Vernooij J.T. van Essen, J.L. Hurink, W. Hartholt,
368 367 366 365 364 363 362 361 360 359 358 357 356 355
2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011
rotables Operating Room Rescheduling Switching Transport Modes to Meet Voluntary Carbon Emission Targets On two-echelon inventory systems with Poisson demand and lost sales Minimizing the Waiting Time for Emergency Surgery Vehicle Routing Problem with Stochastic Travel Times Including Soft Time Windows and Service Costs A New Approximate Evaluation Method for Two-Echelon Inventory Systems with Emergency Shipments Approximating Multi-Objective Time-Dependent Optimization Problems Branch and Cut and Price for the Time Dependent Vehicle Routing Problem with Time Window Analysis of an Assemble-to-Order System with Different Review Periods Interval Availability Analysis of a Two-Echelon, Multi-Item System Carbon-Optimal and Carbon-Neutral Supply Chains Generic Planning and Control of Automated Material Handling Systems: Practical Requirements Versus Existing Theory Last time buy decisions for products sold under warranty
B.J. van den Akker Kristel M.R. Hoen, Tarkan Tan, Jan C. Fransoo, Geert-Jan van Houtum Elisa Alvarez, Matthieu van der Heijden J.T. van Essen, E.W. Hans, J.L. Hurink, A. Oversberg Duygu Tas, Nico Dellaert, Tom van Woensel, Ton de Kok Erhun Özkan, Geert-Jan van Houtum, Yasemin Serin Said Dabia, El-Ghazali Talbi, Tom Van Woensel, Ton de Kok Said Dabia, Stefan Röpke, Tom Van Woensel, Ton de Kok A.G. Karaarslan, G.P. Kiesmüller, A.G. de Kok Ahmad Al Hanbali, Matthieu van der Heijden Felipe Caro, Charles J. Corbett, Tarkan Tan, Rob Zuidwijk Sameh Haneyah, Henk Zijm, Marco Schutten, Peter Schuur M. van der Heijden, B. Iskandar Frank P. van den Heuvel, Peter W. de Langen, Karel H. van Donselaar, Jan C. Fransoo
354 353 352 351 350 349 348 347 346 345 344 343 342 341 339
2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2010 2010 2010
Spatial concentration and location dynamics in logistics: the case of a Dutch provence Identification of Employment Concentration Areas BOMN 2.0 Execution Semantics Formalized as Graph Rewrite Rules: extended version Resource pooling and cost allocation among independent service providers A Framework for Business Innovation Directions The Road to a Business Process Architecture: An Overview of Approaches and their Use Effect of carbon emission regulations on transport mode selection under stochastic demand An improved MIP-based combinatorial approach for a multi-skill workforce scheduling problem An approximate approach for the joint problem of level of repair analysis and spare parts stocking Joint optimization of level of repair analysis and spare parts stocks Inventory control with manufacturing lead time flexibility Analysis of resource pooling games via a new extenstion of the Erlang loss function Vehicle refueling with limited resources Optimal Inventory Policies with Non-stationary Supply Disruptions and Advance Supply Information Redundancy Optimization for Critical Components in High-Availability Capital Goods
Frank P. van den Heuvel, Peter W. de Langen, Karel H. van Donselaar, Jan C. Fransoo Pieter van Gorp, Remco Dijkman Frank Karsten, Marco Slikker, Geert-Jan van Houtum E. Lüftenegger, S. Angelov, P. Grefen Remco Dijkman, Irene Vanderfeesten, Hajo A. Reijers K.M.R. Hoen, T. Tan, J.C. Fransoo G.J. van Houtum Murat Firat, Cor Hurkens R.J.I. Basten, M.C. van der Heijden, J.M.J. Schutten R.J.I. Basten, M.C. van der Heijden, J.M.J. Schutten Ton G. de Kok Frank Karsten, Marco Slikker, Geert-Jan van Houtum Murat Firat, C.A.J. Hurkens, Gerhard J. Woeginger Bilge Atasoy, Refik Güllü, TarkanTan Kurtulus Baris Öner, Alan Scheller-Wolf Geert-Jan van Houtum Joachim Arts, Gudrun Kiesmüller
338 335 334 333 332 331 330 329 328 327 326 325 324
2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010 2010
Analysis of a two-echelon inventory system with two supply modes Analysis of the dial-a-ride problem of Hunsaker and Savelsbergh Attaining stability in multi-skill workforce scheduling Flexible Heuristics Miner (FHM) An exact approach for relating recovering surgical patient workload to the master surgical schedule Efficiency evaluation for pooling resources in health care The Effect of Workload Constraints in Mathematical Programming Models for Production Planning Using pipeline information in a multi-echelon spare parts inventory system Reducing costs of repairable spare parts supply systems via dynamic scheduling Identification of Employment Concentration and Specialization Areas: Theory and Application A combinatorial approach to multi-skill workforce scheduling Stability in multi-skill workforce scheduling Maintenance spare parts planning and control: A framework for control and agenda for future research
Murat Firat, Gerhard J. Woeginger Murat Firat, Cor Hurkens A.J.M.M. Weijters, J.T.S. Ribeiro P.T. Vanberkel, R.J. Boucherie, E.W. Hans, J.L. Hurink, W.A.M. van Lent, W.H. van Harten Peter T. Vanberkel, Richard J. Boucherie, Erwin W. Hans, Johann L. Hurink, Nelly Litvak M.M. Jansen, A.G. de Kok, I.J.B.F. Adan Christian Howard, Ingrid Reijnen, Johan Marklund, Tarkan Tan H.G.H. Tiemessen, G.J. van Houtum F.P. van den Heuvel, P.W. de Langen, K.H. van Donselaar, J.C. Fransoo Murat Firat, Cor Hurkens Murat Firat, Cor Hurkens, Alexandre Laugier M.A. Driessen, J.J. Arts, G.J. v. Houtum, W.D. Rustenburg, B. Huisman R.J.I. Basten, G.J. van Houtum
323 322 321 320 319 318 317 316 315 314 313
2010 2010 2010 2010 2010 2010 2010 2010 2010 2010
Near-optimal heuristics to set base stock levels in a two-echelon distribution network Inventory reduction in spare part networks by selective throughput time reduction The selective use of emergency shipments for service-contract differentiation Heuristics for Multi-Item Two-Echelon Spare Parts Inventory Control Problem with Batch Ordering in the Central Warehouse Preventing or escaping the suppression mechanism: intervention conditions Hospital admission planning to optimize major resources utilization under uncertainty Minimal Protocol Adaptors for Interacting Services Teaching Retail Operations in Business and Engineering Schools Design for Availability: Creating Value for Manufacturers and Customers Transforming Process Models: executable rewrite rules versus a formalized Java program Getting trapped in the suppression of exploration: A simulation model A Dynamic Programming Approach to Multi-Objective Time-Dependent Capacitated Single Vehicle Routing Problems with Time Windows
M.C. van der Heijden, E.M. Alvarez, J.M.J. Schutten E.M. Alvarez, M.C. van der Heijden, W.H. Zijm B. Walrave, K. v. Oorschot, A.G.L. Romme Nico Dellaert, Jully Jeunet. R. Seguel, R. Eshuis, P. Grefen. Tom Van Woensel, Marshall L. Fisher, Jan C. Fransoo. Lydie P.M. Smets, Geert-Jan van Houtum, Fred Langerak. Pieter van Gorp, Rik Eshuis. Bob Walrave, Kim E. van Oorschot, A. Georges L. Romme S. Dabia, T. van Woensel, A.G. de Kok
312 2010 Tales of a So(u)rcerer: Optimal Sourcing Decisions Under Alternative Capacitated Suppliers and General Cost Structures
Osman Alp, Tarkan Tan
311 2010 In-store replenishment procedures for perishable inventory in a retail environment with handling costs and storage constraints
R.A.C.M. Broekmeulen, C.H.M. Bakx
310 2010 The state of the art of innovation-driven business models in the financial services industry
E. Lüftenegger, S. Angelov, E. van der Linden, P. Grefen
309 2010 Design of Complex Architectures Using a Three Dimension Approach: the CrossWork Case R. Seguel, P. Grefen, R. Eshuis
308 2010 Effect of carbon emission regulations on transport mode selection in supply chains
K.M.R. Hoen, T. Tan, J.C. Fransoo, G.J. van Houtum
307 2010 Interaction between intelligent agent strategies for real-time transportation planning
Martijn Mes, Matthieu van der Heijden, Peter Schuur
306 2010 Internal Slackening Scoring Methods Marco Slikker, Peter Borm, René van den Brink
305 2010 Vehicle Routing with Traffic Congestion and Drivers' Driving and Working Rules
A.L. Kok, E.W. Hans, J.M.J. Schutten, W.H.M. Zijm
304 2010 Practical extensions to the level of repair analysis R.J.I. Basten, M.C. van der Heijden, J.M.J. Schutten
303 2010 Ocean Container Transport: An Underestimated and Critical Link in Global Supply Chain Performance
Jan C. Fransoo, Chung-Yee Lee
302 2010 Capacity reservation and utilization for a manufacturer with uncertain capacity and demand Y. Boulaksil; J.C. Fransoo; T. Tan
300 2009 Spare parts inventory pooling games F.J.P. Karsten; M. Slikker; G.J. van Houtum
299 2009 Capacity flexibility allocation in an outsourced supply chain with reservation Y. Boulaksil, M. Grunow, J.C. Fransoo
298
2010
An optimal approach for the joint problem of level of repair analysis and spare parts stocking
R.J.I. Basten, M.C. van der Heijden, J.M.J. Schutten
297 2009 Responding to the Lehman Wave: Sales Forecasting and Supply Management during the Credit Crisis
Robert Peels, Maximiliano Udenio, Jan C. Fransoo, Marcel Wolfs, Tom Hendrikx
296 2009 An exact approach for relating recovering surgical patient workload to the master surgical schedule
Peter T. Vanberkel, Richard J. Boucherie, Erwin W. Hans, Johann L. Hurink, Wineke A.M. van Lent, Wim H. van Harten
295
2009
An iterative method for the simultaneous optimization of repair decisions and spare parts stocks
R.J.I. Basten, M.C. van der Heijden, J.M.J. Schutten
294 2009 Fujaba hits the Wall(-e) Pieter van Gorp, Ruben Jubeh, Bernhard Grusie, Anne Keller
293 2009 Implementation of a Healthcare Process in Four Different Workflow Systems
R.S. Mans, W.M.P. van der Aalst, N.C. Russell, P.J.M. Bakker
292 2009 Business Process Model Repositories - Framework and Survey
Zhiqiang Yan, Remco Dijkman, Paul Grefen
291 2009 Efficient Optimization of the Dual-Index Policy Using Markov Chains
Joachim Arts, Marcel van Vuuren, Gudrun Kiesmuller
290 2009 Hierarchical Knowledge-Gradient for Sequential Sampling
Martijn R.K. Mes; Warren B. Powell; Peter I. Frazier
289 2009 Analyzing combined vehicle routing and break scheduling from a distributed decision making perspective
C.M. Meyer; A.L. Kok; H. Kopfer; J.M.J. Schutten
288 2009 Anticipation of lead time performance in Supply Chain Operations Planning
Michiel Jansen; Ton G. de Kok; Jan C. Fransoo
287 2009 Inventory Models with Lateral Transshipments: A Review
Colin Paterson; Gudrun Kiesmuller; Ruud Teunter; Kevin Glazebrook
286 2009 Efficiency evaluation for pooling resources in health care
P.T. Vanberkel; R.J. Boucherie; E.W. Hans; J.L. Hurink; N. Litvak
285 2009 A Survey of Health Care Models that Encompass Multiple Departments
P.T. Vanberkel; R.J. Boucherie; E.W. Hans; J.L. Hurink; N. Litvak
284 2009 Supporting Process Control in Business Collaborations
S. Angelov; K. Vidyasankar; J. Vonk; P. Grefen
283 2009 Inventory Control with Partial Batch Ordering O. Alp; W.T. Huh; T. Tan
282 2009 Translating Safe Petri Nets to Statecharts in a Structure-Preserving Way R. Eshuis
281 2009 The link between product data model and process model J.J.C.L. Vogelaar; H.A. Reijers
280 2009 Inventory planning for spare parts networks with delivery time requirements I.C. Reijnen; T. Tan; G.J. van Houtum
279 2009 Co-Evolution of Demand and Supply under Competition B. Vermeulen; A.G. de Kok
278 277
2010 2009
Toward Meso-level Product-Market Network Indices for Strategic Product Selection and (Re)Design Guidelines over the Product Life-Cycle An Efficient Method to Construct Minimal Protocol Adaptors
B. Vermeulen, A.G. de Kok R. Seguel, R. Eshuis, P. Grefen
276 2009 Coordinating Supply Chains: a Bilevel Programming Approach Ton G. de Kok, Gabriella Muratore
275 2009 Inventory redistribution for fashion products under demand parameter update G.P. Kiesmuller, S. Minner
274 2009 Comparing Markov chains: Combining aggregation and precedence relations applied to sets of states
A. Busic, I.M.H. Vliegen, A. Scheller-Wolf
273 2009 Separate tools or tool kits: an exploratory study of engineers' preferences
I.M.H. Vliegen, P.A.M. Kleingeld, G.J. van Houtum
272
2009
An Exact Solution Procedure for Multi-Item Two-Echelon Spare Parts Inventory Control Problem with Batch Ordering
Engin Topan, Z. Pelin Bayindir, Tarkan Tan
271 2009 Distributed Decision Making in Combined Vehicle Routing and Break Scheduling
C.M. Meyer, H. Kopfer, A.L. Kok, M. Schutten
270 2009 Dynamic Programming Algorithm for the Vehicle Routing Problem with Time Windows and EC Social Legislation
A.L. Kok, C.M. Meyer, H. Kopfer, J.M.J. Schutten
269 2009 Similarity of Business Process Models: Metics and Evaluation
Remco Dijkman, Marlon Dumas, Boudewijn van Dongen, Reina Kaarik, Jan Mendling
267 2009 Vehicle routing under time-dependent travel times: the impact of congestion avoidance A.L. Kok, E.W. Hans, J.M.J. Schutten
266 2009 Restricted dynamic programming: a flexible framework for solving realistic VRPs
J. Gromicho; J.J. van Hoorn; A.L. Kok; J.M.J. Schutten;
Working Papers published before 2009 see: http://beta.ieis.tue.nl