Scientia et Technica Año XXVIII, Vol. 28, No. 02, abril-junio de 2023. Universidad Tecnológica de Pereira. ISSN 0122-1701 y ISSN: 2344-7214
86
Computational simulation and voracious
algorithm to calculate the least cost route
Simulación computacional y algoritmo voraz para calcular la ruta de mínimo costo
L. A. Lasso-Cardona
DOI: https://doi.org/10.22517/23447214.24737
Scientific and technological research paper
Abstract—Transportation logistics aims to deliver products at the
right time and place at the lowest possible cost. Within this activity
are the problems of optimizing the routing of cargo vehicles that
must travel a minimum cost route for the delivery of goods. The
impact of the use of information technologies in the context of the
supply chain can be measured basically in integration and in the
benefits it brings. The objective of this research is to use
computational simulation to evaluate restrictions of a messaging
system and to show the shortest route found for three package
delivery lines, optimizing the time and allocation of packages for
each line. The project was developed under an empirical and
exploratory methodology, where four phases followed. An
algorithm was designed that used a voracious and stochastic
approach, with an objective function that was evaluated in each of
the n-simulations, to find the shortest route on the three delivery
lines. To test the effectiveness of the algorithm, two test scenarios
were carried out. In both scenarios, it was possible to show that the
greater the number of simulations, the distance found was shorter,
which was the objective of the investigation.
Index Terms computational simulation; route optimization;
stochastic process; supply chain; voracious algorithm.
Resumen La logística de transporte tiene como objetivo la
entrega de productos en el momento y lugar correcto al menor
costo posible. Dentro de esta actividad se encuentran los
problemas de optimización del enrutamiento de vehículos de carga
que deben recorrer una ruta de mínimo costo para la entrega de
mercancías. El impacto que tiene el uso de tecnologías de la
información en el contexto de cadena de suministro se puede medir
básicamente en la integración y en los beneficios que aporta. La
presente investigación tiene por objetivo utilizar la simulación
computacional para evaluar restricciones de un sistema de
mensajería y arrojar la ruta de menor distancia encontrada para
tres líneas de entrega de paquetes, optimizando el tiempo y
asignación de paquetes para cada línea. El proyecto se desarrolló
bajo una metodología empírica y exploratoria, donde siguieron
cuatro fases. Se diseñó un algoritmo que utilizo un enfoque voraz
y estocástico, con una función objetivo que se evaluó en cada una
This manuscript was submitted on February 06, 2023, accepted on April 09,
2023 and published on June 30, 2023.
L. A. Lasso Cardona. Ingeniero de Sistemas, Universidad del Valle,
Colombia. M. Sc. en Gestión de Tecnología Educativa, Universidad de
Santander. (e-mail: luis.lasso@correounivalle.edu.co).
de las n-simulaciones, para encontrar la ruta de menor distancia
en las tres líneas de entrega. Para comprobar la efectividad del
algoritmo se llevaron a cabo dos escenarios de prueba. En los dos
escenarios, se logró evidenciar que, a mayor número de
simulaciones, la distancia encontrada era menor, lo cual era el
objetivo de la investigación.
Palabras claves— algoritmo voraz; cadena de suministro;
optimización de ruta; proceso estocástico; simulación
computacional
I.
INTRODUCTION
OWADAYS, globalization and customer satisfaction are
two fundamental principles that every organization must
have as objectives to fulfill. To achieve this, quality
management throughout the supply chain and the impact it has
to achieve sustainable practices [1], and achieve a competitive
advantage over the market, go hand in hand with the use of
information technologies, as a differentiating element, allowing
to optimize distribution times, transport costs [2], and
improving the quality of delivery service of products, which has
a correlation with customer loyalty [3]. In this order of ideas,
Supply
Chain
Management
(SCM)
aims
to
facilitate
communication between the parts of the business, improve the
delivery of products to the final customer [4] and achieve a
more environmentally friendly organization, on the fact of that
freight vehicle fleets contributed 58% of CO2 emissions for
2017, and that the transport industry in general contributes 83%
of all CO2 emissions today. For this reason, it is clear that
alternatives must also be sought to improve and optimize the
movements of this type of vehicle [5].
For a couple of decades SCM, and its main areas such as
transportation management, resilience and reverse logistics,
have gained interest in various sectors of society, education and
the economy, which seek through their study to identify the
N
87
Scientia et Technica Año XXVIII, Vol. 28, No. 02, abril-junio de 2023. Universidad Tecnológica de Pereira
variables that affect their operation, and thus improve their
methods, allowing organizations to meet the challenges of
modern and dynamic economies, and the increasingly
demanding requirements of customers [6], by adopting
emerging technologies such as Internet of the things (IoT),
Artificial Intelligence (AI) and Big Data that improve the
decision-making process [7-11], and the evaluation of threats
and weaknesses that put the system at risk [12].
Transportation logistics is one of SCM's activities, and it
aims to deliver products at the right time and place at the lowest
possible cost. Generally, transport logistics is one of the factors
that most influences the economic growth of organizations and
a country, allowing better results to be obtained in the global
market [13]. Within this activity are the problems of
optimization of the routing of cargo vehicles (VRP) that must
travel a minimum cost route for the delivery of goods [14], the
programming of vehicles with time windows (VRPTW) that
considers the daily scheduling [15], and trained vehicle routing
(CVRP), which helps identify routes that start and end at a
distribution center [16].
The impact of the use of Information Technology (IT) in the
context of the supply chain can be measured basically in two
aspects: in the integration and in the benefits that it brings [17].
Several investigations show the advantages of using IT to
propose new SCM evaluation models [18], calculate operating
profits [19] and improve management in merchandise
distribution processes in the metal-mechanic sector [20], leather
goods [21], food [22] and textile [23]. Likewise, computer
design and programming techniques have been used to: a)
optimize routes using genetic algorithms [24], b) compensate
for the shortage of delivery vehicles using web-based systems
[25], c) solve the CVRP problem with the help of metaheuristics
[26], d) the planning of distribution routes using a mixed integer
programming model [27] and, e) the optimization of routes of a
multimodal container transport system using dynamic
programming [28].
The objective of this research is to use computational
simulation to evaluate restrictions of a messaging system with
a VRP approach and to show the shortest route found for three
package delivery lines in an area of 150 km
2
, optimizing time
and package allocation for each line. The types of packages are
divided into: documents and boxes. A line that we call “Moto”
is oriented to the delivery of documents. The other two, called
“Box truck 1” and “Box truck 2”, are aimed at delivering boxes.
An algorithm was designed that used a voracious and stochastic
approach, with an objective function that was evaluated in each
of the n-simulations, to find the shortest route on the three
delivery lines. Such algorithm was implemented in an
application in the Java programming language with GUI, which
allowed to randomly generate the data related to the packages
(address and type of package) and enter the entries for the
restrictions, display the map and the order of delivery to be
followed by the drivers of the three lines. To test the
effectiveness of the algorithm, two test scenarios were carried
out. The first scenario consisted of simulating the delivery of 30
packages. To find the least cost route in distance, a test bench
was performed that consisted of running the simulation 1,000,
10,000, 50,000, 500,000, 1,000,000 and 10,000,000 times. The
second scenario consisted of simulating the delivery of 200
packages. In this case, a test bench was carried out with the
same number of simulations of the first scenario, but executing
each simulation size 5 times, and taking the distance and the
execution time of each one, to finally calculate their average. In
both scenarios, it was possible to show that the greater the
number of simulations, the distance found was shorter, which
was the objective of the investigation.
A.
Computational Simulation
Freight transport companies often use computer simulation
to assess risks associated with traffic, packing time, distance to
travel, routes and delivery times, among other factors, and thus
anticipate situations that affect their normal functioning [29]. In
this sense, computational simulation is one of the areas of
Computer Science that provides techniques that help in
situations where it is very difficult to know and evaluate the
behavior of unpredictable systems, such as weather or
environmental conditions [30]. One of the methods applied in
simulation is the voracious algorithms that are commonly used
in optimization problems, and that evaluating an objective
function make short-range decisions based on immediate
information. This allows you to make the optimal choice locally
at each stage, with the expectation of finding a global optimum
[31].
The computational simulation, in combination with
stochastic processes allow to represent the behavior of complex
systems, which makes them a strong technique in solving
problems in sectors such as economics, astronomy, physics, or
industry where deep analysis is required. of all the components
involved in its operation [32]. For example, a simulation would
allow: a) finding the optimal route for a worker to follow in the
order preparation process [33], b) reduce delivery time and total
CO2 emissions from trucks [34-35], c) provide solutions to
intermodal transport that minimize costs and reduce the
negative impact of transport activities, such as costs, time and
environmental [36-37], d) to support Intelligent Transport
Systems (ITS) applications in public transport by combining
optimization models with data from ITS [38] and, e) identify
the distribution routes of electric vehicles (EV) considering the
distribution centers and charging facilities [39].
II.
METHODOLOGY
A.
Problem formulation
Given in a list of addresses that represent the physical location
for package delivery, find the least cost route in distance, using
computational simulation with a stochastic approach to select
each address and calculate the total distance, and store it as a
probable answer. At the end of the simulation show the least
cost route.
B.
Description of the methodology
The project was developed under an empirical and
exploratory methodology. Basically the following phases were
followed:
1.
Research phase: to build the theoretical framework,
information was collected from scientific articles published in
88
Scientia et Technica Año XXVIII, Vol. 28, No. 02, abril-junio de 2023. Universidad Tecnológica de Pereira
the last five years, under the search terms "supply chain", "route
optimization", "computational simulation", "transport
logistics”, “Voracious algorithm” and “stochastic process”, in
databases such as Scopus, Web of Science, IEEE Explore,
Elsevier and Computer Source, and scientific journals, which
allowed to contextualize and delimit the study.
2.
Design phase: based on the collected data, we proceeded to
design the algorithm that ran the n-simulations. Likewise, the
objective function was constructed, which evaluated the
distance of the route generated randomly. To verify the
reliability of the random number generator, the Chi-square test
was performed.
3.
Implementation phase: The Java programming language was
used to implement the algorithm, and the objective function.
The application had a GUI that allowed entering the input
values to randomly generate the data related to the packages
(address and type of package) and enter the entries for the
restrictions, display the map and the delivery order to be
followed the drivers of the three lines.
4.
Test phase: Based on the random data generated in the
previous phase, two test scenarios were carried out. The first
scenario consisted of simulating the delivery of 30 packages,
divided into 9 documents and 21 boxes. The second scenario
consisted of simulating the delivery of 200 packages, divided
into 97 documents and 103 boxes.
C.
Chi-square test
In order to validate the random number generator, the Chi-
square test was applied, which checks the uniformity of the
numbers, measuring the degree of fit between the distribution
of a sample of random numbers and the theoretical uniform
distribution. This is based on the null hypothesis that there is no
difference between the two distributions.
To develop the test, 100 random numbers were generated in
Java, and the statistic was compared with the Chi-square with
alpha of 0.05 and 9 degrees of freedom. Table I shows the
distribution of the numbers and the results.
TABLE I
CHI-SQUARE TEST
Intervals
Oi
Expected frequency
Ei
0.0
0.1
10
10
0.1
0.2
9
10
0.2
0.3
7
10
0.3
0.4
8
10
0.4
0.5
12
10
0.5
0.6
7
10
0.6
0.7
12
10
0.7
0.8
10
10
0.8
0.9
9
10
0.9
1
16
10
Summation:
100
100
Chi-inv:
Since the sum of the statistics is less than Chi-inv, it is not
rejected that the sample numbers follow a uniform distribution,
therefore, the number generator is accepted.
D.
Proposed system
An application was developed in the Java programming
language that used computational simulation to evaluate
restrictions of a messaging system with a VRP approach, and to
show the shortest route found for three package delivery lines
in an area of “150 km
2
”, optimizing time and packet allocation
for each line. The types of packages are divided into: documents
and boxes.
A line that we call “Moto” is oriented to the delivery of
documents. The other two, called “Box truck 1” and “Box truck
2”, are aimed at delivering boxes. An algorithm was designed
that used a voracious and stochastic approach, and an objective
function that was evaluated in each of the n-simulations to find
the shortest route on the three delivery lines. Such an algorithm
was implemented in an application in the Java programming
language, which allowed the data related to the packages
(address and type of package) to be generated randomly and the
entries for the restrictions were entered, the map and the order
of delivery were displayed. the drivers of all three lines must
follow.
The hypothesis to determine the effectiveness of the
proposed system is based on the fact that the greater the number
of simulations, the lower the cost of the route found.
Inputs:
C = number of simulations.
P = number of packages.
D1C y D2C = physical location (address) of
distribution center.
D1T1 y D2T1 = range of addresses in which the “Box
truck 1” line operates.
D1T2 y D2T2 = range of addresses in which the “Box
truck 2” line operates.
The application allows:
Enter the number of simulations to run.
Generate n-quantity of packets randomly and related
data (address and type of packet). The address has the
format “Calle / Carrera ### - ### ###”.
Read the data of the packages to be delivered from a
text file.
Enter the physical location (address) of the distribution
center.
Enter restriction in terms of address ranges for lines
“Box truck 1” and “Box truck 2”.
Determine the restriction zone for the lines “Box truck
1” and “Box truck 2”. If the restriction = 1, the western
zone restriction applies for “Box truck 1” and the
eastern zone restriction for “Box truck 2”. If the
restriction = 2, the restriction north zone for “Box
truck 1” and south zone for “Box truck 2” applies.
In each valid simulation, the total distance of the three
lines is calculated and stored as a probable answer.
Restrictions:
The delivery area is set at “150 km
2
”.
The three delivery lines depart from the distribution
center.
89
Scientia et Technica Año XXVIII, Vol. 28, No. 02, abril-junio de 2023. Universidad Tecnológica de Pereira
Randomly generated package delivery addresses range
from 1 to 150.
The package type is an integer random number
between 0 and 1. 0 = document, 1 = box.
If the package type = 0, it is assigned to the “Moto”
delivery line. Otherwise it is assigned to the lines “Box
truck”.
“Box truck 1” and “Box truck 2” lines must be
balanced in package allocation.
The “Box truck 1” and “Box truck 2” lines can only be
assigned packages according to the restriction of
address ranges.
The “Moto” line has no restrictions regarding address
ranges.
Outputs:
Cost for each delivery line and total cost.
Map with the best calculated delivery route of the three
distribution lines.
List of addresses that the drivers of each line must
follow.
Simulation and optimization algorithm:
Function simulate_route(C, can_ boxes , x_paq_boxes[],
y_paq_boxes[], z_paq_boxes[], text_paq_boxes[]){
random = 0
array x[can_ boxes], y[can_ boxes], z[can_ boxes],
text[can_ boxes]
selected[can_ boxes]
for (i = 0; i < C; i++) {
fill(selected, false)
for (j = 0; j < can_ boxes;) {
random = between (0, can_ boxes)
if (!selected[random]) {
selected [random] = true
x[j] = xPaqCajaT1[random]
y[j] = yPaqCajaT1[random]
z[j] = zPaqCajaT1[random]
text[j] = textPaqCajaT1[random]
j++
}
}
call function_objective(x, y, z, text)
}
}
Objective function:
function_objective(x[], y[], z[], text[]){
x1p1 = x_center_distribution
y1p1 = y_center_distribution
x2p2 = 0
y2p2 = 0
distance = 0
distance_min = Double.MAX_VALUE
for (j = 0; j < can_ boxes; j++) {
x2p2 = x[j]
y2p2 = y[j]
distance+=calc_distance(x1p1, y1p1, x2p2, y2p2)
x1p1 = x2p2
y1p1 = y2p2
}
if(distance_min > distance){
distance_min = distance
save_route_coordinates(x, y)
}
}
III.
RESULTS
To test the effectiveness of the algorithm, two test scenarios
were carried out. The first scenario consisted of simulating the
delivery of 30 packages, divided into 9 documents and 21
boxes, where the “Box truck 1” line was assigned 10 boxes, and
“Box truck 2” 11 boxes. To find the least cost route in distance,
a test bench was performed that consisted of running the
simulation 1,000, 10,000, 50,000, 500,000, 1,000,000 and
10,000,000 times.
The second scenario consisted of simulating the delivery of
200 packages, divided into 97 documents and 103 boxes, where
“Box truck 1” was assigned 51 boxes, and “Box truck 2” 12
boxes. In this case, a test bench was carried out with the same
number of simulations of the first scenario, but executing each
simulation size 5 times, and taking the distance and the
execution time of each one, to finally calculate their average.
It should be noted that the tests were performed with the same
input data and restriction values, making the results more
reliable.
In the graph showing the best calculated route, the
distribution center and the end points of each delivery line are
represented by a circle in fuchsia, the line "Moto" in red, "Box
truck 1" in blue and "Box truck 2 "in green.
A.
First scenario
Inputs:
C = 1,000, 10,000, 50,000, 500,000, 1,000,000 and
10,000,000.
P = 30.
D1C = 75 y D2C = 5. Distribution center location.
D1T1 = 1 y D2T1 = 60. Range of addresses in which
"Box truck 1" operates.
D1T2 = 61 y D2T2 = 150. Range of addresses in which
"Box truck 1" operates.
In this case the applied zone restriction will be: west and east.
Output:
Fig. 1 shows the best route calculated when performing 10
million simulations, with a distance of "Moto" = 32,742, "Box
truck 1" = 33,941 and "Box truck 2" = 40,177. Total “106,860
km”.
90
Scientia et Technica Año XXVIII, Vol. 28, No. 02, abril-junio de 2023. Universidad Tecnológica de Pereira
D1T2 = 61 y D2T2 = 150. Range of addresses in which
"Box truck 1" operates.
Fig 1. Best route - first scenario.
As you can see, all the delivery lines leave from the
distribution center. Box truck lines only operate in the assigned
area. The western and eastern zones are also demarcated.
Table II and Fig. 2 summarize the data obtained in each
simulation and the percentage of improvement in each
simulation.
Output:
After running the test bench, the best calculated route was:
"Moto" = 588,956, "Box truck 1" = 245,764 and "Box truck 2"
= 245,314. Total “1,080,034 km”.
Table III and Fig. 3 summarize the data obtained in each
simulation and the percentage of improvement in each
simulation.
TABLE III
SUMMARY - SECOND SCENARIO
TABLE II
SUMMARY - FIRST SCENARIO
Minimum distance reached
Number of
Simulations
Box truck line
Distance
total
Percentage
improvement
Moto
line
Western
zone
East
zone
1,000
38,269
40,813
51,342
130,424
0
10,000
38,030
39,864
49,031
126,925
2.68
50,000
33,730
36,676
45,948
116,354
10.79
500,000
32,746
33,941
42,937
109,624
15.95
1,000,000
32,742
33,941
40,483
107,166
17.83
10,000,000
32,742
33,941
40,177
106,860
18.07
Fig 2. Improvement percentage between simulations - first scenario.
B.
Second scenario
Inputs:
C = 1,000, 10,000, 50,000, 500,000, 1,000,000 and
10,000,000. 5 times each simulation.
P = 200.
D1C = 75 y D2C = 5. Distribution center location.
D1T1 = 1 y D2T1 = 60. Range of addresses in which
"Box truck 1" operates.
Fig 3. Improvement percentage between simulations - second scenario.
As can be seen, the two scenarios tested the hypothesis, that
the greater the number of simulations, the better the result.
On the other hand, the distance calculated in the first
scenario, although true decreases with each simulation, the
percentage of improvement between the number of simulations
1 million and 10 million is very low, only 0.24% representing
“0.306 km”, which is very low compared to computational cost.
The same is evident for the second scenario, where tests of
greater weight and with a higher computational cost were
carried out, but where it is observed that among the same
number of simulations a better percentage of improvement was
achieved, reflected in “27,430 km” less, than vehicles must
travel.
If the results of the second scenario are taken as a basis, it
can be assumed that, in a week, a courier company with the
Number
of
Simulatio
ns
Average
time of 5
simulati
ons (sec)
Minimum distance reached
Distance
total
%
imp
.
Moto
line
Box truck line
Wester
n zone
East
zone
1,000
0.00
658,08
305,33
208,
83
1,172,26
0
10,000
0.00
618,47
281,84
270,
46
1,170,77
0.13
50,000
0.95
625,56
263,59
256,
45
1,145,61
2.27
500,000
12.05
593,91
261,25
246,
79
1,101,95
6.00
1,000,00
0
24.58
594,92
260,94
251,
59
1,107,46
5.53
10,000,0
00
249.23
588,95
245,76
245,
31
1,080,03
7.87
91
Scientia et Technica Año XXVIII, Vol. 28, No. 02, abril-junio de 2023. Universidad Tecnológica de Pereira
same characteristics as the simulation, would reduce the
displacement of its vehicle fleet by an average of “137.15 km”,
which represents less travel time and savings in fuel
consumption.
IV.
CONCLUSIONS
The results obtained showed the advantages of using
computational simulation techniques to imitate and predict the
behavior of a complex system, where there are many variables
that can affect its normal operation, but being stochastic
processes do not provide the guarantee that the results are
totally reliable. Similarly, if you want to obtain results that
approach solutions with a high degree of effectiveness, it is
necessary to involve an external variable, in this case,
computational power to achieve the best result within a scenario
with many possible solutions that seem the best, and escape
from optimal venues. In this regard, and if the results of the
second scenario are taken as evidence, and the prediction made,
it is well worth increasing the computational capacity factor,
but without neglecting the efficiency that algorithms must have,
to optimize processes within organizations, which will
ultimately be reflected in greater customer satisfaction,
operating capacity and increased profits.
On the other hand, the tests also showed that the application
of the voracious programming technique, the design of the
optimization algorithm and the evaluation method of the
objective function was successful, yielding the expected results,
in a reasonable time.
Finally, it is hoped that the results of the research will help
the academic and scientific community, and that they can be
taken as a basis for future projects that improve the results
obtained, applying techniques such as the Markov Chain, Big
data, Artificial Intelligence and Deep Learning.
REFERENCES
[1] M. Shaiq et al., "Quality Management as Driver of Vertical Integration in
Service Chain: A Study of 3rd Party Logistics Industry", Operations and
Supply Chain Management, 13(3) pp. 224 255. 2020. DOI:
10.31387/oscm0420266
[2] S. Pattnaik, and S. Pattnaik, “Relationships Between Green Supply Chain
Drivers, Triple Bottom Line Sustainability and Operational Performance:
An Empirical Investigation in The UK Manufacturing Supply Chain”,
Operations and Supply Chain Management: An International Journal,
12(4), 198-211. 2019. DOI: 10.31387/oscm0390243
[3] Uvet, H, “Importance of Logistics Service Quality in Customer
Satisfaction: An Empirical Study”, Operations and Supply Chain
Management: An International Journal, 13(1), 1-10. 2020. DOI:
10.31387/oscm0400248
[4] A. Batwa, and A. Norrman, “A Framework for Exploring Blockchain
Technology in Supply Chain Management”, Operations and Supply
Chain Management: An International Journal, 13(3), 294-306. 2020.
DOI: 10.31387/oscm0420271
[5] M. Abdul-Hak, N. Al-Holou, Y. Bazzi and M. Alamir Tamer, “Predictive
Vehicle Route Optimization in Intelligent Transportation Systems”,
International Journal on Data Science and Technology, 5(1), pp. 14-28.
2019. DOI: 10.11648/j.ijdst.20190501.13
[6] H. Yalcin, W. Shi and Z. Rahman, “A review and scientometric analysis
of supply chain management (SCM)”, Operations and Supply Chain
Management: An International Journal, 13(2), 123-133. 2020. DOI:
10.31387/oscm0410257
[7] A. Sharma and P. Khanna, “Relevance of Adopting Emerging
Technologies in Outbound Supply Chain: New Paradigm for Cement
Industry”, Operations and Supply Chain Management: An International
Journal, 13(2), 210-221. 2020. DOI: 10.31387/oscm0410263
[8] M. Akbari, "Logistics outsourcing: a structured literature review",
Benchmarking: An International Journal, 25(5), pp. 1548-1580. 2018.
DOI: 10.1108/BIJ-04-2017-0066
[9] J. Mageto and R. Luke, ‘Skills frameworks: A focus on supply chains’,
Journal of Transport and Supply Chain Management, 14(0), 2020. DOI:
10.4102/jtscm.v14i0.458
[10] J. Alieva and R. Haartman, “Digital Muda - The New Form of Waste by
Industry 4.0”, Operations and Supply Chain Management: An
International Journal, 13(3), 269-278. 2020. DOI:
10.31387/oscm0420268
[11] W. Yudhistyra, E. Risal and I. Raungratanaamporn, “Exploring Big Data
Research: A Review of Published Articles from 2010 to 2018 Related to
Logistics and Supply Chains”, Operations and Supply Chain
Management: An International Journal, 13(2), 134-149. 2020. DOI:
10.31387/oscm0410258
[12] R. Naudé and J. Badenhorst, ‘The challenges behind producing a bottle of
wine: Supply chain risks’, Journal of Transport and Supply Chain
Management, 14(0), 2020. DOI: 10.4102/jtscm.v14i0.471
[13] A. Botha and J. Badenhorst, ‘Risk management in a bulk coal export
logistic chain: A stakeholder perspective’, Journal of Transport and
Supply Chain Management 13(0), 2019. DOI: 10.4102/jtscm.v13i0.424
[14] M. Schneider, T. Gschwind and D. Vigo, “Advances in vehicle routing
and logistics optimization: exact methods”, EURO J Transp Logist, 8,
117–118. 2019. DOI: 10.1007/s13676-018-0139-6
[15] D.E. Mazzuco, et al., “A concept for simulation-based optimization in
Vehicle Routing Problems”, IFAC-PapersOnLine, 51(11), pp. 1720-
1725. 2018. DOI: 10.1016/j.ifacol.2018.08.208
[16] M. Appiah and X. Qiang, “Route Optimization in logistics distribution
based on Particle Swarm Optimization”, International Journal of
Computer Applications. 178. 23-27. 2019. DOI: 10.5120/ijca2019919179
[17] E. Huddiniah and M. ER, “Product Variety, Supply Chain Complexity
and the Needs for Information Technology: A Framework Based on
Literature Review”, Operations and Supply Chain Management: An
International Journal, 12(4), pp. 245-255. 2019. DOI:
10.31387/oscm0390247
[18] N. García, J. Moyano and J. Maqueira, “Planteamiento de un modelo de
evaluación de lean supply chain management”, Revista De Estudios
Empresariales. Segunda Época, (1). 2019. DOI: 10.17561/ree.v2019n1.8
[19] M. Asrol, M. Yani and E. Taira, “Supply Chain Fair Profit Allocation
Based on Risk and Value Added for Sugarcane Agro-industry”,
Operations and Supply Chain Management: An International Journal,
13(2), pp. 150-165. 2020. DOI: 10.31387/oscm0410259
[20] K. Salas, J. Meza, T. Obredor and N. Mercado, “Evaluación de la Cadena
de Suministro para Mejorar la Competitividad y Productividad en el
Sector Metalmecánico en Barranquilla, Colombia”, Información
tecnológica, 30(2), pp. 25-32. 2019. DOI: 10.4067/S0718-
07642019000200025
[21] F. Caballero, O. Walteros and F. Santofimio, “Desarrollo y aplicación del
Supply Chain Management en las empresas de marroquinería en Bogotá”,
Apuntes Contables. 24, 2019. pp. 107-123. DOI:
10.18601/16577175.n24.07
[22] E. Rodríguez, “La gestión de la cadena de suministro sostenible en la
industria alimenticia”, AD-minister, 33, pp. 113 – 134. DOI: 10.17230/ad-
minister.33.6
[23] D. Ruiz, J. Bonilla and C. Masaquiza, “Ajuste estratégico en la cadena de
suministros para la creación de valor de la marca Jean up”, Revista
Universidad y Sociedad, 10(1), pp. 25-32. 2018. Available at:
http://scielo.sld.cu/scielo.php?script=sci_arttext&pid=S2218-
36202018000100025&lng=es&tlng=es
[24] R. Huarote, Y. Vega, H. Aradiel and J. Valverde, “Optimización de ruta
corta usando algoritmo genético generacional”, Ciencia & Desarrollo,
(22), pp. 50-57, 2019. DOI: 10.33326/26176033.2018.22.745
[25] Ma. Hajghasema and A. Abbas, “Optimal Routing in Supply Chain
Aimed at Minimizing Vehicle Cost and Supply”, Procedia Economics
and
Finance,
36,
pp.
353-362.
2016.
DOI:
10.1016/S2212-
5671(16)30047-8
[26] J. Galarcio, et al., “Una nueva metaheurística aplicada al problema de
ruteo de vehículos capacitados (cvrp) para la distribución de productos
perecederos”, Ingeniería E Innovación, 5(1). 2017. DOI:
10.21897/23460466.1107
[27] J. Velarde et al., “Optimization of territories and transport routes for
hazardous products in a distribution network”, Journal of Industrial
92
Scientia et Technica Año XXVIII, Vol. 28, No. 02, abril-junio de 2023. Universidad Tecnológica de Pereira
Engineering and Management, 10(4), pp. 604-622, 2017. DOI:
10.3926/jiem.2107
[28] Y. Congli, “Optimization on Combination of Transport Routes and Modes
on Dynamic Programming for a Container Multimodal Transport
System”, Procedia Engineering, 137, pp. 382-390, 2016. DOI:
10.1016/j.proeng.2016.01.272
[29] C. Osorio and K. Kumar, “Simulation-Based Optimization: Achieving
Computational Efficiency Through the Use of Multiple Simulators”,
Home Transportation Science, 51(2), 2017. DOI: 10.1287/trsc.2016.0673
[30] M. Trigila, A. Gaudiani, E. Luque and M. Naiouf, “Simulación
computacional, ciencia de los datos, cómputo de alto rendimiento y
optimización aplicados a mejorar la predicción de modelos de simulación
que representan la evolución de sistemas complejos”, XX Workshop de
Investigadores en Ciencias de la Computación. 2018. Available at:
http://sedici.unlp.edu.ar/handle/10915/68228
[31] L. Lasso, D. Franco and A. Agudelo, “Voracious and Heuristic
Algorithms: A focus on the Minimum Path Problem”, INGE CUC, 16(2),
2020.
[32] I. MENES et al., “Simulación computacional de procesos de producción,
caso de estudio: Proceso de producción de queso en la planta de lácteos
FCP-ESPOCH”, Revista Publicando, 4(3), pp. 248-265, 2017. Available
at: https://revistapublicando.org/revista/index.php/crv/article/view/847
[33] M. Bučková, M. Krajčovi and M. Edl, “Computer Simulation and
Optimization of Transport Distances of Order Picking Processes”,
Procedia Engineering, 192, pp. 69-74. 2017. DOI:
10.1016/j.proeng.2017.06.012
[34] F. Silva, et al., “Simulation Optimization for Analysis of Sustainable
Logistics Systems”, Pesquisa Operacional, 37(1), pp. 145-171. 2017.
DOI: 10.1590/0101-7438.2017.037.01.0145
[35] J. Peng, “Optimizing the transportation route of fresh food in cold chain
logistics by improved genetic algorithms”, Int. J. Metrol. Qual. Eng. 10,
2019. DOI: 10.1051/ijmqe/2019013
[36] M. Hrušovský, et al., “Hybrid simulation and optimization approach for
green intermodal transportation problem with travel time uncertainty”,
Flex Serv Manuf, 30, 486–516. 2018. DOI: 10.1007/s10696-016-9267-1
[37] Ch. Shaviraa and B. Kara-Sal, “Optimization methodology of intraurban
transport routes of automobile passenger transport”, Economics, 6(72),
2018.
[38] C. Iliopoulou and K. Kepaptsoglou, “Combining ITS and optimization in
public transportation planning: state of the art and future research paths.
EUR”. Transp. Res. Rev. 11(27). 2019. DOI: 10.1186/s12544-019-0365-
5
[39] C. Ma et al., “Distribution path robust optimization of electric vehicle
with multiple distribution centers”, PLoS ONE 13(3), 2018. DOI:
10.1371/journal.pone.0193789
Luis Adrian Lasso Cardona. Systems Engineer,
Universidad del Valle, Colombia. M. Sc. Educational
Technology Management, Universidad de Santander,
Colombia. Associate Professor Faculty of Engineering,
Universidad del Valle, Colombia. Professor Faculty of
Engineering, Unidad Central del Valle del Cauca, Colombia.
ORCID: https://orcid.org/0000-0002-3354-1554