1. INTRODUCTION
One
of the most important processes in logistics is transportation, according to
ILOS (2012), in Brazil the logistics costs were 11.5% of the GNP at that year,
and 7.1 were related to transportation, what means 61.73% of total logistics
costs at 2012 in the country.
As
transportation costs represent the biggest portion of logistical costs, it
represents a great opportunity to achieve cost reduction, this can be reached
by using information technology, more specifically routing softwares, as an
allied for companies that need to
improve their competitively, revising procedures and meliorating resources
usage.
Nevertheless,
investments on information technology by corporations are not always made in an
appropriate and consistent way, that results in non-desirable results.
According to Siqueira, Souza and Viana (2013), small and medium size companies
may not achieve desired results by using information technology due to they do
not use it companywide.
By
analyzing a company processes it is possible to find improvement opportunities,
even if it is a heavy user of information technology. This paper aims to show
the possibility to meliorate the routing process in a logistics operator
located in the city of Suzano, by using a routing software to compare with
manual routing process currently in use.
Based
on the assumptions above, it is possible to propose the research problem as:
what is the impact of using information technologies in routing transportation
activities?
This
way, the main objective of the study is to analyze the delivery vehicles
idleness by using a routing software, comparing the current manual operation with
a simulation created by the software.
The
hypothesis for the research is that the software can perform more efficiently
the creation of routes, that will make possible to have a better utilization of
vehicles, allowing to visualize the idleness and overloads, as well as optimize
the use of the contracted fleet.
This
paper has other five parts, the second part contains the theoretical reference,
the third part is the research method, the fourth is the case study
development, the fifth is the results analysis and the sixth are the final
comments.
2. THEORETICAL REFERENCE
This
chapter presents the theoretical bases to sustain the research, treating
subjects as transportation and distribution, routing and information
technology.
2.1.
Transportation
and Distribution
According
to Emoto and Lima (2007), one of the most challenging tasks to transportation
professionals is the planning of physical distribution, the reason for that is
the mathematical complexity involved in problem solutions to support
transportation decision making.
Physical
Distribution is one of the most important activities to corporations, it is
fundamental to achieve high costumer service levels. A proper planning of
distribution activities generates process efficiency and reliability, as well
as cost reductions and positive impacts in client satisfaction.
As
mentioned by Junior et al. (2012), physical distribution is composed by tasks
such as transportation from producers to final customers, it can be done
straight from producers to customers or through a warehouse network. To make
physical distribution possible, it is necessary a to have professionals with
skills to deal with warehousing, loading and unloading vehicles as well as
appropriate moving equipment’s. The distribution of materials planning sequence
starts by choosing the suitable transportation mean, then the distribution
network and finally the definition of the delivery routes.
2.2.
Routing
According to Emoto and Lima (2007),
routing is a process to define delivery routes or set stop places for vehicle
belonging to a specific fleet, its main goal is to provide the proper materials
to several geographical areas pre-established.
To
Junior et al. (2013), the constrains make it hard to create routing solutions,
some examples as vehicles with different capacities of weigh and volumes,
restrictions to driving time for drivers, maximum speed on the roads,
restricted areas for heavy trucks in some hours of the day or circulation of
dangerous goods.
The
same authors affirm that by utilizing some routing methods, it is possible to
improve resources utilization, achieving higher efficiency and reducing costs
of deliveries, as well as reach a higher service level. With a correct
utilization the routing softwares can provide several routes more suitable to
costumer’s profiles, taking in consideration each of their constrains.
2.3.
Information
Technology
According
to Branco and Gigioli (2014), there are many routing softwares available in the
market to help corporations to plan transportation and distribution. These
softwares can provide satisfactory solutions to problems in this area, reducing
time and efforts when compared with solutions created manually. Routing
softwares work with internal algorithms able to reduce the operations costs and
vehicles utilization.
As proposed by Enomoto and Lima
(2007), in many companies, routing tends to be made manually, this can create
solutions that are not the best option far from the optimum result. Routing
software as computer systems that can provide transportation solutions based on
algorithms and a data base, with its utilization it is possible to obtain
information in a shorter time when compared to manual processing
3. RESEARCH METHODOLOGY
The
research method utilized is exploratory, in a format of case study, as
mentioned by Gil (2010), the exploratory research can take the form of a
bibliographic research and a case study.
Related
to the means, the research was conducted on a case study, that as mentioned by
Yin (2010), has the objective to make an empirical investigation to study a
contemporary phenomenon in a real life context.
This
study research object is a logistics operator enterprise located at the city of
Suzano, as it was not allowed to use the company´s name, from now on it will be
named as Beta Logistics.
As a
protocol for the case study, the research was divided in four steps. The first
was the analytical part when a detailed data collection of shipments occurred
from July to October of 2016 in order to verify the existence of seasonality in
operations, as well as data compilation related to October 2016 to make the
routing simulations of each of the days in study.
The
second step consists basically in data insertion in the routing software using
the data collection fulfilled on the first step for the month of October 2016,
that makes possible to recreate daily delivery routes based on the software
technical features, it will allow to compare with the collected results of
manual routing performed by the company.
It is
necessary to point here that the license provided by the routing software
company was limited to 10 vehicles, due to this limitation, it was considered
to the study the ten vehicles with a higher amount of deliveries. Nevertheless,
checking the samples, it was observed a great incidence in utilization of
sporadic vehicles to perform a short amount of deliveries, that makes
unfeasible to study those routes.
The
third step consists in the analysis of the simulations results comparing to the
real situation of deliveries, the goal here is to investigate the existence of
idleness.
The
fourth step is to apply a quiz to have a qualitative survey with the people
involved with routing routines, this makes possible to check and analyze their
opinion regarding to software functionalities and viability to implement the
routing software and possible improvements in the process. The quiz contains
four questions as can be observed in Figure 01.
Figure 01: Quiz.
4. CASE STUDY
To guarantee the viability of study it was collected
delivery data from July to October of 2016, the data from the first three months
were utilized to evaluate the enterprise´s operations as well as take evidences
of seasonality existence that could distort the analysis, the last month data
was taken in more detailed way to provide the possibility to visualize and
recreate the routes
4.1.
Data
Collection
During
the data collection of the period from July to September, the amount of
deliveries done each day were tabulated, the total of 4,599 deliveries were
performed in 76 working days, considering that there were not operations in
some Saturdays and Sundays during the researched period. With this information
it was possible to obtain an average of 60.51 deliveries per day.
Ins
second hand, searching for a better understanding of operations peculiarities
and to intent identify possible seasonal fluctuations, the information was
organized by month as shown in chart 01. By this way it was possible to observe
a great variation in delivered quantities from one day to another, but no
seasonal fluctuations could be identified in any of the months in analysis.
Chart 01: Daily
demand analysis from July to September of 2016.
Another
cut in information was done in order to analyze the demand profile on weekly
basis, and as in the monthly profile, no seasonal fluctuation could be noticed,
it is not possible to identify this kind of fluctuation along the days, it was
only possible to identify a weak tendency of deliveries increase on Tuesdays
and Fridays, as can be shown in chart 02.
Chart 02: Weekly demand analysis.
The
same variation was identified in the data collected for October of 2016, in
that month it was done 931 deliveries, with a daily average of 44.33
deliveries, this can be seen on chart 03.
Chart
03: October of 2016 demand analysis.
This
variation may justify the sporadic contracted vehicles during the period as
shown on table 01.
Table
01: number of sporadic vehicles per month
|
Working days per
month |
Vehicles amount |
|
1 |
41 |
|
2 |
9 |
|
3 |
4 |
|
4 |
2 |
|
5 |
2 |
Source: created
by the authors
Thus,
to make the research feasible, it was made an option to recreate the routes
only to the vehicles with continuous services, mainly because as mentioned
before, the license provided by the software provider allowed to register ten
vehicles. So, it was selected among the regular vehicles working, the ten which
performed more deliveries during the month, as detailed on table 02.
Table
02: Most utilized vehicles on October of 2016
|
Vehicle |
Working days |
Total deliveries |
Average deliveries/day |
Total weight (Kg) |
Average weight/day (Kg) |
|
VUC1 |
19 |
140 |
7.37 |
60002 |
3158 |
|
VUC2 |
17 |
114 |
6.71 |
40623 |
2389.59 |
|
VUC3 |
12 |
82 |
6.83 |
30162 |
2513.50 |
|
VAN1 |
16 |
74 |
4.63 |
28168 |
1760.5 |
|
TRUCK |
10 |
70 |
7 |
64677 |
6467.70 |
|
TOCO1 |
12 |
62 |
5.17 |
30397 |
2533.08 |
|
TOCO 2 |
18 |
56 |
3.11 |
16764 |
931.33 |
|
TOCO3 |
6 |
34 |
5.67 |
17564 |
2924.33 |
|
VAN2 |
7 |
33 |
4.71 |
82489 |
1178.29 |
Source: created
by the authors.
Finishing
the data collection stage, based on the information, the ten chosen vehicles
made 689 deliveries in total, with 123 different locations. That information
allowed the recreation of routes, so it was possible to analyze the vehicles
performance, as well as idleness and fleet reduction opportunities.
4.2.
The
Routing Software
To
the study it was utilized a routing software named Rout easy, it is a
multi-route software, available to work as an online platform, it allows the
creation of several delivery routes by inserting data manually or by data
charge from a data base in excel platform.
The
registration of delivery sources is very simple, can be done by introducing the
ZIP code from the street, then the user completes the information of building
number, after it is necessary to attribute a name to that origin location.
The
vehicle registration needs a insertion of more variables to reflect the real
conditions of the available fleet. Therefore, the software has eight pre-defined
vehicles models: motorcycle, car, utility, VUC, toco, truck and
eighteen-wheeler. As the user selects each kind of vehicle it is possible to
make adjustments in the configuration such as weigh capacity, maximum volume,
average speed, minimum load, maximum number of deliveries, maximum journey time
and circulation constrains. The number of registered vehicles varies according
to the contract with the customer, to the study it was available to register
the amount of ten delivery vehicles.
After
setting the working parameters of origins and vehicles it is possible to start
the delivery route creation. The data insertion related to customers such as
addresses, weigh, volume, time spent on client facilities and client
constrains, can be done manually or be uploaded to the software using a
standard spreadsheet extracted from the software, containing all the
information about a specific delivery batch as shown in Figure 02.
Figure
02: standard spreadsheet to upload data in the routing software.
After
inserting the data of deliveries in study, the software prepares the routes
following the stablished parameters and create route reports in PDF and MS
Excel formats. All the programed routes are stored by the software and can be
recovered or deleted by the users if necessary.
4.3.
Routes
Rebuild
As
prior mentioned to rebuild the routes there was a data selection based on the
ten most used vehicles, with a total of 689 deliveries during the month, along
the twenty-one working days registered. The data inversion was made by using
the standard spreadsheet shown in Figure 02, this made possible to check the
data accuracy before uploading to the software.
After
checking, the spreadsheets were dismembered in twenty-one spreadsheets named
according to the date of shipments.
Regarding
to parameters fed in the software, some standard data were settled as shown on Table
03. It is important to emphasize that it was assumed an average speed of 30
Km/h, taking in consideration Zandonade and Moretti (2012) study that found out
an average speed in the city of São Paulo and considering that the origin of
shipments is located out of the central belt of the city, it was added 3 Km/h,
justifying the settled.
It
was adopted a daily working journey of eight hours and a displacement time of
one hour, which means that a loaded vehicle leaving Suzano at 7:00 am, making a
lunch stop of one hour, will return maximum at 6:00 pm. The minimum vehicle
load and occupation, as well as stop time in customers were settled based on
peculiarities of the company, on average 20 minutes.
Table
03: routing software parameters
|
Item |
Parameter |
|
Average speed (Km/h) |
30 Km/h |
|
Minimum vehicle occupation (%) |
60% |
|
Maximum trip journey (hours) |
10 hours |
|
Average stop time in customers |
20 minutes |
Source: created
by the authors
The
specifications of vehicles were settled on the software according to each piece
of equipment characteristics, also the daily charges since the company does not
use its own fleet, but rent the vehicles to fulfill its needs, these data are
shown on Table 04.
Table
04: vehicle parameters
|
Vehicle |
Daily fee |
Weight capacity (Kg) |
Allowance to central area in São Paulo. |
|
Utility |
R$ 220.00 |
620 |
Allowed |
|
Van |
R$ 250.00 |
1500 |
Allowed |
|
VUC |
R$ 380.00 |
3500 |
Allowed |
|
Toco |
520.00 |
6000 |
Not allowed |
|
Truck |
590.00 |
12000 |
Not allowed |
Source: created
by the authors
After
setting and parameter and uploading data, routes were created on daily basis so
all the selected deliveries to the study were included. The results are
commented on chapter 5, results analysis.
5. Results Analysis
Regarding
to the quantitative analysis of new routes created by the software, a
comparison with real results from October operations was conducted to provide
an idea of impact caused by the new method.
In a
first moment the analysis showed differences in total deliveries quantities,
after a deeper evaluation it was noticed that the inconsistences were created
due to many deliveries were to companies cities far in the countryside area of
São Paulo State, and the routing software excluded them automatically.
The
reason for those exclusions were the constrains inserted in the software
regarding to maximum working time per day as well as minimum vehicle loading,
so the software excluded since the great distance meant to exceed the maximum
working journey, in other hand, the software did not create new routes to
excluded deliveries due to minimum load for vehicles.
As it
was a significant amount of deliveries excluded, around 10% with destination to
the State countryside, as shown in table 05, all the routes were checked and
recreated, then identified the non-completed deliveries. It made possible to
create new routes to the countryside areas by using more flexible rules to
these cases and adopting a travel speed of 80 Km/h.
Table
05: Percentage of deliveries to São Paulo metropolitan area vs Countryside
|
Date |
% countryside deliveries |
% metropolitan area deliveries |
|
10/03/2016 |
0 |
100 |
|
10/04/2016 |
11.9 |
88.1 |
|
10/05/2016 |
14.29 |
85.71 |
|
10/06/2016 |
2.33 |
97.67 |
|
10/07/2016 |
5.88 |
94.12 |
|
10/10/2016 |
19.44 |
80.56 |
|
10/11/2016 |
9.62 |
90.38 |
|
10/13/2016 |
12.50 |
87.50 |
|
10/14/2016 |
0 |
100 |
|
10/17/2016 |
16.67 |
83.33 |
|
10/18/2016 |
10 |
90 |
|
10/19/2016 |
22.58 |
77.42 |
|
10/20/2016 |
0 |
100 |
|
10/21/2016 |
0 |
100 |
|
10/24/2016 |
80 |
20 |
|
10/25/2015 |
20 |
80 |
|
10/26/2016 |
4.26 |
95.74 |
|
10/27/2016 |
2.13 |
97.87 |
|
10/28/2016 |
13.95 |
86.05 |
|
10/29/2016 |
0 |
100 |
|
10/31/2016 |
83.33 |
16.67 |
|
Total |
10.6 |
89.4 |
Source: created
by the authors
Using
this procedure made possible to accomplish 100% of deliveries and have the
recreation of all routes. The results with the routing software were very
satisfactory, confirming the existence of idleness and proving the
possibilities of fleet optimization.
Another
important information is the significant reduction in the amount of contracted
vehicles , considering the whole month it was possible to reduce from 125
vehicles to 92 in the simulated situation, it is important to emphasize that on
October the 10th and 24th it was necessary to increase
the quantity of vehicles to fulfill the same demand, but even with the increase
the payment values were reduced due to the utilization of smaller and cheaper
vehicles, allied to a better exploitation of the resources.
The
qualitative analysis was performed by introducing the software to key people in
distribution process in the company, their positions were, transportation and
warehousing coordinator, distribution supervisor and the technician responsible
for routing process.
Table
06: saving results comparison current process vs software routing
|
|
Real data October 2016 |
Simulation October 2016 |
Results |
||||
|
Date |
Deliveries |
Vehicles |
Fee |
Vehicles |
Fee |
Vehicles reduction |
Fee reduction |
|
10/03/2016 |
21 |
4 |
R$1,530.00 |
2 |
R$760.00 |
50% |
50.33% |
|
10/04/2016 |
42 |
8 |
R$3,110.00 |
5 |
R$1,960.00 |
37.5% |
36.98% |
|
10/05/2016 |
42 |
7 |
R$2,890.00 |
7 |
R$2,430.00 |
0 |
15.92% |
|
10/06/2016 |
43 |
6 |
R$2,640.00 |
5 |
R$2,160.00 |
16.67% |
18.18% |
|
10/07/2016 |
34 |
8 |
R$3,240.00 |
5 |
R$2,250.00 |
37.5% |
30.56% |
|
10/10/2016 |
36 |
6 |
R$2,430.00 |
7 |
R$2,110.00 |
-16.67% |
13.17% |
|
10/11/2016 |
52 |
10 |
R$4,010.00 |
7 |
R$2,860.00 |
30% |
28.68% |
|
10/13/2016 |
32 |
7 |
R$2,650.00 |
4 |
R$1,370.00 |
42.86% |
48.3% |
|
10/14/2016 |
33 |
5 |
R$2,260.00 |
3 |
R$1,280.00 |
40% |
43.36% |
|
10/17/2016 |
24 |
5 |
R$2,180.00 |
4 |
R$1,500.00 |
20% |
31.19% |
|
10/18/2016 |
40 |
7 |
R$3,020.00 |
6 |
R$2,210.00 |
14.29% |
26.82% |
|
10/19/2016 |
31 |
6 |
R$2,640.00 |
4 |
R$1,440.00 |
33.33% |
45.45% |
|
10/20/2016 |
30 |
7 |
R$2,380.00 |
3 |
R$1,280.00 |
57.14% |
46.22% |
|
10/21/2016 |
28 |
5 |
R$1,980.00 |
3 |
R$1,420.00 |
40% |
28.28% |
|
10/24/2016 |
10 |
2 |
R$1,040.00 |
3 |
R$720.00 |
-50% |
30.77% |
|
10/25/2015 |
30 |
6 |
R$2,270.00 |
5 |
R$1,620.00 |
16.67% |
28.63% |
|
10/26/2016 |
47 |
7 |
R$2,680.00 |
5 |
R$2,090.00 |
28.57% |
22.01% |
|
10/27/2016 |
47 |
7 |
R$2,650.00 |
5 |
R$1,750.00 |
28.57% |
33.96% |
|
10/28/2016 |
43 |
8 |
R$3,110.00 |
6 |
R$2,080.00 |
25% |
33.12% |
|
10/29/2016 |
18 |
3 |
R$1,140.00 |
2 |
R$900.00 |
33.33% |
21.05% |
|
10/31/2016 |
6 |
1 |
R$520,00 |
1 |
R$250.00 |
0% |
51.92 |
|
Total |
689 |
125 |
R$50,370.00 |
92 |
R$34,440.00 |
26.4% |
31.63% |
Source: created
by the authors
There
was a short training to present the software functions, then they prepared
routes for the demand on the day the survey occurred, after that they answered
the quiz from figure 01 and gave their opinion on the software. Both affirmed
that the software is easy to use and efficient on building routes, and they
believe it suits the company needs, and can be used to route the deliveries.
After
all it was requested to attribute a grade varying from zero to ten, where zero
means totally unsatisfactory and ten means fully satisfactory, the average was
9.33.
6. FINAL COMMENTS
Based
on the research results, it is possible to affirm that the routing software
used provided significant earnings on costs and efficiency, as well as quality
in routing.
Nevertheless,
to achieve good results it is necessary well-trained uses with good knowledge
of available functions in the software and companies’ routines, these two
things help to analyze the routes quality and if necessary make adjustments
such relocate deliveries excluded from routes due to the stablished
constrains.
It is
very important that vehicle parameter is accurate in information such as load
capacity and constrains, other important issue is to know customers
characteristics, so the software will operate to fulfill all these parameters
and get better results.
With
more time using the software it is possible that users can be able to give
feedback to improve the utilization and reduce even more the excluded
deliveries.
Finally,
is important to emphasize that the study was conduct with a limited time, the
simulations had as main objective optimize the vehicles reduction, but the
routing software also has options to reduce distances. So for future studies,
it is recommended to compare gains created by using other functionalities, and
if possible to make a longer evaluation of software implementation and check
the learning curve to have a better picture of the advantages obtained.
REFERENCES
BRANCO, F. J. C.; GIGIOLI, O. A. (2014) Roteirização
de transporte de carga estudo de caso: distribuidora de tintas e seu método de
entregas. REV.FAE, Curitiba, v. 17,
n. 2, p. 56 - 81.
ENOMOTO, L. M.; LIMA, R. S. (2007) Análise
da distribuição física e roteirização em um atacadista. Produção, v. 17, n. 1, p. 094-108.
GIL, A. C. (2010) Como Elaborar Projetos de Pesquisa. 5ª ed. São Paulo: Atlas.
JUNIOR,
I. C. L. et al. (2012) Estudo
para implementação de um sistema de roteirização e um novo centro de
distribuição para uma empresa de água mineral do sul de Minas Gerais. IXSEGT Simpósio de excelência em gestão e
tecnologia.
JUNIOR, C. A. M. et al. (2013) O Papel da
roteirização na redução de custos logísticos e melhoria do nível de serviço em
uma empresa do segmento alimentício no Ceará. XX Congresso Brasileiro de Custos - Uberlândia, MG, Brasil, 18 a 20
de novembro de 2013.
LIMA, M. (2014) Custos Logísticos no Brasil. Available:
<http://www.ilos.com.br/web/custos-logisticos-no-brasil/>. Access: 05/10/2016.
YIN, R. K. (2010) Estudo de caso: planejamento e métodos. 4ª ed. Porto Alegre:
Bookman.
SIQUEIRA, E. S.; SOUZA, C. A.; VIANA, A. B. N.
(2013) Uso da Tecnologia de Informação em Empresas de Pequeno e Médio Porte:
uma análise a partir dos dados da pesquisa “TIC Empresas” de 2011. In: Conf-Irm 2013 International Conference on
Information Resources Management, p. 1-14.
ZANDONADE, P.; MORETTI, R. (2012) O padrão
de mobilidade de São Paulo e o pressuposto de desigualdade. EURE (Santiago), v. 38,
n. 113, p. 77-97.