Modelling Mode Choice in Freight Transportation Based on ... · PDF fileModelling Mode Choice...
Transcript of Modelling Mode Choice in Freight Transportation Based on ... · PDF fileModelling Mode Choice...
Modelling Mode Choice in Freight
Transportation Based on Deterministic Logistics
Cost Model
Sevil Köfteci Department of Civil Engineering, Faculty of Engineering, Akdeniz Unversity, Antalya, Turkey
Email: [email protected]
Abstract—Freight transportation is very important sector
for national economies. The well-planned freight
transportation system makes a positive contributions to life
of people. An important part of any numerical analysis of
freight transportation is a ability for forecasting the demand.
Freight transportation demand is far more complex than
passenger transportation demand. It is clear that firm is the
basic decision-making unit in the transportation of freight.
This paper summarizes the results of deterministic logistics
cost analysis based on stated preference technique.
Deterministic logistics cost model is developed at the level of
the individual firm. The dataset used for this study is the
result of 14 interviews with ceramic merchants, which have
been realised in Antalya region, Turkey. In conclusion,
which factors are important for create deterministic
logistics costs model and calculate logistics costs based on
this model were determined in this study.
Index Terms—freight transportation, logistics costs,
deterministic model, demand analysis, stated preference.
I. INTRODUCTION
Interest in modeling and analysis work in freight
transportation has increased greatly over the past few
years for variety reasons. The well-planned freight
transportation system makes a positive contributions to
life of people. An important component of any analysis of
freight transportation is a capability for forecasting the
demand for a certain type of service under given set of
conditions. Logistics is one of the most important
activities in freight transportation. Logistics represents a
collection of activities that ensures the availability of the
right products in the right quantity to the right customer
at the right time [1]. Logistics activities serve as a link
between production and consumption and essentially
provide a bridge between production and market
locations or suppliers separated by distance and time [2].
Different values are added to a product at various stages
of its life cycle. Production and manufacturing adds form
value by converting the raw material or components into
finished parts. Place value is provided through
transportation by moving the product where it is needed.
Time value is provided through storage and inventory
Manuscript received February 1, 2015; revised April 17, 2015.
control ensuring the availability of the product when
needed. Possesion value is added to the product through
marketing and sales [1]. The primary objective of
logistics is to provide appropriate standard customer
service at the minimum possible total logistics cost.
Customer service is measured by the availability of the
required product of acceptable quality within the
specified time [3]. This paper summarizes the results of
deterministic logistics cost analysis based on stated
preference technique. Deterministic logistics cost model
is developed at the level of the individual firm. The
dataset used for this study is the result of 50 interviews
with cement merchants, which were realised in Antalya
region, Turkey. In conclusion, which factors are
important for create deterministic logistics costs model
and calculate logistics costs based on this model were
determined in this study.
II. FREIGHT TRANSPORTATION DEMAND
Freight transportation makes a vital contribution to the
economy and society. Freight transportation helps
production, industry, trade and consumption activities by
ensuring the efficient movement and timely availability
of raw materials and finished goods. Advances in freight
transportation have been a major origin of productivity
growth in the economies of many countries. Forecasting
the developments of future transportation is essential for
freight transportation policy planning. Modelling freight
transportation demand is far more complex than
passenger transportation demand. Modelling studies
related the freight transportation has been carried out
since 1960. The first models about freight demand were
aggregate models dealing with freight transportation.
More recent models, are dissagregate models. Classical
four step freight demand modeling technique is
recognizes that freight demand is derived from economic
activity. It involves comprehensive interrelationships
among economic activity, production and consumption
nodes, distribution or linkages between production and
consumption nodes, mode choice and shipment size
decisions, vehicle trips, and route assignments [4]. Figure
1 illustrates the model components and their key inputs of
classical freight modeling approach [5].
Journal of Traffic and Logistics Engineering Vol. 3, No. 2, December 2015
©2015 Journal of Traffic and Logistics Engineering 180doi: 10.12720/jtle.3.2.180-184
There is a three type of decisions which are
represented as a complex hierarchy of choices in freight
transportation demand analysis studies. These are long
run, intermediate and short run decisions. A firm
ultimately makes a decisions on the purchase of its
transportation services. The long run decisions that a
manager of firm has to make a concerns the location of
the plant.
Figure 1. Model components and data inputs of freight transportation modelling
İntermediate decisions dealing with the production
involving the choice of technology of production and
level of output [6]. The short run decisions involves the
choice of mode, shipment size and point of supply given
the annual use rate of inputs. In terms of production
theory, this is somewhat restrictive, since it allows no
factor substitution between the demand for transportation
and other factors, but it does allow for substitution within
the transportation and logistics cost elements [7].
Therefore, a demand model using the logistics choice
process as its basis must be categorized as a short run
freight demand model. In a short run demand model, the
annual demand for input material is used as only
transportation related choices, such as origin of supplier,
mode of transport. This study then proceeds to develop
the short-run model of freight demand based on logistics
cost elements. The model examined in this paper is
disaggregate, explanatory and considers the firm as the
basic decision making unit.
III. LOGISTICS COST MODEL PARAMETERS
Over the last four decades a substantial researches
about to mode choice on the basis of logistics cost model
for the freight demand analysis was reported in the
literature These costs typically involved in the decision
process are; purchase costs, ordering costs, transportation
costs, in transit-carrying costs, inventory costs, inventory
carrying costs, lost and damage costs, safety (stock-out)
costs. Table I illustrates the mainly cost components used
at these models. In the cells of table, “Y” means “YES”
and “N” means No” Names of authors and their
publications which are include these models are listed
below. Because of the planning and operational decisions
of one cost type typically has an impact on the other cost
type, logistics decisions are usually based on total cost
approach.
1. “Intermodal transportation and inventory cost
model highway-to-rail intermodal user’s manual”
[8].
2. Vieira and F. M. Luiz, “The value of service in
freight transportation” [9].
3. Y. S. Chiang, P. O. Robets, and M. Ben-Akiva,
“Development of a policy sensitive model for
forecasting freight” [10].
4. Y. Sheffi, B. Eskandari, and H. N. Koutsopolous,
“Transportation mode choice based on total
logistics costs” [11].
5. Y. Sheffi, “Carrier/shipper interactions in the
transportation market: An analytical framework”
[12].
6. W. J. Baumol and H. D. Vinod, “An inventory
theoretic model of freight transportation demand”
[13].
7. J. Ozment, “Demand for intermodal
transportation in arkansas” [14].
8. A. Kanafani, “Methodology for mode selectipn
in corridor analysis of freight transportation” [6].
9. G. Chow, “A total logistics cost approach to
measuring collateral benefits of security and
supply chain improvements at international
gateways” 2008, [15].
For example, inventory and transportation decisions
are very closely related. If a slower mode is used for
transportation because it is cheaper, than a higher level of
inventory may be needed. On the other hand, if a faster
mode is used, than a lower inventory level may be
maintened. Therefore a trade-off is need to determine the
optimal inventory level. The primary objectives of
logistics is to provide customer service at the minimum
total cost. Total cost includes the costs incurred to
perform all the logistic functions in order to provide
world-class customer service [3].
The total cost approach has been a core principle in
transportation and logistics decision making activity in
the mid 1950s [15]. In developing the metodology for
mode selection based on logistics cost model, we shall
adopt the perspective of the receiver. The receiver is a
rational economic decision maker who tries to minimize
total logistics costs for the delivered product. A short
explanation of logistics costs parameters are presented
below.
A. Ordering Costs
If the ordering cost is represented by “OC”, the
ordering costs per shipment is represented by “a”, then
calculation of ordering cost as follows:
OC a (1)
B. Transportation Costs
If the transportation cost per shipment is represented
by “TC”, transportation costs per unit product is
represented by “c”, the amount of product carried per
shipment is represented by “u”, then calculation of
transportation cost is as follows:
.TC u c (2)
Journal of Traffic and Logistics Engineering Vol. 3, No. 2, December 2015
©2015 Journal of Traffic and Logistics Engineering 181
C. In-Transit Carrying Costs
If the in-transit carrying cost per shipment is
represented by “ITC”, the amount of product carried per
shipment is represented by “u”, value per unit product is
represented by “p”, annual interest rate is represented by
“f”, transportation costs per unit product is represented by
“c”, The amount of product carried per shipment is
represented by “u”, then calculation of in-transit carrying
costs is as follows:
. . .365
fITC u p t (3)
D. Inventory Costs
Avarage inventory level was recognized %50 in this
study. If inventory cost is represented by “IC”, The
annual storage cost is represented by “M”, annual
demand is represented by “T”, the amount of product
carried per shipment is represented by “u”, value per unit
product is represented by “p”, annual interest rate is
represented by “f”, then Calculation of inventory costs as
follows:
365
[( . ) ( . . . )2 365 2
M u f uIC P
TT
u
If the number of annual shipment (T/u) is represented
by “k” and storage cost per unit (M/T) is represented by d,
calculation of “IC” is rearranged as follow:
[( . ) ( . . )]2 2
u f uIC d P
k (4)
E. Safety Stock Costs
If the safety stock cost is represented by “SC”, the
safety stock level is represented by “n”, then calculation
of safety stock costs as follows:
.. .365. . . .
( ) ( ) [( ) ( )]
.365
Mn
P f n d n P f nTSCT T k k
u u
(5)
F. Loss and Damage Costs
Loss and damage cost includes the actual value of the
material lost and damaged for which the shipper is not
compensated by the carrier [16]. If the loss and damage
cost is represented by “LDC”, the amount of lost and
damaged product is represented by “z”, then calculation
of safety stock costs as follows:
.LDC z P (6)
If total logistics cost is represented by “TLC”,
calculation of total logistics costs is as follows:
TLC OC TC ITC IC SC LDC (7)
. . . 1 .[( ) ( . ) ( ) ( )
365 2
( . ) ( . )
u p f t p f nTLC a u c u d
k k
d p f z p
(8)
With a view to examine process involved in selecting
an appropriate freight demand analysis strategy, it is
useful to represent the decision variables defining
alternative short run analysis strategies as the choice of
origin represented by i, shipment size represented by q
and mode of transportation represented by m. The
criteria for an alternative i,m,q to be a optimal logistics
strategy is satisfaction of the following condition [10]:
imq i m qw w i m q A (9)
where A is the set of possible origin, shipment size and
mode combinations. There is as two general approaches
taken in the account in modeling the logistics decision of
firm. The general difference between these approaches is
related to the level of certainty concerning the
information in the logistics cost function. In the first
approach (deterministic cost models), the logistics cost
function parameters are assumed fully observable.
Therefore, the alternative defined by the choice i, m, q is
selected with certainty if
imq i m qw w i m q A
(10)
A model developed upon this assumption will be
reffered to as a deterministic cost model of short-run
freight demand. The model states:
( / ) 1,
0,
r imq i m qP imq A if w w i m q A
otherwise
where wimq, wi’m’q’ are fully observed. The second
approach assumes that the logistics cost function is not
fully observable. According to this approach, logistics
cost function consists of two parts. The first part is f
observable and the second part is unobservable. In this
study, mode choice was modelled based on deterministic
logistics cost model in freight transportation. The
deterministic logistics cost model assumes that the total
logistics cost function is fully observable. The objective
function for the firm’s logistics strategy is to minimize
the total logistics costs.
IV. EMPIRICAL STUDY
Antalya is the biggest tourism center of Turkey,
located south of Turkey. Therefore, the population of city
is increasing continuously and construction activities are
too much accordingly. Ceramic is the basic elements of
the building construction sector. Some of ceramics used
in the construction sector in Antalya province are
transported from Canakkale Ceramic Factory which is
located at Çan (the town of Çanakkale province). There
are two transportation modes used for carrying of
ceramics from Çan to Antalya. One of them is the road
transportation. Other is the intermodal transportation.
Burdur is transfer station for intermodal transportation.
Fig. 2 and Fig. 3 illustrate the transportation corridor
analysed in this study.
Figure 2. Direct road transportation
Journal of Traffic and Logistics Engineering Vol. 3, No. 2, December 2015
©2015 Journal of Traffic and Logistics Engineering 182
Figure 3. Intermodal transportation
Since this transportation corridor has a big potential
about both transportation types and demand
characteristics knowledge, the surveys were conducted to
transportation managers of 14 ceramics firms whose
shipments were carried by using this transportation
corridor. After surveys were conducted, data set entered
into an Excell spreadsheet. Fig. 4 illustrates the Excel
spreadsheet included dataset table.
Figure 4. The data set sample used in rhe empirical study
Total logistics costs for each transportation operations
made within one year of each company were calculated
by using total logistics cost model (Eq. 8). Total logistics
costs were calculated separately for the intermodal and
direct road transportation. Fig. 5 illustrates graphically
the components of total logistics costs.
Figure 5. Components of total logistics costs
As can be seen in Fig. 5, it is assumed that total
logistics costs consists of three main parts in the study:
Ordering costs
Transportation related costs which is include
transportation costs, in-transit carrying costs and
loss-damage costs
Inventory related costs which is include inventory
costs and safety stock costs.
One can see Fig. 5 that, percentage of transportation
related costs over the other cost components is dominant.
Nevertheless, inventory related costs is the typically other
important logistics cost parameter. This figure shows that,
variables related transportation related costs are an
important factors creating logistics cost model. Cost
components of inventory related costs and transportation
related costs are shown in Fig. 6 and Fig. 7 respectively.
Figure 6. Components of inventory related costs
Fig. 6 shows the subcomponent shares of the
inventory costs such as inventory costs (71%) and safety
stock costs (29 %). This result is not surprising because
stock level in safety stock cost is lower than inventory
cost.
Figure 7. Components of transportation related costs
Components of transportation related costs reported in
Fig. 7 indicate that the transportation costs are largely
composed of direct road transportation costs. It is
followed by intermodal transportation costs. The other
two cost parameters are less effective on transportation
related costs. This implies that the parameters that the
influence the calculation of transportation costs are very
important factors creating transportation related costs.
Transportation costs consists of two parameters:
transportation costs per unit and the amount of product
carried per shipment. The second parameter is used
almost all cost calculations in total logistics cost model.
Therefore, it is assumed that, this parameter not a
determining factor on the cost calculation. As a result,
transportation costs parameter is determined as the most
important factor defining short run freight demand
Çan Bandırma Sığırcı
Station
Burdur
Station
Antalya
42% 7%
18%
33%
71%
29%
Journal of Traffic and Logistics Engineering Vol. 3, No. 2, December 2015
©2015 Journal of Traffic and Logistics Engineering 183
analysis strategy based on deterministic logistics cost
model.
V. CONCLUSION
Freight transportation is very important activity for
national economies. An important part of any quantitative
analysis of freight transportation is a ability for
forecasting the demand for a certain type of condition
under a given set of possibility. This paper summarized
the study on selecting an appropriate freight demand
analysis strategy based on deterministic logistics cost
model. The data set used for this study is the result of 14
interviews about cement sector firms, which have been
realized in the Turkish region of Antalya. The results
confirm that, the logistics cost model is most affected by
transportation cost parameter. Depending on this situation
it can be said that, if the costs of transportation modes
are rearranged, the possibility of firms rely on intermodal
transportation rather than road transportation will
increase. This result is very important authorized persons
dealing with the mode choice activities in freght
transportation.
REFERENCES
[1] R. J. Kasilingam, Logistics and Transportation, Kluwer Academic
Publishers, 1998, pp. 1. [2] İ. Özalp, B. Suvacı, and Z. H. Tonus, “A new approach in logistics
management: Just IN Time-Logistics (JIT-L),” in Proc.
International Conference on Social Sciences, 2010, pp. 4. [3] H. V. Davis and W. H. Drumm, “Logistic costs and customer
service levels,” in Proc. Council of Logistic Management Annual
Conference Proceedings, 1996, pp. 149-159. [4] Holguín-Veras, “On the attitudinal characteristics of motor
carriers toward container availability systems,” International
Journal of Services Technology and Management, vol. 1, no. 2-3, pp. 140-155, 2000.
[5] F. Southworth, “Freight transportation planning: Models and Methods,” Transportation Systems Planning: Methods and
Applications, pp. 4-1, pp. 4-29, 2003.
[6] A. Kanafani, “Methodology for mode selection in corridor analysis of freight transportation,” Research Paper, The University
of California Transportation Center, 1984, pp. 15-29.
[7] J. R. Stock and D. M. Lambert, Strategic Logistics Management, McGraw-Hill International Edition, 2001, pp. 13-26.
[8] Intermodal Transportation and Inventory Cost Model Highway-to-
Rail Intermodal User’s Manual, US Department of Transportation, Federal Railroad Administration, 2005.
[9] Vieira and F. M. Luiz, “The value of service in freight
transportation,” Ph. D. Dissertation, Department of Civil Engineering, Massachusetts Institute of Technology, 1992.
[10] Y. S. Chiang, P. O. Robets, and M. Ben-Akiva, “Development of a
policy sensitive model for forecasting freight,” Center M. I .T. Center for Transportation Studies Final Report, 1977.
[11] Y. Sheffi, B. Eskandari, and H. N. Koutsopolous, “Transportation
mode choice based on total logistics costs,” Journal of Business Logistics, vol. 9, no. 2, pp. 137-154, 1988.
[12] Y. Sheffi, “Carrier/Shipper interactions in the transportation
market: An analytical framework,” Journal of Business Logistics, vol. 7, no. 1, pp. 1-24, 1986.
[13] W. J. Baumol and H. D. Vinod, “An inventory theoretic model of
freight transportatio demand,” Management Science, vol. 116, no. 7, pp. 413-421, 1970.
[14] J. Ozment, “Demand for intermodal transportation in Arkansas,”
Research Paper, 2001 [15] G. Chow, “A total logistics cost approach to measuring collateral
benefits of security and supply chain improvements at
international gateways,” presented at CAAMTBA Conference, 28-29 March 2007, Vancouver, Canada, 2008.
[16] P. Cook, S. Das, A. Aeppli, and C. Martland, “Key factors in road-
rail mode choice in India: Applying the logistics cost approach,” in Proc. Winter Simulation Conferences, 1999, pp. 1281-1283.
Sevil Köfteci was born on 27th August 1974, in Akdeniz University, Antalya, Türkiye.
The list of degrees as shown below:
Ph. D. 2002-2009 Transportation Program in Civil Engineering, Institute of Applied
Sciences, Yildiz Technical University, Istanbul.
M. Sc. 1999-2002 Transportation Program in
Civil Engineering, Institute of Applied Sciences, Akdeniz University, Antalya
B. Sc. 1993-1997 Department of Civil
Engineering, İstanbul Technical University, Istanbul She is an ASST. PROF. DR. in Department of Civil Engineering in
Akdeniz University, Antalya, Türkiye. Her major field of study include
highway and railway pavement.
Journal of Traffic and Logistics Engineering Vol. 3, No. 2, December 2015
©2015 Journal of Traffic and Logistics Engineering 184