Strategic Dispatching System Design for Truckload Transportation

WILDAU TECHNICAL UNIVERSITY OF APPLIED SCIENCES
FACULTY OF ENGINEERING AND NATURAL SCIENCES
Strategic Dispatching System Design for Truckload
Transportation
Academic Master Study programme
“Logistics and Supply Chain Management”
Merve Nur Tas
Wildau, April 2019

i
Contents
List of Abbreviations.................................................................................................................iii
List of Tables.............................................................................................................................iv
List of Figures ............................................................................................................................v
1 INTRODUCTION..............................................................................................................1
1.1 The Trucking Industry................................................................................................1
1.2 Statement of the Problem ...........................................................................................3
1.3 Significance of the Study ...........................................................................................3
1.4 Scope and Delimitation ..............................................................................................5
2 REVIEW OF RELATED LITERATURE AND STUDIES ..............................................6
2.1 TL Network Design....................................................................................................6
2.2 Research Hypothesis ..................................................................................................8
2.3 Conceptual Framework ..............................................................................................9
2.4 Definition of Terms....................................................................................................9
3 METHODOLOGY...........................................................................................................11
3.1 Research Design.......................................................................................................11
3.2 Locating Terminals...................................................................................................11
3.3 Baseline Scenario Creation ......................................................................................12
4 MODEL CONSTRUCTION............................................................................................17
4.1 Introduction to the Model.........................................................................................17
4.2 Route Enumeration...................................................................................................18
4.3 Model Formulation...................................................................................................19
4.4 Problem Size.............................................................................................................20
4.5 Assumptions .............................................................................................................21
5 RESULTS.........................................................................................................................22
6 DISCUSSION ..................................................................................................................23
6.1 KPIs on Dispatching System’s Success ...................................................................23
6.2 Advantages and Disadvantages for The Multi-Terminal Network Dispatching......24
7 CONCLUSIONS..............................................................................................................26
8 REFERENCES.................................................................................................................27
9 APPENDICES..................................................................................................................30
9.1 Appendix A ..............................................................................................................30
9.2 Appendix B...............................................................................................................37

ii
Abstract
Trucking industry is the backbone of the world’s economy and it is also the one of the most
inefficient ones, especially the truckload segment. One of the biggest efficiency problems of
truckload carriers’ faces is the high percentages of driver turnover due to demanding work
conditions. Several attempts have been made to solve the issue during the recent years. This
paper will give a comprehensive summary of the previous works and present an outline for a
multi-terminal dispatching network. A mathematical optimization model is proposed to solve
the multi-terminal network problem and a baseline scenario has created for its testing. Results
are evaluated and compared with the classic PtP approach.
Keywords
Truckload trucking, dispatching, truckload network design, terminal network design, relay
network

iii
List of Abbreviations
AETR: European Agreement Concerning the Work of Crews of Vehicles Engaged in
International Road Transport
ATA: American Trucking Associations
ATRI: American Transportation Research Institute
CMV: Commercial Motor Vehicle
EC: European Council
EEA: European Economic Area
EEC: European Economic Commission
EU: European Union
FMCSA: Federal Motor Carrier Safety Administration
FTL: Full Truck Load
H&S: Hub and Spoke
HOS: Hours-of-Service
KPI: Key Performance Indicator
LTL: Less Than Truck Load
PtP: Point-to-Point
R-TLRDN-MD: Robust Truckload Relay Network Design with Mixed Fleet Dispatching
RN: Relay Network
RP: Relay Point
TL: Truck Load
TLRND-MD: Relay Network Design with Mixed Fleet Dispatching
TLRND: Truckload Relay Network Design

iv
List of Tables
Table 1:North American Freight by Mode: June 2018 (Dollars in Billions) .............................1
Table 2:Modal split of freight transport in total inland freight transport in EU-28 (% of total
tonne-kilometres)........................................................................................................................2
Table 3: TL Transport Terminal Network Design Literature Summary....................................8
Table 4: List of source and destination points..........................................................................12
Table 5:Problem Instance.........................................................................................................15
Table 6: Example Truck Trip ...................................................................................................14
Table 7: Possible Route Types According to the Distance ......................................................19
Table 8: Problem Size ..............................................................................................................20
Table 9:Results.........................................................................................................................22
Table 10:Results for the Whole Network.................................................................................23
Table 11:Results for the FN SD Pair........................................................................................ 23

v
List of Figures
Figure 1: Visualization of Relay Network of a TL Carrier ........................................................4
Figure 2: Conceptual Framework...............................................................................................9
Figure 3: The Dispatching System from the Trailer’s Perspective ..........................................11
Figure 4: Turkey Map with the Source and Destination Points (Google Maps, 2019)............13
Figure 5: The Dispatching System form the Driver's Perspective ...........................................13
Figure 6: International Road Routes in Turkey (Republic of Turkey General Directorate of
Highways, 2018) ......................................................................................................................14
Figure 7: Visualization of an Example Trip with Multi Terminal Network ............................17

1
1 INTRODUCTION
1.1 The Trucking Industry
Trucking is the leading mode of freight transportation in the world: In USA, nearly 71 % of the
nation’s freight is carried by trucks. The USA trucking industry worth $738.9 billion, nearly
70% of it coming from truckload segment (American Trucking Associations, 2018). In EU,
nearly three quarters of total inland freight transport was by road in 2017 (see table 2). In some
countries, the share of road freight in national transport reached extremely high numbers, e.g.,
in United Kingdom by 96.4 % (Eurostat, 2018). The essentiality of the trucking sector can be
summarized by the former slogan of the American Trucking Associations (ATA), “If you
bought it, a truck brought it.” (U.S. House Committee on Public Works and Transportation,
1994).i
Even though increasing emphasis on the importance of sustainable transportation systems in
the world, the dependency to the road transport is still a reality. The numbers are quite clear at
this one, indicating that road transportation is the most used transportation mode in the world
(see table 1) (Bureau of Transportation Statistics, 2018). Not only that, because of the
globalization, increased population, economic activity and increased online sales; number of
trucks on the roads shows an increasing trend over the past years. In U.S., in year 2015, freight
tonnage and value rose 6.5 and 8.2 percent, respectively, compared to year 2012 (Bureau of
Transportation Statistics, 2017).
Table 1:North American Freight by Mode: June 2018 (Dollars in Billions)
(Bureau of Transportation Statistics, 2018).
66.6
15.7
8.0 6.5
4.0
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
Truck Rail Vessel Pipeline Air

2
Table 2:Modal split of freight transport in total inland freight transport in EU-28 (% of total tonne-kilometres).
(Eurostat, 2018)
Truckload trucking industry is a highly fragmented industry with only a few big players and
excessive number of companies with a small-scale of fleets. As in other fragmented industries,
the barrier to entry is very low and exit is easy; one can enter the industry only with one truck
and when they no longer want to be in, the equipment can be resold easily. This leads to
excessive number of players in the market. As of June 2017, only in USA, the number of carriers
has reached to 777,240. However, the data shows that the gap between the small and big players
of the market is massive. 97.3% of these carriers operate with fewer than 20 trucks while 91%
operate with 6 or fewer trucks only (American Trucking Associations, 2018).
Like every industry, trucking industry has its specific problems, one of them being the driver
retention.
Many truckload carriers use the Point-to-Point (PtP) dispatching method. Point-to-Point system
puts the driver in charge during the whole trip, from origin to destination. After the first
delivery, drivers get back home after following the given schedule, which includes more than
one stop most of the time (Campbell, 2005). This results in long trips far away from home for
the drivers, and perception of a low-quality life.
Not only this, with increasing demand for the freight shipments, demand for the drivers is also
increasing. Both of these, increased demand for the drivers and demanding work conditions,
results in extreme numbers in driver turnover, especially in the truckload segment as those are
mostly long-haul shipments. According to American Trucking Association’s (ATA) Trucking
Activity Report, turnover rate is announced as 94% in the first quarter of 2018, 20 points higher
than it is in 2017. In the same report, the turnover rate for less-than-truckload (LTL) carriers is
announced as 10% in 2018, still 2 percent higher compared to year before (American Trucking
Organizations, 2018). The turnover rate is always higher for the TL trucking as working
conditions is much better for the LTL drivers.
As turnover keeps being a major problem for TL carriers, carriers search for solutions varying
from acquiring smaller carriers with drivers on hand to offer compensations to the drivers
(Transport Topics, 2018). According to the Driver Compensation Study by the ATA, the annual
median salaries of a national driver and a private driver showed 15% and 18% increase,

3
respectively, since 2013 (American Trucking Organizations, 2018). However, this does not
seem to fix the problem.
1.2 Statement of the Problem
Driver retention is the first part and more obvious part of the problem. Let’s review what a
truckload trucking service is quickly. The service carrier providing is, picking up of the goods
from one point and to deliver them to another point, in its simplest form. The steps are loading
the goods to the truck and driving until the delivery point and unloading them. Most of the time,
previously described PtP method is used. One driver is responsible from the load from the initial
point to the final destination and bringing the truck back.
A trucking carrier relies on several resources to provide this service. These are namely, the
truck, the driver, and the fuel. While truck is one of the biggest investments of the carriers, the
drivers and the fuel are counted as the two highest operational expenses (American
Transportation Research Institute, 2017). Purchasing a truck, also affects company’s
operational costs with the lease or purchase payments, insurance payments, permits and repair
and maintenance costs.
Resource utilization formula calculates how much of an available resource we can make use of.
It is always an important indicator, as low utilization rate may shed the light on unnecessary
expenses. Companies always work to make the most of their resources, any resource that could
not exploited to the maximum, is a waste. Utilization rate is calculated dividing the used amount
of a resource to its total availability. The formula can be adjusted to the different units, like
hours, capacity etc. For a carrier, truck and driver utilization are the most important ones to
track.
The journey of the truck may be disrupted because of several reasons. First of all, the driver is
going to have to stop for the fuel or battery charge. Secondly, the driver is going to have to get
rest, whether because of personal needs or the driving regulations. This is the point where the
problem arises. A driver is a resource who has to get rest, whereas a truck is not. Whenever the
driver is resting, the truck is waiting idle – loaded or unloaded. This idle time is a reality of the
trucking sector and completely unproductive time for the truck. These waiting times are
changing according to the trip length, and the country’s regulations. In this study European
Commission driving time and rest periods regulation will be used as the base rules. (European
Parliament & The Council, 2006).
A simple, skin-deep solution to the truck utilization problem is team driving, where two drivers
stays in the truck and drive the truck by taking turns, and the truck can continue its way while
one of the drivers is resting. However, it is not much preferred simply because the pay is split
between the two drivers and it requires spending extended times with someone else in such a
small space, which can be quite inconvenient. Besides, it still requires both of the drivers to
leave home for long period of times.
1.3 Significance of the Study
Trucking industry is the backbone of many economies and it does not seem like this will change
any time soon. The increasing trend in demand for trucking forces companies to create creative
solutions to meet the demand and retaining the drivers without an unnecessary shift in the cost
and the service fees.
On his paper, Ronen emphasizes that a truck only generates revenue if it is moving loaded. He
describes loading and unloading as compulsory activities and inevitable but says that waiting
is an unproductive activity for the truck and should be minimized (Ronen, 2005).

4
When the driver completes their driving allowance and has to get rest, instead of waiting for
the resting period to end, if we can manage to change the driver with another one and keep the
truck moving, we can diminish the idle time of the truck, if not eliminate completely. At the
same time, since the truck and the load will continue its way instead of waiting idle, a trip length
from A to B can be reduced significantly. Shortened trip times can mean increased customer
satisfaction, as well as increased customer demand. Decreasing the idle time of the trucks not
only increases efficiency ratios but also decreases the number of trucks has to be owned by the
company, as shipments will be fulfilled in shorter times. But of course, this does not help with
the driver turnover problem.
Now that we decided to use more than one driver for a shipment, we have more options to
advance the dispatching network. Once the trailer is swapped, the driver is no longer has to
drive till the final destination. He or she will be free again and waiting to be assigned to another
route. This allows us to control the range of the drivers sending them till to adjacent terminal
instead of the last point. A multi-terminal network is a dispatching method aims to reduce tour
lengths of drivers by making them drive only in their predetermined range. Therefore, when the
driver hands the trailer over, he or she will be assigned to another trip but will not leave their
range. This allows drivers to go home much more frequently and promises better work
conditions.
Recently, another dispatching system called Relay Network (RN) started to appear in literature.
(Vergara H. , 2012)(See Figure 1). It is quite similar to Hub and Spoke (H&S) network used by
LTL carriers. However, in this network, terminals called relay points (RP) are used to swap
trailers between the drivers instead of goods consodilation. When a lane driver reached to a
relay point, he or she hands the completing the route job to another lane driver and takes over
another shipment. Local drivers complete the route. Some large TL carriers are using terminals
for drivers to exchange the trucks and in this way, they are able to operate in smaller regions
and go home more frequently (Campbell, 2005).
Figure 1: Visualization of Relay Network of a TL Carrier
This study will focus on both the truck, and the driver. The proposed approach will create a
terminal network taking into account the maximum kilometres a driver can travel.

5
In order to be able to apply this idea, some important conditions have to be met. To make the
driver change possible, the carrier should keep available drivers close to the routes of the trucks
or drivers should arrive to the point in an allowed time window. Many routes with different
schedules, will require many changing points, in order to keep the truck idle time as short as
possible. Plus, these changing points has to be in proximity to the original routes, thus the
drivers can change the wheel between them without moving too away from the original route.
The main task is assigning the right number of changing points to the right locations with
necessary number of available drivers, and the optimal solution is dependent to many factors.
In literature, this is called dispatching with relay networks and it is a relatively fresh concept in
trucking.
1.4 Scope and Delimitation
In sake of the easiness in modelling, some assumptions have been made. These will be
explained in detail in Title 4: Model Construction.
This study covers only truckload (TL) shipments. Truckload shipments defined as the
shipments with trailers only dedicated to a single shipper’s cargo.
Truck terminals request a fee for the services provided if they are not owned by the shipper
itself, and in case that they are owned, administrative and other regular expenses, e.g. lease
payments apply. However, in this study, the cost of the truck terminals will be ignored. This
limits the authenticity of a possible financial analysis.
In the model, the necessary stops for the fuel, administration works, and maintenance will be
ignored. Also, the European Commission driving time and rest periods will be used, however,
the 45 minutes breaks in between and other exceptions will be disregarded. These delimitations
make our network solution non-applicable directly to a real-life problem, however, they do not
create an obstacle when making a general analysis of the results and comparing the cases.
The remainder of this thesis is organized as follows: The following section gives a
comprehensive review of the related literature to date about the studies carried out on TL
physical network design. The section 3, Methodology, introduces the methods will be used and
explains the reasoning behind. Section 4 includes the mathematical model for the Multi-
Terminal Network Problem. Section 5 presents and gives an analysis of the results obtained.
And finally, the last section is for the conclusions.

6
2 REVIEW OF RELATED LITERATURE AND STUDIES
2.1 TL Network Design
A Simulation-Based Software System for Evaluating Hub-and-Spoke Transportation Networks
& An Integrated Modelling Framework for Evaluating Hub-and-Spoke Networks in Truckload
Trucking
Taylor et al. present the simulation-based software system HUBNET which is designed to help
constructing a TL hub-and-spoke network and evaluate it. In HUBNET, there are three types
of drivers available: Local, lane and non-network drivers. Local drivers are based at the relay
points (terminals), lane drivers drive the trailers between the points and non-network drivers is
for the customers that would exceed the maximum circuity allowed if sent via relay network
constructed. It has three major considerations while deciding on the terminal locations:
-Demand / Volume
-Having ideally an equal distance between the terminals
-Existing Terminals
In an experiment they performed for two networks in USA, their results show great
improvement in the length of haul, however, increased circuity end increased empty miles have
observed compared to point-to-point versions. It is stated that best results have achieved when
a hybrid network is used, i.e. mixture of RN and PtP, instead of relying on relay network only
(Taha & Taylor, 1994) (Taha, Taylor, & Taha, 1996).
Transportation Relay Network Design
In a very similar study carried out by Hunt, which is also one of the pioneering works in this
area, Hunt created an algorithmic model for designing a relay network for TL carriers. He first
creates the classical PtP network and then puts terminals where drivers can swap the loads along
the way, with regular intervals. The model is created through minimal input: Demand,
maximum and minimum distances between the terminals and the physical network (the road
network). The shortest route algorithm tried to assign the minimum number of terminals
throughout each route in order to keep the load changing times and terminal costs at minimum.
Hunt concluded that some loads causes more extra miles then desired when sent through a relay
network thus they might be sent directly (Hunt, 1998).
Development and Analysis of Alternative Dispatching Methods in Truckload Trucking
Taylor et. al discussed several different alternative dispatching methods in their paper,
including a ‘zone’ model aiming to reduce tour lengths of regional drivers. They tested their
methods in a discrete event simulation program, and later on it is implemented in J. B. Hunt.
They stated that, in J. B. Hunt, the turnover rate is 54% for random over the road drivers
while it is only 22% for the regional drivers with converted routes thanks to the implemented
zone dispatching system. (Taylor, Meinert, Killian, & Whicker, 1999).
Strategic Network Design for Motor Carriers
In his article, Campbell gives an analysis of different operations research models for the motor
carrier network design, focusing on network configuration and terminal locations. He states that
some carriers using relay networks, which drivers exchange trailers in the dedicated terminals
in order to decrease the haul length and allow drivers to return home more frequently
(Campbell, 2005).

7
Strategic Network Design for Multi-Zone Truckload Shipments & Strategic Design and
Analysis of a Relay Network in Truckload Transportation
Üster and Maneshwari developed a mathematical model to solve a multi-zone network problem
for TL trucking. They introduced two types of drivers, who are exchanging the trailers at the
relay points. Their motivation is to bring a solution to the chronic truckload trucking problem
driver turnover by decreasing the haul length and allowing drivers to returning home more
often. The constraints they use in the model included:
-Restriction on the additional kilometres made to reach to the relay points
-Restriction on the allowed distances for local and lane drivers
-Assuring equipment balance on the nodes (Üster & Maneshwari, 2007).
A few years later, Üster and Kewcharoenwong modified the model created to handle larger
problem instances. (Üster & Kewcharoenwong, Strategic design and analysis of a relay network
in truckload transportation, 2011)
Optimization Models and Algorithms for Truckload Relay Network Design
In his paper, Hector A. Vergara defines a network with relay points as a relay network and
introduces the TL Relay Network Design (TLRND) problem. TLRND problem is a cost
minimization problem and it calculates the cost of opening Relay Points and dispatching using
them. In this first version of TLRND, every truckload has to go through at least relay point
before reaching to the destination. Vergara divides drivers into two, lane and local and defines
different distance constraints and different pay schemes for both. (Vergara H. , 2012)
Mixed Fleet Dispatching in Truckload Relay Network Design Optimization
In their paper, Hector A. Vergara and Sarah Root proposes a dispatching system which
combines both point-to-point and relay network methods. Till their work, many researchers
suggested that better results can be obtained through a mixed fleet of PtP and RP drivers over
PtP or RP only, but their work is the first to compare the computational results of both
approaches. They are developing a mathematical model to solve the problem which they called
TL relay network design with mixed fleet dispatching (TLRND-MD). The objective function
minimizes the total costs, for this, they are defining fixed costs for installation of Relay Points,
converting a PtP to a Relay Point and kilometers made. Their constraints cover the following:
-Circuity limitations
-Only open RPs can be visited
-Ensuring equipment balances in nodes
-Minimum Volume to be met by RPs
-Maximum number of loads can be dispatched by PtP method.
Their conclude that both systems, relay network only and relay network together with PtP, are
not just lower in costs, but also outperform the traditional PtP-only dispatching system in
industry performance metrics such as average length of haul and service times for the
truckloads. (Vergara & Root, 2013)

8
Robust Truckload Relay Network Design Under Demand Uncertainty
In one of the most recent works in the literate, Vergara and Mokhtari are presenting a
mathematical model for the design of hybrid relay networks (RN and PtP) aiming to handle the
demand fluctuations and adapts to almost any demand value. They are claiming that the
previous research has failed to incorporate the uncertainty in the nature of the problem
(TLRND) to the model and as a first in the area, they are using a robust optimization approach
to create this model. Robust Truckload Relay Network Design with Mixed Fleet Dispatching
(R-TLRND-MD) problem is introduced, where hybrid dispatching system is used (point-to-
point and relay network). The aforementioned three types of drivers, relay points and nodes are
used to create the network. Problem constraints include:
-Restriction on the number of relay points that can be visited
-Restriction on the additional kilometres made to reach to the relay points
-Assuring equipment balance at the nodes
-Minimum volume required to open a Relay Point
-All orders have to be completed (Vergara & Mokhtari, 2014).
Table 3: TL Transport Terminal Network Design Literature Summary
The Study Topic
(Taha & Taylor, 1994) Incorporating H&S with HUBNET simulation-based
software to construct and evaluate hybrid networks
(Taha, Taylor, & Taha, 1996) Incorporating H&S with HUBNET simulation-based
software to construct and evaluate hybrid networks
(Hunt, 1998) Shortest Path Algorithm, Shortest Path Tree Algorithm,
Spring Algorithm and Linear Programming Relaxation
use to locate RPs
(Taylor, Meinert, Killian, &
Whicker, 1999)
Proposing and evaluating hub and lane models using
SIMNET II simulation package
(Üster & Maneshwari, 2007) Mathematical model to multi-zone dispatching problem
– Heuristic and tabu search framework
(Üster & Kewcharoenwong,
2011)
Strategic Relay Network Design– Bender’s
decomposition-based approach
(Vergara H. , 2012) Introduction of TLRND problem - Branch-and-cut
(Vergara & Root, 2013) Mathematical model to TLRND-MD problem
(Vergara & Mokhtari, 2014) Integer Programming model to R-TLRND-MD problem
2.2 Research Hypothesis
It is hypothesized that when a multi-terminal network is used to swap trailers between drivers
en-route, it is possible to decrease truck waiting times compared to PtP dispatching without
compromising on the delivery schedule.
Disciplines and Sub-Disciplines:
Supply Chain Management; Logistics Management; Operations Research.
The research question:

9
Can using a multi-terminal network improve the length of haul times and increase truck
efficiency for the trucking companies?
2.3 Conceptual Framework
A conceptual framework is created to present the research hypothesis. In this framework, we
have three variables: truck utilization, length of haul, and number of terminals visited per
delivery (number of trailer swaps) (see figure 2).
The framework shows that, as truck utilization increases, length of haul will decrease; and the
amount of the decrease is dependent to the number of times the trailer is swapped between the
drivers during the delivery trip.
2.4 Definition of Terms
Hours of Service (HOS): Hours of Service are regulations which limits the hours a driver can
operate a commercial motor vehicle (CMV), it is issued by Federal Motor Carrier Safety
Administration (FMCSA) in United States. In European Union, similar regulations are present
under different names (See Appendix I). Anyone operating a commercial motor vehicle is
obliged to follow these regulations (FMCSA, 2018). This paper presents the mathematical
model complying with the regulations of European Union; and the term HOS is used to refer to
these rules.
Tractor Unit: A tractor unit, also called a road tractor, is a vehicle that is used to pull the trailer,
it is where the driver cabin and the axle is.
Trailer: A trailer is the long part of the truck, in which the freight is carried. It is detachable
from the truck thus allows to leave the goods or pick up pre-loaded trailer.
Truck Terminal (Relay Points): The driver change (trailer swap) will be realized in so called
“truck terminals.” These terminals provide necessary facilities for drivers to rest and meet their
any type of needs, like shower and food. Terminals are also designed for parking trucks for long
times. Terminals operate 24 hours a day, 365 days a year.
Hub and Spoke Network: Hub and spoke network is a dispatching system is a where several
“spokes” connect to a central “hub”. From one spoke to another, the truck has to go through
from the hub in between. In hubs, products from different sources (spokes) are received,
consolidated and sent to the destination. The system is pioneered in airline industry. To give an
Figure 2: Conceptual Framework

10
example, a passenger willing to go to New York from Prague, makes a stop in Frankfurt, here
New York and Prague are nodes, while Frankfurt acts as a hub.

11
3 METHODOLOGY
3.1 Research Design
The multi-terminal network problem calls for operations research, as it is searching the
optimum / close to optimum allocation of our scarce resource: daily driving time of the drivers.
Therefore, an objection function is built.
When the model is solved, it has to give us two certain values for the two variables: the number
of the terminals required and their location. The terminals should be far enough to the departure
point to allow to current driver to drive maximum hours allowed before the mandatory break
and close enough to stop before the mandatory break. Same rule goes for the distance between
terminals and the terminal-destination line. However, we should also consider that a carrier
usually operates a network, not just a line from A to B. Thus, first what we need to do is create
a baseline scenario with source and destination pairs with associated demand.
3.2 Locating Terminals
If we emphasize the utilization of the truck in the objective function, it will keep assigning
the terminals at frequent intervals. The truck utilization is not the same with the driver
utilization. We can change the driver in every hour and the truck can continue its trip without
stopping, but the driver utilization would be way too low.
Since we are determined to keep the truck utilization at 100%, from the initial point to each and
every destination we should build a continuous terminal network to prevent the truck being idle
at any times. Hence, we need to build a terminal network within maximum 4.5 hours driving
distance to each other, like a fence with certain gaps between each fence board (see figure 3).
Blue Circles represent the terminals.
Orange circle represents the initial source.
Figure 3: The Dispatching System from the Trailer’s Perspective

12
3.3 Baseline Scenario Creation
Inputs of the problem are:
-Sources
-Destinations
-Source - destination pairs (lines that the specific load should be moved)
-Number of drivers available (assumed unlimited)
-Number of vehicles available (assumed unlimited)
-Maximum allowed distance between two terminals, source-terminal and terminal-destination
In this research, the algorithm will try to create a multi-terminal network inside Turkey borders.
Turkey is selected to build the multi-terminal network for two reasons, the first reason is that
Turkey almost solely dependent on road transportation. In 2017, the share of road transportation
according to the freight tonnage was 90% and it never dropped below 88% in last 16 years
(Republic of Turkey Ministry Of Transport And Infrastructure, 2018). Second, Turkey is one
of the countries who follows AETR rules on drivers’ hours, which will make the study easily
applicable to other cases where same rules are followed. AETR rules allows maximum 9 hours
of driving time a day, and 56 hours a week and 90 hours in two weeks (GOV.UK, 2018).
According to the Highways Traffic Regulation, tachograph use is obligatory for the whole
duration of driving and violators gets a fine (Trafik Cezalari, 2019).
3.3.1 Source and Destination Points:
As source and destination points, random points from Turkey border gates with its neighbors,
international ports and important city centers are selected (See table 4). You can see the actual
locations of the points on the map (See table 4) (Google Maps, 2019).
Table 4: List of source and destination points

13
Figure 4: Turkey Map with the Source and Destination Points (Google Maps, 2019)
On the map, ports are represented with blue, border gates are represented with yellow and cities
are represented with brown color.
Border gates, international cargo ports are the busiest points of Turkey with regard to their truck
traffic. In addition to these trading points, 5 large cities are selected to create extra flow. These
cities are selected according to their population. From these points, 6 are randomly selected to
create a smaller problem instance (see Table 5). After selection of these points, 11 source-
destination pairs are determined. Between these pairs, potential terminal points are placed
leaving 40 to 60 kilometers from each other on the shortest distance line between the points
according to the Euclidean distance formula.
All of the selected source and destination points are easily reachable by the international routes
in Turkey (See Figure 5) (Republic of Turkey General Directorate of Highways, 2018).
Highways are the preferred roads for the truckload transportation and will be so in this study.
When constructing the graph, exact locations of the selected points in the map are respected.
mst1 = distance between the source and terminal 1
250 km  dst1, dt1t2, dt2d  292.5 km
Source Terminal 1 DestinationTerminal 2
mst
1
mt2dmt1t2
2
Figure 5: The Dispatching System form the Driver's Perspective

14
The figure above illustrates how the dispatching system works from the driver’s perspective.
There is only one type of driver, who drives the distance between the adjacent nodes on the
network. Drivers assumed to be domiciled at the nodes. The driver represented with blue leaves
from the source node with a loaded trailer and takes it to terminal 1. From there the driver takes
another loaded trailer and turn back to the source node in the same day. The driver represented
with red takes over the trailer from terminal 1 and moves to terminal 2. Distance constraints are
applied.
Figure 6: International Road Routes in Turkey (Republic of Turkey General Directorate of Highways, 2018)

15
Table 5:Problem Instance
Set of SD Pairs Distance (km) Source Node Destination Node
FN – KapıkuleSamsun 680,092 29 14
GN - IpsalaSamsun 694,4 15 14
HN - IstanbulSamsun 503,936 5 14
SN - BursaSamsun 514,6 20 14
FE - KapıkuleAnkara 482,484 29 27
GE - IpsalaAnkara 469,464 15 27
HE - IstanbulAnkara 284,704 5 27
SE - BursaAnkara 269,452 20 27
GS - IpsalaBursa 206,584 15 20
FH - KapıkuleIstanbul 198,028 29 5
GH - IpsalaIstanbul 190,712 15 5
3.3.2 Other Considerations
Other considerations while generating the problem instance are the demand, costs and circuity.
3.3.2.1 Demand
The demand in question here is the number of shipments demanded for each source and
destination match, load information and their order according to their priority. Our model is
deterministic as it is assumed that all the orders are known and fixed at the very beginning. One
shipment for each source-destination pair is taken as the demand. No priority is assigned for
the orders.
3.3.2.2 Costs
Designing and implementing a new dispatching method requires some investment. The main
cost pillars to the Shipper are:
- Opening / installing or renting a terminal
- Hiring / laying off drivers
- Changing the routes
- Additional kilometers due to increased number of points to visit during the shipment
In their mixed fleet approach for dispatching, Vergara and Root built their TLRND-MD
problem on cost minimization. Thus, they have assumed costs for each factor. For the fixed
cost of a relay point installation they determined $10,000, 25% extra costs for each route
visiting a RP instead of going directly to the destination, $1.00 pay for every mile made by
local drivers and $1.30 for others (Vergara & Root, 2013). Even though cost assumptions are
necessary for a cost minimization model, there are many other factors that would affect the
costs for the Shipper and demonstrate them all is beyond the abilities of a mathematical
model. First of all, the costs for relay points would change from one location to another
drastically, because different locations have different land values. They can also be different
in sizes, according to the capacity requirements. Secondly, the cost difference of a delivery
with a PtP method and RP method is not just dependent on the kilometers made. In PtP

16
method, one delivery is made with one driver and a truck assigned to them. However, RP
method would require more than one driver who are hired in different locations. This might
require not only change locations but also employees responsible from human resources in
more than one location. Increased operation complexity would be another reason for the
additional workforce. Even though RP method would add more kilometers to the delivery, it
would drastically reduce the time to customer which would result in higher customer
satisfaction. Last but not least, this method is expected to decrease the driver turnover for the
Shipper which would decrease costs of the replacements. Moreover, evaluating fixed and
operational costs in one objective function might lead misleading conclusions especially when
calculated for short term. Therefore, in this work, building the problem as a utilization
maximization model found to be more fitted to the concept and applied accordingly.
3.3.2.3 Circuity
Since terminals are assumed to be installed on the Euclidean distance line between the source
and destination pair, circuity is not a concern for this work.

17
4 MODEL CONSTRUCTION
4.1 Introduction to the Model
4.1.1 Order Fulfilment Process
Below, an example trip for a load can be seen (See Figure 4). The truck leaves from the source
point and starts its route. During its way, depending on the total kilometers, it stops several
times and a different driver takes over the wheel. In the end, the trailer arrives to the pre-defined
destination point.
Note that the trailer stays the same from point the point while the driver changes.
When we are calculating the truck utilization, first we will calculate its potential output and
divide it to the actual output. Consider a shipment which has to go from A to B, and the distance
from A to B is 900 kilometers. If the truck was not dependent on the driver, if would take
900/65= 13,8 hours to arrive to the destination. But since a driver is only allowed to drive for 9
hours a day, the trip will take 13,8 hours plus resting time of the driver till the next day, which
is 13,8 + 15 = 28,8 hours minimum, assuming the driver continues the next day. During the 15
hours, the truck remains idle (stops for driver change / fuel / meals are neglected). Therefore,
the truck utilization for this case:
13,8 / 28,8 *100 = 47,9 %
It is not even half of the capacity of the truck for this route. Since when any route taking more
than 9 hours is assigned to a driver will decrease the truck efficiency, the model aims to keep
the distances according to driver’s allowances. If a terminal is used to swap the trailer with
another driver after first 9 hours, the trip would take 13,8 + 1 (estimated swapping time at the
terminal) = 14,8 hours. Thus, the truck utilization is: (1 hour of waiting time is counted as not-
utilized):
13,8 / 14,8 * 100 = 93,2 %
Source DestinationTerminal 2Terminal 1
Figure 7: Visualization of an Example Trip with Multi Terminal Network

18
4.1.2 Problem Design:
The model considerations are explained below:
1. The distance between any source point to a terminal, between terminals and terminal
to the destination point cannot exceed 292.5 kilometers.*
2. A terminal opened can be used by more than one shipment pair.
3. If the distance between a source and a destination point is smaller than 292.5
kilometers, the driver does not have to stop at a terminal.
4. When a driver has reached to a terminal, the driver swaps the truck and turns back to
initial point.
5. All orders have to be completed.
*According to the European Union (EU) regulations, the daily driving time per driver shall not
exceed nine hours. The total allowed driving time per driver during any two consecutive weeks
restricted to 9 hours. This means that maximum daily distance is 585 kilometers for a driver
(average truck speed = 65), to allow drivers to go home nightly, maximum trip length is limited
to 292.5 kilometers. Additionally, the regulation obliges the driver to give an uninterrupted
break of not less than 45 minutes after a driving period of four and a half hours; but for the
easiness of the model building, this obligation is ignored (U.S. House Committee on Public
Works and Transportation, 1994).
Because of the increased complexity of the model due to many constraints, the model is divided
into two parts. This division is inspired from the path enumeration algorithm with nodes visited
and length constraints (PENVLC) (Vergara H. , 2012). In the first part, all feasible routes form
source to the destination nodes are generated according to the distance constraint. In the second
part, model is solved according to the these allowed feasible routes.
4.2 Route Enumeration
Before enumerating all possible routes, a directed acyclic graph is constructed according to
selected source and destination points. It is important that the graph is directed to determine the
sink nodes (destinations in this model). All the nodes in the graph are a potential terminal point.
The following notation is made for the nodes:
s 1,2…S sources
d 1,2…D destinations
t 1,2…T potential terminal points

19
Number of possible routes for a source-destination pair rapidly increases as the distance
between the point increases. To overcome this difficulty, it is decided to determine the number
of terminals required to be opened between each pair before the enumeration process.
Accordingly, the table below is constructed. For example, if the distance between a source and
destination pair is 600 kilometers, only the routes with 2 terminals will be generated for that
pair. If the distance between a source-destination pair is smaller than or equal to 292.5
kilometers, that pair is considered as PtP appropriate and the route nodes are recorded as source-
destination only.
Table 6: Possible Route Types According to the Distance
Route Type Route Structure Condition
Direct Shipment s d Distance <= 292.5
Stops in 1 Terminal s t d 292.5 < Distance <=585
Stops in 2 Terminals s t t d 585 < Distance <=877.5
Stops in 3 Terminals s t t t d 877.5 < Distance <=1170
Stops in 4 Terminals s t t t t d 1170 < Distance <=1462.5
Stops in 5 Terminals s t t t t t d 1462.5 < Distance <=1775
Stops in 6 Terminals s t t t t t t d 1775 < Distance <=2047.5
The enumeration process is constrained by the number of terminals between each pair and the
distance between the selected points, source-terminal, terminal-terminal and terminal-
destination.
After enumerating all the feasible routes (r), below properties of each route is recorded:
- Set of nodes visited by each route (Mr)
- Route r visits terminal m or not (Hrm)
4.3 Model Formulation
4.3.1 Notation
Sets
R All feasible routes r
P Source-Destination Pairs p
m 1,2…M all the nodes in graph
Rp Feasible routes for the pair p
Mr Set of nodes visited by each route
Parameters
Om One-time fixed cost for the terminal at m
ap Number of shipments required for the pair p
Hrm 1 If route r visits terminal m,
0 else

20
4.3.2 Variables:
Variables
zr Number of the routes r used (integer)
Ym If terminal is built at m or not (binary)
4.3.3 Objective Function
𝑚𝑖𝑛 ∑ 𝑂 𝑚 𝑦 𝑚
𝑀
𝑚=1
4.3.4 Constraints
∑ 𝑧 𝑟 = 𝑎 𝑝 ∀𝑟 ∈ 𝑅 𝑝
𝑅
𝑟=1
(1)
∑ 𝐻𝑟𝑚 𝑧 𝑟 ≤ 𝑎 𝑝 𝑦 𝑚
𝑅
𝑟=1
∀𝑝 ∈ 𝑃, 𝑚 ∈ 𝑀𝑟 ∶ 𝑟 ∈ 𝑅 𝑝 (2)
𝑦 𝑚 ∈ {0,1} ∀{𝑚} (3)
Objective function:
Minimize the total cost of terminals. This is not for any cost analysis, but to allow the model to
select terminals which are used in more than one route, so that in the end, minimum number of
terminals are opened.
Om = 100, but it can be any value as long as it is bigger than zero.
ap = 1, for each pair p.
Constrains:
(1) Demand is met.
(2) Open the terminals in the mentioned feasible route is route is selected.
(3) Binary Variable
4.4 Problem Size
Aforementioned baseline scenario is used as the problem instance (See Table 5). Problem size
for the given instance is below (Table 8):
Table 7: Problem Size
Number of Nodes in Total 30
Number of Source Nodes 4
Number of Destination Nodes 4
Number of SD Pairs 11

21
4.5 Assumptions
In order to building the problem easier, some assumptions have been made. These are:
It is assumed that the trucks will not stop for fueling or battery charge.
No priority is set for the customers.
It is assumed that all the orders are known and fixed at the very beginning of the two-week
period.
It is assumed that no maintenance or repair work is required during these two weeks.
It is assumed that the number of drivers and trucks are unlimited.
It is assumed that drivers domiciled in close proximity to the terminals, source and destination
points.
It is assumed that every order is equivalent, FTL, and every order requires one truck shipment
only.
For the driving hours, European Commission’s regulations are considered. That being said, the
general rule is set as 9 hours of driving time per day per driver, and 45 minutes long small
mandatory breaks every 4 hours are ignored.
Drivers assumed to be domiciled near terminals, sources and destination points.
It is assumed that swapping process takes 1 hour, and incoming shipment leaves the terminal
after one hour.
It is assumed that loading and unloading at the sources and destination points takes 1 hour each.
A truck’s daily driving time capacity is assumed as 24 hours.
It is assumed that installed terminals have unlimited capacity.
Average speed of a truck is taken as 65 km/h. Thus, a driver can travel maximum 585 kilometers
a day (292.5 to go and 292.5 to come back).
Every driver drives solo (no team drivers).

22
5 RESULTS
Route enumeration process is completed on MS Excel 2016 for Mac and used as data for the
optimization problem (Appendix B).
Optimization problem is solved using IBM ILOG CPLEX Optimization Studio 12.9.0 on a
Toshiba 1.5 GHz PC with processor Intel Core i5 and 4 GB of RAM.
Table 8:Results
Pair Route Route Type
FN 29,6,10,14 S T T D
GN 15,6,10,14 S T T D
HN 5,10,14 S T D
SN 20,10,14 S T D
FE 29,6,27 S T D
GE 15,21,27 S T D
HE 5,27 S D
SE 20,27 S D
GS 15,20 S D
FH 29,5 S D
GH 15,5 S D
According to the selected routes, the terminals will be opened in the nodes 6,10 and 21. Only
with three terminals, it is possible to use a terminal-network dispatching system for these
shipments.

23
6 DISCUSSION
6.1 KPIs on Dispatching System’s Success
After solving the model, the scenarios have evaluated with the KPIs below:
• Kilometers between the Source and Destination
• Time between the Source and Destination
• Truck utilization per load
• Driver Utilization per load
• Number of terminals visited per load
First table presents the result of the all 11 shipments, the second table presents the FN source-
destination pair with two terminal stops along the way.
Table 9:Results for the Whole Network
KPI for Whole Shipments Classic Approach Terminal Network
Total Kilometers between the Source and
Destination
4500,456 4500,456
Total Time between the Source and
Destination (hours)
99,14 59,46
Average Truck utilization per load 89% 100%
Average Driver Utilization per load NA 80%
Average Number of terminals visited per
load
None 0,9
Table 10:Results for the FN SD Pair
KPI for one Shipment between FN Pair Classic Approach Terminal Network
Kilometers between the Source and
Destination
680,092 680,092
Time between the Source and Destination
(hours)
25,46
(1 day on the way)
12,46
Truck utilization per load 41% 100%
Average Driver Utilization per load NA 77%
Number of terminals visited per load None 2
Since it is assumed that the terminals are opened on the Euclidean distance line between the
source and destination points, no additional kilometers occurred, thus, the total kilometers for
the pair did not change. However, since the waiting times are added, there is a big gap
between the time to the destination in the classic approach and the multi-terminal network.
Truck utilization is calculated as it is mentioned in the order fulfilment process. Even though
only two of the pairs requires spending one day on the way, the decrease in the utilization can
be seen. When we look at the second table however, where one shipment is compared, the
difference between the utilization values is significant. It should be noted that the more long-
distance shipments are incorporated in the network, the more the difference will be increased.
For the classic shipment method, the driver utilization is not calculated. In multi-terminal
network, we know that the driver goes directly back to their initial node whether there is a
haul back or not, however, for the PtP network, after completing his or her route to the

24
destination, a driver can be assigned with a different route. Thus, driver utilization calculation
for the classic approach is not applicable.
6.2 Advantages and Disadvantages for The Multi-Terminal Network Dispatching
6.2.1 Advantages
Better Working Conditions – Better Turnover Rates
Whether they stop at the terminals or not, drivers have to rest and sleep at some point. Because
of the long distances they make away from home, many drivers sleep in the bunk provided in
the back of the trucks; sometimes in rest areas, sometimes in customer parking lots (Schneider,
2018). Terminal network allow drivers to return their homes more frequently, as they no longer
have to complete the whole route till the destination and wait for another shipment to turn back
home. More time at home means better working life. Better working life perception of the
drivers returns the carrier as a lower turnover rate.
Shorter Trip Times, Faster Deliveries
Many studies prove that using terminals to swap trailers between the drivers shortens the length
of haul drastically. This translates to faster deliveries, and better customer service.
Much easier to Implement than Hub and Spoke network used by LTL carriers
Since TL carriers use terminals only for swapping the trailers among drivers, and no
consolidation is required, any area that allows this can be used as the terminal, thus little to no
installation is necessary.
Increased Equipment Utilization
This method utilizes the trucks to the maximum.
Decreased Deadhead Cost
Usually point to point delivery results in high deadhead cost due to asymmetric flow between
the source and destination nodes. A multi-terminal network yields better results as there are
more routing options for the drivers.
6.2.2 Disadvantages
Scheduling Complexity
Compared to the assigning one truck and one driver to one order, multi-terminal network
method is highly complicated to schedule. Even after deciding on the optimal terminal
locations, it requires in-time tracking systems in all trucks to ensure no delays.
Higher Costs - at first
Extra kilometers added, terminal costs, scheduling expenses, software expenses will add to the
carriers’ bill. These costs can easily be canceled out with the decreasing costs due to better
driver retention.
Not suitable for small carriers with low demand
When a truck needs a driver change, i.e. the current driver completes its 4.5 hours of driving
allowance, there should be a ready driver close by. And at best scenario, when the driver arrives
to the changing station, they should be taking over another shipment without waiting any further

25
to go back. Of course, the larger and the busier the shipper’s network, the easier to achieve a
continuous flow.
Extra Distance for Reaching to the Dedicated Terminal
Even if it is only a few kilometers, the driver has to leave the original route in order to make it
to the dedicated terminal to meet with the driver who will take the shift.
Added Operation Time for Swapping the Trailers
Meeting at the terminals and trailer change adds additional operation time which does not exist
in traditional point-to-point method.
Visibility Considerations
Carriers’ lack of visibility and/or control on drivers’ HOS and scheduling preferences is another
problem that affects the applicability of the multi-terminal network.
Creating a Possible Bottleneck in the System
Like in hub-and-spoke network, many loads will have to cross from the same point. Even
though only thing to do there is to park & rest & swap trailers between the drivers, it is always
a possibility to create a bottleneck in the system in the lack of careful planning.

26
7 CONCLUSIONS
Road transportation is still the most important transportation mode in the world. It is safe to say
that, everything delivered is carried by trucks at least at some point. TL trucking constitutes
nearly 70% of the trucking industry by value (American Trucking Associations, 2018). Driver
retention is a crucial problem waiting to be solved for TL carriers for years. Even though the
increased compensations offered to the drivers, the status does not seem to be changed. To make
being a truckload driver a more preferable and tolerable job, carriers are trying to find ways to
create more home time for the drivers. Hub and spoke dispatching used by less-than-truckload
firms has inspired the Truckload industry to create a brand-new dispatching system: Relay
Network. A network where drivers drive only to the pre-arranged points, and not to the
destination. The literature on this topic is rather new but it delivers promising results.
Considering these factors, a mathematical optimization model has built. A medium size
problem instance is created to solve. The objective function maximizes the driver utilization
ensuring that HOS is honoured. According to the results obtained, it is possible to make length
of haul shorter by letting the drivers go home nightly. It is not just allowing drivers to go home
much often, it is decreasing the length of haul and increasing the truck utilization significantly.

27
8 REFERENCES
American Transportation Research Institute. (2017). An Analysis of the Operational Costs of
Trucking: 2017 Update. Virginia: American Transportation Research Institute.
American Trucking Associations. (2018). Retrieved from trucking.org:
https://www.trucking.org/News_and_Information_Reports_Industry_Data.aspx
American Trucking Organizations. (2018, June 7). American Trucking Organizations.
Retrieved from https://www.trucking.org/article/Turnover-Rate-at-Large-Truckload-
Carriers-Rises-in-First-Quarter
American Trucking Organizations. (2018, March 27). American Trucking Organizations.
Retrieved from American Trucking Organizations:
https://www.trucking.org/article/New-Survey-Data-Reveals-Increases-in-Driver-
Compensation
Bureau of Transportation Statistics. (2017, June). bts.gov. Retrieved from bts.gov:
https://www.bts.gov/archive/publications/transportation_statistics_annual_report/2016
/chapter_3
Bureau of Transportation Statistics. (2017). Freight Facts and Figures 2017. U.S. Department
of Transportation.
Bureau of Transportation Statistics. (2018). bts.gov. Retrieved from bts.gov:
https://www.bts.gov/content/figure-1-north-american-freight-mode-june-2018
Campbell, J. F. (2005). Strategic Network Design for Motor Carriers. Logistics Systems: Design
and Optimization, 245-278.
Corsi, T. M., & Curtis, M. G. (1987). Changes in Owner-Operator Use, 1977-1985:
Implications for Management Strategy. Transportation Journal, 26(3), 4-16.
Corsi, T. M., & Grimm, C. M. (1989). ATFLs: Driving owner-operators into the sunset. Journal
of the Transportation Research Forum, 29(2), 285-290.
European Automobile Manufacturers Association. (2018). European Freight Transport
Statistics: Limitations, Misinterpretations and Aspirations. Brussels: ACEA.
European Parliament & The Council. (2006). Regulation (EC) No 561/2006. EU: EU.
Eurostat. (2018, August 28). http://appsso.eurostat.ec.europa.eu. Retrieved from
http://appsso.eurostat.ec.europa.eu:
http://appsso.eurostat.ec.europa.eu/nui/submitViewTableAction.do
Eurostat. (2018, July 27). Summary of annual road freight transport by type of operation and
type of transport. Retrieved from Eurostat:
http://appsso.eurostat.ec.europa.eu/nui/show.do?dataset=road_go_ta_tott&lang=en
FMCSA. (2018, May 31). Federal Motor Carrier Safety Administration. Retrieved from
https://www.fmcsa.dot.gov/regulations/hours-of-service
GeoDataSource. (2018). https://www.geodatasource.com/world-cities-database. Retrieved
from https://www.geodatasource.com/.

28
Google Maps. (2019, February). Google Maps. Retrieved from Google Maps:
https://www.google.com/maps/d/u/0/edit?mid=1fRxpMGZibkWjcz7euBhxXsk-
_w_Ne9Tz&ll=38.7005112569078%2C35.337052849999964&z=6
GOV.UK. (2018). GOV.UK. Retrieved from GOV.UK: https://www.gov.uk/drivers-hours/aetr-
rules
Hunt, G. W. (1998). Transportation Relay Network Design. Ph.D Thesis.
Üster, H., & Kewcharoenwong, P. (2011). Strategic design and analysis of a relay network in
truckload transportation. Transportation Science, 45(4), 505-523.
Üster, H., & Maneshwari, N. (2007). Strategic network design for multi-zone truckload
shipments. IIE Transactions, 39(2), 177-189.
Lane, L. L. (1987). Innovation in Trucking: Advanced Truckload Firms. Transportation
Research Record(1154), 11-14.
Republic of Turkey Directorate General of Civil Registration and Citizenship Affairs. (2016,
May 20). Republic of Turkey Directorate General of Civil Registration and Citizenship
Affairs. Retrieved from Republic of Turkey Directorate General of Civil Registration
and Citizenship Affairs: https://www.nvi.gov.tr/hakkimizda/projeler/mernis/il-ilce-
kod-tablosu
Republic of Turkey General Directorate of Highways. (2018). Republic of Turkey General
Directorate of Highways. Retrieved from Republic of Turkey General Directorate of
Highways:
http://www.kgm.gov.tr/Sayfalar/KGM/SiteTr/Projeler/UluslararasiProjeler/uluslararasi
YolGuzargahi.aspx
Republic of Turkey Ministry Of Transport And Infrastructure. (2018). Highway Transportation
Statistics. Republic of Turkey Ministry Of Transport And Infrastructure.
Ronen, D. (2005). Dispatching Continuous Moves. Journal of Transportation Management,
25-37.
Schneider. (2018). Retrieved from Schneider: https://schneiderjobs.com/blog/driver/where-do-
truck-drivers-sleep-advice-for-sleep-road
Taha, T. T., & Taylor, G. D. (1994). An Integrated Modelling Framework for Evaluating Hub-
and-Spoke Networks in Truckload Trucking. Logistics and Transportation Review,
30(2), 141-166.
Taha, T. T., Taylor, G. D., & Taha, H. A. (1996). A simulation-based software system for
evaluating Hub-and-Spoke transportation networks. Simulation Practice and Theory,
3(6), 327-346.
Taylor, G. D., Meinert, T. S., Killian, R. C., & Whicker, G. L. (1999). Development and
analysis of alternative dispatching methods in truckload trucking. Transportation
Research Part E, 191-205.
Trafik Cezalari. (2019, January 1). Trafik Cezalari. Retrieved from Trafik Cezalari:
http://trafikcezalari.web.tr/trafik-cezasi-madde-49-3/
Transport Topics. (2018, June 8). Transport Topics. Retrieved from
https://www.ttnews.com/articles/driver-turnover-puts-pressure-large-carriers-0

29
U.S. House Committee on Public Works and Transportation. (1994). Legislation to preempt
state motor carrier regulations pertaining to rates, routes, and services. Washington:
U.S. Government Printing Office.
U.S. House of Representatives. (2018). transportation.house.gov/about. Retrieved from
transportation.house.gov: https://transportation.house.gov/about/history.htm
USA Social Security Administration. (2012, June 1). Retrieved from Secure.ssa.gov:
http://policy.ssa.gov/poms.nsf/lnx/0302101249
Vergara, H. (2012). Optimization Models and Algorithms for Truckload Relay Network
Design. Theses and Dissertations.
Vergara, H. A., & Mokhtari, Z. (2014). Robust Truckload Relay Network Design Under
Demand Uncertainty. 13th IMHRC Proceedings. Ohio: Georgia Southern University.
Vergara, H. V., & Root, S. (2013). Mixed Fleet Dispatching in Truckload Relay Network
Design Optimization. Transportation Research Part E(54), 32-49.
Walther, S. (2010). Industrializing Transportation Networks: An Analysis and Evaluation of
Alternative Approaches to Cargo Transport Services with Special Reference to the
European Over-the-road Truckload Carrier Industry. Shaker.
Winston, C. (1998). U.S. Industry Adjustment to Economic Deregulation. Journal of Economic
Perspectives, 12(3), 89-110.
Ying, J. S. (1990, May). The Inefficiency of Regulating a Competitive Industry: Productivity
Gains in Trucking Following Reform. The Review of Economics and Statistics, 72(2),
191-201.

30
9 APPENDICES
9.1 Appendix A
REGULATION (EC) No 561/2006 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL
of 15 March 2006
on the harmonization of certain social legislation relating to road transport and amending Council
Regulations (EEC) No 3821/85 and (EC) No 2135/98 and repealing Council Regulation (EEC)
No 3820/85
(Text with EEA relevance)
CHAPTER I
INTRODUCTORY PROVISIONS
Article 1
This Regulation lays down rules on driving times, breaks and rest periods for drivers engaged in
the carriage of goods and passengers by road in order to harmonize the conditions of competition
between modes of inland transport, especially with regard to the road sector, and to improve
working conditions and road safety. This Regulation also aims to promote improved monitoring and
enforcement practices by Member States and improved working practices in the road transport
industry.
Article 2
1. This Regulation shall apply to the carriage by road:
(a) of goods where the maximum permissible mass of the vehicle, including any trailer, or semi-trailer,
exceeds 3,5 tones, or
(b) of passengers by vehicles which are constructed or permanently adapted for carrying more than nine
persons including the driver and are intended for that purpose.
2. This Regulation shall apply, irrespective of the country of registration of the vehicle, to carriage
by road undertaken:
(a) exclusively within the Community; or
(b) between the Community, Switzerland and the countries party to the Agreement on the European
Economic Area.
3. The AETR shall apply, instead of this Regulation, to international road transport operations
undertaken in part outside the areas mentioned in paragraph 2, to:

31
(a) vehicles registered in the Community or in countries which are contracting parties to the AETR, for the
whole journey;
(b) vehicles registered in a third country which is not a contracting party to the AETR, only for the part of
the journey on the territory of the Community or of countries which are contracting parties to the
AETR.
The provisions of the AETR should be aligned with those of this Regulation, so that the main
provisions in this Regulation apply, through the AETR, to such vehicles for any part of the journey
made within the Community.
Article 3
This Regulation shall not apply to carriage by road by:
(a) vehicles used for the carriage of passengers on regular services where the route covered by the
service in question does not exceed 50 kilometers;
(aa) vehicles or combinations of vehicles with a maximum permissible mass not exceeding 7,5 tones used
for carrying materials, equipment or machinery for the driver’s use in the course of his work, and
which are used only within a 100 km radius from the base of the undertaking and on the condition
that driving the vehicle does not constitute the driver’s main activity;
(b) vehicles with a maximum authorized speed not exceeding 40 kilometers per hour;
(c) vehicles owned or hired without a driver by the armed services, civil defense services, fire services,
and forces responsible for maintaining public order when the carriage is undertaken as a
consequence of the tasks assigned to these services and is under their control;
(d) vehicles, including vehicles used in the non-commercial transport of humanitarian aid, used in
emergencies or rescue operations;
(e) specialized vehicles used for medical purposes;
(f) specialized breakdown vehicles operating within a 100 km radius of their base;
(g) vehicles undergoing road tests for technical development, repair or maintenance purposes, and new
or rebuilt vehicles which have not yet been put into service;
(h) vehicles or combinations of vehicles with a maximum permissible mass not exceeding 7,5 tones used
for the non-commercial carriage of goods;
(i) commercial vehicles, which have a historic status according to the legislation of the Member State in
which they are being driven and which are used for the non-commercial carriage of passengers or
goods.
Article 4

32
For the purposes of this Regulation the following definitions shall apply:
(a) ‘carriage by road’ means any journey made entirely or in part on roads open to the public by a vehicle,
whether laden or not, used for the carriage of passengers or goods;
(b) ‘vehicle’ means a motor vehicle, tractor, trailer or semi-trailer or a combination of these vehicles,
defined as follows:
— ‘motor vehicle’: any self-propelled vehicle travelling on the road, other than a vehicle permanently
running on rails, and normally used for carrying passengers or goods,
— ‘tractor’: any self-propelled vehicle travelling on the road, other than a vehicle permanently running
on rails, and specially designed to pull, push or move trailers, semi-trailers, implements or
machines,
— ‘trailer’: any vehicle designed to be coupled to a motor vehicle or tractor,
— ‘semi-trailer’: a trailer without a front axle coupled in such a way that a substantial part of its weight
and of the weight of its load is borne by the tractor or motor vehicle;
(c) ‘driver’ means any person who drives the vehicle even for a short period, or who is carried in a vehicle
as part of his duties to be available for driving if necessary;
(d) ‘break’ means any period during which a driver may not carry out any driving or any other work and
which is used exclusively for recuperation;
(e) ‘other work’ means all activities which are defined as working time in Article 3(a) of
Directive 2002/15/EC except ‘driving’, including any work for the same or another employer, within
or outside of the transport sector;
(f) ‘rest’ means any uninterrupted period during which a driver may freely dispose of his time;
(g) ‘daily rest period’ means the daily period during which a driver may freely dispose of his time and
covers a ‘regular daily rest period’ and a ‘reduced daily rest period’:
— ‘regular daily rest period’ means any period of rest of at least 11 hours. Alternatively, this regular
daily rest period may be taken in two periods, the first of which must be an uninterrupted period of
at least 3 hours and the second an uninterrupted period of at least nine hours,
— ‘reduced daily rest period’ means any period of rest of at least nine hours but less than 11 hours;
(h) ‘weekly rest period’ means the weekly period during which a driver may freely dispose of his time and
covers a ‘regular weekly rest period’ and a ‘reduced weekly rest period’:
— ‘regular weekly rest period’ means any period of rest of at least 45 hours,
— ‘reduced weekly rest period’ means any period of rest of less than 45 hours, which may, subject to
the conditions laid down in Article 8(6), be shortened to a minimum of 24 consecutive hours;

33
(i) ‘a week’ means the period of time between 00.00 on Monday and 24.00 on Sunday;
(j) ‘driving time’ means the duration of driving activity recorded:
— automatically or semi-automatically by the recording equipment as defined in Annex I and Annex IB
of Regulation (EEC) No 3821/85, or
— manually as required by Article 16(2) of Regulation (EEC) No 3821/85;
(k) ‘daily driving time’ means the total accumulated driving time between the end of one daily rest period
and the beginning of the following daily rest period or between a daily rest period and a weekly rest
period;
(l) ‘weekly driving time’ means the total accumulated driving time during a week;
(m) ‘maximum permissible mass’ means the maximum authorized operating mass of a vehicle when fully
laden;
(n) ‘regular passenger services’ means national and international services as defined in Article 2 of
Council Regulation (EEC) No 684/92 of 16 March 1992 on common rules for the international
carriage of passengers by coach and bus ( 1 );
(o) ‘multi-manning’ means the situation where, during each period of driving between any two consecutive
daily rest periods, or between a daily rest period and a weekly rest period, there are at least two
drivers in the vehicle to do the driving. For the first hour of multi-manning the presence of another
driver or drivers is optional but for the remainder of the period it is compulsory;
(p) ‘transport undertaking’ means any natural person, any legal person, any association or group of
persons without legal personality, whether profit-making or not, or any official body, whether having
its own legal personality or being dependent upon an authority having such a personality, which
engages in carriage by road, whether for hire or reward or for own account;
(q) ‘driving period’ means the accumulated driving time from when a driver commences driving following
a rest period or a break until he takes a rest period or a break. The driving period may be continuous
or broken.
CHAPTER II
CREWS, DRIVING TIMES, BREAKS AND REST PERIODS
Article 5
1. The minimum age for conductors shall be 18 years.
2. The minimum age for drivers' mates shall be 18 years. However, Member States may reduce
the minimum age for drivers' mates to 16 years, provided that:

34
(a) the carriage by road is carried out within one Member State within a 50 kilometers radius of the place
where the vehicle is based, including local administrative areas the center of which is situated within
that radius;
(b) the reduction is for the purposes of vocational training; and
(c) there is compliance with the limits imposed by the Member State's national rules on employment
matters.
Article 6
1. The daily driving time shall not exceed nine hours.
However, the daily driving time may be extended to at most 10 hours not more than twice during
the week.
2. The weekly driving time shall not exceed 56 hours and shall not result in the maximum weekly
working time laid down in Directive 2002/15/EC being exceeded.
3. The total accumulated driving time during any two consecutive weeks shall not exceed 90 hours.
4. Daily and weekly driving times shall include all driving time on the territory of the Community or
of a third country.
5. A driver shall record as other work any time spent as described in Article 4(e) as well as any
time spent driving a vehicle used for commercial operations not falling within the scope of this
Regulation, and shall record any periods of availability, as defined in Article 15(3)(c) of
Regulation (EEC) No 3821/85, since his last daily or weekly rest period. This record shall be
entered either manually on a record sheet, a printout or by use of manual input facilities on
recording equipment.
Article 7
After a driving period of four and a half hours a driver shall take an uninterrupted break of not less
than 45 minutes, unless he takes a rest period.
This break may be replaced by a break of at least 15 minutes followed by a break of at least 30
minutes each distributed over the period in such a way as to comply with the provisions of the first
paragraph.
Article 8
1. A driver shall take daily and weekly rest periods.
2. Within each period of 24 hours after the end of the previous daily rest period or weekly rest
period a driver shall have taken a new daily rest period.

35
If the portion of the daily rest period which falls within that 24-hour period is at least nine hours but
less than 11 hours, then the daily rest period in question shall be regarded as a reduced daily rest
period.
3. A daily rest period may be extended to make a regular weekly rest period or a reduced weekly
rest period.
4. A driver may have at most three reduced daily rest periods between any two weekly rest periods.
5. By way of derogation from paragraph 2, within 30 hours of the end of a daily or weekly rest
period, a driver engaged in multi-manning must have taken a new daily rest period of at least
nine hours.
6. In any two consecutive weeks a driver shall take at least:
— two regular weekly rest periods, or
— one regular weekly rest period and one reduced weekly rest period of at least 24 hours. However,
the reduction shall be compensated by an equivalent period of rest taken en bloc before the end of
the third week following the week in question.
A weekly rest period shall start no later than at the end of six 24-hour periods from the end of the
previous weekly rest period.
6a. By way of derogation from paragraph 6, a driver engaged in a single occasional service of
international carriage of passengers, as defined in Regulation (EC) No 1073/2009 of the European
Parliament and of the Council of 21 October 2009 on common rules for access to the international
market for coach and bus services ( 2 ), may postpone the weekly rest period for up to 12
consecutive 24-hour periods following a previous regular weekly rest period, provided that:
(a) the service lasts at least 24 consecutive hours in a Member State or a third country to which this
Regulation applies other than the one in which the service started;
(b) the driver takes after the use of the derogation:
(i) either two regular weekly rest periods; or
(ii) one regular weekly rest period and one reduced weekly rest period of at least 24 hours. However, the
reduction shall be compensated by an equivalent period of rest taken en-bloc before the end of the
third week following the end of the derogation period;
(c) after 1 January 2014, the vehicle is equipped with recording equipment in accordance with the
requirements of Annex IB to Regulation (EEC) No 3821/85; and
(d) after 1 January 2014, if driving during the period from 22,00 to 06,00, the vehicle is multi-manned, or
the driving period referred to in Article 7 is reduced to three hours.

36
The Commission shall monitor closely the use made of this derogation in order to ensure the
preservation of road safety under very strict conditions, in particular by checking that the total
accumulated driving time during the period covered by the derogation is not excessive. By
4 December 2012, the Commission shall draw up a report assessing the consequences of the
derogation in respect of road safety as well as social aspects. If it deems it appropriate, the
Commission shall propose amendments to this Regulation in this respect.
7. Any rest taken as compensation for a reduced weekly rest period shall be attached to another
rest period of at least nine hours.
8. Where a driver chooses to do this, daily rest periods and reduced weekly rest periods away from
base may be taken in a vehicle, as long as it has suitable sleeping facilities for each driver and the
vehicle is stationary.
9. A weekly rest period that falls in two weeks may be counted in either week, but not in both.
Article 9
1. By way of derogation from Article 8, where a driver accompanies a vehicle which is transported
by ferry or train, and takes a regular daily rest period, that period may be interrupted not more than
twice by other activities not exceeding one hour in total. During that regular daily rest period the
driver shall have access to a bunk or couchette.
2. Any time spent travelling to a location to take charge of a vehicle falling within the scope of this
Regulation, or to return from that location, when the vehicle is neither at the driver's home nor at
the employer's operational center where the driver is normally based, shall not be counted as a
rest or break unless the driver is on a ferry or train and has access to a bunk or couchette.
3. Any time spent by a driver driving a vehicle which falls outside the scope of this Regulation to
or from a vehicle which falls within the scope of this Regulation, which is not at the driver's home
or at the employer's operational center where the driver is normally based, shall count as other
work.

37
9.2 Appendix B
SD Pair Set of Routes R Set of Nodes Visited by ri
FN r1 29 2 8 14
FN r2 29 3 8 14
FN r3 29 3 9 14
FN r4 29 4 8 14
FN r5 29 4 9 14
FN r6 29 4 10 14
FN r7 29 5 8 14
FN r8 29 5 9 14
FN r9 29 5 10 14
FN r10 29 5 11 14
FN r11 29 6 8 14
FN r12 29 6 9 14
FN r13 29 6 10 14
FN r14 29 6 11 14
FN r15 29 6 12 14
GN r16 15 16 8 14
GN r17 15 17 8 14
GN r18 15 17 9 14
GN r19 15 30 8 14
GN r20 15 30 9 14
GN r21 15 30 10 14
GN r22 15 5 8 14
GN r23 15 5 9 14
GN r24 15 5 10 14
GN r25 15 5 11 14
GN r26 15 6 8 14
GN r27 15 6 9 14
GN r28 15 6 10 14
GN r29 15 6 11 14
GN r30 15 6 12 14
GN r31 15 7 8 14
GN r32 15 7 9 14
GN r33 15 7 10 14
GN r34 15 7 11 14
GN r35 15 7 12 14
GN r36 15 7 13 14
HN r37 5 8 14
HN r38 5 9 14
HN r39 5 10 14
HN r40 5 11 14

38
SN r41 20 8 14
SN r42 20 9 14
SN r43 20 10 14
SN r44 20 11 14
FE r45 29 6 27
GE r46 29 21 27
HE r47 5 27
SE r48 20 27
GS r49 15 20
FH r50 29 5
GH r51 15 5
i
The committee is currently known as The Committee on Transportation and Infrastructure (U.S. House of
Representatives, 2018).

Strategic Dispatching System Design for Truckload Transportation

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Strategic Dispatching System Design for Truckload Transportation

Similar to Strategic Dispatching System Design for Truckload Transportation (20)

More from Merve Nur Taş

More from Merve Nur Taş (19)

Recently uploaded

Recently uploaded (20)

Strategic Dispatching System Design for Truckload Transportation