Interactive Dynamic Modeling for the Passenger
Flow Bottleneck and Security Checkpoint
Management at an Airport
Yaman BARLAS
M. Serdar CAKACIASLAN
Tolga SAGLIK
Pinar SEKE
Mustafa UGUR
Bogazigi University
Industrial Engineering Department
34342 Bebek Istanbul, Turkey
SESDYN Group: http://www.ie.boun.edu.tr/labs/sesdyn/
Abstract
In this paper, the current structure of the passenger flows in Atatiirk Airport is examined,
aiming an impr in the of flows. It is observed that security
checkpoints are the main causes of the bottlenecks in the passenger flows. In this regard,
passenger flows and their relations with security checkpoints/personnel are modeled
dynamically. The model is set up based on two main flows through the domestic and the
international terminals. After analyzing real hourly flow data of both international and
domestic terminals and estimating their statistical properties, results are used in the model as
input data. Validity of the model is tested under various extreme conditions, against real data
and under different scenarios. In initial simulations, number of active x-rays devices and
personnel needed are formulated as dynamic inputs. After ob
results, an algorithm is developed to distribute necessary number of active x-rays/personnel
to each security checkpoint. Finally, for the airport managers to test their own strategies, a
game version of the model is built. It is expected that by using the model and the simulation
game, decision-making structure of security personnel allocation at the airport will be
improved.
these initial l
Key Words: dynamic system modeling, queue management, airport security checkpoints.
Acknowledgments
This research is supported by Bogazici University Research Grant no 8880.
INTRODUCTION
Atatiirk Airport in Istanbul is the largest airport with the densest air traffic in Turkey. In
comparison with the last year, this year’s five-month-growth is increased by nineteen percent.
Actually, this increase represents the density and also warns about potential problems in
terminals. In addition, the airport has already several critical bottlenecks. As the number of
passengers using this airport increases, these bottlenecks may create some problems. By this
project, we expect to find these possible problems out before the airport management faces
them. Key factors of the problem can be listed as:
e Arrival rates of the entrance points of the airport
© Queue lengths of five checkpoints.
e Total number of x-ray devices available
e Number of desired and active x-rays at these checkpoints.
This project aims to increase the efficiency of the resources of the airport such as x-ray
devices and security personnel, so that there can be less people waiting with the same number
of x-rays. There is a tradeoff between queue lengths and the number of x-rays. The more x-
rays are used, the more security personnel are hired that creates an increase in security
expenses. If the company keeps the number of x-rays low, queue lengths would be longer,
people would wait and the chance of missing their flights would increase, resulting in
customer dissatisfaction.
There are five main checkpoints at the airport and in the model. The goal is to allocate x-ray
devices to these checkpoints in balance, in order to keep queues in acceptable boundaries.
Total number of x-ray devices, i.e. the x-ray capacity of the airport is another key factor at
this point because the active x-rays cannot exceed the physical limits of the airport.
It should be stated that this project is only concerned with departure sections of terminals
because passengers landed do not pass through security checkpoints. Before starting to
explain the model, we should picture the airport structure (departments, gates, security
checkpoints, etc.). There are two main terminals, domestic and international. International
terminal constitutes most portion of the airport with its high passenger flow and huge number
of flights. Thus it is better to begin with international terminal structure.
International Terminal
Atatiirk Airport international terminal departure entrance consists of three different floors.
After people enter the terminal from these entrances, they have to pass through international
departure security checkpoints (primary security). Then, they continue with check-in areas, to
passport desks. Those coming from domestic flights (domestic-to-international transfer) have
to go to passport desks too. With the completion of passport control process, they all have to
pass through airside security checkpoints (secondary security) and passengers make their way
to boarding gates. On the other hand, those coming from international flights (international-
to-international transfer) do not have to pass through passport control. Atatiirk airport has
separate security checkpoint for these passengers. After they are checked in these security
checkpoints, they directly go to their planes to depart.
2
Domestic Terminal
This terminal entrance consists of two different floors. After people enter the terminal from
these entrances, they have to pass through domestic departure security checkpoints (primary
security). Then, they continue with check-in areas to gates. Since the transfer rates of
international-to-domestic and domestic-to-domestic are relatively small, they are negligible.
Before entering the gate areas, all passengers have to pass through airside security
checkpoints (secondary security) and passengers make their way to boarding gates. After they
are checked in these security checkpoints, they directly go to their planes to depart.
RESEARCH OBJECTIVES AND OVERVIEW OF THE MODEL
The purpose of this project is: modeling and analysis of the main passenger flows of the
Atatiirk Airport, by investigating the passenger accumulations at the main security
checkpoints, and eventually developing an algorithmic model to obtain a more efficient way
of distributing the security personnel in the airport, and finally developing a web based game
to allow managers test different decision making policies. Through these objectives, it is
possible to minimize the number of passengers waiting in the security checkpoint queues and
lower the budget planned to be spent for security related costs.
As a research methodology, system dynamics modeling is chosen to represent the general
security structure and passenger flows. Although the input data of the model are deterministic,
in the real world the process is highly stochastic. It is certain that the number of incoming
passengers to the airport can be interpreted more or less accurately through the planned
departure times of scheduled flights, however the expected and actual departure times of the
scheduled flights differ in many cases. Another probabilistic factor is the transfer passengers.
Again due to the same problematic delay structure of arriving flights to the airport, it is hard
to guess the exact time of entrance of transfer passengers to the terminal. For these reasons, it
is not suitable to use a deterministic optimization approach in this case. Furthermore if a
discrete event simulation was used, it would be obligatory to model every entering passenger
to the airport and keep the personal statistics. This process is time consuming, tedious and
creates unnecessary personnel statistics, rather than the general behavior of the working
system. In short, system dynamics method is chosen as the most suitable research method.
The aim of this study is to construct a model that first describes the distribution structure of
the security personnel and then optimizes it with the help of an algorithm. Time unit is
selected as a minute in order to observe meaningfully the dynamic behaviors of the passenger
flows and accumulation processes in the airport. Time horizon, on the other hand, is selected
as a day (in some runs it is selected as three or four days as needed). This time horizon is
sufficient and appropriate to analyze the periodic changes in the system.
The model consists of three main sectors: passenger flows, security personnel distribution and
the algorithmic sector. In the passenger flows sector, the entrance of passengers to the system,
accumulation processes at different areas and exiting from the security related parts of the
airport are modeled. The second sector of the model is security personnel distribution in
which the movement of the security personnel between the security checkpoints is modeled.
The final algorithmic part interrelates the other two parts. In other words, how many security
personnel are required is decided dynamically according to passenger flows and
accumulations at each security checkpoints. In the model, the security checkpoints are
modeled as five nodes: these are domestic terminal departure entrance security checkpoint,
3
domestic terminal airside security checkpoint, international terminal entrance security
checkpoint, international terminal airside security checkpoint and international terminal
transfer line security checkpoint. All of these nodes have the same structure and variables.
The nodes consist of some sections of three main parts of the model.
DESCRIPTION OF THE MODEL
As it is stated before the model consists of three main parts: passenger flows, security
personnel distribution and algorithmic parts. In this section, how they are constructed, which
assumptions are made and how they interact together are explained.
Passenger Flows Segment
This part includes entrance of the passengers, their accumulation in certain stocks and the
time they spend and finally their exit from the system. Boundary of the system is determined
as between the terminal entrance and the secondary (airside) security checkpoint because after
that point passengers are not checked again and there is no need to model that part of the
airport. However this procedure varies in different terminals. The airport consists of two
terminals: domestic and international. Their departure sections are modeled in this project.
There are also domestic-to-domestic and international-to-international transfer lines in the
corresponding terminals, but they are located in different sections of the terminal and also
their inputs are completely different. Thus these transfer lines should be modeled separately.
In addition to that domestic-to-domestic transfer line passenger flow density is negligible so it
is not included in the model.
Domestic Departure Terminal
Passenger inflow to the Domestic departure terminal is input as a graphical function. In this
function, reliable data from Atatiirk Airport is used. Incoming passenger firstly accumulate in
a stock called "Domestic Departure Entrance Security Checkpoint". Their outflow rate from
this stock is depend on number of passengers in this stock and the number of active x-rays
that are used at that moment in the security checkpoint. In this process the average processing
time of an x-ray assumed constant and calculations are made by this value always. The rate is
named as "Domestic Departure Terminal Entrance Rate".
After passing the security checkpoint, passengers spend some time in a common are at the
terminal before they continue to boarding gates. In order to represent the time they spend in
this common area, a third order material delay is used. A third order material delay consists of
three stocks, three outflows and three average times values to calculate the outflow rates.
Names of these three stocks are stated as "Pre Check-In Period", "Check-In Period" and "
After Check-In Period", and the corresponding average times to be spent in these stocks are
assumed 5, 20 and 5 minutes respectively.
Finally passengers reach to the airside security checkpoint after this third order material delay.
In this stock, the principle is the same as in the entrance security checkpoint. Passenger
outflow rate is calculated by average processing time of an x-ray, number of active x-rays and
the number of passengers in the security checkpoint.
Number of Adtive Xray's ‘Avg Time Spent in
Number of Active Xays
inDom Dep fnterance ‘Dom Pre Checkin
in Dom Airside Securty
Avg Time Spent in Dom Chickin _‘AV9 Tne Spent in
Dom After Checkin
Dom Dep Enterance ieee hak
Dm After Checkin
wy
f Dom Common Flow 1
Dom Dep Inflow Rate Dom Dep Enteranod Rate
‘Avg Processing Time of an ‘Avg Processing Time of an
Xray in Dom Dep Enterance ray in Dom Airside Security
Figure 1: Structure of the Domestic Departure Terminal in the Model
International Departure Terminal
The structure of the international terminal is almost the same as the domestic. The main
difference is the passport control stock. In the domestic terminal, passengers proceed to the
airside security checkpoint after they spend some time in the common area. In the
international terminal, however they have to go first to passport control, then to airside
security checkpoint. In addition to the inflow from the common area, passport control stock
has another inflow from the domestic terminal; it is called as "Domestic-to-International
Transfer Flow". The meaning of this flow is that some passengers came from another airport
from Turkey, but they are going to continue their journey to abroad. Hence they first come to
the domestic arrival terminal and then they have to be transferred to international departure
terminal, and their joining point is modeled as passport control stock. The outflow from the
stock is dependent on the average passport control time per person and it is also assumed
constant.
The other only difference from the domestic terminal is that, average passing time of the after
Check-In period is assumed 15 rather than 5 because passengers spend more time in the
international terminal in order to buy a duty stamp for abroad trips.
‘vg Processing Tee of an Avg Te Speat in ‘Avg Tne Spent in|
ay in nt Dep|Enerance Int Pre Chock ‘Avg Time Spent in IK Checkin’ Int ater Checkin
Int Dep Enternce Int After Checkin
‘Avg Passapot Contr)
Time per Person
‘Avg Processing Time of
Day in In Aside Secuty
Ine Pre Che
Lent cha See Number of Active Xeays
in nt Dep Enterance Inn Aiside Security
Figure 2: Structure of the International Departure Terminal in the Model
Another necessary part to be modeled is the international transfer line security. In this line
there are passengers come from abroad and going to leave the country again. They are not
mixed in the common area of the terminal with the other passengers; alternatively their
security check is done in this checkpoint. The stock and flow principles are again the same as
before. The inflow of this stock is another graphical input; the required data is also from the
airport statistics.
Number of Active Xrays in
Int Transfer Line Security
Avg Processing Tim&of an Xray
in Int Transfer Line
Int Transfer Line —=T + Int to Int Transfer Rate
Security Outflow Rate Int Transfer Line Security
Figure 3: Structure of the International Transfer Line Security Checkpoint in the Model
Security Personnel Distribution Segment
In this part the distribution process is represented. As it is stated before, the model consists of
five main security checkpoints. Number of security personnel at each checkpoint and the
remaining unallocated personnel is modeled as stocks. The number of security personnel in
the stock at a security checkpoint determines directly the number of active x-rays at that
checkpoint. The distribution process works as follows. First the algorithm determines the
difference between the desired and the actual number of personnel at each security
checkpoint. If the difference is positive, in other words there is a need for more personnel in a
checkpoint, required number of personnel come to that stock from unallocated stock in a
determined time called security personnel inflow time. In this model all of these inflow times
are assumed seven minutes.
On the other hand, if the difference is negative meaning that more than enough number of
security personnel is present at the checkpoint. In that case, undesired number of personnel is
immediately sent to the unallocated stock. Hence if a personnel goes to a security checkpoint,
s/he may come from any of other checkpoints and the inflow time is like an average time to
spend between any two checkpoints.
Finally it is also possible to change the total number of personnel during the run. It is possible
by making the total number of security personnel variable as a graphical function. When total
number is changed, necessary adjustment is made by a bi-flow between cloud (outside of the
system) and unallocated number of personnel stock. Hence the new entering personnel can
reach a security checkpoint again in seven minutes.
OC
Seeury ton Tne
toin fie Seery
unter Posen unter of Parson
Dit Seine ma os in ‘let aa Scaty Number of Persson
nim Dep Enerce
IntDep Entrance | Irs Secu ra Taare Secy
}
Di of Penn nt
Pet
\ peeaaa Tele rset
Pasco flow Tne {
a . A
‘Change in the Total eS
Parsi tn
Taner Le Secuty
Pec low Tne
Trae ne Scary
Pars ion Tine
{0 Donate Securty
caine: | jouitiovaly
Dom aiid Sent
Damp Ea y
Dom aiid Scuy of Paso in
om et Sct
Tai Nurbr of Perea
Dé ot Perna
‘om Dep Esterance
Natt Pa
inom Dep Ene
Figure 4: Structure of the security personnel distribution part in the mode
Algorithmic Segment
This part describes how the model decides the number of active x-rays and also number of
security personnel at each node (i.e. security checkpoint). The algorithm needs total number
of x-rays and goal queue length per x-ray as inputs at each node. For instance, in International
transfer line security checkpoint there are eight x-ray devices can be used at maximum. These
values are varying between each node. As a secondary input, a goal queue length should be
entered. The algorithm tries to keep the number of passengers per x-ray at that particular
value. In other words, it should activate more x-rays if the actual queue length is more than
the goal. By multiplying the goal queue length of an x-ray and the number of x-rays at that
node, the base queue length of that security checkpoint is obtained. Then, average queue
lengths are calculated. The need coefficients are the ratio of average queue lengths and the
base queue lengths at each node. After that desired number of x-rays at each node is
calculated by multiplying the need coefficients with the total number of x-rays at each node.
This process is done at each node individually, and then total desired number of active x-rays
is calculated, if the number of security personnel is enough for that value available percentage
of desired number of active x-rays stays as one, on the other hand if they are not enough
available percentage is calculated as dividing total available number of x-rays (which is
calculated according to total number of personnel and needed number of personnel to activate
an x-ray) by total desired number of x-rays. How many more (or less) x-ray device is needed
can be calculated, by using this percentage, desired number of active x-rays and the actual
value of active x-rays. After that the difference of working personnel in each node is
calculated via these values and personnel distribution can be done.
Goal Queue Length per
Xray in Int Dep Enterance
Base Queue Length
of Int Dep Enterance
‘Smoothing Time 1
Desired Number of Activ’
rays in Int Dep Enterance
Total Number of Xray
in Int Dep Enterance
In\Qep Enterance
ii Nirmal Wars fit Dep Enterance Rate
in Int Dep Enterance Number of Personnel
in Int Dep Enterance
Number of Retive Xrays
in Int Dep Enterance
Available Percentage of Desired
Number of Active Xrays Needed Personnel per
Xray in Int Dep Enterance
Figure 5: Structure of the algorithmic part in one of the five nodes in the model
MODEL VALIDATION AND OUTPUT ANALYSIS
In order to validate the model, first a steady state analysis and extreme condition tests are
made. Model behaves as expected under both conditions; stocks stay at equilibrium levels in
the steady state analysis and increase gradually in the extreme condition cases such as very
high arrival rates or very low available number of personnel.
Furthermore sensitivity and scenario analysis are made to observe that how the parameter
value changes affect the behavior of the model output. In conclusion of the investigation
critical values for the model parameters are found.
Comparison of the Base Model and Algorithmic Model Outputs
In this section the results of two models, the base model and the model with the distribution
algorithm are compared in terms of the number of people waiting in the queues and the
number of active x-rays at any time. The reason for comparing the number of x-rays is to
show that the model with algorithm uses fewer x-rays than the base model under same input
data.
There are two different simulation runs made for both models in terms of weekdays. First one
is the simulation for Friday, Saturday, Sunday and Monday. On these days air traffic is more
intense than other weekdays. Second simulation is for Tuesday, Wednesday and Thursday.
Friday, Saturday, Sunday and Monday Run
This run has 5760 minutes in total and each 1440 minutes stands for a day consecutively
starting from Friday. Friday is the most crowded day of the week at Atatiirk Airport.
@ 1: Dom Dep Enterance 2: Dom Airside Security
x 400
2 400
f
1 200,
LT
|
erent aoe
1 0
1440.00 2880,00 4320,00 5760.0
Minutes 03:23 10Haz2013 Pax|
Untied
?
Figure 6: Domestic Terminal Queues in Base Model
® 1: Dom Dep Enterance 2: Dom Airside Security
1 300:
2 200
1 150
2 100
=
L
ly
00 1440,00 2880,00 4320,00 5760.0
Minutes 03:10 10Haz2013 Pax}
Untitled
2
Figure 7: Domestic Terminal Queues in Algorithmic Model
® 1: IntDep Enterance 2 Int Airside Security 3: Int Transfer Line Security
4004
300
4
2
3 400
wall | ,
200
|
a Lt Lt
00 144 280,00 4320,00 5760.0
Page 1 Minutes 0323 10 Haz 2013 Pax
SS ge Untitled
NWaeF 2?
Figure 8: International Terminal Queues in Base Model
@ + IntDep Enterance 2: Int Airside Security 3: Int Transfer Line Security
4 200"
2
3 300
4
i A bil
t T r r
0,00 1440,00 2880,00 4320.00 5760,0
Page 1 Minutes 03:10 10 Haz2013 Pax}
LIeF 2? Untied
Figure 9: International Terminal Queues in Algorithmic Model
Observations show that the algorithmic model distributes queues smoother over time horizon.
Algorithmic model avoids sharp jumps where the base model has very dramatic changes. It
can be observed that there are more people in the domestic airside queue of the algorithmic
model. The base model is deterministic and has no feedback structure thus it creates a strange
behavior. However the algorithmic model tries to keep all queues under boundaries.
The total number x-ray comparison was also conducted:
10
® 1:Total Number of xrays Active
1
L y
4
1
1440,00 2880,00 4320,00 760,00
IPage 1 Minutes 14:09 10 Haz 2013 Pad|
Gee 2? Untitled
Figure 10: Total Number of Active X-rays in Base Model
© 1: Total Number of Active Xrays
1 409
V U V
00 1440,00 2880,00 4320,00 5760.0
IPage 1 Minutes 14:18 10Haz 2013 Pad|
Maes 2? Untitled
Figure 11: Total Number of Active X-rays in Algorithmic Model
Tuesday, Wednesday and Thursday Run
This simulation made in order to observe the behavior of the model under less crowded days.
Same results obtained from previous simulation are also valid for these runs. Peaks occur in
the morning at the domestic terminal and these peaks are quite small in the algorithmic model
comparing to the base model.
11
@ 1: Dom Dep Enterance 2: Dom Airside Security
as 504
2 4
1 25 |
2: z
i 2 + _ |
: 1 4
2160,00 3240,00 4320, 0
Page 1 Minutes 14:58 10 Haz 2013 Paz
3aaF ? Untitled
Figure 12: Domestic Terminal Queues in Base Model
® 1:Dom Dep Enterance 2: Dom Airside Security
4 00
2 90
1 100, |
2 45 |
4 rt
H, 4
: ah,
00 4080.00 2160,00 3240,00 4320.01
IPage 1 Minutes 03:12 10 Haz 2013 Paz}
2 Untitled
Figure 13: Domestic Terminal Queues in Algorithmic Model
12
@ 1: int Dep Enterance 2: Int Airside Security
i
2
S
3: Int Transfer Line Security
50
100
a ray 2 2
4 A a eee] ie
A Bh ee rim; fae
,00 1080,00 2160,00 ,00 4320,0
Page 1 Minutes 14:58 10 Haz 2013 Paz
gear ? Untitled
Figure 14: International Terminal Queues in Base Model
®@ 1: intDep Enterance 2: Int Airside Security 3: Int Transfer Line Security
1 110
2 60
3 200
4 = fb"
2 30 | |
3: 100 Nf ry
h
7 i)
1 0
Es pfs 3.
0 1080,00 2160,00 4320.0
Minutes 03:12 10 Haz 2013 Paz|
Untitled
Figure 15: International Terminal Queues in Algorithmic Model
At international airside security checkpoint queue never exceeds 60 people where it reaches
90 people in the base model. A daily cycle can be observed in the graphical result. It is quite
smooth and balanced in contrast to the base model.
@ 1: Total Number of xrays Active
a: 354
,00 1080,00 2160,00 3240,00 4320,0
Page 1 Minutes 11:58 10 Haz 2013 Paz
aar 7? Untitled
Figure 16: Total Number of Active X-rays in Base Model
@ 4: Total Number of Active Xrays
1 30%
1 2054
1
\
si T
00 1080,00 2160,00 3240,00 4320,01
Page 1 Minutes
2 Untitled
Figure 17: Total Number of Active X-rays in Algorithmic Model
The trends between two outcomes have a substantial difference. In the base model the number
of x-rays is set as 29 at 8 am. However the algorithmic model uses only 20-23 x-ray devices.
Behaviors of the active x-ray usages are very similar for both, which is another indicator of
the algorithmic model is sensible and better in practice.
The results of the algorithmic model are promising. With the same input data, one can observe
that the active x-ray trend in algorithmic model is much lower than the base model.
14
GAME DESIGN
By constructing a game, a platform is provided to the authorized people in Ataturk Airport to
carry out interactive experiments about the necessary security personnel numbers and
allocation strategy of these security personnel. The game is constructed in FORIO
SIMULATE programming by converting the STELLA model equations to a web based form.
The game is designed to provide some experience about security personnel allocation to
different checkpoints at airports. Number of security personnel at each checkpoint is
important; it should be enough to avoid long queues at checkpoints, but should not be more
than necessary to minimize unnecessary personnel budget. To this end, a cost score is defined
as a joint function of the average queue length and the number of security personnel used. The
objective of the player is to keep the cost score as low as possible.
In the game, the player simultaneously decides on the total number of personnel and also the
desired number of x-rays at each security checkpoint. Simulation is run for one day; the
decisions are taken once in every two hours. Relevant information about security personnel
allocation problem is given to the player at the beginning. The game interface and outputs of
an example game are shown in Figure 18.
AVERAGE QUEUE LENGTHS im Donestic PERFORMANCE INDICATORS
enranee
ne mesic 18 average
ame Mi Ques’ GOST SCORE
tent
veo was
iu Iteration = aa
‘a0 faa i eonnel
px nteratona Nurnber
ae international 7745 wy Gost
° °
> 120 2¥0 360 aan GOO 720 GH oD 180 tO Tze HO 0 et 210 200 46 o0 720 BD BGO Ton eOOTEAaK0
ime Time
AVAILABLE PERSONNEL NUMBERS
x0 x 2 2 2 2 ° By
OPEN X-RAY LINE NUMBERS
Time Intemational Envance | _IterrationalAsiée | IferaonalTranser | Domestic Bwrance | Domestic airside TOTAL
x0 ° ° 3 ‘ ‘ 2
DESIRED X-RAY NUMBER AND PERSONNEL DESICIONS
S 0 i 0
International Entrance Domestic Entrance
(RUN) ( RUNTOEND
International Airside Domestic Airside
RESET) [__ GANE INFORMATION
International Transfer Total Personnel Number
80} 5}
Figure 18: The Game Interface
CONCLUSION
In this paper, with the purpose of analyzing passenger flow and security check-point
behaviors of Atatiirk Airport, a dynamic model is constructed. After constructing the model,
validation is done with extreme condition and sensitivity analyses in order to proceed to
scenario analysis. In scenario analysis part, several different cases are simulated to predict
alternative solutions, in case of possible changes in passenger behaviors and security system
15
structure. Next, simulations and output analysis are completed with real data inputs. Finally,
the model is turned into a web based simulation game.
Following extreme condition and sensitivity analyses, it is concluded that the model shows
realistic and meaningful behaviors and results. This indicates that the model is well structured.
By comparing with the behaviors of real data, validity of the model is confirmed. Key factors
of the model are five queue lengths, number of active x-ray devices in these queue locations,
and number of security personnel working at these queues. Inputs and construction of the
model is designed to focus on these three key factors. In order to simplify the construction,
the model is separated into two parts, domestic and international terminals, by taking into
account inter-relations with each other.
In domestic departure entrance, domestic-to-international and international-to-international
transfers and international departure entrance points, graphical real data inputs are entered to
the model. Since domestic-to-domestic and international-to-domestic transfer rates are too
small compered with the other two transfer rates, these two transfers are not reflected in the
model. Effects of changes in total number of security personnel, graphical inputs, smoothing
time and personnel inflow time to security checkpoints are investigated in the scenario
analysis part.
An algorithm that dynamically finds the number of active x-ray machines and allocated
personnel is next designed. The algorithm decides about these two decision variables
optimally, by keeping the queue lengths at the checkpoints as close to their desired levels as
possible. Results are shown to be promising, using real input data.
Finally, an interactive game version of the model is constructed via FORIO Online
Simulations Software. Thus, a web based interactive platform is offered to managers of the
Airport to carry out their own policy experiments about the security personnel numbers and
allocation strategy of these security personnel.
REFERENCES
Barlas, Y. 2002. “System Dynamics: Systemic Feedback Modeling for Policy Analysis in
Knowledge for Sustainable Development: An Insight into the Encyclopedia of Life
Support Systems.” UNESCO Publishing — Eolss Publishers 1131-1175.
Sterman, J. 2000. “Business Dynamics. Systems Thinking and Modeling for a Complex
World. “ McGraw-Hill, U.S.A.
Farringtori, P. A. H. B. Netnbliard, D. T. Stiirrock, arid G. W. Evans. Analysis and
Simulation of Passenger Flows in Airport Terminal. Proceedings of the 1999 Winter
Simulation Conference .
Joanna E. Manataki, Konstantinos G. Zografos. 2010. “Assessing airport terminal
performance using a system dynamics model.” Journal of Air Transport
Management 16, 86-93.
Sterman, J. 1988. People Express Management Flight Simulator. Massachusetts Institute of
Technology, Cambridge.
16
APPENDIX A: COMPLETE DIAGRAM OF THE MODEL
| \ '
| i | ae
| wii |
| a
| y i
i;
17
APPENDIX B: COMPLETE LIST OF MODEL EQUATIONS
Avg_Queue_Length_of Dom_Airside_Security(t)= Avg_Queue_Length_of Dom_Airside_Security(t - dt) +
(Adjustment_5) * dt
INIT Avg_Queue_Length_of_Dom_Airside_Security = 3
INFLOWS:
Adjustment_5 = (Dom_Airside_Security-Avg_Queue_Length_of_Dom_Airside_Security)/Smoothing_Time_5
Avg_Queue_Length_of Dom_Dep_Entrance(t) = Avg_Queue_Length_of Dom_Dep_Entrance(t - dt) +
(Adjustment_4) * dt
INIT Avg_Queue_Length_of_Dom_Dep_Entrance = 50
INFLOWS:
Adjustment_4 = (Dom_Dep_Entrance-Avg_Queue_Length_of_Dom_Dep_Entrance)/Smoothing_Time_4
Avg_Queue_Length_of _Int_Airside_Security(t)= Avg Queue_Length_of_Int_Airside_Security(t - dt) +
(Adjustment_2) * dt
INIT Avg_Queue_Length_of_Int_Airside_Security = 5
INFLOWS:
Adjustment_2 = (Int_Airside_Security-Avg_Queue_Length_of_Int_Airside_Security)/Smoothing_Time_2
Avg_Queue_Length_of_Int_Dep_Entrance(t) = Avg_Queue_Length_of_Int_Dep_Entrance(t - dt) +
(Adjustment_1) * dt
INIT Avg_Queue_Length_of_Int_Dep_Entrance = 100
INFLOWS:
Adj t_1 = (Int_Dep_E vg _Queue_Length_of_Int_Dep_Entrance)/Smoothing_Time_1
Avg_Queue_Length_of_Int_Transfer_Line_Security(t) = Avg_Queue_Length_of_Int_Transfer_Line_Security(t
- dt) + (Adjustment_3) * dt
INIT Avg Queue_Length_of_Int_Transfer_Line_Security = 0
INFLOWS:
Adjustment_3 = (Int_Transfer_Line_Security-
Avg_Queue_Length_of_Int_Trans Line_Security)/Smoothing_Time_3
Dom_After_Check-In(t) = Dom_After_Check-In(t - dt) + (Dom_Common_Flow_2 - Dom_Common_Flow_3) *
dt
INIT Dom_After_Check-In = 20
INFLOWS:
Dom_Common_Flow_2 = Dom_Check-In/Avg_Time_Spent_in_Dom_Check-In
OUTFLOWS:
18
Dom_Common_Flow_3 = Dom_After_Check-In/Avg_Time_Spent_in_Dom_After_Check-In
_Airside_Security(t) =Dom_Airside_Security(t - dt) + (Dom_Common_Flow_3 -
le_Security_Outflow_Rate) * dt
INIT Dom_Airside_Security = 3
INFLOWS:
Dom_Common_Flow_3 = Dom_After_Check-In/Avg_Time_Spent_in_Dom_After_Check-In
OUTFLOWS:
Dom_Airside_Security_Outflow_Rate =
MIN(Dom_Airside_Security,Number_of_Active_Xrays_in_Dom_Airside_Security)/Avg_Proce:
an_Xray_in_Dom_Airside_Security
ing _Time_of_
Dom_Check-In(t) = Dom_Check-In(t - dt) + (Dom_Common_Flow_1 - Dom_Common_Flow_2) * dt
INIT Dom_Check-In = 40
INFLOWS:
Dom_Common_Flow_1 = Dom_Pre_Check-In/Avg_Time_Spent_in_Dom_Pre_Check-In
OUTFLOWS:
Dom_Common_Flow_2 = Dom_Check-In/Avg_Time_Spent_in_Dom_Check-In
Dom_Dep_Entrance(t) = Dom_Dep_Entrance(t - dt) + (Dom_Dep_Inflow_Rate - Dom_Dep_Entrance_Rate) *
dt
INIT Dom_Dep_Entrance = 50
INFLOWS:
Dom_Dep_Inflow_Rate = GRAPH(TIME)
(0.00, 4.49), (60.0, 0.779), (120, 0.0462), (180, 0.766), (240, 2.08), (300, 11.0), (360, 26.2), (420, 9.97), (480,
10.3), (540, 12.3), (600, 17.9), (660, 12.4), (720, 9.46), (780, 11.0), (840, 12.6), (900, 13.5), (960, 14.4), (1020,
9.25), (1080, 10.0), (1140, 14.5), (1200, 8.10), (1260, 7.54), (1320, 7.72), (1380, 9.84), (1440, 6.89), (1500,
3.29), (1560, 2.32), (1620, 2.74), (1680, 5.00), (1740, 6.07), (1800, 6.88), (1860, 8.45), (1920, 8.14), (1980,
8.23), (2040, 10.2), (2100, 11.8), (2160, 12.9), (2220, 11.5), (2280, 9.55), (2340, 10.5), (2400, 11.8), (2460,
11.8), (2520, 10.3), (2580, 12.9), (2640, 13.0), (2700, 7.42), (2760, 6.36), (2820, 9.64), (2880, 5.51), (2940,
2.63), (3000, 1.86), (3060, 2.19), (3120, 4.00), (3180, 4.86), (3240, 5.51), (3300, 6.76), (3360, 6.51), (3420,
6.58), (3480, 8.14), (3540, 9.46), (3600, 10.3), (3660, 9.18), (3720, 7.64), (3780, 8.38), (3840, 9.46), (3900,
9.41), (3960, 8.24), (4020, 10.3), (4080, 10.4), (4140, 5.94), (4200, 5.09), (4260, 7.71), (4320, 6.89), (4380,
3.29), (4440, 2.32), (4500, 2.74), (4560, 5.00), (4620, 6.07), (4680, 6.88), (4740, 8.45), (4800, 8.14), (4860,
8.23), (4920, 10.2), (4980, 11.8), (5040, 12.9), (5100, 11.5), (5160, 9.55), (5220, 10.5), (5280, 11.8), (5340,
11.8), (5400, 10.3), (5460, 12.9), (5520, 13.0), (5580, 7.42), (5640, 6.36), (5700, 9.64), (5760, 9.64)
OUTFLOWS:
Dom_Dep_Entrance_Rate =
MIN(Number_of_Active_Xrays_in_Dom_Dep_Entrance,Dom_Dep_Entrance)/Avg_Processing_Time_of_an_X
ray_in_Dom_Dep_Entrance
19
Dom_Pre_Check-In(t) = Dom_Pre_Check-In(t - dt) + (Dom_Dep_Entrance_Rate - Dom_Common_Flow_1) * dt
INIT Dom_Pre_Check-In = 20
INFLOWS:
Dom_Dep_Entrance_Rate =
MIN(Number_of_Active_Xrays_in_Dom_Dep_Entrance,Dom_Dep_Entrance)/Avg_ Processing _Time_of_an_X
ray_in_Dom_Dep_Entrance
OUTFLOWS:
Dom_Common_Flow_1 = Dom_Pre_Check-In/Avg_Time_Spent_in_Dom_Pre_Check-In
Int_After_Check-In(t) = Int_After_Check-In(t - dt) + (Int_Common_Flow_2 - Int_Common_Flow_3) * dt
INIT Int_After_Check-In = 20
INFLOWS:
Int_Common_Flow_2 = Int_Check-In/Avg_Time_Spent_in_Int_Check-In
OUTFLOWS:
Int_Common_Flow_3 = Int_After_Check-In/Avg_Time_Spent_in_Int_After_Check-In
Int_Airside_Security(t) = Int_Airside_Security(t - dt) + (Int_Airside_Security_Inflow_Rate -
Int_Airside_Security_Outflow_Rate) * dt
INIT Int_Airside_Security = 5
INFLOWS:
Int_Airside_Security_Inflow_Rate = MIN(Passport_Control/(5*Avg_Passport_Control_Time_per_Person) ,
28/Avg_Passport_Control_Time_per_Person)
OUTFLOWS:
Int_Airside_Security_Outflow_Rate =
MIN(Int_Airside_Security,Number_of_Active_Xrays_in_Int_Airside_Security)/Avg_Processing_Time_of_an_
Xray_in_Int_Airside_Security
Int_Check-In(t) = Int_Check-In(t - dt) + (Int_Common_Flow_1 - Int_Common_Flow_2) * dt
INIT Int_Check-In = 50
INFLOWS:
Int_Common_Flow_1 = Int_Pre_Check-In/Avg_Time_Spent_in_Int_Pre_Check-In
OUTFLOWS:
Int_Common_Flow_2 = Int_Check-In/Avg_Time_Spent_in_Int_Check-In
Int_Dep_Entrance(t) = Int_Dep_Entrance(t - dt) + (Int_Dep_Inflow Rate - Int_Dep_Entrance_Rate) * dt
INIT Int_Dep_Entrance = 100
INFLOWS:
20
Int_Dep_Inflow_Rate = GRAPH(TIME)
(0.00, 12.2), (60.0, 5.85), (120, 4.13), (180, 4.87), (240, 8.90), (300, 10.8), (360, 12.2), (420, 15.0), (480, 14.5),
(540, 14.6), (600, 18.1), (660, 21.0), (720, 22.9), (780, 20.4), (840, 17.0), (900, 18.6), (960, 21.0), (1020, 20.9),
(1080, 18.3), (1140, 23.0), (1200, 23.1), (1260, 13.2), (1320, 11.3), (1380, 17.1), (1440, 9.19), (1500, 4.39),
(1560, 3.09), (1620, 3.65), (1680, 6.67), (1740, 8.09), (1800, 9.18), (1860, 11.3), (1920, 10.9), (1980, 11.0),
(2040, 13.6), (2100, 15.8), (2160, 17.2), (2220, 15.3), (2280, 12.7), (2340, 14.0), (2400, 15.8), (2460, 15.7),
(2520, 13.7), (2580, 17.2), (2640, 17.3), (2700, 9.90), (2760, 8.48), (2820, 12.9), (2880, 7.35), (2940, 3.51),
(3000, 2.48), (3060, 2.92), (3120, 5.34), (3180, 6.47), (3240, 7.34), (3300, 9.01), (3360, 8.68), (3420, 8.78),
(3480, 10.9), (3540, 12.6), (3600, 13.8), (3660, 12.2), (3720, 10.2), (3780, 11.2), (3840, 12.6), (3900, 12.5),
(3960, 11.0), (4020, 13.8), (4080, 13.9), (4140, 7.92), (4200, 6.78), (4260, 10.3), (4320, 9.19), (4380, 4.39),
(4440, 3.09), (4500, 3.65), (4560, 6.67), (4620, 8.09), (4680, 9.18), (4740, 11.3), (4800, 10.9), (4860, 11.0),
(4920, 13.6), (4980, 15.8), (5040, 17.2), (5100, 15.3), (5160, 12.7), (5220, 14.0), (5280, 15.8), (5340, 15.7),
(5400, 13.7), (5460, 17.2), (5520, 17.3), (5580, 9.90), (5640, 8.48), (5700, 12.9), (5760, 12.9)
OUTFLOWS:
Int_Dep_Entrance_Rate =
MIN(Int_Dep_Entrance,Number_of_Active_Xrays_in_Int_Dep_Entrance)/Avg Processing_Time_of_an_Xray_
in_Int_Dep_Entrance
Int_Pre_Check-In(t) = Int_Pre_Check-In(t - dt) + (Int_Dep_Entrance_Rate - Int_Common_Flow_1) * dt
INIT Int_Pre_Check-In = 20
INFLOWS:
Int_Dep_Entrance_Rate =
MIN(Int_Dep_Entrance,Number_of_Active_Xrays_in_Int_Dep_Entrance)/Avg_Processing_Time_of_an_Xray_
in_Int_Dep_Entrance
OUTFLOWS:
Int_Common_Flow_1 = Int_Pre_Check-In/Avg_Time_Spent_in_Int_Pre_Check-In
Int_Transfer_Line_Security(t) = Int_Transfer_Line_Security(t - dt) + (Int_to_Int_Transfer_Rate -
Int_Transfer_Line_Security_Outflow_Rate) * dt
INIT Int_Transfer_Line_Security = 0
INFLOWS:
Int_to_Int_Transfer_Rate = GRAPH(TIME)
(0.00, 0.00), (60.0, 0.00), (120, 0.132), (180, 0.00), (240, 0.647), (300, 21.8), (360, 19.4), (420, 4.42), (480,
2.84), (540, 4.44), (600, 12.6), (660, 12.4), (720, 5.54), (780, 1.60), (840, 2.81), (900, 6.78), (960, 9.39), (1020,
19.3), (1080, 5.69), (1140, 6.49), (1200, 1.12), (1260, 5.62), (1320, 20.1), (1380, 7.93), (1440, 6.89), (1500,
3.29), (1560, 2.32), (1620, 2.74), (1680, 5.00), (1740, 6.07), (1800, 6.88), (1860, 8.45), (1920, 8.14), (1980,
8.23), (2040, 10.2), (2100, 11.8), (2160, 12.9), (2220, 11.5), (2280, 9.55), (2340, 10.5), (2400, 11.8), (2460,
11.8), (2520, 10.3), (2580, 12.9), (2640, 13.0), (2700, 7.42), (2760, 6.36), (2820, 9.64), (2880, 5.51), (2940,
2.63), (3000, 1.86), (3060, 2.19), (3120, 4.00), (3180, 4.86), (3240, 5.51), (3300, 6.76), (3360, 6.51), (3420,
6.58), (3480, 8.14), (3540, 9.46), (3600, 10.3), (3660, 9.18), (3720, 7.64), (3780, 8.38), (3840, 9.46), (3900,
9.41), (3960, 8.24), (4020, 10.3), (4080, 10.4), (4140, 5.94), (4200, 5.09), (4260, 7.71), (4320, 6.89), (4380,
3.29), (4440, 2.32), (4500, 2.74), (4560, 5.00), (4620, 6.07), (4680, 6.88), (4740, 8.45), (4800, 8.14), (4860,
8.23), (4920, 10.2), (4980, 11.8), (5040, 12.9), (5100, 11.5), (5160, 9.55), (5220, 10.5), (5280, 11.8), (5340,
11.8), (5400, 10.3), (5460, 12.9), (5520, 13.0), (5580, 7.42), (5640, 6.36), (5700, 9.64), (5760, 9.64)
21
OUTFLOWS:
Int_Transfer_Line_Security_Outflow_Rate =
MIN(Int_Transfer_Line_Security,Number_of_Active_Xrays_in_Int_Transfer_Line_Security)/Avg_Processing_
Time_of_an_Xray_in_Int_Transfer_Line_Security
Number_of_Personnel_in_Dom_Airside_Security(t) = Number_of_Personnel_in_Dom_Airside_Security(t - dt)
+ (Personnel_inflow_to_Dom_Airside_Security - Personnel_outflow_from_Dom_Airside_Security) * dt
INIT Number_of_Personnel_in_Dom_Airside_Security = 25
INFLOWS:
Personnel_inflow_to_Dom_Airside_Security =
Dif_of_Personnel_in_Dom_Airside_Security/Personnel_Inflow_Time_to_Dom_Airside_Security
OUTFLOWS:
Personnel_outflow_from_Dom_Airside_Security = -Dif_of_Personnel_in_Dom_Airside_Security
Number_of_Personnel_in_Dom_Dep_Entrance(t) = Number_of_Personnel_in_Dom_Dep_Entrance(t - dt) +
(Personnel_outflow_from_Dom_Dep_Entrance - Personnel_inflow_to_Dom_Dep_Entrance) * dt
INIT Number_of_Personnel_in_Dom_Dep_Entrance = 5
INFLOWS:
Personnel_outflow_from_Dom_Dep_Entrance =
Dif_of_Personnel_in_Dom_Dep_Entrance/Personnel_Inflow_Time_to_Dom_Dep_Entrance
OUTFLOWS:
Personnel_inflow_to_Dom_Dep_Entrance = -Dif_of_Personnel_in_Dom_Dep_Entrance
Number_of_Personnel_in_Int_Airside_Security(t) = Number_of_Personnel_in_Int_Airside_Security(t - dt) +
(Personnel_inflow_to_Int_Airside_Security - Personnel_outflow_from_Int_Airside_Security) * dt
INIT Number_of_Personnel_in_Int_Airside_Security = 5
INFLOWS:
Personnel_inflow_to_Int_Airside_Security =
Dif_of_Personnel_in_Int_Airside_Security/Security_Inflow_Time_to_Int_Airside_Security
OUTFLOWS:
Personnel_outflow_from_Int_Airside_Security = -Dif_of_Personnel_in_Int_Airside_Security
Number_of_Personnel_in_Int_Dep_Entrance(t) = Number_of_Personnel_in_Int_Dep_Entrance(t - dt) +
(Personnel_inflow_to_Int_Dep_Entrance - Personnel_outflow_from_Int_Dep_Entrance) * dt
INIT Number_of_Personnel_in_Int_Dep_Entrance = 5
INFLOWS:
Personnel_inflow_to_Int_Dep_Entrance =
Dif_of_Personnel_in_Int_Dep_Entrance/Personnel_Inflow_Time_to_Int_Dep_Entrance
22
OUTFLOWS:
Personnel_outflow_from_Int_Dep_Entrance = -Dif_of_Personnel_in_Int_Dep_Entrance
Number_of_Personnel_in_Int_Transfer_Line_Security(t) =
Number_of Personnel_in_Int_Transfer_Line_Security(t - dt) +
(Personnel_inflow_to_Int_Transfer_Line_Security - Personnel_outflow_from_Int_Transfer_Line_Security) * dt
INIT Number_of_Personnel_in_Int_Transfer_Line_Security = 20
INFLOWS:
Personnel_inflow_to_Int_Transfer_Line_Security =
Dif_of_Personnel_in_Int_Transfer_Line_Security/Personnel_Inflow_Time_to_Int_Transfer_Line_Security
OUTFLOWS:
Personnel_outflow_from_Int_Transfer_Line_Security = -Dif_of_Personnel_in_Int_Transfer_Line_Security
Number_of Unallocated_Personnel(t) = Number_of Unallocated_Personnel(t - dt) +
(Personnel_outflow_from_Int_Dep_Entrance + Personnel_outflow_from_Int_Airside_Security +
Personnel_outflow_from_Int_Transfer_Line_Security + Personnel_inflow_to_Dom_Dep_Entrance +
Personnel_outflow_from_Dom_Airside_Security + Change_in_the_Total_Number_of_Personnel -
Personnel_inflow_to_Int_Dep_Entrance - Personnel_inflow_to_Int_Airside_Security -
Personnel_inflow_to_Int_Transfer_Line_Security - Personnel_outflow_from_Dom_Dep_Entrance -
Personnel_inflow_to_Dom_Airside_Security) * dt
INIT Number_of_Unallocated_Personnel = 30
INFLOWS:
Personnel_outflow_from_Int_Dep_Entrance = -Dif_of_Personnel_in_Int_Dep_Entrance
Personnel_outflow_from_Int_Airside_Security = -Dif_of_Personnel_in_Int_Airside_Security
Personnel_outflow_from_Int_Transfer_Line_Security = -Dif_of_Personnel_in_Int_Transfer_Line_Security
Personnel_inflow_to_Dom_Dep_Entrance = -Dif_of_Personnel_in_Dom_Dep_Entrance
Personnel_outflow_from_Dom_Airside_Security = -Dif_of_Personnel_in_Dom_Airside_Security
Change_in_the_Total_Number_of_Personnel = Total_Number_of_Personnel-
(Number_of_Personnel_in_Dom_Airside_Security+Number_of_Personnel_in_Dom_Dep_Entrance+Number_of
_Personnel_in_Int_Airside_Security+Number_of_Personnel_in_Int_Dep_Entrance+Number_of_Personnel_in_I
nt_Transfer_Line_Security+Number_of Unallocated_Personnel)
OUTFLOWS:
Personnel_inflow_to_Int_Dep_Entrance =
Dif_of_Personnel_in_Int_Dep_Entrance/Personnel_Inflow_Time_to_Int_Dep_Entrance
Personnel_inflow_to_Int_Airside_Security =
Dif_of_Personnel_in_Int_Airside_Security/Security_Inflow_Time_to_Int_Airside_Security
Personnel_inflow_to_Int_Transfer_Line_Security =
Dif_of_Personnel_in_Int_Transfer_Line_Security/Personnel_Inflow_Time_to_Int_Transfer_Line_Security
23
Personnel_outflow_from_Dom_Dep_Entrance =
Dif_of_Personnel_in_Dom_Dep_Entrance/Personnel_Inflow_Time_to_Dom_Dep_Entrance
Personnel_inflow_to_Dom_Airside_Security =
Dif_of_Personnel_in_Dom_Airside_Security/Personnel_Inflow_Time_to_Dom_Airside_Security
Passport_Control(t) = Passport_Control(t - dt) + (Int_Common_Flow_3 + Dom_to_Int_Transfer -
Int_Airside_Security_Inflow_Rate) * dt
INIT Passport_Control = 7
INFLOWS:
Int_Common_Flow_3 = Int_After_Check-In/Avg_Time_Spent_in_Int_After_Check-In
Dom_to_Int_Transfer = GRAPH(TIME)
(0.00, 3.20), (60.0, 4.78), (120, 0.00), (180, 0.15), (240, 0.05), (300, 0.57), (360, 0.17), (420, 1.42), (480, 5.23),
(540, 7.32), (600, 0.4), (660, 4.87), (720, 5.22), (780, 5.82), (840, 3.20), (900, 9.33), (960, 4.42), (1020, 2.27),
(1080, 3.68), (1140, 5.50), (1200, 4.37), (1260, 5.03), (1320, 3.58), (1380, 0.88), (1440, 8.35), (1500, 3.99),
(1560, 2.81), (1620, 3.32), (1680, 6.07), (1740, 7.36), (1800, 8.34), (1860, 10.2), (1920, 9.87), (1980, 9.98),
(2040, 12.3), (2100, 14.3), (2160, 15.6), (2220, 13.9), (2280, 11.6), (2340, 12.7), (2400, 14.3), (2460, 14.3),
(2520, 12.5), (2580, 15.7), (2640, 15.8), (2700, 9.00), (2760, 7.71), (2820, 11.7), (2880, 6.68), (2940, 3.19),
(3000, 2.25), (3060, 2.66), (3120, 4.85), (3180, 5.89), (3240, 6.67), (3300, 8.19), (3360, 7.89), (3420, 7.98),
(3480, 9.87), (3540, 11.5), (3600, 12.5), (3660, 11.1), (3720, 9.26), (3780, 10.2), (3840, 11.5), (3900, 11.4),
(3960, 9.99), (4020, 12.5), (4080, 12.6), (4140, 7.20), (4200, 6.17), (4260, 9.35), (4320, 8.35), (4380, 3.99),
(4440, 2.81), (4500, 3.32), (4560, 6.07), (4620, 7.36), (4680, 8.34), (4740, 10.2), (4800, 9.87), (4860, 9.98),
(4920, 12.3), (4980, 14.3), (5040, 15.6), (5100, 13.9), (5160, 11.6), (5220, 12.7), (5280, 14.3), (5340, 14.3),
(5400, 12.5), (5460, 15.7), (5520, 15.8), (5580, 9.00), (5640, 7.71), (5700, 11.7), (5760, 11.7)
OUTFLOWS:
Int_Airside_|
28/Avg Pz
ecurity_Inflow_Rate = MIN(Passport_Control/(S*Avg_Passport_Control_Time_per_Person) ,
ort_Control_Time_per_Person)
Available_Percentage_of_Desired_Number_of_Active_Xrays =
IF(Total_Desired_Number_of Active_Xrays<Total_Available_Number_of_Xrays) THEN | ELSE
Total_Available_Number_of_Xrays/Total_Desired_Number_of_Active_Xrays
Avg_Passport_Control_Time_per_Person = 0.636
Avg_Processing_Time_of_an_Xray_in_Dom_Airside Security = 0.395
Avg_Processing_Time_of_an_Xray_in_Dom_Dep_Entrance = 0.313
Avg_Processing_Time_of_an_Xray_in_Int_Airside_Security = 0.395
Avg_Processing_Time_of_an_Xray_in_Int_Dep_Entrance = 0.313
Avg_Processing_Time_of_an_Xray_in_Int_Transfer_Line_Security = 0.313
Avg_Time_Spent_in_Dom_After_Check-In = 5
Avg_Time_Spent_in_Dom_Check-In = 20
Avg_Time_Spent_in_Dom_Pre_Check-In = 5
24
Avg_Time_Spent_in_Int_After_Check-In = 15
Avg_Time_Spent_in_Int_Check-In = 20
Avg_Time_Spent_in_Int_Pre_Check-In = 5
Base_Queue_Length_of_Dom_Airside_Security =
Goal_Queue_Length_per_Xray_in_Dom_Airside_Security*Total_Number_of_Xrays_in_Dom_Airside_Securit
»
Base_Queue_Length_of_Dom_Dep_Entrance =
Goal_Queue_Length_per_Xray_in_Dom_Dep_Entrance*Total_Number_of_Xrays_in_Dom_Dep_Entrance
Base_Queue_Length_of_Int_Airside_Security =
Goal_Queue_Length_per_Xray_in_Int_Airside_Security*Total_Number_of_Xrays_in_Int_Airside_Security
Base_Queue_Length_of_Int_Dep_Entrance =
Total_Number_of_Xrays_in_Int_Dep_Entrance*Goal_Queue_Length_per_Xray_in_Int_Dep_Entrance
Base_Queue_Length_of_Int_Transfer_Line_Security =
Goal_Queue_Length_per_Xray_in_Int_Transfer_Line_Security*Total_Number_of_Active_Xrays_in_Int_Trans
fer_Line_Security
Desired_Number_of_Active_Xrays_in_Dom_Airside_ Security =
Total_Number_of_Xrays_in_Dom_Airside_Security*Need_Coefficient_of_Dom_Airside_Security
Desired_Number_of_Active_Xrays_in_Dom_Dep_Entrance =
Need_Coefficient_of_Dom_Dep_Entrance*Total_Number_of_Xrays_in_Dom_Dep_Entrance
Desired_Number_of_Active_Xrays_in_Int_Airside_Security =
Need_Coefficient_of_Int_Airside_Security*Total_Number_of_Xrays_in_Int_Airside_Security
Desired_Number_of_Active_Xrays_in_Int_Dep_Entrance =
Need_Coefficient_of_Int_Dep_Entrance*Total_Number_of_Xrays
| Int_Dep_Entrance
Desired_Number_of_Active_Xrays_in_Int_Transfer_Line_Security =
Need_Coefficient_of_Int_Transfer_Line_Security*Total_Number_of_Active_Xrays_in_Int_Transfer_Line_Sec
urity
Dif_of_Number_of_Xrays_in_Dom_Airside_Security =
MAX(1,MIN(Total_Number_of_Xrays_in_Dom_Airside_Security,Available_Percentage_of_Desired_Number_
of_Active_Xrays*Desired_Number_of_Active_Xrays_in_Dom_Airside_Security))-
Number_of_Active_Xrays_in_Dom_Airside_Security
Dif_of_Number_of_Xrays_in_Dom_Dep_Entrance =
MAX(1,MIN(Total_Number_of_Xrays_in_Dom_Dep_Entrance,Available_Percentage_of_Desired_Number_of
_Active_Xrays*Desired_Number_of_Active_Xrays_in_Dom_Dep_Entrance))-
Number_of_Active_Xrays_in_Dom_Dep_Entrance
Dif_of_Number_of_Xrays_in_Int_Airside_Security =
MAX(MIN(Available_Percentage_of_Desired_Number_of_Active_Xrays*Desired_Number_of_Active_Xrays_
in_Int_Airside_Security,Total_Number_of_Xrays_in_Int_Airside_Security),1)-
Number_of_Active_Xrays_in_Int_Airside_Security
Dif_of_Number_of_Xrays_in_Int_Dep_Entrance =
MAX(MIN(Total_Number_of_Xrays_in_Int_Dep_Entrance,Available_Percentage_of_Desired_Number_of_Act
25
ive_Xrays*Desired_Number_of_Active_Xrays_in_Int_Dep_Entrance),1)-
Number_of_Active_Xrays_in_Int_Dep_Entrance
Dif_of_Number_of_Xrays_in_Int_Transfer_Line_Security =
MAX(1,MIN(Total_Number_of_Active_Xrays_in_Int_Transfer_Line_Security,Available_Percentage_of_Desir
ed_Number_of_Active_Xrays*Desired_Number_of_Active_Xrays_in_Int_Transfer_Line_Security))-
Number_of_Active_Xrays_in_Int_Transfer_Line_Security
Dif_of_Personnel_in_Dom_Airside_Security =
Dif_of_Number_of_Xrays_in_Dom_Airside_Security*Needed_Security_per_Xray_in_Dom_Airside_Security
Dif_of_Personnel_in_Dom_Dep_Entrance =
Dif_of_Number_of_Xrays_in_Dom_Dep_Entrance*Needed_Security_per_Xray_in_Dom_Dep_Entrance
Dif_of_Personnel_in_Int_Airside_Security =
Dif_of_Number_of_Xrays_in_Int_Airside_Security*Needed_Security_per_Xray_in_Int_Airside_Security
Dif_of_Personnel_in_Int_Dep_Entrance =
Dif_of_Number_of_Xrays_in_Int_Dep_Entrance*Needed_Personnel_per_Xray_in_Int_Dep_Entrance
Dif_of_Personnel_in_Int_Transfer_Line_Security =
Dif_of_Number_of_Xrays_in_Int_Transfer_Line_Security*Needed_Security_per_Xray_in_Int_Transfer_Line_
Security
Goal_Queue_Length_per_Xray_in_Dom_Airside_Security = 12
Goal_Queue_Length_per_Xray_in_Dom_Dep_Entrance = 15
Goal_Queue_Length_per_Xray_in_Int_Airside_Security = 10
Goal_Queue_Length_per_Xray_in_Int_Dep_Entrance = 15
Goal_Queue_Length_per_Xray_in_Int_Transfer_Line_Security = 18
Needed_Personnel_per_Xray_in_Int_Dep_Entrance = 3.5
Needed_Security_per_Xray_in_Dom_Airside_Security = 3.5
Needed_Security_per_Xray_in_Dom_Dep_Entrance = 3.5
Needed_Security_per_Xray_in_Int_Airside_Security = 3.5
Needed_Security_per_Xray_in_Int_Transfer_Line_Security = 3.5
Need_Coefficient_of_Dom_Airside_Security =
Avg_Queue_Length_of Dom_Airside_Security/Base_Queue_Length_of_Dom_Airside_Security
Need_Coefficient_of_Dom_Dep_Entrance =
Avg_Queue_Length_of Dom_Dep_Entrance/Base_Queue_Length_of_Dom_Dep_Entrance
Need_Coefficient_of_Int_Airside_Security =
Avg_Queue_Length_of _Int_Airside_Security/Base_Queue_Length_of_Int_Airside Security
Need_Coefficient_of_Int_Dep_Entrance =
Avg_Queue_Length_of_Int_Dep_Entrance/Base_Queue_Length_of_Int_Dep_Entrance
Need_Coefficient_of_Int_Transfer_Line_Security =
Avg_Queue_Length_of_Int_Transfer_Line_Security/Base_Queue_Length_of_Int_Transfer_Line_Security
26
Number_of_Active_Xrays_in_Dom_Airside_Security =
Number_of_Personnel_in_Dom_Airside_Security/Needed_Security_per_Xray_in_Dom_Airside_Security
Number_of_Active_Xrays_in_Dom_Dep_Entrance =
Number_of_Personnel_in_Dom_Dep_Entrance/Needed_Security_per_Xray_in_Dom_Dep_Entrance
Number_of_Active_Xrays_in_Int_Airside_Security =
Number_of_Personnel_in_Int_Airside_Security/Needed_Security_per_Xray_in_Int_Airside_Security
Number_of_Active_Xrays_in_Int_Dep_Entrance =
Number_of_Personnel_in_Int_Dep_Entrance/Needed_Personnel_per_Xray_in_Int_Dep_Entrance
Number_of_Active_Xrays_in_Int_Transfer_Line_Security =
Number_of_Personnel_in_Int_Transfer_Line_Security/Needed_Security_per_Xray_in_Int_Transfer_Line_Secur
ity
Personnel_Inflow_Time_to_Dom_Airside_Security = 7
Personnel_Inflow_Time_to_Dom_Dep_Entrance = 7
Personnel_Inflow_Time_to_Int_Dep_Entrance = 7
Personnel_Inflow_Time_to_Int_Transfer_Line_Security = 7
Rounded_Number_of_Active_Xrays_in_Dom_Airside_Security =
ROUND(Number_of_Active_Xrays_in_Dom_Airside_Security+0.5)
Rounded_Number_of_Active_Xrays_in_Dom_Dep_Entrance =
ROUND(Number_of Active Xrays_in_Dom_Dep_Entrance+0.5)
Rounded_Number_of_Active_Xrays_in_Int_Airside_Security =
ROUND(Number_of Active Xrays_in_Int_Airside_Security+0.5)
Rounded_Number_of_Active_Xrays_in_Int_Dep_Entrance =
ROUND(Number_of_Active_Xrays_in_Int_Dep_Entrance+0.5)
Rounded_Number_of_Active_Xrays_in_Int_Transfer_Line_Security =
ROUND(Number_of_Active_Xrays_in_Int_Transfer_Line_Security+0.5)
Security_Inflow_Time_to_Int_Airside Security = 7
Smoothing Time_1 =5
Smoothing _Time_2=5
Smoothing Time_3=5
Smoothing Time_4=5
Smoothing Time_5=5
Total_Available_Number_of_Xrays = Total_Number_of_Personnel/3.5
Total_Desired_Number_of_Active_Xrays =
Desired_Number_of Active Xrays_in_Dom_Airside_Security+Desired_Number_of_Active_Xrays_ in Dom_D
ep_Entrance+Desired_Number_of_Active_Xrays_in_Int_Airside_Security+Desired_Number_of_Active_Xrays_
in_Int_Dep_Entrance+Desired_Number_of_Active_Xrays_in_Int_Transfer_Line_Security
27
Total_Number_of_Active_Xrays =
Number_of Active Xrays_in_Dom_Airside_Security+Number_of Active Xrays_in_Dom_Dep_Entrance+Nu
mber_of Active Xrays_in_Int_Airside_Security+Number_of Active Xrays_in_Int_Dep_Entrance+Total_Num
ber_of_Active_Xrays_in_Int_Transfer_Line_Security
Total_Number_of_Active_Xrays_in_Int_Transfer_Line_Security = 8
Total_Number_of_Personnel = 120
Total_Number_of_Xrays_in_Dom_Airside_Security = 12
Total_Number_of_Xrays_in_Dom_Dep_Entrance = 9
Total_Number_of_Xrays_in_Int_Airside_ Security = 21
Total_Number_of_Xrays_in_Int_Dep_Entrance = 16
28
