matthew pan extended essay candidate #: d 0130 013...

Matthew Pan Extended Essay Candidate #: D 0130 013 Introduction of Problem Nowadays, airline travel is almost always about time. You feel like you are in a 100 meter dash down the concourse because your plane departs in two minutes. You don’t have enough time to get from arrival gate A to departure gate G because the immigrations officer needed a full ten minutes to check your passport. You have a board meeting in New York in two hours, but you’re stuck in the Pearson International Airport in Toronto for the night because of a snowstorm. At the airport, you may have felt as if time was working against you. Passenger airlines also feel this way because they are constantly running on a tight schedule twenty-four/seven. To airlines, time equals money. Every minute that a plane is late for takeoff means lost revenue for an airline. Through research, it has been found that gate delays, the extra time that an aircraft remain at the airport gate, costs domestic carriers in the United States approximately $22.38 a minute. This figure adds up to $220 million/year system-wide in lost revenue for airlines – a lot of money, even for a large airline.1 To reduce gate delays, most airlines are now looking at new methods to reduce the time it takes to perform longest inter-flight operations- one of these being the boarding passengers. The object of this essay is to determine an efficient boarding procedure for midsized (200 seats or less) passenger aircraft in terms of how fast all passengers can be boarded while retaining, or even increasing, passenger safety and satisfaction. This paper will focus on human organization during the boarding process of an aircraft within existing configurations of the aircraft and airport terminal. Importance of Topic One of the first questions that one may ask when contemplating this topic is: why should anyone be concerned about how passengers are boarded onto an airplane? The answer to this question may come easily to those who have traveled frequently on international or cross-continental flights. Boarding orders are important in that they determine passenger satisfaction and safety, as these factors affect the image of the airline. Imagine if no boarding method existed when boarding a Boeing 747 with approximately 400 seats and only two isles. It is undeniably clear that if no boarding order existed, there would be chaos. People would be running, pushing and shoving just to get the “best” seats in the plane (namely the seats located forward of the engines of the aircraft because of the reduced noise). Aisles would be clogged up hundreds of times because of passengers loading carry-on luggage into the overhead storage bins. In the worst case-scenario, injuries or death may occur. One may only need to look to an event that occurred in this year to see the terrifying things that could happen in crowded spaces such as an airplane. A stampede took place in February when over 30,000 tourists visited a lantern show in China. The accident left 37 people dead and 24 others injured due to trampling. The 1 Funk, M. (2003, November). The Visualization of the Quantification of the Commodification of Air Travel Or: Why Flying Makes You Feel Like a Rate in a Lab Cage. Popular Science. Pp. 69

Matthew Pan Extended Essay Candidate #: D 0130 013 cause? A person accidentally tripped in a crowded area, causing a horrifying chain reaction. Organized boarding methods help dramatically reduce the risk of something like this from occurring onboard an aircraft. Another reason for the importance of the study of passenger boarding procedures is because it is a major factor in determining how efficient and how profitable an airline is. As mentioned before, the more time that a plane has to spend on the tarmac, the less profit is earned by the airline operating that aircraft. According to a study performed by Boeing, passenger “enplaning” (passengers boarding an aircraft) along with “deplaning” (passengers leaving an aircraft) is one of the largest factors in determining “turntime” (also referred to as “turnaround time”), the time required to unload an airplane after its arrival at the gate and to prepare for its departure again. This is indicated by the Figure 1, which shows relative lengths of times of turntime operations. Notice time span of the activity, “Board Passengers” compared to other operations charted within the figure.

PASSENGER SERVICES

Position Passenger Bridges or Stairs

Deplane Passengers

Service Cabin

Service Galleys

Board Passengers

Remove Passenger Bridges or Stairs

LUGGAGE/CARGO HANDLING

Forward Cargo Compartment

Aft Cargo Compartment

AIRCRAFT SERVICING

Fuel Aircraft

Service Lavatories

Service Portable Water

UNLOAD

UNLOAD LOAD

LOAD 50% CARGO VOLUME

50% CARGO VOLUME

Push Back

Position/Remove Equipment

TURN TIME OPERATIONS AND RELATIVE TIME SPAN FOR EACH ACTIVITY

Figure 1. Relative Lengths of Time for Turntime Operations.1

Matthew Pan Extended Essay Candidate #: D 0130 013

Faster Enplaning and Deplaning

Faster Turnaround Time

More Flights Annually

More Paying Passengers Annually

More Profit for Airline

Figure 2. How Faster Enplaning and Deplaning Affects Airline Profitability

An airline almost always measures its productivity and profitability by turntime. Figure 2 has been included to show one of the major ways that an airline can profit from having faster passenger enplaning and deplaning methods. If turn time were to increase on the other hand, gate delays would occur, leading to losses in profits for the airline operating the grounded aircraft. When considering the information that has just been presented, it is evident that the consideration of boarding methods is very important to both the passenger and the airline. It is surprising however, that the lack of public research material seems to prove otherwise. Hours of research in libraries and on the internet have only yielded a small stack of materials. This lack of information is also noted in the only two university papers found on this topic. Only a few main resources have been used for this in preparation and compilation of this essay; seven are known to be publicly available. Overall, this has presented a challenge.

Matthew Pan Extended Essay Candidate #: D 0130 013 Project Strategy and Procedure This project was carried out much like in the fashion of a science experiment, and thus, a specific procedure was followed.

1. Find references in information databases of research materials under theses headings:

• Aircraft Boarding Processes, Procedures, Methods or Strategies • Turn Time • Passenger Boarding Instruments to be used:

- Calgary Public Library Database - University of Calgary Library Database - Internet Search Engines

Criteria for the selection of sources: - Reliability of sources. - Magnitude of content relating to subject of investigation. - Magnitude of experimentation.

2. Create bibliography of sources. 3. Read over sources for details/evidence and information relating to subject of

investigation. 4. Analyze material. 5. Interview Air Canada and other airlines’ officials (e.g. gate agents). 6. Find boarding processes used by different airlines. 7. Design new and logical boarding processes. 8. Test all boarding processes using computer simulation. 9. Analyze boarding processes, their feasibility, practicality and other non-quantifiable

factors. 10. Select most efficient boarding order with lowest combination of seat interferences,

aisle interferences and boarding time. 11. Implement boarding procedure within an airline as a pilot project. 12. Collect results and formulate conclusions.

It should be noted that steps 5 and 11 of this procedure failed. Air Canada did not grant my wishes to interview gate agents and other officials for reasons unknown. Implementation of this project was attempted at Air Canada, but failed due to increasing security restrictions of airports. 2 2 Several attempts to contact officials of the airline ended in failure. The airline had implicitly stated through e-mails that the company was not in a position to assist me at that time, as the airline was, and still is undergoing a state of financial crisis. Nevertheless, steps were taken to see if this was truly not an option. Multiple trips were taken to Air Canada’s offices at the Calgary International Airport to see if I was able to conduct interviews with Air Canada personnel. These wishes were repeatedly turned down due to various reasons. Also, as expected, requests for conducting tests aboard Air Canada’s aircraft were also turned down due to security reasons. All other airlines that I contacted were surprisingly unable to provide information as well, and directed me to their corporate website, which was of no help either.

Matthew Pan Extended Essay Candidate #: D 0130 013 Research Before performing simulations, research was done to see how others approached the problem of aircraft boarding. Four main sources of information were found on the subject. It is should be noted that all of these papers used computer simulations and used data from actual passenger boardings. Van Landeghem from the Ghent University presented: “A Simulation of Passenger Boarding Times in Airplanes”3 which stated that the fastest boarding strategy consisted of passengers boarding individually by seat and row number. The next fastest procedure was a back-to-front “alternate half-row” boarding system. In this procedure, a group consists of a half row of seats (either ABC or DEF), which would be boarded alternately (e.g. if a half-row group located at the back of the plane in row 26 was boarded first, the next group to board would be the half-row located in row 24). This method was cited to take 15.8 minutes. There is, however, one problem with the data presented by the paper: only five replications were performed for each of the boarding procedures that were tested, which may render the data highly inaccurate. The second document that was studied was: “The Role of Computer Simulation in Reducing Airplane Turn Time”4 prepared for Boeing by Marelli, S. et al. This study used a specially designed computer simulation program called “P.E.D.S.” (Passenger Enplaning/Deplaning Simulation) to find a way to reduce turn time. The program used a technique called “discrete event simulation” to simulate a boarding method. This simulation method uses the concept that a boarding process is a set of interrelated elements or events. Each activity that occurs during a boarding process starts, continues for a specific time, and stops at specific intervals. The program uses mathematical probability for random events and passenger behavior that may occur inside a passenger aircraft. P.E.D.S. predicted that it would take approximately 22 min to board a Boeing 747-200 (which is slightly larger in terms of the number of seats than the Airbus A320 used in this paper’s simulation). It is unclear, however, what boarding procedure was used to obtain this data. A team from the Arizona State University (ASU) was the author of two studies entitled: “The Aircraft Boarding Problem” 5 and “Development of Efficient Boarding Strategies at American West Airlines”6 prepared for America West Airlines. These papers utilized a simulation program based on what they call an “interference model.” In this model, the total passenger boarding time is dependant on events called “interferences”. This model will be explained later, in depth, as the simulation program used in this paper is the same

3 Landeghem, H. Van. (2000). A Simulation Study of Passenger Boarding Times in Airplanes. 4 Marelli, S., Mattocks, G., Merry, R. (1998). The Role of Computer Simulation in Reducing Airplane Turn Time. 5 Hogg, G.L, van den Briel, M.H.L., Villalobos, J.R. (2003, May). The Aircraft Boarding Problem. 6 Hogg, G.L, van den Briel, M.H.L., Villalobos, J.R. (2004, February). Development of Efficient Boarding Strategies at American West Airlines.


program used by the ASU research team. The results obtained from their simulation program showed that a “reverse pyramid” boarding strategy could reduce aircraft’s turntime by 3-5 minutes when compared to a traditional back-to-front boarding approach. Background Information Selection of Aircraft The primary type of aircraft that this paper and all experimentation documented in this paper will focus on is a generic Airbus A320. This airplane will have of three rows of four first class seats and 23 rows of six coach class seats. Figure 3 has been included for reference. Note the number of first class seats (12), the number of economy seats (126), seat column letters, and the total number of rows as these numbers and letters will be used quite frequently within this paper. The typical seating configuration of an A320 has been used for a number of reasons. First, mid-sized passenger aircraft, such as the A320 being used for this paper, are the most common-sized aircraft in most airline fleets, and therefore render this study applicable to a wider majority of airlines compared to modelling a smaller or larger aircraft. Also, the computer simulation program used for this paper is based upon the Airbus A320. Lastly, designing boarding procedures for any aircraft larger than the one being used by this paper would be an arduous task -especially with the resources available at the time of the creation of this paper. Although the results of this study would work best in improving boarding methods in mid-sized aircraft, with a little modification to these methods, the results could apply to a larger range of aircraft sizes.

Primary Exit

Figure 3. Seating Configuration for an Airbus A320 used in this Essay.

Econom

y First C

lass

A B C D

A B C D E F

Matthew Pan Extended Essay Candidate #: D 0130 013 Interferences Passenger boarding delays usually occur because of three reasons: gate agents cannot process passengers fast enough to keep a steady flow of passengers moving into the aircraft, unexpected random events (such as a terrorist threat onboard a grounded aircraft), and boarding interferences - conflicts between passengers during the boarding procedure. The problem of keeping steady flow passengers boarding the aircraft due to the slow processing procedures of gate agents can easily be solved with the introduction of one or more gate agents into the system. The occurrence of random events that delay cannot be reduced easily. Boarding interferences, however, are more complex in nature and can rarely be avoided or corrected…but can be definitely reduced. An “interference” model will be used in this paper to determine the boarding method that uses the least amount of time on average.

Within a passenger aircraft, two types of boarding interferences can occur during the boarding of an aircraft, both of which cause delays at the gate and have the potential to cause a decrease in passenger satisfaction. One of these interferences is called “aisle interference”. This type of boarding interference occurs when a passenger loads carry-on luggage into the overhead storage compartments of the aircraft. To perform this action however, the passenger loading the carry-on luggage must stand in the narrow aisle and momentarily act as an obstacle to those waiting to get to their seats. Unless passengers can squeeze past the person loading his or her carry-on luggage, the boarding of passengers is momentarily halted.

Passenger loading luggage into overhead bins (Obstruction).

Passengers waiting to pass obstructing passenger.

Figure 4. Example of an Aisle Interference.

TO FRONT OF PLANE


The other type of interference that can occur on passenger aircraft is called “seat interference”. Seat interferences arise when a passenger’s route to his or her seat is blocked by another passenger sitting in the same row, but closer to the aisle as rows are usually clustered together with not much legroom for passengers. The situation worsens if one has to get past two or more passengers that are already sitting in their seats located closest to the aisle within a half-row. Passengers waiting to be seated may do one of the following: ask the passengers already seated to stand up and move into the aisle while they move into their seat further away from the aisle, or they may choose to simply attempt to move into the seat without asking the already seated passengers to move.

With either of these two types of boarding interferences, it is obvious that if the frequency of interferences increases during the boarding of an aircraft, the boarding time increases and thus gate delays occur. Therefore, one of the methods to reduce boarding time is to find or develop a procedure that tries to reduce the amount of seat and aisle interferences. The Simulation Simulations of boarding methods for this paper were performed using a model built within ProModel 2001 by Menkes H.L. van den Briel of the Arizona State University. The team from ASU had obtained videotapes of actual airplane boardings, and recorded data essential to the creation of an enplaning simulation. From these videotapes they found:

• The arrival rate of passengers entering the aircraft from the terminal (time between passengers or more or less the speed of the gate agent- assuming one gate agent is present at boarding) was an exponential distribution with an average of 9.0 seconds (the exponential distribution accounted for late arrivals at the terminal).

• Row speed, the speed of passenger travelling from one row to the next, was a

triangular distribution with a low of 0.6 seconds, an average of 0.95 seconds, and a high of 1.3 seconds.

Passengers needing to pass seated passenger in the same half-row to reach their own seat (Seat Interference)

No S

eat Interference

Figure 5. Examples of Seat Interferences.


• 60% of the passengers had a triangularly distributed luggage speed, or the speed that passengers place their carry-on luggage into the overhead bins, with a low of 3.2 seconds, an average of 7.1 seconds and a high of 38.7 seconds. The other 40% of the passengers had a luggage time of zero seconds, as they had no large carry-on luggage place in the overhead luggage compartments or decided to place their luggage beneath their seats.

• Also related to luggage speed, a “pass ratio” was determined. Sometimes,

when a person was loading his or her luggage into the overhead bins, the stalled passengers would slide past or “pass” the aisle interference without using much time. On average, one passenger out of every ten would do this.

• Seat interference time was found to be a triangular distribution with a low of

7.4 seconds, an average of 9.7 seconds, and a high 15.5 seconds.

• An average of 29 parties (groups of people that have requested that they sit in adjacent seats and board at the same time e.g. a family, a group of friends/colleagues) when were found to board a plane. It was found that an average of twenty parties of two and nine parties of three boarded. Larger parties were rare and were not integrated into the study.

The simulation that the team from ASU built afterwards, which is used extensively by this paper, was built using this data. The model is based on the interferences mentioned previously as an indication of the time it takes for passengers to board. When interferences occurs within a boarding procedure simulation, a predetermined penalty is applied to the overall time. It needs to be noted that the simulation is based on a few assumptions. These include:

• A “call-off” system will be will be used. This is where a gate agent calls out groups to be boarded next using a P.A. system.

• All first class passengers are to be boarded first in every boarding method. Therefore, this study will mainly apply to the boarding of passengers in the economy class.

• Passengers do not sit in the wrong seat and they do not walk past the row containing their seat.

Each boarding method simulation was run 100 times, which allowed for a confidence interval of less than 60 seconds for each different boarding procedure tested. The Results/Observations

Matthew Pan Extended Essay Candidate #: D 0130 013 37 different boarding orders were simulated and tested under the following categories:

1. Random (RAND): All passengers in the economy class are boarded together as one group. First class passengers are boarded first as one group, followed by coach class passengers.7

2. By Block (B): Passengers are boarded by groups with each group

containing a number of contiguous rows. 3. By Half-Row/Half Block (HR): Same as “By block”, except the block spans

only half a row (ABC or DEF). 4. By Seat Row (R): A single row of seats (ABC DEF) equals a group. 5. By Seat Column (C): A single column of seats (A, B, C, D, E, or F) equals

a group. 6. By Reverse Pyramid Type (RP): Based on the boarding order created by

the ASU team. All boarding procedures that use the same, general, outside-in / back-to-front hybrid method used by the Reverse Pyramid approach was placed under this category. The original Reverse Pyramid boarding process is labeled RP1.

7. By U-Boarding Type (U): This category features new and experimental

boarding procedures where seats belonging to a group forms a U-shape. 8. By Seat (S): Each passenger forms its own group and is called individually

by row and number.8 9. By Airline Pattern (AP): Alaska (AP1), Delta (AP2) and United Airlines’

(AP3) recently implemented boarding processes. 10. By Special Pattern (SP): Any other pattern that does not fit any of these

categories. Consisted of a two nonsymmetrical boarding processes9 The next three pages contain the observations recorded from the simulations.

7 Simulation on this boarding procedure was only to confirm the results of Van Langedem’s study, “A Simulation of Passenger Boarding Times in Airplanes” as it is deemed impractical and hazardous before the simulation trials. 8 Simulation on this boarding procedure was only to confirm the results of Van Langedem’s study, “A Simulation of Passenger Boarding Times in Airplanes” as it is deemed impractical before the simulation trials. 9 SP1 provided by Menkes H.L. van den Briel. He stated that this model was slightly better than his “Reverse Pyramid” boarding process in terms of expected interferences.

Matthew Pan Extended Essay Candidate #: D 0130 013 Data Table 1 – Boarding Procedure vs. Minimum, Average and Maximum Times for 100 Replications Boarding Procedure Minimum Time (s) Average Time (s) Maximum Time (s) RAND 1238.7600 1437.8778 1706.4000B1 1430.8800 1672.5972 1914.7200B2 1377.2400 1581.2592 1798.0200B3 1313.2200 1506.1008 1733.1600B4 1292.8800 1491.6810 1710.6000B5 1222.7400 1473.6930 1751.4000B6 1172.2800 1443.1560 1734.8400B7 1219.6800 1436.7552 1755.7200B8 1284.6000 1532.7888 1822.6800B9 1307.7000 1516.0098 1753.6800HR1 1259.9400 1535.2098 1776.1200HR2 1215.3000 1436.8200 1701.7800HR3 1220.7600 1466.1552 1695.6000HR4 1106.8800 1413.2232 1659.8400HR5 1195.5600 1425.8976 1627.2600HR6 1104.6600 1417.5402 1754.2800R1 1628.5200 1946.3418 2240.2200R2 1215.7800 1427.4150 1690.6800C1 1162.8600 1383.2184 1649.5800C2 1095.8400 1376.0664 1691.3400RP1 1173.0600 1387.8024 1611.8400RP2 1096.6200 1364.9976 1570.6200RP3 1184.3400 1385.4390 1650.3600RP4 1174.2600 1374.4260 1684.2000RP5 1141.6800 1379.9208 1608.9600RP6 1136.5200 1380.2970 1680.1200U1 1144.6200 1375.0644 1681.5600U2 1054.9800 1388.8308 1750.8600U3 1178.9400 1385.3988 1610.5800U4 1136.4000 1388.4912 1606.9800U5 1098.0000 1386.3438 1637.5200S1 1097.7600 1370.0682 1657.5600AP1 1173.6000 1430.5452 1669.4400AP2 1213.3200 1451.5686 1689.7200AP3 1221.2400 1480.2780 1709.1600SP1 1160.9400 1387.5726 1657.9200SP2 1121.3400 1378.8498 1684.6200

Matthew Pan Extended Essay Candidate #: D 0130 013 Data Table 2 – Boarding Procedure vs. Minimum, Average and Maximum Number of Seat Interference Occurrences for 100 Replications

Boarding Procedure Minimum Seat Interference

Average Seat Interference

Maximum Seat Interference

RAND 55 71.95 91B1 55 71.81 85B2 47 73.17 88B3 54 72.21 88B4 56 72.22 95B5 60 73.36 89B6 52 72.75 87B7 56 74.76 84B8 56 72.46 95B9 56 74.13 88HR1 56 71.77 92HR2 54 72.09 92HR3 55 72.08 92HR4 52 71.89 92HR5 53 71.53 92HR6 58 72.39 87R1 56 72.26 91R2 56 71.98 90C1 0 2.77 6C2 0 2.94 6RP1 0 2.94 5RP2 1 3.07 6RP3 0 3.03 6RP4 1 3.23 6RP5 0 2.89 6RP6 1 3.02 5U1 21 31.06 41U2 33 43.97 55U3 15 23.61 51U4 15 24.06 31U5 13 19.37 27S1 1 2.94 6AP1 51 72.56 89AP2 55 71.13 84AP3 56 72.69 88SP1 0 3.01 6SP2 0 2.94 6

Matthew Pan Extended Essay Candidate #: D 0130 013 Data Table 3 – Boarding Procedure vs. Minimum, Average and Maximum Number of Aisle Interference Occurrences for 100 Replications

Boarding Procedure Minimum Aisle Interference

Average Aisle Interference

Maximum Aisle Interference

RAND 39 52.04 68B1 39 52.38 64B2 40 53.04 68B3 38 51.57 65B4 39 52.27 63B5 42 52.74 68B6 40 53.25 65B7 35 53.41 69B8 41 53.32 67B9 43 54.18 67HR1 32 43.31 54HR2 38 49.55 63HR3 33 47.18 59HR4 38 50.16 64HR5 35 51.73 63HR6 39 51.62 63R1 44 54.31 67R2 48 60.19 70C1 30 42.05 55C2 28 42.02 57RP1 30 42.64 62RP2 28 43.24 57RP3 31 41.96 58RP4 30 41.68 59RP5 28 41.95 59RP6 26 41.28 53U1 34 47.04 57U2 35 49.46 64U3 33 45.97 61U4 30 45.54 60U5 33 44.71 59S1 2 6.83 11AP1 41 54.33 69AP2 41 54.72 69AP3 40 54.65 67SP1 26 41.01 57SP2 26 41.58 55

Matthew Pan Extended Essay Candidate #: D 0130 013 [Insert FINALRESULTS.xls GRAPH PREVIOUS TO THIS PAGE] Visual Layout of Boarding Processes

RANDOM (RAND)

B1 B2 B3

B4 B5 B6 B7


B8 B9 HR1 HR2

HR3 HR4 HR5 HR6


R1 R2 C1 C2

RP1 RP2 RP3 RP4


RP5 RP6 U1 U2

U3 U4 U5


24 70 116 93 139 47

46 92 138 115 69 23

22 68 114 137 91 45

6

7 53 99 122 76 30

55 101 124 78 32

33 79 125 102 56 10

110 64 18

44 90 136 113 67 21

65 19 111 88 134 42

135 20 66 112 89 43

39 85 131 108 62 16

17 63 109 132 86 40

41 87 133

82 36

37 83 129 106 60 14

15 61 107 130 84 38

128 105

35 81 127 104 58 12

13 59

11 57 103 126 80 34

31 77 123 100 54 8

9

52 98 121 75 29

74 28 97 120 51 5 4 96 50

25 71 117 3 49 95 118 72 26

27 73 119

2 48 94

AP1 (ALASKA

AIRLINES)

AP2 (DELTA

AIRLINES)

AP2 (UNITED

AIRLINES)


SP1

SP2

Matthew Pan Extended Essay Candidate #: D 0130 013 Analysis The results obtained from the simulation were graphed and then compared. The graphs can be found following this page. Individual Boarding Procedure Statistics

• Procedure with shortest elapsed time: RP2 o Average time of 100 replications: 1364.9976 seconds (22.749960

min)

• Procedure with the lowest number of seat interferences: C1 o Average seat interference occurrence of 100 replications: 2.77

• Procedure with the lowest number of aisle interferences: S1 o Average aisle interference occurrence of 100 replications: 6.83

Category Statistics

• Categories that have considerably short average elapsed time compared to other boarding procedures: RAND, C, RP, U, S and SP

• Categories with considerably low average of seat interference occurrences

compared to other boarding procedures: C, RP, S, SP

• Categories with lower average of seat interference occurrences compared to other boarding procedures: C, RP, S, SP

Overall Simulation Statistics

• Mean of average boarding times: 1449.072989 seconds (24.15121648 min) • Mean of average occurrence of seat interference: 45.84

• Mean of average occurrence of aisle interference: 47.54

• Confidence Interval Range for 100 replications of each boarding procedure:

> 60 seconds.

Matthew Pan Extended Essay Candidate #: D 0130 013 Evaluation of Boarding Categories Due to word count restrictions, only select boarding method categories that perform well and/or are interesting will be discussed. Others can be found in appendix 4. By Column

“By column” boarding, according the data collected from the simulation, seems to be an excellent method category for boarding. As one might be able to interpret from the graphs, boarding by column performs well in all statistical tests (when excluding the results of the “by seat” boarding method). The boarding processes included within this category are relatively simple which benefits everyone. Also, The very small and low range of seat interferences (maximum: 6, minimum: 0, average: 3) allows for passengers to be seated comfortably without having to slide past a seated passenger. The below-average number of aisle interferences also contributes to passenger approval in that less onboard line-ups will occur while enplaning. Overall, the “by column” category of boarding methods performs very well in terms of low interferences.

By Reverse Pyramid Type

Based on a single boarding strategy designed by the Arizona State University, this type of boarding category had the purpose of optimizing the original reverse pyramid method. Boarding by “reverse pyramid type” proved statistically to be one of the best boarding procedure categories tested. It was the category that contained the fastest boarding procedure (RP2) with a time of 1364.9976 seconds (22.75 minutes). Also, the category had relatively low values for the number of seat and aisle interferences, which will result in, as stated previously, greater levels of customer satisfaction. However, one person has claimed the reverse pyramid method would make “the general public… be more confused.”10 Overall, this category is one of the best boarding categories tested with RP1 and RP2 as best boarding procedures within the category in terms of time and number of interferences.

Seating

The process of boarding passengers individually was only included in the simulation to confirm Van Landegham’s findings. Technically, this pattern did perform the best

10 Mattern, H. (2003, May 1). AmWest Hopes new Boarding Procedure Will Save Time for All Involved.


within the collection of simulated boarding procedures, as also described by Van Landeghem in his study11, but one must look at the feasibility and the practicality of the boarding method. Boarding by individual seat would require every single passenger to line up in a certain specific order, which cannot be expected to happen. Also, if such a boarding method were to be implemented, waiting passengers might be frustrated at the “apparent” long-winded method and gate agents would have to call out for a passenger approximately every nine seconds or less (which makes concentration on processing passenger’s boarding passes increasingly difficult). Therefore, this paper will exclude the “by seat” boarding method from the official conclusion and report of findings.

Airline Pattern

This category’s purpose was to see how new patterns developed by airlines compared with the other boarding procedures. In general, they performed poorly when judged against boarding methods in categories C, RP, U, and SP.

Special Patterns

This category consisted of two boarding procedures that did not fit with any other boarding category. One of these boarding methods (SP1) was provided by Menkes H.L. van den Briel of the Arizona State University, who had stated that SP1 performed almost as well or even slightly better than his reverse pyramid boarding method that he tested for American West. SP2, being an experimental design, performed exceptionally well as it almost matched up with and even exceeded some of the results of SP1 (eg. average time). Both of these boarding procedures worked surprisingly well and even better than the reverse pyramid category in terms of aisle interferences.

11 Landeghem, H. Van. (2000). A Simulation Study of Passenger Boarding Times in Airplanes.

Matthew Pan Extended Essay Candidate #: D 0130 013 Conclusions and Implications Instead of finding one definite boarding method that was the fastest and most passenger-aware, there were several that performed almost equally well in all statistical simulations. Because of decreased numbers of interferences than all other boarding methods (except “by seat”), passenger safety and satisfaction was improved, which met the objectives of this study. The most effective boarding procedures included the following:

• RP1 • RP2 • C1 • C2 • SP1 • SP2

Further studies into this topic could possibly include optimization of these selected boarding processes and implementation into an airline. As this study did not include many other possible boarding procedures due to lack of resources and time.

Matthew Pan Extended Essay Candidate #: D 0130 013 Successful Boarding Procedures

SP1

SP2

RP1 RP2

C1 C2

Matthew Pan Extended Essay Candidate #: D 0130 013 Appendix Random

The results of the random boarding method were quite surprising, which was also noted by Van Landeghem in his paper12. This boarding procedure had a relatively low boarding time, which beat 13 of the 37 (35 percent) boarding methods studied.

Some advantages of using this boarding method include:

• The random boarding method need to use a call-off system. The only two announcements that a gate agent will need to make is to announce for first class passengers to board and then all other passengers. This means less time used in time is used for call-out, and the gate agent may be able to concentrate more on checking boarding passes.

• Short turn time. Some disadvantages include:

• The “first come first serve” mentality that most people will have in this situation will push them to try to get onto the plane and into their seat as quickly as possible. To accomplish this goal, some people may push and shove while many others will attempt to “budge” ahead of others in a lineup (which frequently happens at venues such as a theatre or a carnival ride). As a result, other passengers may get frustrated.

• Senior Citizens and parents with children may have a hard time trying to

board in a random method, leading to passenger dissatisfaction.

• As mentioned in the “Importance of Topic” section of this paper, it is quite possible that things might become chaotic due to a random event such as somebody tripping in the aisle while others are rushing to get to their seats.

• The high number of seat and aisle interferences may make passengers disenchanted with the airline, as they are placed in uncomfortable situations.

Overall, this boarding method is not effective according to the data, due to the impact and quantity of its non-quantifiable disadvantages and the high occurrences of seat and aisle interferences, despite a low boarding time. However, many no-frills discount airlines use this boarding procedure which may have contributed to their success. By Block 12 Landeghem, H. Van. (2000). A Simulation Study of Passenger Boarding Times in Airplanes.


This boarding category was one of the most utilized by the modern airline industry until major airlines began to experiment with other boarding methods recently. This category of boarding methods contained relatively average boarding times and high values for seat interference and aisle interference occurrences. Because boarding by block is a relatively simple and straightforward procedure, gate agents and flight attendants are able to easily memorize the boarding “by block” method and where groups sit. This makes them more efficient and possibly faster at processing passengers. Also, passengers will be less confused as to where they sit and which group they are in.

A disadvantage of using this boarding method is the relatively high number of seat and aisle interferences, which, like in the random boarding strategy, may contribute to an increase in customer dissatisfaction. Boarding “by block” is an acceptable boarding procedure to use when reviewing simulation data and non-quantifiable, although not one of the best.

By Half-Row and Half-Block

The idea of boarding by half-row and half-block was a concept proposed by Van Landeghem who stated that the best boarding time outside of the “by seat” boarding method was by half-row. His simulation predicted time for boarding using the “half-block” procedure was 15.8 minutes. The simulation used for this essay clocked the procedure to take about 25.6 minutes. Although it was anticipated that the results the simulation would differ from the results obtained by Van Landeghem, it unexpected that the predicted boarding times would differ by as much as ten minutes. Like boarding by block, this category of boarding procedures is relatively simple and straightforward to remember…which benefits passengers, gate agents and flight attendants. Also similar to boarding by block, there may be passenger dissatisfaction due to high numbers of seat and aisle interferences.

U Boarding

The U Boarding category is actually much like the reverse pyramid in design and concept, and therefore has much of the same results. The only major difference is that this category has many more seat interferences than the reverse pyramid category.

Matthew Pan Extended Essay Candidate #: D 0130 013 By Row

This category contains the procedure with the longest boarding time. R1 was clocked to take, on average, 1946.3418 seconds (32.43903 minutes). Even though the average time that it takes to board an aircraft using this boarding method is considerably longer than any other simulated strategy, it does not necessarily mean that boarding by row is not a very practical method. In fact, the other method in the “by row” category, R2, is only four minutes over the average. Concerning seat and aisle interferences, both procedures in the “by row” category do poorly as one might be able to see from the graphs. The only advantages and disadvantages that may have been considered to appear when using a system such as this are the same as previously discussed categories. However, because boarding by row has many groups, gate agents may have a harder time processing passengers while they are attempting to call-off the next group. Based on the results and inferences made, this boarding method category is not great, but still usable. However, more research and actual testing needs to be performed before a complete conclusion can be made concerning the “by row” boarding category.

Matthew Pan Extended Essay Candidate #: D 0130 013 References Funk, M. (2003, November). The Visualization of the Quantification of the Commodification of Air Travel. Popular Science.263 #5, 67-74. Gillie, J. (2004, February 13). Alaska ground crews take up the Tango. Tribnet. Retrieved March 4, 2004: http://www.tribnet.com/business/story/4737282p-4684895c.html Hogg, G.L, van den Briel, M.H.L., Villalobos, J.R. (2004, February). Development of Efficient Boarding Strategies at American West Airlines. Hogg, G.L, van den Briel, M.H.L., Villalobos, J.R. (2003, May). The Aircraft Boarding Problem. Landeghem, H. Van. (2000). A Simulation Study of Passenger Boarding Times in Airplanes. Marelli, S., Mattocks, G., Merry, R. (1998). The Role of Computer Simulation in Reducing Airplane Turn Time. Aero Magazine. Technology Product Development. Retrieved February 2, 2004 from the Internet: http://www.boeing.com/commercial/aeromagazine/aero_01/t/t01/index.html Mattern, H. (2003, May). AmWest hopes new boarding procedure will save time for all involved. Azcentral. Retrieved February 5, 2004 from the internet: www.azncentral .com/specials/specials42/articles/ Pilcher, J. (2004, January 24). Boarding System Works. The Cincinnati Inquirer. United Airlines. (2003, June 10). United Airlines Begins New and Simple Boarding Process. United Press Release. Retrieved February 5, 2004 from the internet: www.united.com/press/detail/0,1442,51084-1,00.html Acknowledgements Ms. C. Antonuk Mr. G. Donaldson Mr. D. Cantrill Special Thanks goes to: Menkes H.L. van den Briel

matthew pan extended essay candidate #: d 0130 013...

Documents