MSc Thesis_Devleoping a Simulation Model for Airport Check-In

Mihir Baxi

DEVELOPING A MODEL TO ANALYZE

IMPACTS OF SELF-SERVICE AND WEB

CHECK-IN AT AIRPORTS

School of Engineering

MSc Airport Planning and Management

CRANFIELD UNIVERSITY

School of Engineering

MSc Thesis

Academic Year 2006-2007

Mihir Baxi

DEVELOPING A MODEL TO ANALYZE IMPACTS OF

SELF-SERVICE AND WEB CHECK-IN AT AIRPORTS

Supervisor: Mr. Ralph Anker

September 2007

This thesis is submitted in partial fulfillment of the requirements for the degree of

Master of Science in Airport Planning and Management

© Cranfield University 2007. All rights reserved. No part of this publication may be

reproduced without the written permission of the copyright owner.

Developing a Model to analyze impacts of Self-service and Web Check-in at airports

Abstract

i

CRANFIELD UNIVERSITY Department of Air Transport

ABSTRACT

This research deals with the development of a simulation model to understand the

impacts of the self-service and web check-in and estimate the requirements for the

same. The research is divided in two parts.

Firstly, it deals with collecting the relevant data from the airports and analyzing the

data to identify the key parameters that affect the new check-in process. Secondly

these parameters and understanding are used to develop a simulation model to

estimate the resources and analyze the impacts of various what-if scenarios.

It was realized from case studies that there are many variables that affect the process

but the arrival profile and the processing time can explain the queuing patterns at the

airports. The simulation model developed is based on this understanding. To keep the

model flexible arrival profiles have been kept as a variable input.

As a result of research, a simple tool which could be used on any computer to analyze

and estimate the requirements for the self-service and web check-in at airports has

been developed successfully.


Acknowledgement

iii


ACKNOWLEDGEMENT

Joining Cranfield University was one of the toughest decisions taken by me, which

would not have been possible without inspirations from Dipan and Murali, who have

influenced my actions a lot and deserve much more than thanks.

I would like to thank my thesis supervisor Mr. Ralph Anker for his insightful discussions

which always gave me new ideas. Thanks, Mr. Richard Moxon for taking pain to

arrange the visit to the airport.

A special thanks to Elizabeth Hegarty at LCY, Martyn Davies at MAN, and Nicky Stubbs

and Pete Hiller at LHR and to all others who made this research possible.

Lastly, I would like to show my deepest gratitude to my parents and brother who have

unlimited confidence in my capabilities to succeed and given me strength to do so.

Hope I meet their expectations.

Finally, I think this work will be incomplete if I do not mention the name of the person

who has given unconditional support and believed in my stupefying talks for whole

year. Mansi, this was never possible without your dedication and sacrifices.


Table of Contents

v


TABLE OF CONTENTS

Abstract .................................................................................................................... i

Acknowledgement .................................................................................................. iii

Table of Figures ....................................................................................................... xi

Table of Tables ...................................................................................................... xiii

Acronyms............................................................................................................... xv

1.0 Introduction................................................................................................... 1

1.1 Capacity Constraints at Airports.................................................................................... 1

1.2 New Technologies and IATA Initiatives ......................................................................... 2

1.3 Problem Statement ....................................................................................................... 4

1.4 Research Objectives ...................................................................................................... 5

1.5 Research Methodology ................................................................................................. 5

1.6 Thesis Outline................................................................................................................ 6

2.0 Literature Review........................................................................................... 7

2.1 Research Theses............................................................................................................ 7

2.1.1 Reducing the air travel hassle factor through Self-service Check-in process

improvements....................................................................................................................... 7

2.1.2 Simulation of Passenger flow in Self-service Check-in........................................... 8

2.1.3 Common User Self-service Check-in: Benefits to the air transport Industry.......... 9

2.2 Books and Peer Reviewed Papers ............................................................................... 10

2.2.1 Books and Standard References.......................................................................... 10

2.2.2 Peer Reviewed Papers ......................................................................................... 10

2.3 Conclusions ................................................................................................................. 12

3.0 Self-Service and Web Check-In ..................................................................... 15

3.1 History of Self-service Kiosks....................................................................................... 15

3.2 Self-service at The Airports ......................................................................................... 16

3.3 Check-in Process.......................................................................................................... 17

3.3.1 The Traditional Check-In Process......................................................................... 17

3.3.2 Check-In with Kiosks ............................................................................................ 18

3.4 Check-in Configuration................................................................................................ 19

3.4.1 One Step .............................................................................................................. 20

3.4.2 Two Step.............................................................................................................. 21

3.4.3 Three Step............................................................................................................ 21

3.5 Conclusions ................................................................................................................. 21


Table of Contents

vii


4.0 Case Studies................................................................................................. 24

4.1 Methodology............................................................................................................... 24

4.2 Case Study 1- London City Airport, London ................................................................ 25

4.2.1 Arrival Profile....................................................................................................... 26

4.2.2 Processing Time................................................................................................... 27

4.2.3 Queuing Time ...................................................................................................... 30

4.3 Case Study 2- Manchester Airport .............................................................................. 32

4.3.1 Arrival Profile....................................................................................................... 32

4.3.2 Processing Times ................................................................................................. 33

4.3.3 Queuing Time ...................................................................................................... 36

4.4 Case Study 3- London Heathrow Airport .................................................................... 37

4.4.1 Arrival Profile....................................................................................................... 38


4.4.3 Queuing Time ...................................................................................................... 40

4.5 Conclusions ................................................................................................................. 41

5.0 Simulation Model......................................................................................... 43

5.1 The Approach .............................................................................................................. 43

5.1.1 Arrival Profile....................................................................................................... 43


5.1.3 Service Standards ................................................................................................ 46

5.2 Description of Model................................................................................................... 47

5.3 Assumptions................................................................................................................ 49

5.4 Validation .................................................................................................................... 50

5.4.1 Comparison to Existing Data ............................................................................... 50

5.4.2 Comparison to Existing Standards....................................................................... 52

5.4.3 Concluding Comments......................................................................................... 53

5.5 Conclusions ................................................................................................................. 54

6.0 Application of Model.................................................................................... 55

6.1 Existing Situation......................................................................................................... 55

6.1.1 As Is Model .......................................................................................................... 55

6.1.2 Scenario 1 ............................................................................................................ 56

6.1.3 Scenario 2 ............................................................................................................ 57

6.1.4 Scenario 3 ............................................................................................................ 57

6.1.5 Discussions .......................................................................................................... 58

6.2 Other Experiments ...................................................................................................... 59

6.2.1 Scenario 4 ............................................................................................................ 59

6.2.2 Scenario 5 ............................................................................................................ 60

6.2.3 Discussions .......................................................................................................... 61

6.3 Conclusions ................................................................................................................. 62


Table of Contents

ix


7.0 Discussions .................................................................................................. 63

7.1 Dedicated versus Common Use .................................................................................. 63

7.2 CUSS versus Web Check-In.......................................................................................... 64

7.3 Emerging Technologies ............................................................................................... 65

7.4 Bag Drop-off ................................................................................................................ 65

7.5 Conclusions ................................................................................................................. 66

8.0 Conclusions.................................................................................................. 67

8.1 Overall Discussion ....................................................................................................... 67

8.2 Statement of Research Value...................................................................................... 69

8.3 Further Work and Research ........................................................................................ 69

8.4 Final Conclusions......................................................................................................... 71

Works Cited ........................................................................................................... 73

Further Reading...................................................................................................... 77

ANNEXURE A

Results from Statistical Analysis for Processing Times

ANNEXURE B

Simulation Model: Description and Users Guide

ANNEXURE C

Results from Simulation Model: Scenario 1 – Scenario 5


Table of Figures

xi


TABLE OF FIGURES

Figure 1-1 Penetration of CUSS at airports Compiled from: (IATA(b) n.d.) ................ 3

Figure 3-1 Passenger Processing Flow........................................................................... 17

Figure 3-2 Check-in through Kiosk.................................................................................. 19

Figure 3-3 Typical Stand Alone Kiosk Photo by Mihir Baxi .......................................... 20

Figure 4-1- Self-service Kiosks for Air France and CityJet Photo by Mihir Baxi .......... 25

Figure 4-2 Arrival Profile for Air France, LCY .................................................................. 26

Figure 4-3 Arrival Profile for VLM, LCY ........................................................................... 27

Figure 4-4 Processing Time per Pax - Self-service Kiosks, LCY ....................................... 28

Figure 4-5 Processing Time per Pax - Bag Drop-off, LCY ................................................ 29

Figure 4-6 Processing Time per Pax- Check-in Counters, LCY ........................................ 29

Figure 4-7 Queuing at Air France Kiosks, LCY ................................................................. 30

Figure 4-8 Queuing at Air France Bag Drop-off, LCY ...................................................... 31

Figure 4-9 Queuing at VLM Check-in Counters, LCY....................................................... 31

Figure 4-10 Self-service Kiosks at MAN Photo by Mihir Baxi...................................... 32

Figure 4-11 Arrival Profile - Monarch, MAN................................................................... 33

Figure 4-12 Processing Time per Pax- Self-service Kiosk, MAN ..................................... 34

Figure 4-13 Processing Time per Pax - Bag Drop-off, MAN............................................ 35

Figure 4-14 Processing Time per Pax - Check-in Counters, MAN................................... 35

Figure 4-15 Queuing at Monarch Check-in Counters, MAN........................................... 36

Figure 4-16 Self-service Kiosks at LHR Photo by Mihir Baxi....................................... 37

Figure 4-17 Arrival Profile for Lufthansa, LHR ................................................................ 39

Figure 4-18 Processing Time per Pax - Self-service Kiosks, LHR..................................... 40

Figure 4-19 Processing Time per Pax - Bag Drop-off, LHR.............................................. 40

xii Developing a Model to analyze impacts of Self-service and Web Check-in at airports Table of Figures


Figure 5-1 Arrival Profiles ................................................................................................. 1

Figure 5-2 Process Model for Simulation ....................................................................... 48

Figure 5-3 Simulation result of wait times for MAN ...................................................... 51

Figure 5-4 Actual wait times at MAN ............................................................................. 52

Figure 5-5 Arrival Profile for Simulation representing IATA assumptions ..................... 53

Figure 7-1 Check-in Mode Forecast by IATA .................................................................. 65


Table of Tables

xiii


TABLE OF TABLES

Table 2-1 Processing Times Compiled by Author, based on (Lejarraga 2004) ............ 9

Table 4-1 - Details Observed at the Airports.................................................................. 24

Table 4-2 Flight Schedule Air France, LCY....................................................................... 26

Table 4-3 Flight Schedule VLM, LCY................................................................................ 26

Table 4-4 Processing Times at LCY.................................................................................. 28

Table 4-5 Passenger Wait Times for AF and VLM, LCY................................................... 30

Table 4-6 Flight Schedule Monarch, MAN...................................................................... 33

Table 4-7 Breakups for Passenger Check-in for Monarch, MAN.................................... 34

Table 4-8 Processing Times for Monarch, MAN............................................................. 34

Table 4-9 Passenger Wait Times for Monarch, MAN ..................................................... 36

Table 4-10 Flight Schedule Lufthansa, LHR .................................................................... 38

Table 4-11 Processing Times for Lufthansa Airlines, LHR............................................... 39

Table 4-12 Passenger Wait Times at LHR, Lufthansa Airlines ........................................ 41

Table 5-1 Processing Times and Distribution profile...................................................... 45

Table 5-2 Inputs in the Simulation Model ...................................................................... 50

Table 5-3 Comparison with the existing and simulation results .................................... 51

Table 5-4 Results for the simulation of 2500 TPHP........................................................ 53

Table 6-1 Total Process Time for As-is model ................................................................ 55

Table 6-2 Summary of the Results.................................................................................. 56

Table 6-3 Total Process Time for Scenario 1 .................................................................. 56

Table 6-4 Summary of Utilization of Servers.................................................................. 57



xiv Developing a Model to analyze impacts of Self-service and Web Check-in at airports


Table 6-7 Service Standards for the Scenario 4 and 5.................................................... 59

Table 6-8 Summary of the results for Scenario 4 & 5 .................................................... 60

Table 8-1 Processing Times for all Check-in Modes ....................................................... 67


Acronyms

xv


ACRONYMS

ADRM - Airport Design Reference Manual, 9th Edition, IATA

AF - Air France

ANOVA - Analysis of Variance

ATM - Automatic Teller Machine

BA - British Airways

BAA - British Airport Authority

BCBP - Bar Coded Boarding Passes

CUSS - Common User Self-Service

CUTE - Common User Terminal Equipment

DLR - Dockland Light Railway

ET - Electronic Ticket

EU - European Union

IATA - International Air Transport Association

IRSS - Intelligent Resource Simulation Systems

LCY - London City Airport

LHR - London Heathrow Airport

MAN - Manchester Airport

MWT - Maximum Wait Time

Pax - Passengers

RFID - Radio Frequency Identification

StB - Simplifying the Business

STD - Standard Time of Departure

TPHP - Typical Peak Hour Passengers


Introduction

1


1.0 INTRODUCTION

The aviation industry has grown at an unprecedented rate recently due to various

reasons. The EU aviation market has shown growth because of liberalization and low

cost airlines, while the Asian market is growing faster than ever because of strong

economic growth in Asia. In 2006, the airports of emerging economies have shown the

growth in double digits. The busiest hubs of India Mumbai and Delhi, have shown 20%

growth in the last year. A few European airports like Dublin, Hamburg and Oslo have

also shown double-digit growth rates. At the global level, air traffic showed 5% growth

(Pilling 2007a).

The Global Traffic Forecast, ACI1 has predicted traffic growing at 4.6% in the following

decade and for the next 20 years the growth rate is to be 4% (Pilling 2007a). The

Airbus Global Forecast shows that on average the traffic is supposed to grow at 4.8%

per annum until 2025 (Airbus 2006). The report also forecasts a doubling of

frequency, which might create a problem for the airports world over for infrastructure

and air space management.

To cope up with the growth airports have to expand the terminal facilities and meet

new standards of operational efficiency. The growth in traffic requires a huge

investment by airports to develop new terminals and airside facilities. The capacity at

airports is constrained if new projects are not implemented on time. The new

technologies are evolving to relieve the congestion at airports, requiring airports to

change their operational strategy and expansion plans.

1.1 CAPACITY CONSTRAINTS AT AIRPORTS

Air traffic is growing continuously, but the development of airports is constrained by

the funding and space available to expand the terminal. The new projects have also to

take care of the new environmental constraints along with the operational constraints

already existing at a site like Heathrow Terminal 5. There has been a situation of

unacceptable queues at Heathrow and the only solution is more facilities (ATI News

2007).

Pan European air navigation agency Eurocontrol is warning that lack of airport capacity

will constrain the growth in traffic by 2025 (Kaminski-Morrow 2004). Assuming no

capacity constraints on air traffic and airport capacity, traffic is supposed to grow at

1 Airports Council International is a forum of airports worldwide and they publish reference documents

relevant to the industry.

2 Developing a Model to analyze impacts of Self-service and Web Check-in at airports

Introduction


least 2.5 times of the 2003 level (Toole and Thompson 2005). Airport capacity at

present seems unlikely to meet these demands across the system.

To meet new capacity demands the airports will need the investment of US $55 billion

in Europe for infrastructure until 2025 (Toole and Thompson 2005). ACI also suggest

that the current charge structure at the airports will not be sufficient to finance the

projects. A similar situation exists at US airports where the airports are favoring a

change in policy for the passenger facility charge to finance the new infrastructure

projects (Hughes 2005). Airports in the US estimate they will need US $71.5 billion for

the capital improvements projects from 2005 to 2009 and they will fall short of US $3-4

billion a year (Hughes 2005).

The expansion of airports is a complex exercise and needs to address issues like

operational constraints and the environment along with funds. In the light of

environmental restrictions on a new development, projects take longer to complete

and have to follow stringent regulations for completion. As a result, airport capacity is

constrained and we find longer queues at the terminals. The new security

requirements and the frequent changes increase inconvenience to passengers and

there is a need for improving the process.

1.2 NEW TECHNOLOGIES AND IATA INITIATIVES

In response to the situation, airports have started implementing new technologies at

the terminal for convenience of the passenger. The new solutions strive to improve

operational efficiency and reduce queues at the airport. The new technology like self-

service and web check-in are being installed at many airports to increase Check-in

capacity. Technologies like biometrics and Iris recognition are being implemented for

increased security measures, while radio frequency identification is being tested for

improved baggage tracking and management.

IATA is taking various initiatives to improve the passenger experience, while taking into

consideration the industry wide point of view. In the 2004 Annual General Meeting

IATA won the airlines backing to improve efficiency and reduce costs. The agenda of

Simplifying the Business (StB) was to improve the basic elements of travelers’ journeys

while cutting through the expenses of the conventional legacy of an expensive and

complex system (SITA(a) n.d.). The projects were selected to achieve increased

efficiency for the airlines and cut down the costs of operations, while improving the

customer service and making it possible to implement them on a larger scale (IATA(a)

n.d.).


Introduction

3


The programme Simplifying the Business (StB) includes five projects in its initial stage,

which are as follows:

1. E-Ticketing (ET)

2. Bar Coded Boarding Pass (BCBP)

3. IATA e-Freight

4. Common Use Self-service Check-in (CUSS)

5. Radio Frequency Identification for Aviation (RFID)

The Role of IATA is to educate and bring awareness to the industry about the common

vision for simplified business model. It will also encourage adaptation of common

standards and provide support and necessary market intelligence to all the

stakeholders.

Most of the technologies are on their way to implementation at the airports. The e-

Ticket has been widely implemented by airlines as they see it as greatly beneficial. The

savings from e-ticket are obvious and the advent of low cost airlines has speed up the

process of implementing this initiative.

Secondly, the CUSS kiosks are being installed at a number of airports and it allows

passengers to check-in for multiple airlines. The cost saved per passenger being

checked in by CUSS is US$ 2.50 (IATA(b) n.d.). Many airlines are already using the

dedicated self-service Kiosks and IATA is pursuing airports and airlines to adapt CUSS

as the industry standards. The Figure 1-1 shows the penetration of CUSS worldwide.

Figure 1-1 Penetration of CUSS at airports Compiled from: (IATA(b) n.d.)


Introduction


In addition it is being advocated that web check-in will gain more importance in the

coming years. All major airlines now allow passengers to check-in through the web,

which is enabled due to BCBP. Each project has its own benefits, but each also plays

an important role in enabling airports and airlines to provide the passenger with a full

self-service model. Thus it is essential to link all the three systems of e-ticket, CUSS

and BCBP to realize significant profit and efficiency.

1.3 PROBLEM STATEMENT

In recent years the airlines have started using innovative Check-in technologies. The

dedicated Self-service Check-in kiosks are now an integral part of the airport facility

and the new e-ticket also allows the use of the internet for check-in. These processes

enable airlines to reduce the time and number of staff required, thus saving

substantial costs in operations. These changes in the process have allowed airports to

handle more passengers in the same space (Weiss 2006) and the winding queues have

been reduced. But there are many issues related to the implementation of the self-

service check-in.

Further to this, IATA is advocating CUSS, similar to standardized ATMs at banks. New

CUSS standards will allow the sharing of kiosks among airlines similar to CUTE and also

give access to the technology for the smaller airlines. The advantages of adopting

CUSS have been discussed in many papers and conferences alike and will be discussed

later in detail.

The impacts these new check-in technologies have on the design and operation of the

airports and how far it has been successful in reducing queues is not quite evident.

There is a need to understand the process and its implication on the effects that it will

have on operations and implementation. Also it is essential to study what kind of

changes will be required for moving from one technology to another.

Also, there are no set standards for installing the system at the airports. Because

technology is new, IATA does not have any standard procedures or thumb rules to

estimate the requirements. There are many sophisticated simulation tools available

that need a lot of inputs and are costly for an airport to acquire. The author feels that

simpler tools, easily accessible to all will be sufficient for the problem and be more

valuable. It is thus necessary to establish the common standards for estimating

requirements for self-service and web check-in facilities.


Introduction

5


1.4 RESEARCH OBJECTIVES

The research objectives are as follows:

• To establish the key parameters that affect the self-service and web check-in

process and the factors that influence them.

• To develop generic standards which could be used to determine the efficiency

and estimate the resources.

• To develop a simulation model with the understanding gained for estimating

resources for the whole check-in system for any airport.

• The model should be simple enough to understand and should use commonly

available software, so that it is accessible to all.

• To add value to the industry understanding of the self-service and web check-in

process in a tangible way for future use.

1.5 RESEARCH METHODOLOGY

In order to achieve the stated objectives, the author will conduct case studies at a

number of different airports to understand the process and also collect the primary

data to be used in the simulation model. The purpose of visiting different airports is to

understand the variations in the process that exist at all airports and gather as much

data as possible.

The method of data collection will basically include observing the passenger in the

process and other key issues at the airports. The data collected will be analyzed to

understand the differences and various factors that might affect the check-in process

at airports. The collected data will also help in establishing key parameters for the

simulation model.

The key parameters will be used as inputs for the simulation model to be developed in

MS Office Excel. The simulation model will be validated against the existing situation

and industry standards wherever they exist. This validated simulation model can be

used for various experiments to explain the behavior of the check-in system as a whole

and provide the airport planners with a tool to estimate requirements for check-in

systems.


Introduction


1.6 THESIS OUTLINE

This chapter explains the current position of self-service and web check-in in the

industry and discusses the problems in determining the requirements and operations

of new technology at the airport. The second chapter, literature review, looks into the

various existing literature on and around the subject. It covers the literature explaining

the general understanding and review of papers on check-in simulation and resource

allocation.

Chapter three explains the overall understanding of the process and steps involved in

the self-service check-in. It enables the reader to understand the process of check-in

with reference to self-service and web check-in. The fourth chapter discusses the case

studies carried out at three airports and documents the critical observations for arrival

profiles, processing times and queuing patterns. The next chapter deals with the

process of developing the simulation model and discusses the parameters that affect

the process. The critical assumptions and the validation of the model are also

discussed in the same chapter.

The Sixth chapter explains the application of model and key observations regarding the

implementation of the new check-in process. This chapter explains the impacts of the

new process with the help of various scenarios to estimate the requirements at airport

in various situations. Chapter seven discusses the other technological developments

that might affect the use of the model and results. The concluding chapter

summarizes the learning from the research and brings all the understanding together.


Literature Review

7


2.0 LITERATURE REVIEW

It is essential to find out what has already been done regarding the research that is

being carried out to understand the scope and need of the research. The self-service

and web check-in have been installed recently at the airports and there is not much

research on the resource allocation models or the provision of space required for the

implementation of such technology. However, there are a large number of papers and

articles showing the importance and benefits of self-service and web check-in at

airports. The literature review is divided in two parts; the first part discusses the

theses that are available on similar topics and the second section looks at what is

available as the standard reference on the subject and discusses various papers dealing

with the simulation and resource allocation modeling for Check-in counters.

2.1 RESEARCH THESES

There are three theses available on the similar topic of interest; two of them are from

Cranfield University and one from Massachusetts Institute of Technology. The

research findings of each thesis and important points are discussed in details.

2.1.1 Reducing the air travel hassle factor through Self-

service Check-in process improvements

This research thesis mainly deals with human factors involved in the implementation

of the self-service check-in at airports. The thesis looks into the role of roving check-in

agents and the usage of the self-service facilities. The author has analyzed the process

at three US airports, namely Houston, Cleveland and Newark and has brought up

interesting facts in terms of the operations of the self-service check-in by Continental

Airlines.

The research identified some key issues in the implementation of the services, as

usage of the kiosks was dependent on the activity of the agent assigned to help the

passengers. One more aspect was how efficiently do the agents perform and

encourage the use of kiosks by passengers. It was observed that at times the agents

performed the whole process for the passenger. In addition, the use of machines and

the number of passengers waiting in line were dependent on the number of kiosks

assigned to each check-in agent.


Literature Review


The author also conducted a survey amongst the check-in agents and passengers using

the self-service kiosks. The findings for the attitude of check-in agents towards the

self-service kiosks varied with the size of airport as they had different roles at different

airports. At spoke airports, the self-service kiosks were seen as technology that

enables the agents to perform the role more efficiently as they were responsible for

many other activities. While at the main hub airports it was seen as a threat to the job

and was not taken positively as check-in agents were only responsible for check-in.

The passenger survey showed some positive results and it was seen that self-service is

seen as a positive move by passengers. The passengers were ready to take control of

the process and more than happy to check-in by themselves. Most of the passengers

also showed the inclination to use the self-service rather than normal check-in if both

took the same time.

2.1.2 Simulation of Passenger flow in Self-service

Check-in

The main purpose of this thesis was to create a simulation tool to estimate the

requirement for the Self-service check-in process for the low-cost carrier easyJet. The

author carried out observations at East Midlands Airport, which was the only airport

with self-service check-in for easyJet at the time of the research thesis.

The author collected detailed information on the arrival patterns and profiles of the

passenger. easyJet was insisting on 100% check-in using kiosks and the author

collected information regarding the processing times for using kiosk and baggage drop-

off. The model was developed using the Witness simulation software, used by BAA

and many other airlines for the evaluation of the process at airports.

A generic model was developed so that it can be used at any airport to estimate the

requirements for check-in by easyJet. The interface was designed in Excel for the

convenience of the users. The check-in process that was considered for modeling

included the passengers tagging the bags printed by the kiosks by themselves and then

dropping the bags at baggage drop off. The number of baggage drop-offs critically

defined the overall processing times.

The author observed that the processing times were normally distributed and were

different in two cases, with queues and without queues. The transaction times

considered were as follows:


Literature Review

9


Description Processing Times

Self-service Kiosk 1.6 Minute without queue 1.26 Minute with Queue

Tagging the Bag 0.45 Minute*Number of bags

Time for Bag Drop-off 0.35 + 0.18 Minute* Number of bags

Table 2-1 Processing Times Compiled by Author, based on (Lejarraga 2004)

The thesis dealt with estimating the resources for the self-service check-in process

dedicated to airlines with known flight schedules for the given time and day. The

model developed was a management and optimization tool to determine the required

resources.

2.1.3 Common User Self-service Check-in: Benefits to

the air transport Industry

The thesis discusses the benefits of CUSS to the air transport industry. The author has

compiled information regarding the benefits of the CUSS and showed to what extent it

can be beneficial to airports. The thesis gives a good overall view for the position of

the CUSS in the industry.

CUSS architecture has been explained in detail and various applications and interfaces

required for the implementation are included in the discussion. The author has also

calculated the costs for installing CUSS in the existing airport terminal at Belfast City

Airport. The analysis of the cost shows clear benefits to the airlines in terms of

manpower and for airports in terms of the cost of the infrastructure. The cost of

installing per kiosk is calculated at around GBP 29,680. There would be additional costs

for maintaining the inventory and training the staff.

Raymond also looked into the queuing systems and the time that the system will

require for the same number of passengers. The implementation of the CUSS shows a

significant improvement in the processing and waiting times for the passenger.

The author showed CUSS was successful in performing satisfactorily in different acute

situations. The processing time considered for check-in was 1.85 minutes and for the

kiosk it was 1.15 minutes, which was referred from other sources.

The author did not try to develop some model to estimate the requirements but

performed analysis to show the benefits to the stakeholders. The author has discussed

the benefits at large but also points out that there are implications on airports for

implementing CUSS, especially in fast bag drop-off and the security check area.


Literature Review


2.2 BOOKS AND PEER REVIEWED PAPERS

As explained the technology of self-service is very new and there are no papers

particularly on the use of the technology and resource allocation. This particular

aspect is not discussed in detail in the latest edition of IATA – ADRM2 and in the books

that are considered authoritative texts on the design of airports. There are many

papers available which look in to the resource allocation and simulation of the check-in

counters both dedicated and using CUTE3. To understand the process some of the

crucial ones are discussed in detail here.

2.2.1 Books and Standard References

The IATA – ADRM describes the factors to be considered for the design of the check-in

area for the check-in desks with CUTE. The manual provides some standard thumb

rules based on the queuing theories and which are very useful in sizing the overall

terminal at the initial stages. However, the manual as such is not adaptable and the

inclusion of new technology at the airport cannot be explained by the standard

formulas as the behavior of the passengers and space requirement changes. It is not

possible to estimate the requirement of fast bag drop-off for the self-service kiosks.

Richard de Nuefville and Amedeo Odoni in the book Airport Systems mentions that the

standards established for check-in will not work in the 21st Century as the technology is

continuously changing. Nuefville and Odoni (2003) mention that the electronic ticket

and kiosk check-in will reduce the space requirements for the check-in hall. They go

even further saying that the kiosk may eliminate the need of the traditional check-in

hall. It also mentions the benefits of the common use equipments as not all the

airlines will need extensive infrastructures like the one that is operating at the airport

for the whole day. CUSS can be beneficial to the airports in the same ways and give

the same facilities to all airlines.

2.2.2 Peer Reviewed Papers

Most of the papers dealing with the optimization and resource allocation suggest the

Check-in process is more appropriately represented by simulation rather than queuing

theory. Dijk and Sluis (2006) have looked into the Check-in computations and

2 ADRM- Airport Development Reference Manual 9

th Edition 2004 – Standard text that gives thumb rules

for sizing various facilities in the Terminal. 3 CUTE- Common User Terminal Equipment, the facilities at the airports are shared between the airlines

to reduce the space and resources required.


Literature Review

11


optimization by simulation and integer programming. The paper describes a process

for optimization, which involves two steps

• A computation and optimization of the number of desks for an individual flight.

• A minimization of the total number of the desks and staffing hours.

The paper deals with the stochastic aspects of check-in computations in the first step

and uses simulation to establish the initial requirements. Dijk and Sluis suggest using

deterministic techniques for optimization of the allocated resources. The paper favors

simulation for the check-in computation mainly for three reasons, which has uncertain

behavior.

• The number of actual travelers

• The traveler arrival times

• The check-in times for a traveler.

The paper uses the simulation results to estimate the hourly requirements of the desks

for check-in and also suggest using variable desk opening like Joustra and Dijk (2001) to

minimize the staffing hours and shared use for the available desks. The results of the

observation was that for most flights only one desk is required for the last hour if the

desks are opened before two and half hours before the departure times, unlike all the

desks being kept open at airports.

Joustra and Djik (2001) on similar lines show why the simulation is more appropriate

for check-in at airports and supports it with a case study at Amsterdam Airport. The

queuing at airports is explained as the function of the strong fluctuations and peak

over the day in the number of arriving passengers. The queuing theory is incapable of

explaining this variability and thus is not appropriate for calculating the required

check-in desks. The paper considers the variations for check-in like

• Common versus dedicated check-in

• Dynamic versus static opening and closing

• Extension of the Check-in period

• Overflow for economy class passengers

• Bank lining.

The authors consider the impact of various aspects on the queuing and resources

required for each aspect. It is noted that the queuing theory can be applied to the use

of common user Check-in as the arrival patterns will be fairly steady for the given set

of counters. The important point that is brought up is that it is essential to use

simulation if the results are to be reliable.


Literature Review


In another paper Park and Ahn (2003) explain in detail the model required to optimize

the assignment of check-in based on the passenger arrival pattern. The main structure

of the paper is similar and the factors considered for the calculations are the same as

other papers. The authors mention that arrival patterns of the passenger are

dependent on the flight departure times (time of the day), type of aircraft and the load

factors. It also depends on the type of operations including charter or scheduled. The

check-in requirements are not the same for all flights. Long haul international flights

might be handled two hours prior to the Standard Time of Departure (STD), whereas it

might be only thirty minutes for a domestic flight. The authors explained that the peak

hour at the airport would be earlier than the airside peak hour as the first passenger

will check-in three to four hours before the STD.

Authors have based the model on Seoul Gimpo International Airport (GMP). The

observations at GMP show that the average processing time for the passengers is 96

seconds. The passengers without baggage were processed in 68 seconds. To

understand the passenger arrival pattern a survey was conducted at the airport. Only

1.5 % of passengers arrived at the airport 170 minutes before STD and 90% of them

arrived 30 minutes prior to STD. Authors showed how these cumulative rates of the

arrival affect the check-in requirements.

The other interesting paper dealing with the check-in resource allocation is by Chun

and Mak (1999). The model is based on Hong Kong Kai Tak Airport. Intelligent

Resource Simulation System (IRSS) as described by the authors was used to estimate

the check-in requirements. IRSS uses historic data and the airport database for

different types of flights for arrival patterns and processing times for the passengers.

The input is given in the form of flight schedule and IRSS calculates the counters

required for each flight individually. The allocation model is a sophisticated one and

needs a huge amount of data. The model enables the user to model various “what-if”

scenarios easily and assess the possibilities for change. It also generates graphs and

tables to compare different scenarios to understand the results easily.

2.3 CONCLUSIONS

This chapter looks at the literature that is available on CUSS and related subjects. The

author has tried to collect different viewpoints and different aspects of CUSS. It is

important to understand the needs of new technology and the parameters that will

affect the use of the new system. CUSS will definitely benefit the airports as can be

seen in the discussion, but will have some implications for implementation. There are

no papers or trade articles discussing in particular the disadvantages of CUSS; neither


Literature Review

13


do they discuss problems related to fast bag drop-off or any other consequence. The

passengers are happy to embrace the new technology in general and airlines too are

moving towards the use of CUSS from dedicated kiosks.


Literature Review


The papers on the resource allocation discuss the parameters that affect the check-in

process and highlight the need of simulation against queuing theory. The important

parameters that affect check-in process are

• Passenger Arrival Pattern

• Processing Times at Check-in

The models described in this chapter consider check-in by common use terminals and

thus will be valid for CUSS. The common theme throughout the papers is to minimize

the resources and maintain the service standards, which the author thinks, will be well

addressed by self-service and web check-in.


Self-Service and Web Check-In

15


3.0 SELF-SERVICE AND WEB CHECK-IN

The concept of self-service is not a new concept with the passengers as they come

across self-service machines at various places like banks, shopping malls, railway

stations, etc. The airline industry has adopted the model of self-service by switching to

e-ticket where the customer completes the whole transaction on the computer

without any human interface. The self-service kiosks for checking in have been a

recent development at airports. The American airlines like Continental and Delta have

been pioneers in using this new technology for efficient passenger processing. The

self-service kiosks have become an integral part of these airline operations. The

European airlines realized the importance of the self-service and adapted the model.

In this chapter, the brief history of the self-service is discussed and the process is

described to understand the use of self-service and web check-in.

3.1 HISTORY OF SELF-SERVICE KIOSKS

The use of self-service is not a new concept; banks have been using it for a long time

now. The banking industry adopted the ATM concept for reducing costs and providing

better services for the customers. The first ATM was installed as early as 1967 by

Barclays Bank in London, UK (BBC News n.d.). The banks started installing ATM

machines in the bank buildings first and where a cash dispensing machine was not

linked to the account directly. With the spread of internet connectivity the ATM

machines have become a part of the urban landscape and available at parks, shopping

malls or airports with many more services on offer than just cash dispensing.

The adoption of self-service is gaining importance in other industries for two main

reasons, increased efficiency and reduced costs and labor. The customers are

technology ready as they have access to computers and internet at home, and they

feel comfortable interacting with the machine. It has been observed that customers

are now more open to experiment with kiosks (Murphy 2007). Other industries like

retail, finance, hotels, etc are considering use of self-service kiosks. Even libraries are

trying out self-service technologies to issue books. The customers are now ready to

make bigger transactions with kiosks and many models for self-service have been put

to the test (Maras 2006).

It will be interesting to note an example of the fully automated RedBox DVD Rental

kiosk (Murphy 2007). The kiosk offers new DVDs for rent and can hold more than 500

DVDs with new titles being added every week. It has been observed that it is not the

choice of titles but the idea of instant access to the service that is more important in




selecting to use the kiosk. It is worth noting that eventually people will not need discs

as they will be able to download full movies electronically and the kiosks might

become useless. Thus, the use of self-service might be an intermediate solution for

providing services as new technology and ways of distribution emerge in the market.

3.2 SELF-SERVICE AT THE AIRPORTS

Self-service kiosks were first introduced by Continental Airlines in 1995 at US airports

(Miller 2003). Since then the self-service has become an integral part of providing

services for passengers. Most schedule airlines now provide the option for self-service

kiosk check-in at major airports. The low-cost airlines like easyJet insist on 100 %

check-in through kiosks at smaller airports to reduce labor costs (Lejarraga 2004). The

cost of check-in through kiosks is just $0.16 as against $3.68 with normal check-in with

an agent (Weiss 2006). The airports have realized the benefits of the kiosks and IATA is

now promoting the installation of CUSS instead of dedicated check-in.

The Airport IT Trends4 for 2006 shows a rise in deployment of CUSS kiosks as against

the dedicated one. There are only 8% airports planning to deploy dedicated kiosks as

against 60% to implement CUSS. The airports and airlines have understood the

importance of the shared facilities. The airports gain higher throughput and reduce

the clutter of airline specific kiosks (Conway 2006). Vancouver Airport has increased

throughput by 250% by installing CUSS. The airport has seen a 25% increase in

domestic traffic, but the airport now employs 30% fewer check-in staff. Thus has been

able to postpone the expansion plans until 2012 (Conway 2006). Another example is

Las Vegas McCarran Airport, which has installed 100 CUSS kiosks at the airport and

installed some off airport as well (Weiss 2006). McCarran airport claims to be a 100%

common use airport. It has invested $1-2 million in CUSS, which saves the airport from

a building expansion of $20 million (Conway 2006).

Similar to the banking industry now the check-in kiosks are moving away from the

terminal buildings. The kiosks at McCarran Airport are located in parking areas or

hotel lobbies (Conway 2006). The passengers could check-in from their hotel lobby or

from a convention center. It is simpler for the passengers with no baggage, whereas

with baggage collection it requires a lot of effort from the airport to make it work. The

off-site check-in helps free up space in the terminal during peak hours. The offsite

check-in kiosks are now being installed at railway stations and airport hotels in Europe

too.

4 ACI-SITA Airport IT Trends is a worldwide survey of around 200 airports investment strategy for IT

Infrastructure.



17


3.3 CHECK-IN PROCESS

Airports provide facilities for processing the passengers efficiently and comfortably.

Boarding the aircraft involves following systematic steps and the capacity of the

terminal depends on the capacity of each system. In the last couple of years, the

process has changed for various reasons like the change in security regime. The

airports need to accommodate new technology like kiosks, Biometrics, RFID, etc. This

section explains the traditional check-in process and the modifications in the process

by installing self-service kiosks.

3.3.1 The Traditional Check-In Process

Airport passenger processing is a systematic process and the airport capacity is as good

as the weakest link. The traditional passenger processing is shown in the Figure 3-1 in

blue. The passenger arrives at the airport and approaches the check-in counter. The

check-in process is a one-step process where he/she can interact with the check-in

agent and decide on seats and drop bags. After check-in, the passenger proceeds to

the security check where the hand baggage and personal belongings are scanned. The

check-in baggage in most of the European airports is scanned in-line unlike the new

security requirement at US airports where the bags need to be scanned before taking

them to check-in. Passengers are now in secured area and can shop and relax. Prior to

boarding the aircraft they queue up in boarding lounge where the identity of the

person is verified once again. If the flight is international, passengers will have to go

through the Immigration process.

Figure 3-1 Passenger Processing Flow




3.3.2 Check-In with Kiosks

The new concept of self-service kiosks has divided the check-in process in two parts:

getting the boarding pass and dropping the bags at bag drop-off. The passenger

arrives at the airport and proceeds to the kiosk, which issues the boarding pass based

on the information provided by the passenger. The passenger then proceeds to the

fast baggage drop-off if he/she has any baggage otherwise can move to the security

check. The main steps in the check-in process through kiosk are shown in Figure 3-2.



19


Figure 3-2 Check-in through Kiosk

It is easy to follow the instructions on the kiosk and very direct in most of the cases.

The speed of checking in depends on the familiarity of the passenger with the kiosks

and on the speed of the machine itself in processing the request. The kiosks are also

equipped with the passport reader to enable international passenger check-in.

The fast bag drop-off is generally manned and the baggage is tagged at the drop-off.

Some airlines like easyJet expect passengers to tag the bags from the baggage tag

printed through kiosks, which has more chance of mistakes. The baggage drop off

facilities at most airports is airline specific and there are very few airports with

common use bag drop-off.

All major airlines now allow Web Check-in as an option where a passenger can print

bar coded boarding passes at home. In that case, the passenger without baggage can

directly move to the security check and board the aircraft without any hassles at the

airports. The passengers with baggage can drop the bags at the baggage drop-off and

proceed to the security check; this is represented in Figure 3-1. Web Check-in has seen

a rise in passenger acceptance as it gives them a higher comfort level. It is believed

that both the self-service and Web check-in will cater for most of the passengers in the

near future.

3.4 CHECK-IN CONFIGURATION

It is understood that the new technology will require new kinds of arrangements and

changes in the existing facilities. The standard check-in process as seen is a single step

procedure where boarding pass, bag tag and bag drop are all is done at one place

whereas with kiosk the process can be either two or three steps. Although, kiosks as

seen in Figure 3-3 are compact and require less space in comparison to same number

of check-in desks.

The stand alone kiosks are placed in a group of four or five kiosks and the location of

kiosks is crucial to its success. With so many kiosk types available and with the

changing requirements the configuration of the check-in kiosks depends on a large

number of factors. The main aspects to be considered in the placement of kiosks are

• Visibility

• Accessibility and movement of passenger traffic

• Comfort and privacy of the passenger




Figure 3-3 Typical Stand Alone Kiosk Photo by Mihir Baxi

There are many possible solutions for the layout of kiosks and it has been observed

that for the efficient use of kiosks it is essential to have some roving agents to help the

customers increase their transaction speed (Miller 2003). Some of the basic

arrangements discussed by Miller (2003) and Tomber (2007) are presented here to

understand the requirements but are classified by steps in the process.

3.4.1 One Step

The implementation of CUSS kiosks with baggage drop can make the self-service

process a single step process. The passenger will print his/her boarding pass and

baggage tag at the kiosk and drop his bag on the baggage belt next to the kiosk. The

process is almost similar to the traditional check-in except that the passenger has to

complete the process on his/her own, and it takes almost same time. There are roving

agents available to help the passenger. Two different approaches are adopted: the

roving agents are behind the counters assisting the passenger with bag tag and bag

drop-off. Similarly, the roving agents can be in front of the counters to assist more

passengers and this kind of arrangement is more efficient and can process more

passengers.



21


3.4.2 Two Step

The advantage of using self-service kiosks is that the check-in process can be divided in

two steps and it can process 40 to 50 passengers per hour. The location of kiosks

decides the passenger flow pattern and convenience. The two steps include printing a

boarding pass and dropping the bags along with bag tags at baggage drop off. The

baggage drop off are manned and the agents print the tag and attach it to bags. This is

more convenient as the passengers tend to make errors in attaching tags, which might

prove costly in overall operations. This system has a larger throughput as the

passengers without bags can be filtered and need less processing time.

3.4.3 Three Step

The process can be further sub divided and one more step can be introduced: the

passengers can tag their bags themselves and then can drop them to the bag drop.

The difference here is that there are different platforms provided for tagging the bags

and it is independent from the kiosk or bag drop. The processing rates are almost

similar to the two step process but the space required might be less for the same

number of passengers. easyJet Airline implemented a similar process at East Midland

Airport. The main issue is the tagging of bags correctly and dropping them to bag

drop. The roving agents are available to help the passengers. The staff required will

be fewer as the passengers serve themselves.

3.5 CONCLUSIONS

This chapter explains the evolution of the Self-service kiosks in general and the airline

industry. The development of the check-in kiosks can be compared to the banking

industry, as similar to ATMs the check-in kiosks are now moving away from the

terminals. It can be seen that the common use will be more beneficial in that

situation, almost similar to withdrawing money from any ATM on the network.

However, it can be seen in the discussion that we are moving toward paperless world

and the electronic medium is spreading as the acceptable medium. Thus, kiosks might

become secondary technology in the near future.

This chapter also highlights the steps and process involved in using kiosks. The speed

and accuracy of the check-in depends on the passengers and speed of the machines

and steps involved in printing the boarding pass. There are various possibilities with

the process and all of them have been discussed in the chapter. It is evident that all




the processes will have different impacts on the airport design and operations.



23


The arrival pattern of the passenger might not be significantly different but the

processing times will be different in all three type process. The airport thus has to

decide what the best way is to implement the self-service check-in. It is important to

note that the processing time will also be influenced by various other factors like

availability of agents, technology readiness of passenger, location of kiosks, etc. Thus

to ensure the proper allocation of resources in terms of kiosks the process to be

implemented should be considered.


Case Studies


4.0 CASE STUDIES

In order to understand the use of check-in technologies the process was observed at

three airports. The observations at all the airports are discussed in detail in this

chapter. The three airports used for data collection and observation have different

profiles in order to encompass as much variability as possible in the collected data.

The airports observed are London City Airport (LCY), Manchester International Airport

(MAN) and London Heathrow Airport (LHR). The chapter discusses the data collected

and analyzes the information to understand the process in a real life situation. At all

the airports observed the self-service check-in was a two step process similar to the

one discussed in 3.4.2.

4.1 METHODOLOGY

The main aim of collecting the data and observing the passengers at the airport was to

understand the factors that affect the check-in process. There are many ways to

collect the relevant information but it is important to select the right method for

observation and collection to understand the passenger behavior without affecting the

quality of the data required. To understand the process and measure the efficiency of

the kiosks and fast bag drop-off in processing the passenger, it was necessary to find

out the serving times and wait times for the same at airports. It was considered more

suitable to gather this information by observing the passenger in process and making

notes of the time taken for each step of the process. The interviewing of the

passenger would not give the exact times and there is no opportunity to measure the

performance of the system.

Details Observed at Airports

Group Size

Baggage per Passenger

Arrival Time

Queuing Time

Processing Time

Type of Flights

Flight Schedule

Proportion of passengers using CUSS

Table 4-1 - Details Observed at the Airports

The key aspects that were observed at the airports are as shown in Table 4-1. The data

allows one to understand the differences between profiles and type of operations

taking place at each airport. The other supporting information like flight schedules and

the number of passengers travelling and the proportion of passengers for each type of

check-in mode was gathered from the airports where available.


Case Studies

25


4.2 CASE STUDY 1- LONDON CITY AIRPORT, LONDON

London City Airport (LCY) is a small airport in the Dockland area of London city. The

airport handled 2.37 million passengers in 2006 (AirTransport Intelligence 2007). The

main airlines operating from the airport are regional airlines like VLM, Scott Airways,

CityJet, etc. The airport serves domestic and short haul European destinations. The

terminal has 26 check-in counters and around 12 self-service check-in kiosks belonging

to different airlines. The self-service kiosks belong to British Airways (BA), Air France

(AF)/ CityJet, SAS, BMI, etc. BA and AF kiosks are located in front of the entrance to

the terminal with a very good visibility.

The operations for two airlines AF and VLM were observed for collecting data. AF and

CityJet are both operated by Air France and use only self-service for check-in from the

six kiosks that are available at the terminal to check-in the passengers. Two of the

kiosks are placed in the corridor approaching the terminal from the DLR station. The

rest of the four are in the terminal and have sufficient queuing spaces. There are four

check-in desks available in total to AF and CityJet, which are used as baggage drop-off

points.

Figure 4-1- Self-service Kiosks for Air France and CityJet Photo by Mihir Baxi

VLM is one of the biggest regional airlines operating from LCY and uses traditional

check-in counters to process the passengers. They have four check-in counters at the

airport and one of them is used as a Business Class check-in.

The passengers travelling were mostly on business and very few passengers had bags

to check-in. Most of the passengers were travelling alone, the maximum group size

was two, and the majority of passengers were above 35 years of age.


Case Studies


4.2.1 Arrival Profile

LCY is busy during the early hours of the morning and the check-in counters open at

6.00 AM. AF operates short-haul intra European flights and the first flight of the day is

at 7.20 AM to Dublin. The check-in kiosks are open for check-in throughout the day for

any flight. The flight schedule for the observed flights is given in Table 4-2. These

same flights were observed for two days. Similarly, three flights for VLM were also

observed for one day and are shown in Table 4-3. LCY is the hub for VLM and it

operates short-haul intra European flights catering to business destinations.

Destination Time Seats

Dublin 7:20 100

Paris Orly 8:00 100

Dublin 8:35 100

Table 4-2 Flight Schedule Air France, LCY

Destination Time Seats

Manchester 11:00 50

Brussels 11:00 50

Amsterdam 11:05 50

Table 4-3 Flight Schedule VLM, LCY

Figure 4-2 and Figure 4-3 show the arrival profiles for the two airlines. The airport as

suggested mostly handles business passengers and it was reflected in the arrival profile

of the passengers. Only 20% of passengers arrive at check-in 90 minutes before the

STD of the flight. It can be seen from the figures that 40% of the passengers arrive in

the last 45 minutes of departure time.

Figure 4-2 Arrival Profile for Air France, LCY


Case Studies

27


The large numbers of passengers coming between 45 to 75 minutes before STD (Figure

4-2) is because of the flight schedule, which has three flights at an interval of half an

hour each. Though the self-service kiosks are available quite early, very few people

arrive at the airport early to check-in. The passengers using the airport are business

travelers and they prefer to arrive as late as possible to maximize the utilization of

their time. The fact that LCY is a small terminal, there are significantly fewer queues at

the airport and the passenger can be processed faster, which allows the passenger to

check-in as late as 10 minutes before STD. The arrival profile reflects this behavior of

the passengers. Most of the passengers do not have bags for check-in so they do not

need to use the bag drop-off, which speeds up the process in general.

Figure 4-3 Arrival Profile for VLM, LCY

4.2.2 Processing Time

As mentioned AF uses self-service kiosks for processing all the passengers, whereas

VLM uses traditional check-in counters for the same. Passengers using kiosks need to

go to baggage drop-off if required to check in bags. The processing time for each

process is shown in Table 4-4. The characteristics for each method are discussed in

detail in this section.

4.2.2.1 Self-service Kiosks

As seen from Table 4-4 the processing time for the kiosk is 1.98 minute with standard

deviation of 0.96 minute per passenger against the belief that it takes less than a

minute for the kiosk. Further to this, the following was observed at the airport


Case Studies


Figure 4-4 Processing Time per Pax - Self-service Kiosks, LCY

• The minimum time was 0.56 minutes and on average it took at least a minute

to complete a transaction and print a boarding pass.

• The processing times for the customer who had some experience of using a

kiosk was significantly less than average.

• Most of the passengers needed assistance in completing the process and there

were four roving agents helping passengers.

Average Maximum Minimum Standard

Deviation

Confidence

Level

Sample

Size

Self-service Kiosks 1.98 6. 25 0.56 0.96 0.15 157

Baggage Drop-off 2.40 4.66 0.40 0.85 0.20 32

Check-in Counters 1.31 6.28 0.23 1.39 0.50 78

Group Size 1.28 Pax 2 Pax 1 Pax

Number of Bags 0.60 Bags

Table 4-4 Processing Times at LCY

All times in minutes per Passenger

4.2.2.2 Baggage Drop-off

There were four bag drop-offs available to drop the bags. As seen from Table 4-4 the

baggage drop takes 2.40 minutes with a standard deviation of 0.85 minutes per

passenger. The observation might not be a correct representation of reality as it was

not possible to make systematic observations. There was a lack of signage and the

passengers coming to the airport were not quite aware of the AF process of check-in

through kiosks and used to queue up at the baggage drop assuming it was a full service

check-in, resulting in chaos and lot of balking from queue. Thus the sample size (32)

observed was very small and the results could not be accepted confidently.


Case Studies

29


Figure 4-5 Processing Time per Pax - Bag Drop-off, LCY

Figure 4-6 Processing Time per Pax- Check-in Counters, LCY

4.2.2.3 Check-in Counters

Check-in counters were used by VLM and there were a maximum four counters open

at the time of observation. The average time of processing per passenger is 1.31

minute with a deviation of 1.39 minutes. It could be seen that the average processing

time is smaller than kiosks but has a huge standard deviation. This is a result of the

efficiency of the check-in agent and the interaction with the customer. This human

element causes significant variations in check-in times. Passengers without bags were

processed faster and took 30 to 45 seconds to complete the transaction.


Case Studies


4.2.3 Queuing Time

The other important aspect that was observed in the process is the waiting times for

each passenger. The maximum and average queuing times for each method are

shown in Table 4-5. It could be seen from the table that there are no long queues at

either kiosks or check-in counters and the passengers are processed quite efficiently.

Max Queuing

Time

Average Queuing

Time

Self-service Kiosks 5.50 Min 0.56 Min

Baggage Drop-off 4.66 Min 2.36 Min

Check-in Counters 7.80 Min 1.86 Min

Table 4-5 Passenger Wait Times for AF and VLM, LCY

Figure 4-7, Figure 4-8 and Figure 4-9 show the number of passengers queuing at the

kiosks, bag drop-off and check-in counters respectively. The main points and

observations for the wait time were as follows

• It is quite evident from the comparisons that the kiosks though processing a

greater number of passengers wait times are lower.

• The maximum number of passengers waiting for the services is 8 for kiosks

against 19 in the case of check-in counters.

Figure 4-7 Queuing at Air France Kiosks, LCY

• The baggage drop data as explained is not that reliable but it could be

concluded that the efficient operation of kiosks will need organized bag drop-

off arrangements.


Case Studies

31


• Queuing at VLM Check-in counters was a result of the unavailability of the

check-in agent and a similar situation arose for AF where passengers were

confused and waited for help from the roving agent, who might have been busy

with other passengers.

Figure 4-8 Queuing at Air France Bag Drop-off, LCY

Figure 4-9 Queuing at VLM Check-in Counters, LCY


Case Studies


4.3 CASE STUDY 2- MANCHESTER AIRPORT

Manchester Airport (MAN) is the biggest airport in the north and handled 27.6 million

domestic and international passengers in 2006 (AirTransport Intelligence 2007) with

three terminals at the airport. Terminal T1 is the busiest of all and handles airlines like

Aer Rianta, Ryan Air, Monarch Airlines, Thomson Fly, Lufthansa, etc. The terminal has

107 check-in counters and a number of CUSS kiosks, which are shared by three or four

airlines.

Figure 4-10 Self-service Kiosks at MAN Photo by Mihir Baxi

Monarch Airlines operates scheduled and charter flights from T1 at MAN to European

tourist destinations. The airline has 10 check-in desks, 2 for frequent fliers and

business class and the rest for normal check-in. There are 5 CUSS kiosks and 3 bag

drop-offs in addition to the traditional check-in. It also allows for web check-in for

passengers. The check-in counters open three hours before STD and the flights are

available at the kiosk at the same time.


Monarch Airline operates six schedule flights in the later part of the day; the schedule

for the observed flights is shown in Table 4-6. The flights as seen are mostly leisure

destinations and most of the passengers on the flights are leisure passengers going on

vacation. These flights were observed for two consecutive days.


Case Studies

33


Time Destination Seats

14:05 Jerez 180

14:40 Faro 220

14:45 Tenerife 180

15:00 Lanzarote 180

15:25 Malaga 220

15:55 Palma 220

Table 4-6 Flight Schedule Monarch, MAN

The arrival profile for the flights can be seen in Figure 4-11. Most of the passengers

are leisure travelers and it can be seen from the figure that almost 20% of the

passengers arrived at the airport even before the opening of the counters. It can also

be seen that 80% of the passengers arrive more than 35 to 40 minutes before

departure. The passengers also arrive in big groups with one bag per passenger at

least. The average group size was 3.15 passengers. The main observation is that

leisure travelers arrive early at the airport to avoid queues and for their own

convenience.

Figure 4-11 Arrival Profile - Monarch, MAN

4.3.2 Processing Times

The airport as discussed allows the use of self-service and Web check-in for check-in

along with traditional desks. The process was observed for all three methods and the

processing times were measured for each mode. The breakup of the passengers

processed by each method is shown in Table 4-7. The results of the observation for

the process are shown in Table 4-8.


Case Studies


23-Jul-07 24-Jul-07

Pax % Pax %

Self-service Kiosks 114 11% 82 8%

Web Check-in 107 10% 83 8%

Check-in 858 80% 891 84%

Total 1079 1056

Table 4-7 Breakups for Passenger Check-in for Monarch, MAN

Compiled by Author, based on Information from MAN

As seen in Table 4-8 the average processing time per passenger is 1.23 minutes by

kiosks with a standard deviation of 0.88 minutes, while for check-in counters, it is 1.48

minutes with a standard deviation of 0.83 minutes. The sample size for the self-service

kiosks is smaller but is 28% of the total population; thus it could be considered

representative of the full population. The overall result for the kiosk and bag-drop

cannot be accepted confidently, as the sample sizes are small. The following are the

other observations that were made regarding the process at the airport


Deviation

Confidence

Level (95%)

Sample

Size

Self-service Kiosks 1.23 3.95 0.16 0.88 0.30 37

Baggage Drop-off 1.13 3.16 0.11 0.73 0.28 28

Check-in Counters 1.48 5.48 0.50 0.83 0.13 150

Group Size 3.15 Pax 8 Pax 1 Pax


Table 4-8 Processing Times for Monarch, MAN

All times are in Minutes per Passenger

Figure 4-12 Processing Time per Pax- Self-service Kiosk, MAN


Case Studies

35


Figure 4-13 Processing Time per Pax - Bag Drop-off, MAN

Figure 4-14 Processing Time per Pax - Check-in Counters, MAN

• All passengers approached the check-in in groups of 3 or 4.

• Passengers were unaware that they could check in through kiosks and had to

be pulled from the normal queue to use the kiosks.

• The kiosk displayed the numbers of the counters being used as Bag Drop-off,

which were visible and had appropriate signage.

• Most of the passengers required assistance and the roving agents played a key

role in getting people to use the kiosks.

• The arrangement of the kiosks was facing the approach of the passengers and

was placed so that they were easily accessible and visible.


Case Studies


4.3.3 Queuing Time

The wait times were observed to assess the efficiency of the system. There were not

many people waiting at the self-service or the baggage drop but the queue at the

check-in counters was very long. The results of the observations are shown in Table

4-9 and Figure 4-15.

Maximum

Queuing Time

Average

Queuing Time

Self-service Check-in 0.96 0.00

Bag Drop-off 26.83 7.13

Check-in 33.46 24.10

Table 4-9 Passenger Wait Times for Monarch, MAN

The queuing at the check-in counters reflects the arrival profile of the passengers; the

initial surge of passengers creates a high demand, which the system is unable to meet

and thus the queues persist until the end of the process. The other reason for the long

queuing is the passengers arriving in groups for leisure trips. There was virtually no

queuing to use the self-service kiosks. The following were the observations at the

airport

Figure 4-15 Queuing at Monarch Check-in Counters, MAN

• The crucial part was the operational aspect of the kiosks; on both days there

was a time when the kiosks broke down and affected the overall operation.

• On Day 2, it was observed that only two kiosks were working for the majority of

the time and it was difficult to manage the passenger queues.


Case Studies

37


• The reason for the no queuing at the kiosks was because most of the

passengers were not aware of their use and the roving agents were responsible

for fetching the passengers from the queues of check-in counters.

• The check-in counters could not be opened earlier than three hours before the

flight because of the technical issue of data transfer between systems (Davies

2007).

• There were significant wait times for the Bag Drop-off, which could further

discourage passengers from using the kiosks.

4.4 CASE STUDY 3- LONDON HEATHROW AIRPORT

London Heathrow Airport (LHR) is the busiest airport in the UK. There are four

terminals which handled 67.5 million passengers in 2006 (AirTransport Intelligence

2007). Terminal 2 mostly handles European flights. Terminal 2 caters to the airlines

like Lufthansa, Air France, Iberia, Olympic Airlines, etc.

Figure 4-16 Self-service Kiosks at LHR Photo by Mihir Baxi

Lufthansa Airlines has 6 check-in counters and 9 self-service kiosks to process the

passengers. The kiosks can be used by the other airlines in the alliance. There are 4

fast bag drop-offs for serving the passengers. The check-in counters open two hours

before STD. The kiosks are stand-alone in the front of the check-in counters to attract

more passengers and there are roving agents to help passengers. The Quick bag drop-

off could be open or closed depending on the intensity of traffic (Hiller 2007).


Case Studies



The airport operates various destinations throughout the day; the flights observed for

the day are shown in Table 4-10. The flights as mentioned earlier are to short-haul

European destinations. The flights extend from late afternoon to the end of the day

and there are a lot of business travelers in addition to the leisure passengers going on

vacation in the mid afternoon flights.

Time Destination Pax Travelling

16:30 Dusseldorf 64

17:05 Frankfurt 98

17:40 Stuttgart 58

18:05 Frankfurt 108

18:25 Dusseldorf 91

18:25 Hamburg 92

18:35 Cologne Bonn 70

19:00 Frankfurt 85

19:10 Munich 101

Table 4-10 Flight Schedule Lufthansa, LHR

Compiled by Author, based on Information from LHR

It has been observed that 60% of the passengers arrive more than 60 minutes before

the STD. The other 40% of passengers are dispersed and arrive at regular intervals.

The business passengers are generally travelling single and there were occasional big

groups of leisure travelers but most often the group size was 2. It can be seen in

Figure 4-17 that there are very few passengers arriving in the last 20 minutes but else

there is a steady flow of passengers.


Case Studies

39


Figure 4-17 Arrival Profile for Lufthansa, LHR


Lufthansa Airlines allows check-in with traditional and self-service kiosks but only the

data for self-service kiosks and fast bag drop-off were observed. Table 4-11 shows the

results of the observations.

The average processing time per passenger at the kiosk was 2.16 minutes with a

standard deviation of 0.95 minutes, while processing time at the bag drop-off was

1.46 minutes with a standard deviation of 1.11 minutes.


Deviation

Confidence

Level

Sample

Size

Self-service Kiosks 2.16 6.16 0.48 0.95 0.13 191

Baggage Drop-off 1.46 9.16 0.31 1.11 0.21 101

Group Size 1.34 Pax 3 1


Table 4-11 Processing Times for Lufthansa Airlines, LHR

All times are in Minutes per Passenger


Case Studies


Figure 4-18 Processing Time per Pax - Self-service Kiosks, LHR

The other observations that were made regarding the process are as follows

• The location of the kiosks made them very accessible and easily visible before

the passengers could see the check-in counters.

• There were passengers who could complete the transactions without any help

from the roving agents unlike the other two airports.

• The longer average time might be the result of the steps at kiosk or the speed

of the machine in printing the boarding pass.

Figure 4-19 Processing Time per Pax - Bag Drop-off, LHR

4.4.3 Queuing Time


Case Studies

41


The results of the observations are shown in Table 4-12. It was observed that there

were no queues for the use of the kiosks on the day of the observation. There were

some queues at the bag drop-off but not significantly long and the average wait time

per passenger was 0.65 minutes. The queuing at the bag drop was sometimes the

result of packing and unpacking the baggage to meet the weight or security

requirements.

Maximum

Queuing Time

Average

Queuing Time

Self-service Check-in 0.00 0.00


Table 4-12 Passenger Wait Times at LHR, Lufthansa Airlines

4.5 CONCLUSIONS

The case studies show that there are significant variations in the results depending on

the profile or operations of the airport. It can be observed that the queuing at the

airport is affected by three aspects namely

• Arrival Profile

• Group Size and Bags

• Processing Times

The arrival profiles as seen from the case studies vary at all airports. The passengers at

LCY arrive nearer to the STD while at MAN the passengers arrive early at the airport.

LCY is a small airport and mostly handles business passengers who prefer to spend as

little time as possible at the airport. It could be seen that the profile at LHR is uniform

except for the later part when there are significant number of business passengers

arriving at the airport. The arrival profiles will affect the use of resources and the

queues develop at peak times when the arrival rate is bigger than processing time.

The processing time is directly proportional to the group size and the number of bags

per passenger. The group size and number of bags depend on the type of passengers

travelling. Leisure passengers will travel in a bigger group with more bags as could be

seen at MAN as against business passengers who will travel alone at most with one bag

as observed at LCY.

The processing times for all three methods were observed and interesting results could

be seen. It takes longer to check in from kiosks at all airports. It was noticed that it

takes at least one minute to check in from kiosks even if everything goes right. It was


Case Studies


also noticed that there is a difference between processing times for check-in kiosks at

different airports. This could be a result of various reasons. The time at the kiosk is

dependent on the number of interaction steps and clarity of the steps. The other

aspect in the use of kiosks is the typing errors in giving input and correction time. A

further analysis and discussion on processing time per passenger is carried out in the

next chapter.

In addition to this it was observed that most of the passengers at LCY and MAN needed

assistance to use the kiosks. It was seen that many times the roving agents performed

the whole process without encouraging the passenger to use the kiosk on their own.

The situation needs improvement for increased efficiency, which could be achieved by

deploying kiosks similar to Fast Ticket dispatch machines at London Underground,

which have a “Call for Assistance” button on the screen for the passengers who get

stuck or confused, thus requiring assistance from the agent only when required whilst

encouraging the passenger to use the kiosk on his/her own.


Simulation Model

43


5.0 SIMULATION MODEL

The previous chapters explain the ideal process and the real conditions at the airports.

The complexity of the process is evident from the discussion and estimating the

requirement for the changing process will be interesting to explore. Estimating the

requirement on the basis of the queuing theories might not give exact results for the

large number of passengers as discussed by Joustra and Dijk (2001). It is also suggested

by Yan, Tang, and Chen (2003) to use simulation for assessing the requirements for the

Check-in process. The process gained another dimension with the introduction of the

kiosks and web check-in. The model developed can be a useful tool for estimating the

requirements for the airport in the planning and design stage and also for studying the

impact of the new process on the airport. While developing the model the utmost

care was taken to incorporate as many variables as possible. The process and

assumptions are discussed in this chapter.

5.1 THE APPROACH

As discussed earlier, the process of check-in has evolved over the last few years and

there are no standard tools to estimate the requirements for the combined resources

required by the airport to check in the passengers. Thus, the simulation tool provides

such a tool to airport planners and operators. It was also realized that it was necessary

to keep the model as flexible as possible to make it useful. The main factors that affect

the efficiency and flexibility are identified and discussed below.


There are many papers explaining the importance of the arrival profile at an airport

and the factors that affect it. It is an accepted fact that no two airports have the same

arrival profile. The arrival profiles depend on a lot of factors and sometimes there are

drastic differences in the behavior of the passengers at the same airport for different

flights. The case studies discussed in the earlier chapter also support this view as three

distinct arrival profiles could be seen at three airports. The resources can be allocated

reflecting the changes in the arrival of passengers to maintain the required service

standards.

The main factors that affect the arrival profile at an airport (or for any flight) are as

enumerated below:


Simulation Model


• Time of departure - Early Morning, Afternoon or Evening

• Type of flight - Scheduled or Charter, Low Cost or Full service, etc

• Destination being served - Business or Leisure

• Airport access mode available - Underground Metro Rail or Car

• Check-in Open Times

While developing the model it was realized that the arrival profiles should be flexible in

order to keep the model widely applicable. Thus, the simulation model has three

profiles as shown in Figure 5-1 based on the case studies, but new profiles could be

added after observations for the analysis of the airport being modeled.

The profiles depict three distinct behaviors of the passengers, namely leisure, business

and a mix of both for scheduled airlines. As seen from figures the leisure profile has a

lot of passengers coming in early while the business profiles has it later. It is essential

to determine the type of traffic and arrival patterns at the airport to estimate the

requirement correctly.


The results from the case studies show that there are variations in the processing

times at the three airports. The passengers arrive in a different size of groups with

different ratios of bags per passenger. The passenger group size and number of bags

depend on the type of passengers travelling. Leisure passengers will arrive in bigger

groups with more bags in general as seen at MAN, while LCY and LHR show a smaller

group size and lower proportion of bags per passenger where there is significant

proportion of business passengers.

The processing times for groups were converted in to processing time per passenger.

These processing times for one airport were then compared to other airports to

Figure 1a - Leisure Figure 1b - Business Figure 1c- Mix

Figure 5-1 Arrival Profiles


Simulation Model

45


measure the variance in the data and to see if the data among the airports varies due

to the group size. The results of the analysis were used in the model to generate the

processing times for the passengers joining the system. The details of the statistical

tests and results are given in Annexure A.

5.1.2.1 Check-in Counters

The use of check-in counters was observed at two airports, LCY and MAN. The single

factor Analysis of Variance test (ANOVA) (Refer Annexure A) was carried out to find if

the difference in the data is significant in these two independent samples collected

randomly. It was concluded that there is no significant variation in the data. The mean

and standard deviation for the combined data is given in Table 5-1. The processing

time includes the variation for passengers with or without bags. In addition to this it

was noticed that the service times for the counters were exponentially distributed.

5.1.2.2 Self-service Kiosks

The process times for kiosks were observed at all three airports. The results of the

ANOVA test showed that there was no significant influence of the group size on the

processing time per passenger. The mean and standard deviation for the combined

data is given in Table 5-1. It could be seen that the processing time per passenger is

higher than the check-in counters but has a slightly bigger standard deviation and it

was observed that the service times from kiosks were normally distributed.

5.1.2.3 Bag Drop-off

The data for bag drop-off was collected from all the three airports. As with check-in

and kiosks the results from the ANOVA test suggested that there was no significant

variation. It can be assumed with fair confidence that there is no significant effect on

the per passenger processing time by the number of bags and group size. The mean

and standard deviation are shown in Table 5-1. The process times were normally

distributed as against the exponential distribution for the counters. The difference in

the distribution is because there is mostly one bag per person and they already have a

boarding pass whereas at the check-in counters the passenger may have a bag or not

and the processing times will be significantly different in both situations.

Average Time Standard Deviation Distribution

Check-in Counters 1.43 0.83 Exponential

Self-service Kiosks 1.93 0.96 Normal with Skew

Bag Drop-off 1.51 1.13 Normal with Skew

Table 5-1 Processing Times and Distribution profile


Simulation Model


5.1.3 Service Standards

The service standard at airports in terms of the check-in process is the wait time and

the overall processing time per passenger. The resources at any airport are provided

to give adequate service standards and allow for queuing to some acceptable limit. It

is sensible to maintain less wait time at the bag drop-off than check-in counters, thus

the resources required will be larger. The number of passengers arriving at the airport

is dependent on the arrival profile for the airport or flight. The resources could be

adjusted to match the service standards; one more counter could be opened or closed

to maintain similar standards throughout the process. There would be more resources

required in the case of leisure profile where there are a lot of passengers waiting for

the services initially. The resources are designed keeping in mind the peak hour traffic,

but the peak hour in itself has variations, explained by the arrival profiles, thus it

makes sense to determine optimum resources catering for most of the traffic with

acceptable service standards.

During the development of the model efforts were made to understand the impacts of

such aspects on the overall performance. While developing the model it was seen that

one way to optimize the resources was to provide similar standards throughout the

whole process. In normal circumstances, first few passengers would not be required

to wait as the system would be completely idle, while it would be completely busy at

times. Passengers coming in late would have longer wait times than ones coming early

and the average time would still be below required standards. Thus, the numbers will

not represent the actual situation where some of the passengers would have to wait

for a longer time than is acceptable. It was desirable to spread these longer wait times

in to the system to achieve a lower variation.

In a normal first come first serve queuing model the passenger coming first will be

served immediately at arrival if the system is idle. In the simulation model developed,

the passengers are provided the services as late as possible, within the limits of the

Service Standards (MWT) provided by the user. This helps to spread the wait times to

the time where there is none; as a result the number of passengers waiting is more but

the average wait time is lower and the passengers waiting more than MWT are at an

acceptable level.


Simulation Model

47


5.2 DESCRIPTION OF MODEL

The simulation model is developed in MS Excel 2007. The model is kept simple and

users need to input all the parameters in the first sheet and can see various results like

wait times, processing times, usage of resources, etc on different sheets in the

workbook. A further description of the model and the key features are explained in

the Annexure B for reference. The self-service process considered for the model is a

two step process as discussed in 3.4.2. This section describes the overall process that a

passenger follows in the model.

Peak hour passengers at the airport are determined and are the key input for the

model. The passengers using each type of check-in technology are assumed and the

appropriate type of passenger profile is selected. The passengers arrive at the airport

in the pattern generated by the arrival profile. The passengers are segregated as per

the break-up suggested by the user for the appropriate check-in method.

There are three different queues in the system. The model is developed with three

different bank queues serving the multiple servers for check-in counters, kiosks and

bag drop-off. The passengers segregated for each process join the required queue.

The passenger has to wait in the queue till the server is available to process him/her.

The different processing times are assigned to each check-in methods.

The bag drop-off has a special arrangement and the passengers already checked in

through the web approach the bag drop-off directly on arrival at the airport.

Passengers from the self-service kiosks required to check baggage join the same bank

queue for bag drop-off. After finishing the process the passenger leaves the system for

a further process required to board the aircraft.


Simulation Model


Figure 5-2 Process Model for Simulation


Simulation Model

49


5.3 ASSUMPTIONS

The simulation model is a simplified representation of the reality. The model was

designed on the basis of some assumptions to simplify the process and also to

represent the results correctly. The simulation model is limited to 2500 peak hour

passengers and 25 servers in each method of check-in and bag drop-off.

The following were the assumptions made in the development of the model.

• The whole system is Common Use.

• All the passengers arrive individually at the airport.

• Passengers do not arrive before the check-in open times.

• The Arrival profile for all three methods of check-in is the same.

• The inter-arrival rate is the Poisson distribution following the arrival profile.

• All counters, self-service kiosks and bag drop-off open at the same time. The

kiosks do not serve passengers before the check-in open times.

• All three queues and the system are unique and independent.

• The passengers join the last position in the queue and do not leave the system

or change the method of check-in from counters to kiosk or vice versa.

• There is no separate provision for special counters like Business class, Frequent

fliers, etc. at any stage in the system.

• The processing times at all airports will have similar distribution profiles.

• Passengers using the kiosk and bag drop-off do not take a break and walk times

are not considered.

• All the kiosks are assumed to be at the airport.

• The number of bags per passenger is ignored and taken into account by the

variation of process times.

• No technical breakdowns, change in staff or maintenance have been

considered in the model.


Simulation Model


5.4 VALIDATION

In the development of any model one of the most important aspects is the validation

of the model. There are a number of ways for the validation of the model, but it is not

possible to use all the methods to validate the model (Sargent 2005). The simulation

model developed cannot be compared to any other existing model due to the lack of

access. There is no possibility of conducting experiments in real time and validating

the model. The existing data collected in all cases is just representative and cannot be

used for validation of the model except for one occasion.

The data collected at MAN could be modified to fit as input to the model and the

results can be compared to the available data. The other method of validation will be

comparing it to thumb rules given in ADRM-IATA. The model was tested against these

assumptions to see if it gave similar results.

5.4.1 Comparison to Existing Data

The data for the airport was collected and the proportions for the various check-in

methods are also available and the resources used for each of them are known. The

simulation model could be tested against the available data to see how the model

performed or represented the same.

Pax %

Self-service Kiosks 82 8%

Web Check-in 83 8%

Check-in 891 84%

Total 1056

Units

Kiosks 5

Check-in Counters 7

Bag Drop-off 3

Table 5-2 Inputs in the Simulation Model

The key inputs for the simulation model are shown in Table 5-2. The check-in open

times were considered as 180 minutes. At the airport there was a separate desk for the

frequent fliers, which is not possible in the model. Thus, 7 desks are considered for the

comparison.

Check-in Queue Reality Simulation

Maximum Queuing Time 33.46 35.20

Average Queuing Time 24.10 23.91


Simulation Model

51


Table 5-3 Comparison with the existing and simulation results

The simulation was run with this input and the key output, the queuing times were

compared with the real data collected. The comparison of the results is shown in

Table 5-3.

The data for the kiosks and the bag drop-off was not similar to what happened in

reality. At the airport, the passengers for the kiosks were fetched from the queue of

counters and did not follow any arrival profile. The simulation model as explained

cannot represent this and thus gives different results. Figure 5-3 and Figure 5-4 show

the similarity of the queuing patterns at the airport.

The observed data was collected for the usage of the two counters over a period of the

time and the author observed that the queue lengths were similar at all the counters.

Figure 5-3 Simulation result of wait times for MAN

The results from the experimentation suggest that the simulation model is able to

represent the reality with a fair accuracy. The wait time is the only output which is

comparable as there are no other factors that are available from the real data.


Simulation Model


Figure 5-4 Actual wait times at MAN

5.4.2 Comparison to Existing Standards

Apart from comparing the model with the existing data it is also important to compare

the model with the industry standards. The comparison to only existing collected data

is not sufficient as the collected data might not be representative of all the situations

and there might be some error in the collection of the data. Also the model was

developed from the same data source so it was necessary to benchmark the model

with the external source.

To overcome the above problem the simulation model was compared to the results

generated from thumb rules given in ADRM-IATA. Chapter F of ADRM deals with

airport capacity and provides the thumb rules for calculating the resources required

for the airport. The ADRM calculates the check-in counters requirements based on the

peak 30 minute demand in the peak hour. It also segregates the requirement for the

business class check-in.

ADRM gives an example for the calculation of the check-in counters. The same

example is considered for the comparison to the simulation model for simplification

and authenticity, though the requirements specified in the example need some

modifications to make them comparable to the simulation model. The problem

requirements are given below

• Peak hour Passenger- 2500 Pax

• Economy Passengers - 100%

• Peak 30-min Demand - 906 Pax

• Maximum Queuing Time - 30 Minutes


Simulation Model

53


The check-in counters required for a 90 second average processing time per passenger

were 26.5 counters.

The simulation model can calculate requirements for 2500 peak hour passengers. It

was necessary to develop an imaginary arrival profile for the calculation to be relevant

to ADRM as shown in Figure 5-5. The other assumptions were the same as the

calculations. The results from the simulation model are shown in Table 5-4. The

simulation model gives MWT with 24 counters.

Peak Hour Passenger 2500

Check-in Counters 24

Average Wait Time 16.96

Max Wait Time 29.66

Table 5-4 Results for the simulation of 2500 TPHP

As seen from the result the model compares with the assumptions of IATA. The result

enables to confirm that the model is based on sound assumptions.

Figure 5-5 Arrival Profile for Simulation representing IATA assumptions

5.4.3 Concluding Comments

The above tests help in checking the reliability of the model and to see how close the

model is to the reality. Although the model is able to show reliable results for both the

situations there are some more points which need further consideration.


Simulation Model


There are no industry wide standard rules for estimating the resources for the kiosks

and bag drop-off and the model was not directly validated for estimating the

requirements for the same. There is no sufficient data about existing conditions for

kiosks and bag drop-off to validate it with simulation. But the results for the check-in

counters could be successfully validated and as the whole the model is based on

similar assumptions the results for the kiosks should be reliable.

It would have been beneficial to carry out some more validation tests for the kiosks

and bag-drop off directly to add more reliability to the model.

5.5 CONCLUSIONS

This chapter explains the process of the model development. The approach for

developing the model as explained was to keep it as flexible as possible. The important

aspects in the process were grasped and put together in the model to represent the

reality as close as possible. The main factors that affect the efficiency and utilization of

the resources are the arrival profile of passengers. But the provision of resources is

guided by the service standards to be provided and the MWT allowed for the

passengers. The model gives an opportunity to observe the queuing patterns at the

airport in relation to the arrival profiles and adjust the resources to meet the service

standards.

This chapter also discusses the validation process for the simulation model. The model

was validated in two ways against the actual collected data and the industry standards.

It was established that the model depicts the reality with a fair accuracy and is based

on the sound assumptions to use in real life. It was not possible to validate the

performance of the model in relation to the kiosks and bag drop-off. The main reason

was the lack of industry standards and data to support the results of simulation for the

kiosks and bag drop-off. The research had a limited time scale and it was not possible

to collect the data again for validation.

It was concluded that as the assumptions for the check-in counters are validated and

the whole simulation model is based on similar principles, the results for the kiosks and

bag drop-off should be fairly accurate. It can be concluded that the model will be

useful in estimating the results for an airport.


Application of Model

55


6.0 APPLICATION OF MODEL

All the previous chapters have discussed the check-in process with self-service kiosk

and development of the model to study the impact of the technology on the airport

design and operations. The process was observed at the existing airports and the

information and understanding gathered was used to develop the simulation model.

The model was validated to see if it reflects the reality and is based on valid

assumptions and principles. The next stage is to test the model in different conditions

and determine the utility of the model. It is essential to see if the model can simulate

various scenarios and what can be derived from the results. The details of how to use

the model and the structure are discussed in Annexure B.

The model was used to simulate the existing and hypothetical situations to compare

the results in various scenarios. The simulation model was used to test different

scenarios to find out what would be the optimum solution for a given situation. The

details of all the experiments are discussed in this chapter.

6.1 EXISTING SITUATION

As discussed earlier, in the data collected at all three airports, the data for MAN has lot

more details and could be used as input for the simulation model. The existing

situation was simulated and various scenarios were developed to see how the

resources affect the queuing and overall operations at the airport. The existing

breakup of the passengers using different check-in methods and different resources

used at the airport are shown in Table 6-2. The model was run for 1060 peak hour

passengers with a check-in opening 180 minutes before STD.

6.1.1 As Is Model

The first situation considered was the exact replication of the existing conditions. It

could be seen that the queues are as long as in the case studies. Amongst the kiosks,

only two kiosks process 80% of passengers. Thus it could be said that the kiosks are

underutilized. The bag drops are being used uniformly and there are no queues longer

than 5 minutes. The average total times for all the three are as shown in

Table 6-1.

Table 6-1 Total Process Time for As-is model

Average Total Process time

Check-in 21.11

Kiosks 1.80

Fast Bag Drop-off 2.23




As-Is Scenario 1 Scenario 2 Scenario 3

Peak Hour Passengers 1060

Passenger Break Up

Check-in 84% 72% 72% 60%

Kiosks 8% 20% 20% 25%

Web Check-in 8% 8% 8% 15%

Resources

Check-in Counters 7 7 6 6

Kiosks 5 5 5 5

Bag Drop-off 3 3 4 4

Waiting Time Check-in

Maximum Waiting Time 30.08 23.94 29.75 20.46

Average Waiting Time 20.29 7.86 20.85 5.33

More than 5 Min 93% 49% 93% 32%

More than 10 Min 89% 27% 90% 26%

More than 20 Min 58% 14% 62% 1%

Waiting Time Kiosks

Maximum Waiting Time 0 4.43 5.57 7.68

Average Waiting Time 0 0.65 0.81 1.33

More than 5 Min 0% 0% 4% 9%

More than 10 Min 0% 0% 0% 0%

More than 20 Min 0% 0% 0% 0%

Waiting Time Bag Drop-off

Maximum Waiting Time 4.61 24.62 9.12 19.08

Average Waiting Time 0.69 6.61 2.06 6.74

More than 5 Min 0% 50% 18% 45%

More than 10 Min 0% 22% 0% 30%

More than 20 Min 0% 9% 0% 0%

Table 6-2 Summary of the Results

6.1.2 Scenario 1

In this scenario the resources are kept constant but the possibility of 20% of the

passengers using the kiosk has been evaluated. The kiosks are utilized in the more or

less uniform pattern. The queuing at counters has reduced and there are very few

passengers waiting more than 20 minutes. But the change has an adverse impact on

the bag drop-off and the passengers have to queue for dropping the bag. The average

wait time is 6.61 minutes and MWT is 24.62 minutes.

Average Total Processing Time

Check-in 14.87

Kiosks 2.32


Table 6-3 Total Process Time for Scenario 1



57


As-Is Scenario 1 Scenario 2 Scenario 3

Check-in Counters

Server 1 14% 14% 17% 19%

Server 2 14% 17% 18% 19%

Server 3 14% 13% 17% 17%

Server 4 14% 14% 17% 18%

Server 5 16% 13% 17% 13%

Server 6 15% 14% 15% 15%

Server 7 14% 15%

Kiosks

Server 1 50% 34% 26% 27%

Server 2 31% 22% 26% 20%

Server 3 12% 17% 21% 21%

Server 4 6% 15% 15% 17%

Server 5 1% 11% 12% 15%

Bag Drop-off

Server 1 39% 31% 28% 28%

Server 2 31% 32% 26% 26%

Server 3 29% 36% 25% 25%

Server 4 20% 21%

Table 6-4 Summary of Utilization of Servers

6.1.3 Scenario 2

To measure the sensitivity of the bag drop, in this scenario the proportion of

passengers is kept the same but one bag drop has been added, leaving the counters at

6. The results are interesting because the queues at the counters are at the same level

as in the as-is model. There is a significant improvement at the bag drop and no

passenger has to wait more than 10 minutes. Thus it might not be a good idea to

increase the number of bag drops.


Check-in 13.44

Kiosks 2.58



6.1.4 Scenario 3

This scenario was to see what would be the impact if the proportion of passenger

using the other two increases significantly. There are 60% of the passengers using

check-in by counters and the other 40% kiosks and web check-in.




It could be seen that the maximum queuing time at both counters and bag drop-off

has increased but there are very few passengers who have to wait for more than 20

minutes. The average total processing times are as shown in Table 6-6.


Check-in 7.76

Kiosks 3.30



6.1.5 Discussions

The as-is model shows the in efficiencies that already exist at the airport. The

scenarios were developed assuming that it is not possible to increase the resources.

Monarch Airline had 10 check-in counters including the bag drop-off.

Scenario 1 shows that if the proportion of passengers using the self-service kiosks

increases the queues built up at the bag drop-off. This will make the overall process

time longer for the passengers. To reduce the queues at the bag drop in Scenario 2 the

number of bag drop-offs was increased by one. The analysis models shows that the

process at bag-drop off speeds up but the queues built up at check-in counters. Thus

the change does not improve the overall situation.

Scenario 3 was developed to see how the further increase in the proportion of the self-

service and web check-in passengers affect the queues. The average and MWT

increased for both counters and bag drop-off. But there are only a few passengers

waiting more than 20 minutes and the average total processing time is reasonable. It

was noticed that the queues at the counters are at the earlier part due to the arrival

pattern (leisure), while the queues at the bag drop-off are at the later part as it is the

second step in the process. Thus, one more bag drop-off can be added by reducing the

number of counters when there is a smaller queue at counters probably.

It could be seen in all the scenarios that the Self-service kiosks are enough to cater to

the assumed traffic levels. Thus it could be concluded that the main contribution to

the queues is the result of the lack of counters. Scenario 1 seems to be the

appropriate break-up of passengers and resources, which will allow the airport to

achieve maximum efficiency. The queues at the bag drop-off can be reduced as

suggested by Scenario 3. While the further increase in the proportion of passengers as

seen in Scenario 3 will benefit the airport operations and increase efficiency.



59


6.2 OTHER EXPERIMENTS

The simulation model was used to analyze various what-if scenarios to study the

impacts of the evolution of technology on the resources at the airport. The input used

in various scenarios is summarized in Table 6-8.

Scenario 4 is based on Leisure arrival profile and Scenario 5 on Business. The peak

hour passengers are constant in all the experiments. The experiments show the

different resources that would be required to meet the different proportions of

passengers by each check-in method.

The leisure passengers would have lower service standards as against the business

passengers. In all the scenarios the service standards to be achieved were kept

constant to estimate the resources for each scenario and are shown in Table 6-7.

Check-in Kiosks Bag Drop-off

Leisure 20 10 10

Business 10 5 5

Table 6-7 Service Standards for the Scenario 4 and 5

The results from the simulation model are shown in Annexure C. Table 6-8

summarizes the main results and key figures from the simulation model results. The

critical observations are discussed for each of the scenarios.

6.2.1 Scenario 4

These set of scenarios demonstrate the impact of changes in the proportion of

passengers on the resources required for leisure passengers. The key issues that were

observed in obtaining the results is that due to the arrival pattern maximum resources

are required in the earlier part of the process and are seen in results of simulation in

Annexure C. The increase in passenger by self-service is more likely in the near future,

and scenario 4B reflects the same. It could be seen that the number of kiosks required

gets doubled to meet the same service standards. It could be seen that the total

number of counters and bag drop-off required have increased. The required check-in

counters have been reduced but to maintain the same standards the number of bag

drop-off have increased substantially.

Similarly in Scenario 4C, a further increase in the proportion of passengers using fast

bag drop-off means a larger number of bag drop-off is required, but the need is not

significantly higher. It should also be noticed that the check-in counters required are

very minimal. The total counters and bag drop-off required are 31 in Scenario 4C.




Scenario

4A

Scenario

4B

Scenario

4C

Scenario

5A

Scenario

5B

Scenario

5C

Peak Hour Passengers 2000 2000 2000 2000 2000 2000

Passenger Arrival Profile Leisure Leisure Leisure Business Business Business

Check-In Open Times 120 120 120 120 120 120

Passenger Break Up

Check-in 60% 40% 20% 60% 40% 20%

Kiosks 20% 40% 40% 20% 40% 40%

Web Check-in 20% 20% 40% 20% 20% 40%

Passengers using Bag drop-off

Kiosks 100% 50%

Web Check-in 80% 50%

Resources

Check-in Counters 16 11 6 19 13 6

Kiosks 10 20 20 11 20 20

Bag Drop-off 14 22 25 10 14 18

Waiting Times Check-in

Maximum Waiting Time 19.12 18.42 17.61 11.89 11.86 10.81

Average Waiting Time 11.83 12.06 9.88 4.72 4.84 3.68

More than 5 Min 90% 89% 85% 45% 41% 34%

More than 10 Min 78% 69% 50% 16% 14% 5%

More than 15 Min 19% 32% 16% 0% 0% 0%

Waiting Times Kiosks



More than 5 Min 22% 23% 20% 0% 0% 5%

Waiting Times Bag Drop-off



More than 5 Min 14% 23% 10% 5% 1% 0%

Table 6-8 Summary of the results for Scenario 4 & 5

6.2.2 Scenario 5

The scenario is similar to 4 but uses the business arrival profile. In scenario 5A the

number of counters required is larger in order to maintain the high level of service

standard. The use of bag drop-off is limited but the number of bag drop once again

reflects the service level to be provided. With an increase in the proportion of

passengers using kiosks the total number of counters and bag drop-off is reduced to

27. The kiosks required are doubled and four more bag drop-offs are required.



61


In scenario 5C the proportion of passengers using the Web check-in is increased to

40%. The impact is seen on the counters. Now only 6 counters are required to

maintain the same level of service standards. It could also be seen that the total

resources required for the counters and bag drop-off is only 24.

6.2.3 Discussions

It could be seen from Table 6-8 that the arrival profile has a very large impact on the

utilization of the resources. The service standards in terms of queuing times

determine the resources required. The results demonstrate that the change from one

technology to another does not actually improve or reduce the resources required if

there are a lot of passengers using bags. The total number of check-in counters and

bag drop-off required remains almost constant for Scenario 4, nor does the service

standards improve significantly with a change in technology.

The airport may gain capacity increase in the case of a lot of business passengers as

there will be spare resources. The number of counters required reduces to 6+18 with

the increase in passengers using alternative technology. The number of bag drop-off

required is not as high as in the case of Scenario 4. Scenario 5 leverages the benefit of

breaking the process in two steps, namely the printing boarding process and bag drop-

off.

It was also noticed that 19 check-in counters are required in Scenario 5 as against 16 in

Scenario 4 for the same number of passengers, as the service standard is higher. In

contrast to this, for kiosks the same number of kiosks is sufficient in both scenarios.

Further to this only 18 bag drop-off are required in Scenario 5 as against 25 in Scenario

4. Thus, fewer resources are required for the total passengers in Scenario 5 than in

Scenario 4, but with fewer bags.

In the process of determining the exact resources required, various other observations

were made. They are as follows

• There is no gain in the efficiency of the process by just improving only one

element of the process. The increase in the number of kiosks increases the

throughput and the queues at the bag drop-off get longer.

• There is a lag in the use of the bag drop-off as seen from all the figures and

there is a scope of optimization and proper operational strategy to use

resources efficiently.

• The number of resources was selected keeping in mind the optimum usage of

all the resources and average total process time.




• It was seen that one additional kiosk reduces the wait time significantly, while

the addition of one counter increased the efficiency marginally.

6.3 CONCLUSIONS

This chapter explains the application of the model to the real life situation and shows

the utility of the model. The simulation model helps to understand the check-in

process and allocate resources as required for the whole system. From the

experiments carried out with the help of the model the process of evolution of

technology could be analyzed. The implementation of new technology can be

evaluated with the results from the simulation results.

It was realized as a result of experimentation that the self-service and web check-in

process are beneficial for the airports operating passengers with fewer bags. If there

are lots of passengers with bags it is easier to process the passengers at check-in

counters. But having said that it was also seen that the counters required are the same

and changing or adapting to the new technology should not be an issue within the

context of space and resources required. The model suggests that efficiency of the

self-service and web check-in is completely dependent on the bag drop-off.

The model was used only to evaluate some simple scenarios, but the author realizes

that it could be used for various situations. The model could also be applied and used

for

• Estimating the resources at a new airport or for a particular flight.

• Determining resources for various “what-if” scenarios in terms of arrival

profiles and the proportion of the passengers using different technologies.

• Allocating resources at existing airports

• Studying the queuing patterns and impacts of changes of resources on queues.

• Predicting the peak queue times and allocating resources accordingly.

Thus the model could be used for various situations to either improve the efficiency of

the system or estimate resources. The model acts as a tool and adds value to the tacit

understanding already existing in the industry for the use of kiosks and fast bag drop-

off.


Discussions

63


7.0 DISCUSSIONS

During the duration of the research, the author came across several other pieces of

information and it was thought necessary to discuss some of the observations and

information so as to understand the process better. The information will also help

understand the impact of the new check-in process on the airports. The main aspect

covered in the discussion is related to the evolving technology and the seamless

integration of the various aspects in to the system.

7.1 DEDICATED VERSUS COMMON USE

This debate of dedicated versus common use is always going on between airports and

airlines. The airlines will prefer to maintain their own identity and service standards

and customer connection; for this they need to have dedicated resources. The airports

on the other hand argue that it is more convenient for them to operate common use

resources as that would require fewer resources and consequently less infrastructure

and cost.

The installation of CUSS is no exception. There are many major airlines operating

dedicated kiosks at a number of airports. As seen earlier the use of CUSS is spreading

at airports. The important point that is raised is should the bag drop be common use

or not. There are very few airports like Vancouver (Pilling 2005) which has a common

use bag drop-off facility. At most of the airports the bag drop-off is airline dedicated

so the passenger has to look for a specific location for bag drop-off.

It was observed that the kiosks at MAN were CUSS kiosks, but were more or less

dedicated to Monarch Airlines, as they were located near the counters of Monarch

Airlines and hosted by the airline roving agents. Thus the kiosks were effectively

dedicated to Monarch Airlines. A similar situation existed at LHR, where all alliance

airlines could use the kiosks but which generally did not happen.

Thus, the success of CUSS is dependent on the successful integration of bag drop-off in

the system. There are some solutions available for the same and are discussed later in

the chapter. The model for simplicity assumes that kiosks and bag drop-off are

common use.


Discussions


7.2 CUSS VERSUS WEB CHECK-IN

The other important aspect which will affect the requirements at the airport is the

domination of Web check-in. The use of e-ticket and BCBP has enabled the airlines to

use the internet for checking in the passengers. The experience from the other

industries like banking shows that more and more customers are switching to

paperless mode and airline industry could not be an exception. BA and KLM believe

that online check-in will gain importance in the near future (Conway 2006).

But there are many issues to be addressed before that happens. The current Web

check-in applications has three to four hour cut-off times for checking in. This situation

created a barrier at MAN, as the normal counters and kiosks could not be opened

before this cut-off time due to technical reasons (Davies 2007). Many airlines do not

allow passengers with bags to check-in through web. The other aspect is that the web

check-in is to cater to different group of passengers (SITA(b) n.d.) and takes 4 to 6

minutes to complete the transaction.

At any airport there will be passengers who will not have direct access to the internet

and a printer, thus there will always be a need for check-in at airports and CUSS might

be a more efficient way forward. Figure Figure 7-1 shows the IATA estimates for the

breakup of check-in technologies. The author believes that though the Web check-in

will gain importance it will not be able to replace Self-service check-in at airports. Also,

an increase in web check-in means more bag drop-off at the airports as shown by the

model. The resources required for various methods of the check-in modes can be

evaluated using the model as shown in the previous chapter.


Discussions

65


Figure 7-1 Check-in Mode Forecast by IATA

7.3 EMERGING TECHNOLOGIES

The simulation model was developed keeping in mind the current self-service process

at the airports. The author realizes that the process is evolving and there are new

technologies being tested in order to further gain efficiency. It could be seen that the

current process is developed by dividing the traditional check-in process in two steps.

The efficiency of the system is mainly due to the filtering of passengers needing bag

drop off facilities. The BCBP has enabled passengers to print their boarding passes at

home using the internet. This aspect has been included in the model and only the

passengers needing bag drop-off approach the counters.

The advantage of 2D BCBP is that the medium of print becomes irrelevant and it can be

printed on any format i.e. paper or digital. This has enabled the development of a

check-in process using a mobile phone. This enables passengers to check-in using their

mobile phone and there is no need for kiosks or online check-in. The main advantage

is that the passenger does not need to print the boarding pass, as the BCBP can be

saved on the mobile phone.

This technology is being used extensively in Japan and Finland and is very popular

amongst the business travelers (Baxter 2007). The process can use either SMS or a

special hardware linked to the phone. The other possibility is using Bluetooth and

wireless enabled PDAs to check in at airports on entering the terminals (Zimmerman,

et al. 2001) with special travel cards linked to PDA.

It is evident that the process again takes advantage of filtering the passengers which

will not require using bag drop. This will enable the airports to increase capacity but it

should be kept in mind that providing sufficient space for the bag drop-off in a two

step process is essential for efficiency. This kind of process has not been included in

modeling as this will require further data collection which was not possible.

7.4 BAG DROP-OFF

As seen throughout the thesis a key step in the process is bag drop-off. It was

observed that in two steps printing the boarding pass and bag drop-off, later is much

more complicated. Thus, it requires more thought and consideration. It could also be

seen from the discussions that most of the innovation and technological developments

are for the first step of the process.


Discussions


Currently, at most airports the bag tags are printed at the dedicated bag drop-off

counters from the BCBP. At some airports it is possible to print the bag tags from the

kiosks and passenger can tag and weigh their bags and drop them at the bag drop-off.

The simulation model is based on the first method, as at all the airports in case studies

there was a similar process in place.

The common use bag drop-off is not a common practice and as explained earlier limits

the use of CUSS kiosks. The problems with the passengers self tagging the bags are

discussed by Raymond (2005). The issue is further complicated by regulations which

say once the bag is tagged it should be in the ownership of the airline (Conway 2006).

Thus, it is clear that the bag drop-off process requires further improvements.

The author believes that the breakthrough technology will be something similar to the

baggage re-claim belts where the passengers just drop the bags with the tags already

printed and attached. The main hurdles in such cases are the control of size and

weight of bags and the correct tagging of bags. The use of RFID might make the

sorting a little easier and automatic.

The evolution of the bag drop-off might reduce the processing times and the model

might need to be adjusted for the same. Although the model will not stand true for a

three step process, it can be used for a one step process after some modifications.

7.5 CONCLUSIONS

This chapter explains the factors that might affect the simulation model and the

exceptions which could not be explained by the simulation model. It could be seen

that new technologies keep on emerging to improve the efficiency. These new

technologies will affect all the aspects of the process from the arrival profiles of

passengers and the processing times per passenger. The simulation model could be

used in a number of such occasions after some modifications, but obviously it cannot

solve all the scenarios.


Conclusions

67


8.0 CONCLUSIONS

This chapter concludes the research and discusses the important outcomes of the

research along with the problems encountered during the research. This chapter also

suggests the areas of further work required to add value to the model and the other

problem areas for research.

8.1 OVERALL DISCUSSION

The research was initiated with an understanding that there is a need for a tool to

estimate the resources required for the changing technology. The self-service kiosks

are seen as a positive move and it has been believed that it improves the check-in

process and reduces queues at airports, although the author has not come across any

academic research or case studies showing an improvement in the situation apart from

the interviews and the claims from the airports in the trade journals and periodicals.

The self-service check-in process is based on separating the two different processes of

printing boarding pass and bag drop. This in particular is the reason for the

improvement in the throughput of the system.

The author undertook case studies to collect the data in the real life situation. The

results of the case study regarding the processing times were particularly interesting.

It is evident from the table that the total processing time for the kiosk plus bag drop is

far more than the normal check-in process.

Average Time Standard Deviation

Check-in Counters 1.43 0.83

Self-service Kiosks 1.93 0.96


Table 8-1 Processing Times for all Check-in Modes

Thus, the passenger using both needs to spend more time in the system for two

reasons: more processing time and more waiting time in the queues for two processes.

To maintain passenger comfort it will be necessary to maintain lower wait times at

both the services, implying the need for more resources.

It was seen that the arrival profiles at all three airports were different as suggested by

the various papers in the literature review. The main factor that affects the arrival

profiles is the passenger type travelling on the flights. The author believed that the

arrival profiles for all three check-in methods would be different, but in the airports

that were selected for case studies this could not be verified.


Conclusions


The aim of developing the simulation model was to understand the queuing patterns

and transfer all tacit knowledge in to some useful tool for estimating the resources.

Efforts were made to use all the understanding of the process in developing the model.

But it was realized that there are very few important parameters that actually affect

the process. It was necessary to keep all the input as flexible as possible to keep the

model applicable to a large number of scenarios. It was concluded from that the arrival

profiles were the only parameter which varies significantly at all airports and it was

decided that the user should be able to modify it easily.

The variation in processing times observed was not that significant and the statistical

test also supports the observations. Thus, the processing time distributions were

developed on the basis of the observations and cannot be modified by the user.

Although if it is required to modify the processing time distribution for reasons like a

change in technology or process improvement they could be adjusted without any

major restructuring of the model.

Any simulation model needs validation to make it reliable. The developed model was

validated by two methods; first with the existing data and observations and secondly

with industry standard practices. It was not possible to validate the data for the kiosks

and web check-in as the author was not able to collect sufficient data and there are no

industry standards in place to compare. It was realized that it would have benefited

the model if it could have been validated with real life situations. But it was not

possible to collect the data after making the simulation model as the data collection

was delayed due to security reasons heightened by the terrorist attacks that took place

during the duration of the project. In hindsight, it would have been ideal if the model

could have been tested at any of the observed airports for reliability.

While developing the scenarios and testing the model it was understood that the

queuing pattern at the airport is the function of the arrival profile. The effects on the

queuing pattern due to change in the resources could be analyzed with the help of the

model. It was observed that the simulation model can be a useful tool for planning the

requirements for the check-in process at airports.


Conclusions

69


8.2 STATEMENT OF RESEARCH VALUE

The learning and the value of this research can be divided in two parts as follows:

Firstly, the research enables one to understand the key parameters that affect the new

check-in process using self-service and web check-in. The research compiles all the

facts about the self-service and web check-in process for the existing situation. It

verifies and documents the process times and arrival profiles for each method by

collecting data at three airports. The research also highlights the differences that exist

at different airports in operations and passenger profiles.

Secondly, it gives the airport planners a tool to estimate the requirements for the

check-in process. The simulation model developed could be used for understanding

the queuing patterns and measure the efficiency of the system in terms of the number

of passengers waiting beyond the MWT, the percentage of passengers being processed

by each server and the total process time per passenger.

The simulation model developed has a very wide utility in the airport planning and

operations of check-in systems. The results from the simulation model enable the

users to assign the resources required for each method of check-in. The various

scenarios analyzed in the research give an insight in the impacts of implementing the

new check-in process. Any airport could use a similar process to analyze the

requirements for the transitional stage in the implementation of new check-in

technologies.

Thus the simulation tool helps the planners to make a better decision and support the

knowledge with tangible results. The model could identify or predict the bottlenecks

that might exist in the system. The model allows the visualization of the impacts of

increasing or decreasing the resources and to reach an optimum solution.

8.3 FURTHER WORK AND RESEARCH

The research conducted has helped in understanding the process, but it has also raised

further questions which might need further research and analysis. After the

development of the model the author realized that the following points could also

have been addressed by the model.


Conclusions


• In the model it would have been interesting to allow different opening times

for the counters and kiosks as in real life the kiosks opening time is

independent of the check-in open times.

• There should be a more detailed output like system idleness and queue

lengths, which was implicitly available from the model, but required further

effort and time.

The author realized that the model could be further developed for the optimization of

the resources. Some efforts were made to optimize the use of the resources by

controlling the wait time per passenger but were not successful and have not been

included in the discussions and scenario analysis. But it was understood that flexible

opening times for the counters and bag drop-off are key to optimizing the use of

resources, and thus could be developed as an added feature in the model.

It should be noted that the model uses a single bank queue for the kiosks, whereas the

kiosks stand alone or in groups requiring separate lines. The impact of various

configurations and layouts at the airport their contribution to efficiency could also be

analyzed.

The author believes that the key to improving further efficiency lies in the process

improvements at the bag drop-off. It will be interesting to look at various new

developments and understand the impacts of the self tagging process in a single or

three steps.

One of the most important aspects which could not be determined by the author was

the analysis of the arrival profiles for different check-in modes. There is a need for

research to establish the relation between the check-in mode and arrival patterns of

passengers at the airport. Does the passenger arrive early as the kiosks are available at

the airport irrespective to check-in times or does the passenger come late to the

airport as it is perceived that check-in with kiosks is faster? Or is the arrival profile

independent of the check-in method? This understanding could help in the allocation

of resources and the location of the kiosks in the terminal and the use of bag drop-off.

The use of kiosks also adds flexibility of location and the off-airport use of self-service

check-in is gaining popularity. The benefits of such a process are quite evident, but it

will be interesting to see how many passengers use such facilities, and a cost-benefit

analysis and factors for locating the kiosks and bag drop-off at such places could be

undertaken.


Conclusions

71


8.4 FINAL CONCLUSIONS

This chapter has put in place all the conclusions and important points that need to be

considered for further research. The statement of research value shows that the

stated objectives have been successfully met. The research demonstrates that the

arrival profile is one of the important factors at any airport, which determines the

queuing patterns. The total processing time for kiosk check-in and bag drop-off is

significantly more than the traditional check-in process.

The author was successful in developing the simulation model which could become an

important tool estimating the requirements for the check-in process at the airport.

The model is able to demonstrate the understanding gained in the case studies, which

could be used for increasing the efficiency of the system. The model as shown could be

useful in analyzing various what-if scenarios and in seeing the impact of changing

resources. Thus the model empowers the airport planner with a tool for estimating

the resources for the check-in process with self-service and web check-in.


Works Cited

73


WORKS CITED

A Joint ACI, Airline Business and SITA Survey. "The Airport IT Trends Survey, Executive

Summary." http://www.sita.aero. 2006.

http://www.sita.aero/NR/rdonlyres/F5953D0C-D08E-484A-8D34-

9E05D2CA4AA3/0/AirportIT_Booklet06.pdf (accessed June 2007).

Airbus. "Global Market Forecast, The future of flying- 2006-2025." 2006.

AirTransport Intelligence. ATI -AirTransport Intelligence. 2007. http://www.rati.com

(accessed August 2007).

ATI News. "BAA Cheif: Passenger experience is unacceptable." http:\\www.rati.com.

Edited by Keiran Daly. 2007. http://www.rati.com/news/item.asp?id=226256

(accessed June 2007).

Baxter, Andrew. "Innovations: Airport Check-in: Board your flight by mobile phone."

FT.com, May 14, 2007: 1.

BBC News. BBC NEWS | Scotland | ATM inventor honoured 40 years on.

http://news.bbc.co.uk/1/hi/scotland/4135269.stm (accessed June 2007).

Chun, Hon Wai, and Raymond Wai Tak Mak. "Intelligent Resource Simulation for an

Airport Check-in Counter Allocation System." IEEE Transactions on Systems, Man, and

Cybernetics—PART C: Applicaiton and Reviews 29, no. 3 (August 1999).

Conway, Peter. "Common vision." Airline Business, January 2006: 40-42.

Davies, Martyn, interview by Mihir Baxi. "During the visit at MAN." Personal. (July 23,

2007).

Dijk, Nico M. van, and Erik van der Sluis. "Check-in Computation and optimization by

simulation and IP in combination." European Journal of Operational Research 171

(March 2006): 1152-1168.

Hiller, Pete, interview by Mihir Baxi. "During the visit at LHR." Personal. (August 06,

2007).

Hooke, Raymond W. "Common use Self-Service kiosks: How will they benefit the air

transport industry?" Unpublished MSc Thesis, Department of Air Transport, School of

Engineering, Cranfield University, Cranfield, 2005.

Hughes, David. "Growing Pains." Aviation Week & Space Technology, August 8, 2005:

45-47.

IATA. "Chapter F - Airport Capacity." In Airport Development Reference Manual, by

IATA, 178-212. IATA, 2004.

IATA(a) . "StBProgrammeSummary." www.iata.org/stbsupportportal.

http://www.iata.org/NR/rdonlyres/00A6CD7D-9329-42E0-A8A1-

242749B3BCC7/0/April2007StBProgrammeSummary.pdf (accessed April 2007).


Works Cited


IATA(b). "StB_FactSheet_CUSS.pdf." http://www.iata.org/stbsupportportal.

http://www.iata.org/stbsupportportal/stb_fact_sheets.htm (accessed April 2007).

Joustra, Paul E., and Nico M. Van Dijk. "Simulation of Check-in at Airport." Proceedings

of the 2001 Winter Simulation Conference. 2001.

Kaminski-Morrow, David. Eurocontrol warns of severe airport constraints by 2025.

London, December 15, 2004.

Lejarraga, Amaya Barona. "Simulation of Passenger flow in a Self Check-in Procees."

Unpublished MSc Thesis, Department of Enterprise Integration, School of Industrial

and Manufacturing Science, Cranfield University, Cranfield, 2004.

Maras, Elliot. "Self-serve kiosks address a changing customer." Automatic

Merchandiser, March 2006: 20.

Miller, Jared Keith. "Reducing the Air Travel Hassle Factor Through Self-Service Check-

in Process Improvements." MSc Thesis, Massachusetts Institute of Technology, 2003.

Murphy, Samantha. "Kiosk 411." Chain Store Age, January 2007: 68.

Neufville, Richard de, and Amedeo Odoni. "Overall design of passenger buildings."

Chap. 15 in Airport Systems: Planning Design, and Management, by Richard de

Neufville and Amedeo Odoni, 605-654. McGraw-Hill, 2003.

Park, Yonghwa, and Seung B. Ahn. "Optimal assignment for check-in counters based on

passenger arrival behaviour at an airport." Transportation Planning and Technology 26,

no. 5 (October 2003): 397-416.

Pilling, Mark. "Efficiency drive." Airline Business, May 21, 2005: 70.

Pilling, Mark. "Table Toppers." Airline Business, June 2007a: 6.

Sargent, Robert G. "Verification and Validation of Simulation Model." Proceedings of

the 2005 Winter Simulation Conference. 2005.

SITA(a). "IATA's Simplifying the Business."

http://www.sita.aero/News_Centre/Simplifying_the_Business/default.htm. SITA.

http://www.sita.aero/NR/rdonlyres/95C8A775-CE4C-4B2F-ACAD-

02659B8067B6/0/SITA_SimplifyingTheBusiness_PP_A4.pdf (accessed June 2007).

SITA(b). "Will Internet Check-in replace CUSS Check-in." http:\\www.sita.aero.

(accessed June 2007).

Tomber, Dave. "Check-in Evolution." Paasenger Terminal World, June 2007: 52-56.

Toole, Kevin O, and Jackie Thompson. "Capacity Crunch." Airline Business, June 21,

2005: 50.

Weiss, David. "Analysis: Kiosk Uptime, Revenue." Airport Business, July 2006: 9-11.


Works Cited

75


Yan, Shangyao, Chin-Hui Tang, and Miawjane Chen. "A model and a solution algorithm

for airport common use check-in counter assignments." Transportation Research Part

A, no. 38 (October 2003): 101-125.

Zimmerman, Thomas, Steve Ihde, Karsten Aalders, Scott Wright, Michel Wirth, and

Stephen Hopkirk. "Travel Card: Airport Self-Check In using a wireless PDA." Intelligent

Transportation System Conference Proceedings. Oakland, 2001.


Further Reading

77


FURTHER READING

Amer, Suzie. "Checking In, From far away." Succesful Meeitngs, February 2004: 19.

Berkley, Blair J, and Amit Gupta. "Indentifying the information requirements to deliver

quality service." International Journal of Service Industry Management 6, no. 5 (1995):

16-35.

Cao, Yuheng, Aaron L Nsakanda, and Irwin Pressman. "A Simulaiton Study of the

Passenger Check-in System at the Ottawa International Airport."

DeNise, Antoinette. "Do-it-yourself Check-in is spreading." Meeting News, October 10,

2005: 32.

Doshi, Naren, and Robert Moriyama. "Application of Simulation Models in Airport

Facility Design." Edited by C.-H. Chen, J. L. Snowdon, and J. M. Charnes E. Yücesan.

Winter Simulation Conference. 2002.

Field, David. "Outside the box." Airline Business, Aug 2003: 48-50.

Horonjeff, Robert, and Francis X. Mckelvey. "Planning and Design of the Terminal

Area." Chap. 10 in Planning and Design of Airports, by Robert Horonjeff and Francis X.

Mckelvey, 431-516. Mc-Graw Hill, 1994.

Howes, John. "On the Cusp of CUSS." www.kioskeurope.com. 2006.

http://www.kioskeurope.com (accessed June 2007).

John A Lawrence, Jr, and Barry A Pasternack. "Queuing Models." Chap. 9 in Applied

Management Science: Modeling, Spreadsheet Analysis, and Communication for

Decision Making, by Jr John A Lawrence and Barry A Pasternack, 501-550. John Wiley &

Sons, Inc., 2002.

Kiran, Ali S, Tekin Cetinkaya, and Serafettin Og. "Simulation modelling and analysis of a

new International Terminal." Proceedings of the 2000 Winter Simulaiton Conference.

2000.

Odoni, Amedeo R., and Richard De Neufville. "Passenger Terminal Design."

Transportation Research- Part A 26A, no. 1 (1992): 27-35.

Ragsdale, Cliff T. "Queuing Theory." Chap. 13 in Spreadsheet Modeling and Decision

Analysis: A practical Introduction to Management Science, by Cliff T. Ragsdale, 668-

702. Thomson South-Western, 2004.

Rapport, Greg. "The time is right for Self-Service Kiosks." Convenience Store Decision,

August 2006: 32-34.

Schwartz, Adele C, and Douglas W Nelms. "Do-it-yourself check-in." Air Transport

World, 2000, Summer ed.: 19-20.

Slack, Frances, and Jennifer Rowley. "Kiosk 21: a new role for information kiosks?"

International Journal of Information Management 22 (2002): 67-83.


Further Reading

79


Snowdon, Jane L., Soad El-Taji, Mario Montevecchi, Edward MacNair, C. Adam Callery,

and Scott Miller. "Avoiding the blues for Airline travelers." Edited by D.J. Medeiros, E.F.

Watson, J.S. Carson and M.S. Manivannan. Winter Simulation Conference. 1998. 1105-

1112.

Sobie, Brendan. "Jumping the Queue." Airline Business, January 2007: 54.

Yan, Shangyao, Chin-Hui Tang, and Miawjane Chen. "A model and a solution algorithm

for airport common use check-in counter assignments." Transportation Research Part

A, no. 38 (October 2003): 101-125.

MSc Thesis_Devleoping a Simulation Model for Airport Check-In

Documents