Analysis of Passenger Group Behaviour and Its Impact on Passenger

Flow using an Agent-based Model

Lin Cheng, Clinton Fookes, Vikas Reddy and Prasad K.D.V. Yarlagadda

Queensland University of Technology, 2 George St, Brisbane, QLD, Australia

Keywords: Group Dynamics, Agent-based Model, Airport, Passenger Flow, Simulation.

Abstract: Group interaction within crowds is a common phenomenon and has great influence on pedestrian behaviour.

This paper investigates the impact of passenger group dynamics using an agent-based simulation method for

the outbound passenger process at airports. Unlike most passenger-flow models that treat passengers as

individual agents, the proposed model additionally incorporates their group dynamics as well. The

simulation compares passenger behaviour at airport processes and discretionary services under different

group formations. Results from experiments (both qualitative and quantitative) show that incorporating

group attributes, in particular, the interactions with fellow travellers and wavers can have significant

influence on passenger’s activity preference as well as the performance and utilisation of services in airport

terminals. The model also provides a convenient way to investigate the effectiveness of airport space design

and service allocations, which can contribute to positive passenger experiences. The model was created

using AnyLogic software and its parameters were initialised using recent research data published in the

literature.

1 INTRODUCTION

Revenue of airports nowadays is gradually

transferring from aviation related sectors to non-

aviation sectors (retail revenues) and also from

traditional airline sources (lease arrangements) to

passengers (fees collected from ticket sales)

(Harrison et al., 2012, Schultz et al., 2010a). A

positive passenger experience is likely to result in

repeat visits, which not only helps further generate

airport’s financial profit, but also satisfies the needs

of other stakeholders such as operating airlines,

retailers, passengers and visitors (Popovic et al.,

2010). Hence, the passenger experience has become

a major factor that influences the success of an

airport.

In this context, passenger flow simulation has

become a significant approach in designing and

managing airports (Schultz et al., 2007,

Kleinschmidt et al., 2011). Although it has been

proven that social interactions greatly influence

crowd behaviours and decision making, far too little

attention has been paid to group behaviour when

developing passenger flow models (Qiu and Hu,

2010, Singh et al., 2009, Ma et al., 2012). This paper

aims to evaluate the impact of group dynamics on

passenger flow in an international departure terminal

using an agent-based model for the check-in process.

The remainder of the paper is organised as follows.

Section 2 reviews previous work related to airport

design and simulation techniques. Section 3

demonstrates the construction and configuration of

the agent-based model in the context of an

international airport. Section 4 provides the

simulation results and analysis and Section 5

summarises the major findings.

2 RELATED WORK

In recent years, there has been an increasing amount

of literature on airport passenger terminal design,

analysis and modelling. Tosic (1992) offered a

comprehensive review about global airport terminal

models. The review introduced the features and

proposed applications of the models, along with

their strengths and weaknesses. Generally, model

inputs are the physical layouts of the building, flight

schedules, arrival time of passengers and processing

rates. The evaluations of the model usually consist

of the queue length, utilities and waiting times at all

733

Cheng L., Fookes C., Reddy V. and Yarlagadda P..

Analysis of Passenger Group Behaviour and Its Impact on Passenger Flow using an Agent-based Model.

DOI: 10.5220/0005086807330738

In Proceedings of the 4th International Conference on Simulation and Modeling Methodologies, Technologies and Applications (SIMULTECH-2014),

pages 733-738

ISBN: 978-989-758-038-3

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

facilities. Although some literature had considered

passengers as groups of people, how the group

dynamic influences group behaviour at each activity

and the overall system performance were not

illustrated nor analysed. In spite of this, those

models provided valuable references for future

model designs.

Schultz et al. (2010b) investigated passenger

dynamics in the airport terminal by analysing field

data from Dresden International Airport. The

research pointed out that approximately 50% of

passengers were travelling in groups and the group

size has significant influence on passenger speed.

Other factors that influence passenger speed are

gender, travel purpose (business/ leisure) and the

amount of carry-on baggage.

Popovic et al. (2009) presented an observation

technique that investigated how passenger activities

mediate people’s experience in the airport. In the

study, detailed passenger behaviour in the airport

was recorded. It was found that passengers travelling

in groups had a considerable waiting time at the

security process. The video showed that after the

security screening point, people wait for their group

members in the middle of the walkway to passport

control. The findings of the study provide valuable

information for modelling passenger behaviour and

group dynamics in this paper. Using the same

observation technique, Livingstone et al. (2012)

reported results of passenger landside retail

experience in airports. Through the data collection

from 40 passengers, researchers found that the

existence of passenger’s travel companions can

influence passenger’s landside dwell time and

shopping behaviour in discretionary activities. The

limitation of the observation technique is that

passengers who participated in the research were

aware that they were being recorded. Furthermore,

the low efficiency of video recording and data

processing restricted the technique to only a small

number of people.

Ma, et al. (2011, 2012) introduced an individual

agent decision model to simulate stochastic

passenger behaviour in airport departure terminals.

Using Bayesian networks, the conditional

probabilities of passengers’ advanced traits

(shopping preference, hunger level, technology

preference, etc) were calculated through the basic

traits (age, gender, nationality, flight class, etc.). By

considering the restriction factors (such as remaining

time and walking distance) passengers in the

simulation can behave autonomously based on the

results of Bayesian network inferences. However,

the simulation did not explain how the group

dynamics influence the passengers’ decision making

process and what will happen if passengers were in a

group where group members have very different

behaviour in their advanced traits.

Cheng et al. (2014) conducted a case study to

demonstrate how the agent-based passenger flow

model can be used to examine the efficiency of an

airport evacuation strategy. By comparing

evacuation time of individual passengers and

passengers in groups, the impact of group dynamics

during an airport evacuation process was analysed.

The simulation results shows that group dynamics

can significantly impact passenger behaviour during

airport evacuation processes and consequently

affects the total evacuation time.

3 MODEL REALISATION

3.1 Airport Environment

In agent-based modelling, three key elements need

to be identified and modelled: agents, their

environment, and their interactions with other agents

and the environment (Macal and North, 2010). The

model environment is an airport departure terminal,

which is divided into landside and airside.

Figure 1: The International airport departure processes.

Chec

-in Securit

Customs Boardin

Discretionary Activities

Landside Airside

SIMULTECH2014-4thInternationalConferenceonSimulationandModelingMethodologies,Technologiesand

Applications

734

The landside of the terminal is open to the public,

while the airside of the terminal is only accessible

for passengers.

Figure 1 illustrates a high-level description of the

passenger departure processes in the model.

Passenger activities are categorised into processing

and discretionary activities (Kraal et al., 2009).

Processing activities are mandatory for passengers

before boarding the plane. On the landside of the

terminal, passengers check-in for their flights, and

pass through security screening and border

processing before entering airside and boarding.

Discretionary activities are considered as any other

activities undertaken by passengers during non-

processing time (Kraal et al., 2009, Livingstone et

al., 2012). Discretionary activities can happen

between two sequential mandatory activities as

shown in Figure 1. Examples of discretionary

activities in the proposed model include random

walking, store browsing, having food and using

other airport services. Retail shops and airport

services are located at both landside and airside to

emulate the real-world scenario.

In order to guarantee there is ample time left for

security measures, passengers are often advised to

arrive at the airport three hours before the standard

flight departure time. In the model, the flight check-

in process starts at 2.5 hours prior to flight

departures and closes on 25 minutes before the

departure time. A row of check-in service counters

(eight counters per row) are assigned to the check-in

process of each flight. Among the eight counters,

there are two counters for business class passengers

and six counters for economy passengers.

3.2 Pedestrian Configuration

Pedestrians in the model are categorised into

passengers and wavers. Passengers are those who

will board on the plane, while wavers are fellow

companions who accompany the passengers to the

airport but do not board the flight. Age, gender,

residential status and travel purpose are four basic

characteristics of passengers in the model. These

four factors can influence advanced passenger

characteristics such as mobility and shopping

preference.

Table 1 summarises the distribution of airport

passengers’ age and gender provided by the global

passenger survey conducted by IATA (2013).

According to the country of residence, passengers in

the model are divided into Australian resident and

overseas visitors. The Australian Bureau of Statistics

(ABS) provided the information of departure

passengers’ country of residence and passengers’

main reasons for their journey in 2012-2013

financial year (ABS, 2013). These four basic

characteristic factors: age, gender, country of

residence and travel purpose will be initialised to

each agent according to the percentage rate showed

in the Table 1. The age and gender are assigned to

each agent when the agent enters the system. Since

passenger groups usually share common features of

country of residence and travel purpose, these two

factors are initialised to each agent after the

pedestrian group has finished assembling and will

assume passengers in the same group have a

common country of residence and travel purpose.

Based on these four basic characteristic factors,

passenger groups in the model are initialised with

different speeds and activity preferences, which

enable agents to act autonomously in the simulation.

Table 1: Basic passenger characteristics in the airport.

Passenger

characteristics

Source Detailed factors Percentage in

total passengers

Age range IATA global passenger survey

(IATA, 2013)

< 25 10%

25 - 34 31%

35 - 44 23%

45 - 54 18%

55 - 64 11%

>65 7%

Gender IATA global passenger survey

(IATA, 2013)

Male 59%

Female 41%

Country of residence Australian Bureau of Statistics

(ABS, 2013)

Australian resident 58%

Overseas visitor 42%

Travel purposes Australian Bureau of Statistics

(ABS, 2013)

Business 15%

Leisure 85%

AnalysisofPassengerGroupBehaviourandItsImpactonPassengerFlowusinganAgent-basedModel

735

3.3 Pedestrian Group Interaction

Pedestrians in the airport are predominantly driven

by specific goals: passengers want to finish airport

processes and board their flights; and wavers

accompany passengers in the airport and send them

off. In this paper we focus explicitly on the

interactions within pedestrian groups. There are

some basic rules that govern the relationship and

interactions of a group:

 During movement, pedestrians in the same

group generally move toward the same

destination.

 All group members will try to keep a uniform

speed, except during situations such as

avoiding obstacles and collision with other

pedestrians.

 If group members fall behind due to any

reason, other group members will slow down

until the stalled group member catches up.

 At mandatory processes such as check-in,

passengers who finish the process faster need

to wait for all other group members to

complete the process before moving on.

 If time is allowed for discretionary activities,

pedestrians in the same group will generally

undertake the same activity together once the

activity is chosen.

The model takes the departure flight schedule

and passenger numbers for each flight as input. The

attributes of the agents such as group size, speed,

flight schedule and class, and shopping preference

are initialised. The interactions between the agents

and the environment are defined. Based on the flight

schedule, the agents are appropriately introduced in

the environment. Detailed micro-activities in each

process, for instance, the passenger behaviour at

check-in process is modelled based on observational

data collected by the Human System team of the

Airports of the Future project (Kraal et al., 2009).

4 RESULTS AND ANALYSIS

The 3D simulation environment of an international

airport departure terminal is shown in Figure 2. The

model is built within the AnyLogic 6.8 platform to

simulate the daily operation of the airport. Activities

of each agent in the system were updated

successively according to preset characteristics

within a discrete-event structure of the AnyLogic

simulation software.

To evaluate the effect of group dynamics on

facilitation and overall congestion at the check-in

area, we ran simulations under three different

scenarios. These are passengers travelling: (a) alone;

(b) in groups of varying size; (c) in groups of

varying size with wavers. Figure 3 illustrates the

screenshots taken for the same flight and timeline of

the simulation. From the model observation, it was

noted that passengers who travel in groups will wait

for group members in the pathway after finishing the

check-in process. This waiting behaviour of

passenger groups can cause congestion in the

pathway behind the check-in area and slow down the

passenger flow. More severe congestion can be seen

in the scenario where passenger groups are

accompanied by wavers [Figure 3 (c)].

Data collected from three different simulations at

the check-in process show that passenger group

dynamics influence the check-in queue time and

dwell time (Figure 4). The check-in dwell time is the

average time elapsed between passengers entering

the check-in area and leaving it with their

companions (if there are any), while check-in queue

time is the average time elapsed between passengers

entering the queuing area and getting served by the

check-in staff. From the table in Figure 4, we note

passengers travelling in groups or with wavers spend

approximately 3 minutes to regroup after the

process. This leads to a longer dwell time in the

check-in process.

(a) (b)

Figure 2: Airport departure terminal simulation environment (a) landside of the terminal; (b) airside of the terminal.

SIMULTECH2014-4thInternationalConferenceonSimulationandModelingMethodologies,Technologiesand

Applications

736

(a) (b) (c)

Figure 3: Facilitation and overall congestion at check-in for three different scenarios. Passenger travelling: (a) alone; (b) in

groups; (c) in groups with wavers.

Figure 4: Regroup, queue and dwell times at the check-in process for the three different scenarios.

The model results also suggest that the time

passengers spend in queuing can be influenced by

group structure. It can be seen that passengers

travelling alone spend approximately 2 minutes less

in the queue when compared to passengers travelling

in groups. A possible explanation for witnessing

such a trend could be the congestion caused by

people waiting to regroup with their fellow travellers

around the queuing area. In essence, ignoring group

dynamics in agent based modelling may yield results

that may not accurately represent the real-world

observations.

5 CONCLUSIONS

Through the simulation of the check-in process, it is

shown that agent-based modelling can be used to

analyse group dynamics of pedestrians in a complex

environment. The results in this study suggest that

the group dynamics can potentially lead to

congestion and longer check-in dwell times. Such

scenarios can lead to potential flight delays and thus

contribute to a lower level of service (LOS) and poor

passenger experience. Furthermore, they may also

leave the passengers with less time for discretionary

activities which may not be favourable for airport

retail operators. Therefore, from the airport

management perspective, it would be beneficial if

terminal operators could run the simulation

beforehand by inputting flight schedules and

passenger quantity details into the model. The

simulation results will provide valuable information

for airport operators to respond proactively to any

potential congestion. Furthermore, the advanced

modelling incorporating group dynamics provides a

more powerful long-term planning tool and airport

design analysis tool.

ACKNOWLEDGEMENTS

The Airports of the Future research project is

supported under the Australian Research Council’s

Linkage Projects funding scheme (LP0990135). The

authors would like to thank Philip Kirk who

provided some of the model parameters used in the

AnalysisofPassengerGroupBehaviourandItsImpactonPassengerFlowusinganAgent-basedModel

737

simulation. The authors also acknowledge the

contribution of the industry stakeholders involved in

this project. More details on the project can be found

at: www.airportsofthefuture.qut.edu.au.

REFERENCES

ABS 2013. Overseas Arrivals and Departures. Australian

Bureau of Statistics.

Cheng, L., Reddy, V., Fookes, C. & Yarlagadda, P. K. D.

V. 2014. Impact of passenger group dynamics on an

airport evacuation process using an agent-based

model. In: akhgar, B. & arabnia, H. R. (eds.) 2014

International Conference on Computational Science

and Computational Intelligence. Las Vegas, Nevada,

USA: IEEE Computer Society.

Harrison, A., Popovic, V., Kraal, B. J. & Kleinschmidt, T.

2012. Challenges in passenger terminal design : a

conceptual model of passenger experience. In:

Israsena, P., Tangsantikul, J. & Durling, D. (eds.)

Proceedings of the Design Research Society (DRS)

2012 Conference. Chulalongkorn University,

Bangkok: Department of Industrial Design, Faculty of

Architecture, Chulalongkorn University.

IATA 2013. IATA Global Passenger Survey.

Kleinschmidt, T., Guo, X., Ma, W. & Yarlagadda, P. K.

2011. Including airport duty-free shopping in arriving

passenger simulation and the opportunities this

presents. In: Jain, S., Creasey, R. R., Himmelspach, J.,

White, K. P. & Fu, M. (eds.) Proceedings of the 2011

Winter Simulation Conference. IEEE.

Kraal, B. J., Popovic, V. & Kirk, P. J. 2009. Passengers in

the airport : artefacts and activities. Design : Open

24/7.

Livingstone, A., Popovic, V., Kraal, B. J. & Kirk, P. J.

2012. Understanding the airport passenger landside

retail experience. In: Israsena, P., Tangsantikul, J. &

Durling, D. (eds.) DRS 2012 Bangkok – Research:

Uncertainty, Contradiction and Value. Chulalongkorn

University, Bangkok: Department of Industrial

Design, Faculty of Architecture, Chulalongkorn

University.

Ma, W., Fookes, C., Kleinschmidt, T. & Yarlagadda, P.

2012. Modelling passenger flow at airport terminals :

individual agent decision model for stochastic

passenger behaviour. SIMULTECH 2012. Rome.

Ma, W., Kleinschmidt, T., Fookes, C. & Yarlagadda, P. K.

D. V. 2011. Check-in processing : simulation of

passengers with advanced traits. In: Fu, M. & White,

K. P. (eds.) Proceedings of the 2011 Winter

Simulation Conference. IEEE.

Macal, C. M. & North, M. J. 2010. Tutorial on agent-

based modelling and simulation. Journal of

Simulation, 4, 151-162.

Popovic, V., Kraal, B. & Kirk, P. J. 2010. Towards airport

passenger experience models. Proceedings of 7th

International Conference on Design & Emotion.

Popovic, V., Kraal, B. J. & KirK, P. J. 2009. Passenger

experience in an airport : an activity-centred approach.

IASDR 2009 Proceedings. COEX, Seoul: Korean

Society of Design Science.

Qiu, F. & Hu, X. 2010. Modeling group structures in

pedestrian crowd simulation. Simulation Modelling

Practice and Theory, 18, 190-205.

Schultz, M., Lehmann, S. & Fricke, H. 2007. A discrete

microscopic model for pedestrian dynamics to manage

emergency situations in airport terminals. In: Waldau,

N., Gattermann, P., Knoflacher, H. & Schreckenberg,

M. (eds.) Pedestrian and Evacuation Dynamics 2005.

Springer Berlin Heidelberg.

Schultz, M., Schulz, C. & Fricke, H. 2010a. Passenger

dynamics at airport terminal environment. Pedestrian

and Evacuation Dynamics 2008. Springer.

Schultz, M., Schulz, C. & Fricke, H. 2010b. Passenger

Dynamics at Airport Terminal Environment.

Pedestrian and Evacuation Dynamics 2008, 381-396.

Singh, H., Arter, R., Dodd, L., Langston, P., Lester, E. &

Drury, J. 2009. Modelling subgroup behaviour in

crowd dynamics DEM simulation. Applied

Mathematical Modelling, 33, 4408-4423.

Tosic, V. 1992. A Review Of Airport Passenger Terminal

Operations Analysis And Modeling. Transportation

Research Part a-Policy and Practice, 26A, 3-26.

SIMULTECH2014-4thInternationalConferenceonSimulationandModelingMethodologies,Technologiesand

Applications

738