A Microsimulation Modeling of Pedestrian Characteristics in

Bangkok Transit System Case Study

Jumrus Pitaksringkarn

and Suhail Shaik

Department of Civil Engineering, School of Engineering, King Mongkut’s Institute of Technology Ladkrabang,

Bangkok 10520, Thailand

Tranportation Engineer, S2R Consulting Company Limited, Bangkok 10250, Thailand

Keywords: Social Force Model, Walking Speed, Microsimulation, Bangkok Transit System.

Abstract: Although there are many international standards of walkway design, walking behaviors are different in each

country/region. To determine the pedestrian characteristics the concept of the social force model, which is

analogous to resolving forces in Newtonian mechanics was adopted. Behavioral data related to pedestrian

walking speed were collected by using a digital camera at BTS (Bangkok Transit System) station and

manually extracted needed factors like pedestrian speed, density and others. After completing the calibration

and validation process using a VISSIM microsimulation technique, pedestrian walking speed is analyzed on

the basis of density. The analysis shows that the walking speed of pedestrians is 75.07 m/min, which is slower

than the U.S. pedestrians. It is also found that the walking speed and the body size directly affect the pedestrian

flow rate. A similar traffic microsimulation model has also been applied to analyze the pedestrian capacity

that is calibrated by adjusting pedestrian speed. Due to the smaller body size of Asians compared to Americans,

the flow rate observed in this study is higher. In particular, the pedestrian capacities per one-meter width of

uni-direction and bi-direction are 91 peds/m/min and 78 peds/m/min, respectively.

1 INTRODUCTION

Many in Bangkok has a formula “Walk-BTS-Walk”

to avoid Bangkok’s chronic traffic jam. It is just

because of the rapid development of high-density

residential communities around BTS (Bangkok Mass

Transit System) stations and high-density shopping

malls, offices, and commercial buildings located

around that or other stations additionally BTS offers

ease accessibility to airports and bus terminals. A

survey study in Bangkok (Rastogi et al., 2003) has

found that 62 % of passengers living near transit

stations within 1 km prefer to walk. Similarly, in

Mumbai, India (Pongprasert & Kubota, 2017) found

that 85 % of passengers living near transit stations are

comfortable to walk up to the distance of 1250

meters. Evidently, walking is one of the most

economical and effective modes of transportation for

short-distance trips, at the same time, a simple fact is

that most journeys start or end with a walking trip. As

Bangkok is a rapidly developing Metropolitan city,

https://orcid.org/ 0000-0002-1273-7603

https://orcid.org/ 0000-0002-2390-1587

with increasing inward migration, there is a rise in

traffic congestion problems. It has a population of

more than 15 million people and has an increasing

tendency each year. The increase in population is

directly proportional to travel demand. And data from

the Office of Transport and Traffic Policy and

Planning in Thailand (OTP) indicates that more than

half of all travel demand are using private vehicles.

The key solution is a sustainable transportation

system that focuses on safety, eco-friendliness, and

dependency reduction on limited resources. For

example, a walking system will reduce the usage of

private vehicle/motorcycle transportation. Therefore,

proper and standardized walkway design system is

necessary.

When it becomes necessary to provide, making

decision on selecting an inappropriate concept can be

costly to rectify. For example, to design a pedestrian

facility, foreign criterion was adopted in Chinese

metro stations without any further studies to

understand the difference between Chinese and

Pitaksringkarn, J. and Shaik, S.

A Microsimulation Modeling of Pedestrian Characteristics in Bangkok Transit System Case Study.

DOI: 10.5220/0011067300003191

In Proceedings of the 8th International Conference on Vehicle Technology and Intelligent Transport Systems (VEHITS 2022), pages 353-359

ISBN: 978-989-758-573-9; ISSN: 2184-495X

353

Westerners. This leads to a peerless connection

between demand and capacity of pedestrians during

rush hours in some Shanghai, China, metro stations.

Such that designing effective and appropriate walking

facilities that suit local pedestrian characteristics

considered as a challenge by dense and regular

passenger traffic. For example, a typical illustration is

metro transfer stations. So consciously, to sidestep the

issue, in-depth study of local pedestrian flow

characteristics is compulsory (Ye et al., 2008).

Application of Western standards for pedestrian

facility design in Asian countries is not always

effective way to design and, in some cases,

inappropriate, as every location is bound by its own

unique set of environmental and physical constraints.

As the study conducted by (Y Tanaboriboon, Record,

& 1991) summarizes, in designing of pedestrian

facilities walking speed parameter is of utmost

importance. Therefore, more attention should be

given to pedestrian facilities by doing studies on

pedestrian flow characteristics and walking behavior

to find the optimal design for pedestrian spaces.

Hence, this study will explore the walking speed in

Thailand by using the case study of BTS Sky Walk

Bangna – BITEC. The traffic microsimulation model

has been applied to analyze the pedestrian

characteristics which are calibrated through adjusted

pedestrian speed.

2 LITERATURE REVIEW

2.1 Pedestrian Walking Speed

The findings of the various studies show that Asian

pedestrians walk slower compared to Western

pedestrians as shown in Table 1. (Y Tanaboriboon &

Guyano, 1991) conducted a study in Thailand and

obtained a mean walking speed of 73 m/min, which

is relatively comparable to the 74 m/min walking

speed of Singaporean pedestrians (Yordphol

Tanaboriboon et al., 1986) but significantly slower

than the 81 m/min speed of the U.S. pedestrians

(Fruin, 1971).

In the United States, study conducted by (Navin

& Wheeler, 1969) in the University of Missouri,

Columbia found students were walking at an average

speed of 79 m/min. Similar study conducted in the

United States by (Gupta, 1986) and yielded a result of

slightly higher mean walking speed of 88 m/min.

In Asia, (Gupta, 1986) conducted a study in Delhi

and found a mean pedestrian walking speed of 72

m/min. (Yu, 1993) found a mean pedestrian walking

speed of 73 m/min in China. (Gerilla et al., 1995)

found a mean pedestrian walking speed of 70.6 m/min

in the Philippines.

In addition, (Fruin, 1971) also studied the

distribution of free-flow pedestrian walking speeds at

Port Authority of New York bus terminal and

Pennsylvania Train Station. The results are shown in

Figure 1.

Table 1: Mean walking speeds in various studies.

Region City, Country

Mean

Walking

Speed

(m/min)

Author

Asian

Bangkok,

Thailand

73.0

Tanaboriboon

and Guyano

(1991)

Singapore 74.0

Tanaboriboon

et al. (1986)

Delhi, India 72.0 Gupta (1986)

China 73.0 Yu (1993)

Metro Manila,

Philippines

70.6 Gerilla (1995)

Western

New York,

United States

81.0 Fruin (1971)

Columbia,

United States

79.0

Navin and

Wheeler

(1969)

Pittsburgh,

United States

88.0 Hoel (1968)

Figure 1: Speed distribution according to (Fruin, 1971).

2.2 Human Ellipse of Pedestrians

Human Ellipse (Body Ellipse) has a significant

impact on pedestrian characteristics. According to (Y

Tanaboriboon et al., 1991) study, Thai pedestrians

walk slower than Western pedestrians, but Thai

pedestrian capacity per 1-meter width is greater.

While designing a pedestrian facility, current

practices adopt pedestrian space requirements

mentioned in US-Highway Capacity Manual (HCM-

VEHITS 2022 - 8th International Conference on Vehicle Technology and Intelligent Transport Systems

354

2010), which serves as the standard for developing a

level of service criteria for a pedestrian facility

design.

Dimensions of a pedestrian are bound to be

different, when two different regions are compared.

As per the study conducted by (Singh et al., 2016)

concluded that compared to Western people Asians

are generally shorter with relatively less broader

shoulders as shown in Figure 2.

Singh, N., (Singh et al., 2016) studied Asian

people and found Asian ellipse dimensions to be

0.3476m × 0.5082m (Body Depth × Shoulder Width)

while the dimension provided by USHCM 2010 is

0.4572m × 0.6096m (Body Depth × Shoulder Width).

Therefore, it was concluded that Asians have 23.97%

relatively less body height and their shoulder are

16.63% less when compared to the Americans.

Figure 2: The study of Asian Ellipse dimensions (Singh et

al., 2016).

Table 2: Average dimensions of pedestrians for different

countries (Dynamics Ltd., 2005).

Population

M = Male; F

= Female

Width

(cm)

Depth

(cm)

Area in

Rectangle

)

Area in

Ellipse

)

France M 51.50 28.00 0.14 0.11

F 47.00 29.50 0.14 0.11

USA

M 51.50 29.00 0.15 0.12

F 44.00 30.00 0.13 0.10

Great

Britain

M 51.00 32.50 0.17 0.13

F 43.50 30.50 0.13 0.10

India M 45.50 23.50 0.11 0.09

F 39.00 25.50 0.10 0.08

Japan M 41.00 28.50 0.12 0.09

F 42.50 23.50 0.10 0.08

Hong

Kong

M 47.00 23.50 0.11 0.09

F 43.50 27.00 0.12 0.09

3 METHODOLOGY

Figure 3 shows the framework employed for this

study to estimate pedestrian characteristics, also

shows intermediate steps of data collection methods,

analysis, and an overview of validating the developed

model.

Figure 3: Research framework.

3.1 Data Collection

This study was conducted in Bangkok, Thailand. As

this study required an area with higher pedestrian

density to perform speed studies, so, BTS Sky Walk

Bangna – BITEC as shown in Figure 4, was selected.

Pedestrians were timed manually over a measured

length, and then their respective speeds were

calculated. A portable video camera was used to

collect all the data on pedestrian traffic. The camera

was placed in a fixed position to obtain a view that

encompassed all the selected study area. The surveys

were conducted for 7 hours during peak and off-peak

periods. Speed data for calibrating models and

density data for validating model has been collected.

In addition, the processed results of the walking speed

are shown in Figure 5.

Figure 4: Site location.

A Microsimulation Modeling of Pedestrian Characteristics in Bangkok Transit System Case Study

355

Figure 5: Results of walking speed (m/s=meter/second).

3.2 Pedestrian Simulation

This study used the traffic microsimulation modeling

software for pedestrians, PTV Viswalk, to analyze the

pedestrian capacity with Social Force Model theory

(Helbing & Molnár, 1995).

Fundamental notion of Social Force Model is to

model the elementary forces that influence motion of

pedestrians (as shown in Figure 6), this modelling is

analogous to how Newtonian mechanics is used to

resolve forces acting on a physical object. Total force

evaluated comprises of social, psychological, and

physical forces, which eventually results in an

entirely physical parameter acceleration of

pedestrian.

The forces discussed above come into play due to

the desire of pedestrians to reach a destination and

these forces are the results of the influence of various

factors such as obstacles, walls, other pedestrians etc.,

(PTV VISION, 2016). In the same way, pedestrian

behavior can be categorized into three different levels

as in (Hoogendoorn & Bovy, 2002):

 At strategic level of minutes to hours, a

pedestrian plans his or her route and generates a

list of stops/destinations.

 At tactical level of seconds to minutes, a

pedestrian chooses the route between the

destinations. Thereby he takes the network into

account.

 At operational level of milliseconds to seconds,

the pedestrian performs the actual movement to

avoid oncoming pedestrians, navigates through

a dense crowd, or simply continues the

movement toward his/her destination.

The Social Force Model encompasses both

operational and tactical levels of pedestrian behavior.

Therefore, this study discusses the aspects of the

strategic level of pedestrian behavior using factors

such as pedestrian volume and pedestrian speed.

Figure 6: Social Force Model (Helbing and Molnár, 1995).

Among all the other microscopic pedestrian

model SFM has the ability to model all the

interactions and tends to be more realistic when

reproducing the pedestrian walking environments

(Helbing et al., 2005). Also, SFM has been

considered by a majority of researchers (Teknomo,

2016) -(Helbing et al., 2006) - (W & A, 2007) -

(Lakoba et al., 2016) the reason being, social force

model accurately captures most of the phenomenon

resulting due to complex interactions between

pedestrians compared to other similar models.

4 DATA ANALYSIS AND

RESULTS

4.1 Calibration and Validation Process

The models were calibrated by adjusting two

parameters. First is the desired walking speed

collected from the site (shown in Figure 7), second is

the Asian body ellipse (shown in Figure 8). In order

to obtain a realistic pedestrian behavior, simulation

run was visually validated and necessary adjustments

were made accordingly, as shown and discussed in

Figure 3.

For the output to be considered satisfactory the

statistical GEH (Geoffrey E. Havers) equation value

of output was constrained to a value of GEH < 5.0 for

85 % of all modeled volumes. Further, the summary

of the validation process results is shown in Table 3.

=





()



()



(1)

Where E is the estimated traffic volume from the

simulation model, V is the observed traffic volume.

VEHITS 2022 - 8th International Conference on Vehicle Technology and Intelligent Transport Systems

356

Figure 7: Calibration of walking speed in the traffic

microsimulation model.

Figure 8: Calibration of Asian body ellipse in the traffic

microsimulation model.

Table 3: Result of the Validation Process.

From TO Observed Simulation GEH

11:00 11:30 1,262 1,238 0.6788

11:30 12:00 1,184 1,170 0.4081

12:00 12:30 1,432 1,440 0.2111

12:30 13:00 1,231 1,215 0.4575

13:00 13:30 1,681 1,717 0.8734

13:30 14:00 1,701 1,723 0.5317

14:00 14:30 1,750 1,739 0.2634

14:30 15:00 896 908 0.3996

15:00 15:30 1,326 1,324 0.0549

15:30 16:00 1,291 1,289 0.0557

16:00 16:30 1,413 1,440 0.7149

16:30 17:00 2,313 2,320 0.1454

17:00 17:30 1,834 1,851 0.3960

17:30 18:00 1,672 1,702 0.7304

18:00 18:30 1,555 1,556 0.0254

4.2 Simulation Results and Pedestrian

Flow Characteristics

After the model calibration, the analysis of pedestrian

flow rates at various pedestrian densities were

analyzed. For the analysis to provide greater

resolution, pedestrian volumes from 1,000

pedestrians/hr to 30,000 pedestrians/hr with an

increment of 1,000 pedestrians/hr were used.

It is noticed that the average pedestrian walking

speed is evidently higher in U.S when compared to

Asian walking speeds, which results in higher

average flow rates. However, when pedestrian

volumes near the walkway capacity as shown in

Figure 10 and Figure 11 in terms of pedestrian

density, the body size factor gains more importance.

Hence, resulting in the maximum Thai Pedestrian

flow rate being greater than U.S flow rates when

nearing critical volumes.

Under the free flow conditions as shown in Figure

9, it is evident that a pedestrian is relatively free to

decide on his/her walking speeds due to less

obstructive forces, this scenario was also observed in

simulation trials and it further validates the approach

of modeling pedestrian behavior using social force

model.

But when the pedestrian volume approaches the

capacity, the flow rate of this study is relatively

higher than US due to the smaller body size of Asians

on average. In addition, when comparing between

uni-directional and bi-directional walking, it is found

that the capacities of the flow rate are different. The

pedestrian flow rates of uni-directional and bi-

directional walking are 91 peds/m/min and 78

peds/m/min, respectively. Figure 12 compares the

variation observed between uni-directional and bi-

directional pedestrian flows vs area module observed

in this study and past studies, which indicates the

model proposed in this study have the potential to be

used to model the pedestrian behavior accurately.

Figure 9: PTV VISSIM 3D Animation (Low Density).

A Microsimulation Modeling of Pedestrian Characteristics in Bangkok Transit System Case Study

357

Figure 10: PTV VISSIM 3D Animation (High Density).

Figure 11: PTV VISSIM 2D Animation (High Density).

Figure 12: Pedestrian flow area module relationship.

5 CONCLUSION

The methodology of this study proves to be robust

enough to model the pedestrian behaviors accurately,

and the validation technique employed provides a

measure to compare and validate the observed and

simulated pedestrian volumes of the skywalk in

Bangkok. The results of this case study demonstrates

that Thai pedestrian walking speed is 75.07 m/min,

which is higher than 73.0 m/min of the previous

walking speed study results conducted in Bangkok in

1991 but still slower than U.S. pedestrian walking

speed.

The pedestrian flow per one-meter width of uni-

direction and bi-direction is 91 peds/m/min and 78

peds/m/min, respectively. It is also clearly indicated

that a uni-directional walkway would provide a

higher pedestrian flow rate than a bi-directional

walkway.

In addition, the maximum Thai pedestrian flow

rate obtained in this study is greater than the one

obtained in the U.S. It could indicate that the Thai

pedestrian body size is smaller than the U.S.

Therefore, this study concludes that walking speed

and body size significantly affect the pedestrian flow

rate. Most importantly, the methodology and results

of this study would provide useful information for

better planning and robust design of pedestrian

facilities in Bangkok. Similarly, the other cities with

similar pedestrian flow characteristics could also

adopt this methodology where walkways reach their

capacities and the importance of body sizes comes

into the foreground.

For future research, in addition to factors

considered in this study, it would be essential to

consider socio-economic factors like age, gender,

occupation, etc. Also, a comparison of different types

of pedestrian locations at midblock crosswalks,

signalized and unsignalized crosswalks, and

sidewalks would also be a great prospect for studying

VEHITS 2022 - 8th International Conference on Vehicle Technology and Intelligent Transport Systems

358

the pedestrian flow characteristics and behaviors

using the proposed methodology.

ACKNOWLEDGMENTS

The corresponding author wishes to express his

thanks to the Department of Civil Engineering,

School of Engineering, King Mongkut’s Institute of

Technology Ladkrabang for allowing him to conduct

this research. The authors would also like to thank

S2R Consulting Co. Ltd., and Mr. Satapana

Nammuang whose comments and suggestions helped

to improve this paper.

REFERENCES

Dynamics Ltd. (2005). Graphic - Levels of Service.

www.crowddynamics.com

Fruin, J. J. (1971). Pedestrian planning and design.

Gerilla, G. P., Hokao, K., & Takeyama, Y. (1995).

Proposed level of service standards for walkways in

Metro Manila. Journal of the Eastern Asia Society for

Transportation Studies, 1(3), 1041–1060.

Gupta, R. . (1986). Delhi 2010 AD: Cycle - An important

mode even after the 20th century. Int. Conf. on

Transportation System Studies, 625–632.

Helbing, D., Buzna, L., Johansson, A., & Werner, T.

(2005). Self-organized pedestrian crowd dynamics:

Experiments, simulations, and design solutions.

Transportation Science, 39(1), 1–24.

Helbing, D., Johansson, A., Mathiesen, J., Jensen, M. H., &

Hansen, A. (2006). Analytical Approach to Continuous

and Intermittent Bottleneck Flows. Physical Review

Letters, 97(16).

Helbing, D., & Molnár, P. (1995). Social force model for

pedestrian dynamics. Physical Review E, 51(5), 4282.

Hoogendoorn, S. P., & Bovy, P. H. L. (2002). Normative

Pedestrian Behaviour Theory and Modelling.

Transportation and Traffic Theory in the 21 St Century,

219–245.

Lakoba, T. I., Kaup, D. J., & Finkelstein, N. M. (2016).

Modifications of the Helbing-Molnár-Farkas-Vicsek

Social Force Model for Pedestrian Evolution, 81(5),

339–352.

Navin, F. P., & Wheeler, R. J. (1969). Pedestrian Flow

Characteristics. Traffic Engineering, Inst Traffic Engr,

39.

Pongprasert, P., & Kubota, H. (2017). Switching from

motorcycle taxi to walking: A case study of transit

station access in Bangkok, Thailand. IATSS Research,

41(4), 182–190.

Rastogi, R., Engineering, K. K. R.-J. of T., & 2003,

undefined. (2003). Travel characteristics of commuters

accessing transit: Case study. Ascelibrary.Org, 129(6),

684–694.

Singh, N., Parida, P., Advani, M., & Gujar, R. (2016).

Human Ellipse of Indian Pedestrians. Transportation

Research Procedia, 15, 150–160.

Tanaboriboon, Y, & Guyano, J. (1991). Analysis of

pedestrian movements in Bangkok.

Tanaboriboon, Yordphol, Hwa, S. S., & Chor, C. H. (1986).

Pedestrian characteristics study in Singapore. Journal of

Transportation Engineering, 112(3), 229–235.

https://doi.org/10.1061/(ASCE)0733-947X(1986)112:

3(229)

Teknomo, K. (2016). Microscopic Pedestrian Flow

Characteristics: Development of an Image Processing

Data Collection and Simulation Model.

https://arxiv.org/abs/1610.00029v1

W, Y., & A, J. (2007). Modeling crowd turbulence by

many-particle simulations. Physical Review. E,

Statistical, Nonlinear, and Soft Matter Physics, 76(4 Pt

2).

Ye, J. H., Chen, X., Yang, C., & Wu, J. (2008). Walking

Behavior and Pedestrian Flow Characteristics for

Different Types of Walking Facilities. Transportation

Research Record, 2048, 43–51.

Yu, M. F. (1993). Level of service design standards for non-

motorized transport in Shanghai, China. Asian Institute

of Technology, Bangkok.

A Microsimulation Modeling of Pedestrian Characteristics in Bangkok Transit System Case Study

359