Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 187 (2017) 206 – 211

10th International Scientific Conference Transbaltica 2017:

Transportation Science and Technology

The Influence of Human Factor on Congestion Formation

on Urban Roads
Jan Lizbetin, Ladislav Bartuska*
Department of Transport and Logistics, Faculty of Technology, The Institute of Technology and Business in České Budějovice, Czech Republic

Abstract

Road urban transport and traffic engineering in general are long-term discussed topics. Traffic problems are solved from many
aspects, such as transport infrastructure, transport organization, high share of car traffic, a negative impact on air quality,
emissions etc. Finally, the problem of traffic congestion on urban roads is also highly topical. With increasing economic
development of individual states the numbers of cars are increasing as well − the traffic volumes are increasing at the expense of
road capacity. The paper deals with the formation of congestion particularly on urban roads, where the driver reaction time
influences the size of congestion as well. In the first part of the paper the issues of traffic flow and congestion formation are
theoretically discussed. The second part of the paper outlines in practice an influence of human factor (the driver’s Perception −
Response Time) on the congestion formation.
© 2017
© 2017The TheAuthors.
Authors. Published
Published by Elsevier
by Elsevier Ltd.is an open access article under the CC BY-NC-ND license
Ltd. This
Peer-review under responsibility of the organizing committee of the 10th International Scientific Conference Transbaltica 2017:
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Transportation
Peer-review underScience and Technology.
responsibility of the organizing committee of the 10th International Scientific Conference Transbaltica 2017

Keywords: traffic congestion, traffic flow, urban road infrastructure, perception response time of driver

* Corresponding author.
E-mail address: bartuska.vste@seznam.cz

1877-7058 © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of the 10th International Scientific Conference Transbaltica 2017
doi:10.1016/j.proeng.2017.04.366
 Jan Lizbetin and Ladislav Bartuska / Procedia Engineering 187 (2017) 206 – 211 207

1. Introduction

Definition of traffic congestion according to the authors Cerna and Cerny is as follows [1]: “Traffic congestion is
clustering the vehicles on road section within a traffic network which reduces the throughput and significantly slows
the velocity of the traffic flow”.
Congestion can be recurrent (regular, occurring on a daily, weekly or annual cycle) or non-recurrent (traffic
incidents, such as accidents and disabled vehicles). Economist William Vickrey identified six types of
congestion [2]:
• Simple interaction on homogeneous roads: where two vehicles travelling close together delay one another.
• Multiple interaction on homogeneous roads: where several vehicles interact.
• Bottlenecks: where several vehicles are trying to pass through narrowed lanes.
• “Trigger neck” congestion: when an initial narrowing generates a line of vehicles interfering with a flow of
vehicles not seeking to follow the jammed itinerary.
• Network control congestion: where traffic controls programmed for peak-hour traffic inevitably delay off-peak
hour traffic.
• Congestion due to network morphology, or polymodal polymorphous congestion: where traffic congestion
reflects the state of traffic on all itineraries and for all modes. The cost of intervention for a given segment of
roadway increases through possible interventions on other segments of the road, due to the effect of triggered
congestion.
Formation of congestion is closely related to the density of traffic (see next chapter). When traffic volume
approximates the capacity of the road, congestion are generated as well. In cities, this condition becomes mostly
during the morning and afternoon peak hours, when the demand for transport is most significant. Figure 1 shows a
morning and afternoon peak traffic within the identified traffic volumes during the day on urban road section [3].

Fig. 1. Day-long traffic volumes determined by authors on urban road section. Source: Authors.

2. Traffic flow theory

For the purposes of this article, the authors demonstrate the macroscopic approach to the traffic flow theory also
known as the constant time headway (CTH) model. When applying this model, the idea from the field of traffic
safety is considered, i.e. that each reference vehicle keeps a constant distance from the vehicle ahead. Based on this
model the characteristics of the traffic flow can be generally determined: velocity v, density k, and traffic volume q
(flow) [4].
208 Jan Lizbetin and Ladislav Bartuska / Procedia Engineering 187 (2017) 206 – 211

Density (k) is defined as the number of vehicles per unit length of the roadway. Inverse of density is spacing (s),
which is the center-to-center distance between two vehicles. Traffic volume or flow (q) is the number of vehicles
passing a reference point per unit of time, vehicles per hour. The inverse of flow is headway (h), which is the time
that elapses between the i-th vehicle passing a reference point in space and the (i + 1)-th vehicle. In
congestion, h remains constant.
We assume that the road capacity (or traffic volume) q can be expressed as a time period in proportion to the sum
of the length of vehicles and the distance between those vehicles. Neglecting the length of the vehicle, the hourly
capacity (hourly traffic volume) can be expressed as [4]

3600
q= = const . , for h = 2 s it is q = 1800 [cars/hour]. (1)
h

This safety distance is given as time reserve expressing reaction time of the driver. However, in order to have a
realistic model it is necessary to take into account the non-zero length of the vehicle together with the length of
distance between the two vehicles. The common practice to analyze mixed traffic flow is to convert all vehicles into
equivalent numbers of passenger car units (PCUs) which correspond to the PCU length of 4−5 meters. At the speed
v we can describe in simple relation the length (distance) of vehicles [5]

d = d min + vh . (2)

Density is subsequently the inverse of vehicle distances according to

1
k (v )= . (3)
d min + vh

After transferring to the opposite relation velocity dependence on density is obtained according to

1⎛ 1 ⎞
v(k )= ⎜ − d min ⎟ . (4)
⎜ ⎟
h⎝ k ⎠

To express the intensity depending on the density we obtain a linear progression

1
q(k )= v(k )⋅ k = (1− d min k ) . (5)
h

We know from experience that the density of traffic flow is changed in a way that the place with the same density
is transmitted along the traffic flow in the direction of vehicles movement or in the opposite direction. For example,
after starting of the flow of vehicles during the traffic signal “go” at a controlled intersection it is possible to track
the movement of place with increased density, which is created by slowing down and then again accelerating
vehicles. In the traffic flow there are some longitudinal waves. Small changes in the traffic flow spread in this flow
in so called kinetic waves [6, 7].

3. Practical research on the influence of human factor on congestion formation

Psychology is inherent to the transport processes, accidents, or congestion. Ninety percent of all accidents are
caused by human factor. The remaining ten percent are other causes mostly combined with human factor. Accident
rate is closely related to congestion, especially in towns and their centers.
 Jan Lizbetin and Ladislav Bartuska / Procedia Engineering 187 (2017) 206 – 211 209

Traffic psychology is engaged in teaching and training drivers, rehabilitation of drivers, and driver behavior,
which is not always in line with how they should behave towards other road users [8].
The perception-response time is in terms of traffic psychology the period which is calculated from the emergence
of an unexpected event until the driver’s response. According to the literature, this period lasts about 2 seconds. It
varies according to the characteristics of the driver (age, physical fitness, health, and other factors) [9].
The accordion effect is a slang term, which means a slow driver reaction time to start or deceleration of the
vehicle ahead. This creates congestion where the human factor is to blame. This phenomenon has been studied at
busy intersections in České Budějovice during the morning and afternoon rush hours, the time with the daily highest
levels of congestion. Drivers’ reaction times to starting of the vehicles in front of them were measured. Table 1
shows the measured reaction times of drivers in the morning rush hour.

Table 1. Driver perception-response times in the morning rush hour. Source: authors.
driver reaction time 1 [s] driver reaction time 2 [s] driver reaction time 3 [s]
1 1.02 1 1.75 1 0.55
2 1.75 2 3.71 2 1.07
3 2.12 3 1.4 3 0.45
4 0.98 4 1.52 4 1.56
5 2 5 1.75 5 2.89
6 1.2 6 2.17 6 1.22
7 3.1 7 1.23 7 0.81
8 1.65 8 0.95 8 1.15
9 4.68 9 2.22 9 1.1
10 1.4 10 1.45 10 2.12
∑ 19.9 ∑ 18.15 ∑ 12.92
average 1.99 average 1,815 average 1.292
Average of reaction times 1.699
Sum of reaction times 50.97

Measurements took place in the morning peak at the intersection of Husova třída and Dlouhá Louka in České
Budějovice. Every time ten drivers’ reaction times were measured. For the accuracy of the result, the measurement
was repeated three times. For each measurement the total driver reaction time and average reaction time were
calculated. Of these three averages overall average of driver reaction time was calculated. Measurement began from
the time of switching on the “Caution” traffic light signal until starting of the first driver [10]. Subsequently, from
starting of the first driver to starting of the second driver etc. The overall average reaction times in the morning peak
were 1.699 seconds. The total sum of thirty drivers’ reaction times was 50.97 seconds.
If the driver reaction time was less than 1.699 seconds, more vehicles would move and congestion would reduce.
For the calculation we determine that the reaction time of each driver is e.g. 1.4 seconds and the calculation will be
at the number of ten vehicles [11].
In situation A, 10 vehicles in the column would successively start to move in 16.99 seconds. In situation B, 10
vehicles in the column would successively start to move in 14 seconds.
The difference of the reaction time is 2.99 seconds, which means that two more cars would drive through the
intersection. At the number of 30 cars, where the sum of the reaction times is 50.97 seconds (as measured), six more
cars would drive through [12].
Measurements took place in the afternoon peak at the intersection of Husova třída and Dlouhá Louka in České
Budějovice. Again, ten driver reaction times were measured. As with the previous table, measurements were
repeated three times. Amounts of reaction times and the average reaction times of drivers resulted from these
measurements and these averages were averaged into total average [13]. Measurement began from the time of
switching on the “Caution” traffic light signal until starting of the first driver. Subsequently, from starting of the first
210 Jan Lizbetin and Ladislav Bartuska / Procedia Engineering 187 (2017) 206 – 211

driver to starting of the second driver etc. The overall average of the reaction times in the afternoon peak is 1.594
seconds, and the sum of thirty drivers’ reaction times is the total of 47.83 seconds.

Table 2. Driver perception-response times in the afternoon rush hour. Source: authors.
driver reaction time 4 [s]
1 1.07
2 0.4
3 3.01
4 1.42
5 0.87
6 2.89
7 1.27
8 1.12
9 1.29
10 1.45
∑ 14.79
average 1.479
Average of reaction times
Sum of reaction times
driver reaction time 5 [s]
1 1.5
2 0.38
3 1.54
4 2.46
5 1.78
6 0.99
7 1.65
8 2.67
9 1.9
10 3.05
∑ 17.92
average 1.792
driver reaction time 6 [s]
1 0.78
2 1.28
3 2.96
4 1.15
5 0.65
6 1.69
7 1.22
8 2.09
9 1.14
10 2.16
∑ 15.12
average 1.512
1.594
47.83

If the driver reaction time was shorter, e.g. 1.4 seconds, more vehicles would move and congestion would be
reduced. At the number of 30 cars, where the sum of reaction times is 47.83 seconds, three more automobiles would
drive through.

4. Conclusions

Congestion in urban areas is currently one of the most pressing problems in transport. One of the reasons for the
formation of congestion in cities is delayed perception-response time of drivers who, according to surveys, are
waiting in the column for the commencement of a stationary vehicle in front of them and do not follow consistently
traffic in the column, or rather the traffic light signalling at the intersection [14, 15]. By the delayed reaction of
drivers a so called accordion effect is caused, which negatively and unnecessarily reduces throughput of the critical
point, i.e. the intersection, and thereby contributes to increasing congestion, or rather its extension. Human factor
thus has a significant effect on the formation of congestion, however, not a priority one [16, 17]. To a great extent,
traffic management system in the city is involved in the creation of congestion in the city as well, especially traffic
signals and their settings at intersections. Several measurements could significantly reduce congestion in the city:
the implementation of various telematic systems as well as the correct setup and synchronization of traffic light
signalling at intersections (i.e. creating a “green wave”); increasing the capacity of roads and construction of other;
traffic regulation; limiting the right of entry to certain areas, or charging of the traffic within the city areas [18].

References

[1] Cerna, J. Cerny, Teorie řízení a rozhodování v dopravních systémech (in Czech), Pardubice: Institut Jana Pernera. 2004. ISBN 80-86530-15-9.
[2] William S. Vickrey, Congestion Theory and Transport Investment, American Economic Review 59(2) (May 1969) 251−260. Available from
Internet: http://ideas.repec.org/a/aea/aecrev/v59y1969i2p251-60.html
[3] L. Bartuska, V. Biba, R. Kampf, Modeling of Daily Traffic Volumes on Urban Roads. Proc. of ICTTE. City Net Scientific Research Center
Ltd. Belgrade (2016) 900−905. ISBN 978-86-916153-3-8.
[4] P. Holcner, J. Macur, T. Apeltauer, Pattern Formation in Traffic Microscopic Model, Traffic and Granular Flow 07, Springer Verlag GmBH
(2007) 249−254. ISBN 3540770739.
 Jan Lizbetin and Ladislav Bartuska / Procedia Engineering 187 (2017) 206 – 211 211

[5] P. Holcner, Safe distance between vehicles and traffic flow simulation (in Czech), Stavebni obzor, published by Faculty of Civil Engineering
8 (2009) 251−255. CTU in Prague, VUT, VSB, Praha. ISSN 1210-4027.
[6] L. Bartuska, O. Stopka, J. Lizbetin, Methodology for Determining the Traffic Volumes on Urban Roads in the Czech Republic. Transport
Means – Proc. of the 19th Intern. Scientific Conference on Transport Means. Kaunas: Kaunas University of Technology. 2015, pp. 215−218.
ISSN 1822-296X.
[7] V. Rievaj, O. Stopka, J. Vrabel, L. Mokrickova, C. Schmidt, The impact of air resistence on the fuel consuption in real conditions within the
transport operation, Communications – Scientific Letters of the University of Zilina 2 (2016). ISSN 1335-4205.
[8] M. Sucha, V. Rehnova, M. Koran, D. Cernochova, Dopravní psychologie pro praxi: Výběr, výcvik a rehabilitace řidičů (in Czech). Praha:
Grada Publishing. 2013. ISBN 978-80-247-4113-0.
[9] M. Tomasikova, F. Brimercik, A. Nieoczym, Vehicle Simulation Model Criterion, Logi − Scientific J. on Transport and Logistics 6(1) (2015)
130−135. ISSN 1804-3216.
[10] I. Simkova, V. Konecny, Key performance indicators in logistics and road transport, Logi − Scientific Journal on Transport and Logistics
5(2) (2014) 87−96. ISSN 1804-3216.
[11] D. Hraskova, V. Bartosova, Managerial Reporting: A Flexible Tool for Managing Transport Company, Advances in Social and Behavioral
Sciences 5 (2014) 97−100. ISSN 2339-5133.
[12] T. Kliestik, Optimization of Transport Routes Based on Graph Theory as a Part of Intelligent Transport Systems. In: 17th International
Conference on Transport Means, Transport Means − Proceedings of the International Conference, Kaunas (Lithuania): Kaunas University of
Technology (2013) 308−311. ISSN 1822-296X.
[13] L. Rovnanik, J. Gnap, Application of reverse network system in urban waste management, Logi − Scientific Journal on Transport and
Logistics 6(1) (2015) 106−114. ISSN 1804-3216.
[14] I.Futyu, G. Gal, Implementing the qualitative consistency of traffic flows in equilibrium models, Logi − Scientific Journal on Transport and
Logistics 4(1) (2013) 40−52. ISSN 1804-3216.
[15] W. Gardziejczyk, M. Motylewicz, Noise level in the vicinity of signalized roundabouts, Transportation Research Part D: Transport and
Environment 46 (2016) 128−144. ISSN: 13619209. doi: 10.1016/j.trd.2016.03.016
[16] A. Osama, T. Sayed, M. H. Zaki, K. Shaaban, An inclusive framework for automatic safety evaluation of roundabouts, Journal of
Transportation Safety and Security (2016) 1−18. ISSN: 19439962. doi: 10.1080/19439962.2015.1107795
[17] M. Gogola, M. Veternik, The analysis of the environmental impact of road traffic in city of Martin, Logi − Scientific Journal on Transport
and Logistics 7(1) (2016) 49−60. ISSN 1804-3216.
[18] R. Kampf, J. Lizbetin, L. Lizbetinova, Requirements of a transport system user, Communications – Scientific Letters of the University of
Zilina 14(4) 106−108. ISSN 1335-4205.
[19] A. Marczuk, W. Misztal, T. Słowik, W. Piekarski, M. Bojanowska, I. Jackowska, Chemical determinants of the use of recycled vehicle
components. Przemysł Chemiczny 94(10) (2015) 1867−1871. doi:10.15199/62.2015.10.46