2008 IEEE Intelligent Vehicles Symposium

Eindhoven University of Technology

Eindhoven, The Netherlands, June 4-6, 2008

Vehicle/Pedestrian Conflict Analysis and

Exclusive Right-Turn Phase Setting Study
Yuelong Su, Danya Yao, Member, IEEE, Yi Zhang, Member, IEEE, Li Li, Member, IEEE, Zheng Wei

walkers are interrupted by right turning vehicles [9]; see

Abstract—Many Asian cities, i.e. Beijing, China, are facing Fig.1.
increasing conflicts between the booming vehicles and bicyclists/ From the viewpoint of pedestrians, to apply exclusive
pedestrians in the five years. One important alleviating method right-turn phase is a natural way to boost traffic efficiency
is to introduce exclusive right-turn phases which have rarely
been designed or set up for signalized intersections before. To
and reduce accidents. However, to add exclusive right-turn
estimate how this method can improve traffic efficiency, the phase certainly hinder vehicle traffic flow. Thus, we have to
conflicts between right-turn vehicles and other road users are find a good compromising point between vehicle traffic flow
first studied and modeled through vision monitoring. Then, traffic efficiency and pedestrian traffic efficiency.
different settings of phasing scheme are compared based on the To solve this problem, we need to answer two questions
simulation tests that are designed to incorporate the conflicts that lack detailed analysis at first:
characteristics. The results show that the vehicle/pedestrian
traffic efficiency varies with several conditions, i.e. vehicle/
1) what vehicle/pedestrian traffic flow efficiency could be
pedestrian arriving rate. This indicates that some appropriate without exclusive right-turning phase;
adaptive phasing strategies should be used to better serve traffic 2) what vehicle traffic flow efficiency could be with
requirements. exclusive right-turning phase.
In this paper, these two questions are answer by studying
Index Terms—Exclusive right-turn phase, conflict modeling. the vision-based traffic monitoring data. It should be pointed
out that the frequently encountered pedestrian intrusion
events are also taken into count when calculating the vehicle
I. INTRODUCTION traffic efficiency of exclusive right-turn phase [9]-[10].

E XCLUSIVE phase refers to a designated left- or right-

turn phasing time in one traffic signal control period used
only by vehicles making those turns [1]- [8]. It is designed to
Another problem should be emphasized is to directly test
different signal control strategies at a real intersection is very
difficult and troublesome. One main method is to optimize
timely separate the pedestrians, bicyclists and vehicles which phase schemes through simulations, since the performance
have to move across an intersection through the same area so according to different parameters can be obtained easily, and
as to avoid the collision accidents. If there are not so many the varying tendency of passing capacity can be observed
pedestrians/bicyclists or left/right-turn vehicles, there is no much more obviously. Thus, a surrogate simulation system is
need to set up exclusive left/right-turn phase so that we can built here. The results show that the vehicle/pedestrian traffic
dispatch more vehicles/pedestrians meanwhile. efficiency compromising point varies with several conditions,
However, many Asian cities, i.e. Beijing, China, are facing which indicates that some adaptive phase switching strategies
increasing conflicts between the bicyclists/pedestrians and should be applied.
the booming vehicles in the last five years. For example, the To give an detailed explanation of the proposed method,
lack of exclusive right-turn phase at many intersections in the rest of this paper is organized as follows: Section II
Beijing leads to significant congestions. Statistically, in 240 describes how the vehicle/pedestrian/bicyclists conflicting
green periods of the monitored intersections, 30% of crossing data are collected and statistically studied, Section III studies
some exclusive right-turn phasing schemes and Section IV
presents the simulation based comparison results of these
Manuscript received December, 15th, 2007.
Y. Su is with Department of Automation, Tsinghua University, Beijing, schemes. Finally, the paper is concluded in Section V.
China 100084 (e-mail: suyl05@mails.tsinghua.edu.cn).
D. Yao with School of the Information Science and Technology, Tsinghua II. CONFLICTS ANALYSIS
University, Beijing, China 100084 (e-mail: yaody@mail.tsinghua.edu.cn).
Y. Zhang with Department of Automation, Tsinghua University, Beijing,
China 100084 (e-mail: zhyi@mail.tsinghua.edu.cn).
A. Conflicting Data Collection
L. Li is with Department of Automation, Tsinghua University, Beijing, In order to obtain the conflicting data, one typical urban
China 100084 (corresponding author, phone: +86 (10) 62782071; e-mail: signalized intersections containing an exclusive right-turn
li-li@mail.tsinghua.edu.cn).
Z. Wei is with Department of Automation, Tsinghua University, Beijing, lane but without exclusive right-turn signals are monitored by
China 100084 (e-mail: maorithu@gmail.com). using video camera.

978-1-4244-2569-3/08/$20.00 ©2008 IEEE. 733

 Fig.1. Conflicts between right-turn vehicles and pedestrians.

may be caused by unusual pedestrian/vehicle behavior, are

neglected to ensure the statistical reliability.

Table I. Conflicting proportions of right-turn vehicles.

Fig.2. Indirect conflicts of right-turn vehicles.

The conflicts happened at peak hours (5:30 pm to 6:30 pm)

and off-peak hours (9:00 am to 10:00 am) are both recorded.
The time of right-turn vehicles passing intersection with
and/or without disturbance are calculated carefully.
The right-turn vehicles are divided into three categories in
this paper: pass the intersection without any conflicts, with
direct conflicts and with indirect conflicts.
If the first right-turn vehicle in the queue is interfered by
pedestrians and bicyclists, the other queuing vehicles in the
same queue would decelerate or stop, resulting in traffic jam
in the right-turn lane. Such a situation is defined as indirect
conflicts of right-turn vehicles; see Fig.2.

B. Statistical Results of Conflicting Data

Some fundamental statistical results of the conflicts are
given in Table I and Fig.3.
As shown in Fig.3, comparing to that without disturbance,
the average passing time of right-turn vehicles encountered
disturbance increased sharply from 6 seconds to 10 seconds.
In other words, it is noteworthy that 70% of the passing time
results from pedestrians' and bicyclists' disturbances, if there
is no protection of exclusive right-turn phase. This estimation
has been validated through other traffic data, and applied to Fig.3. The average passing time with and without disturbance.
the simulation of setting exclusive right-turn phase.
The abnormal samples, i.e. nodes 2 and 69 in Fig.3 which

734
 III. EXCLUSIVE RIGHT-TURN PHASE DESIGN Table II. Conventional phase schemes for different intersection types
Phasing scheme for the Cross-shaped intersection in Fig.4(a)
A. Problem Descriptions
The exclusive right-turn phase design must consider a wide
variety of the prevailing conditions, including the (geometry)
type of intersection, the traffic channeling scheme and the
signal control strategy, etc.. The diagrams of the popular
Cross-shaped/T-shaped intersections are shown in Fig.4-5,
combining with channeling scheme of lanes. The associated Phasing scheme for the Cross-shaped intersection in Fig.4(b)
phase designs including pass through, left- and right- turn are
illustrated as Table II, where the exclusive right-turn phase
has been integrated into the whole phasing design.

Phasing scheme for the T-shaped intersection in Fig.5

The moving direction of vehicles

B. Improved Phasing Scheme

(a) For many cross-shaped intersections in Beijing, the preset
phasing scheme as given in Table II should be optimized. The
deficiencies lie in two parts:

1) the waiting time of right-turn vehicles can be shortened,

since the right-turn vehicles can be discharged only once in
the whole period;
2) the low flux of left-turn vehicles frequently results in the
waste of exclusive left-turn green time.

Considering these facts, a new phase design is illustrated in

Table III, where the right-turn vehicles can be discharged
twice in one period.

Table III. Optimized phase schemes for the Cross-shaped intersection

Phasing scheme for the Cross-shaped intersection in Fig.4(b)
(b)
Fig.4. The channeling schemes diagram of different Cross-shaped
intersections.

IV. COMPARISON STUDY OF THE DIFFERENT EXCLUSIVE

RIGHT-TURN PHASE DESIGN

A. Simulation Framework
Based on what are discussed above, the traffic efficiency
Fig.5. The channeling schemes diagram of T-shaped intersections.
varieties with/without exclusive right-turn phase are tested

735
through simulations first. This is mainly because simulation is phasing time is assumed 40 seconds, the accumulative total
the most economical and effective method to estimate the passing intersection's time equals to:
influence of exclusive right-turn phase.
The simulating calculation consists of two steps. The first (500/10)*40s+(500/10)*30s=3500s
step is to determine the delay of right-turn vehicles, which are
disturbed by pedestrians and bicyclists without phasing Thus, the saving time with exclusive right-turn phase equal
protect. In the simulation, the passing intersection's time of to:
right-turn vehicles with disturbance is about 70% more than
that without disturbance according to statistic. In either case, 4075s-3500s=575s
the needed minimal green time of discharging vehicles in a
Let's consider more complex simulation situations, where
queue are calculated on the basis of our previous research
four parameters can be changed: the saving times of passing
results [9]-[10]; see Table IV. In the second step, the capacity
intersection with exclusive right-turn phase can be calculated
at the signalized intersection with/without setting exclusive
totally, these parameters including vehicles waiting for right
right-turn phase is analyzed; see Fig.6.
turning (W), discharging vehicles in one right-turn phase (D),
Table IV. The needed minimal green time for queues with different lengths the average passing intersection’s time per vehicle (T) and the
Vehicles through green phasing time (G). The computing results are
in the 5 6 7 8 9 10 shown in Table V, in which the negative values represent that
queue
time of passing intersection will be enhanced if there exists
Minimal
Green 13s 16s 18s 20s 23s 25s exclusive right-turn phase. Here in Table V, the probability of
Time conflicts between right-turn vehicles and other road users is
assumed 0.9, this proportion accords with the real statistical
traffic data observed in Beijing. If the conflicting percent is
changed to 50%, the computing results are shown in Table
VI.

Table V. The computing results of simulations (with conflicting

probability set as 0.9) (Unit: second)

Fig.6. The framework of the simulation

Table VI. The computing results of simulations (with conflicting

B. Simulation Example probability set as 0.5) (Unit: second)

For example, given 500 vehicles waiting for right turning

and assume the average passing time per vehicle is 5 seconds
without conflict. According to previous statistical values, the
probability of conflicts between right-turn vehicles and
pedestrians is 90%, resulting in 70% increase of intersection
passing time. So if there is not exclusive right-turn phase, the
accumulative total passing time equals to:

500*10%*5s+500*90%*5s*(1+70%)=4075s

Without changing any parameters above, suppose there is C. Efficiency Analysis

exclusive right-turn phase, assuming 10 queuing vehicles will Through varieties of conflicting probability, the efficiency
be discharged in one right-turn phase, the right-turn green of setting exclusive right-turn phase can be studied according
phasing time equals to 30 seconds (the need minimal green to a wide variety of conditions, including different conflicting
time + passing intersection's time = 25s+5s) on the basis of
probabilities, different discharging right-turn vehicles in one
the method shown in Fig.6. Additionally, if the through green
green phase and different intersection widths.

736
 1) Efficiency with Respect to Conflicting Probability

By changing conflicting probability from 0.2 to 0.9, the

tendency of saving time can be observed as shown in Fig.7.
Parameters of discharging vehicles in one right-turn phase,
the average passing intersection's time per vehicle and the
through green phasing time are chosen as 10, 5s, and 40s
respectively. In short, the saving time because of applying
exclusive right-turn phase is decreased with the reduction of
conflicting probability, if the conflicting probability is lower
than 60%, setting exclusive right-turn phase will extend the
needed green time of total right-turn vehicles, so the saving
times become negative. Four different color lines represent
different traffic flow of right-turn vehicles varying from 350s
to 500s. Fig.8. Variation of saving time with respect to discharging vehicles queue
length.

3) Efficiency with Respect to Average Passing Time

Through changing average passing intersection’s time of

right-turn vehicles from 5s to 10s, the tendency of saving
times can be observed as shown in Fig.9. Parameters of the
conflicting probability, discharging vehicles in one right-turn
phase and the through green phasing time are chosen as 0.9, 8,
and 30s respectively. The saving times for right-turn vehicles
are increased sharply with the extension of intersection’s
width, more wide intersection needing more excessive
passing time for right-turn vehicles. Four different color lines
represent different traffic flow of right-turn vehicles varying
from 350s to 500s.

Fig.7. Variation of saving time with respect to conflicting probability.

2) Efficiency with Respect to Discharging Queue Length

Through changing discharging vehicle length in one right-

turn phase from 5 to 10, the tendency of saving time can be
observed as shown in Fig.8. Parameters of the conflicting
probability, the average passing intersection's time per
vehicle and the through green phasing time are chosen as 0.9,
7s, and 30s respectively. The saving times for right-turn
vehicles are increased sharply with the enhancement of
discharging vehicles in one right-turn phase, four different
color lines represent different traffic flow of right-turn
vehicles varying from 350s to 500s. This phenomenon can be
explained by considering the stop-and-go characteristics of
vehicles: the accumulative start-up lost time will increase Fig.9. Variation of saving time with respect to average passing time.
with the reduction of discharging vehicles in one right-turn
phase.
V. CONCLUSIONS
From the above simulation tests we can see that if the
conflict probability occurring in the crosswalk between
pedestrian and right-turning vehicles is higher than 0.6,
setting exclusive right-turn phase should be considered. It is

737
apparent that more pedestrians lead to higher conflicting [9] Y. Su, Z. Wei, D. Yao, and Y. Zhang, "Start-up lost time and headway
model of queuing motor vehicles," Journal of Highway and
probability. So, based on observation, when the vehicle Transportation Research and Development, 2007.
arriving rate is above 500 vehicles/hour and the pedestrian [10] Y. Su, Z. Wei, S. Cheng, and D. Yao, et. al., "Departure headways of
arriving arte is above 800 person/hour, exclusive right-turn mixed traffic flow at signalized intersections: Distributions, simulations
and validations," Transportation Research Board Annual Meeting,
phase is suggested.
Proceeding CD, 2008.
The saving times for right-turn vehicles are increased
sharply with the enhancement of discharging vehicles in one
right-turn phase, because this exclusive right-turning curtail
the accumulative start-up lost time. But except for right-turn
vehicles, the vehicles' waiting times of other directions
(through and left-turn) should be balanced in the design of
signal control strategy. In the future research, optimizing
solutions to more complicated signal phase timing will be
studied.
The efficiency of exclusive right-turn phase will increase
with the width of intersection, since the vehicles got more
chance to conflict with pedestrians. If the exclusive right-turn
phase is used in the ramping areas of the in-city expressways
in Beijing, each right-turning vehicle needs at least 7 seconds
to pass the intersection. Thus, it is expected that to apply
exclusive right-turn phase in all such intersections or build
more pedestrian over-crossings or underground passages may
have an immediate and substantial impact on Beijing traffic
efficiency.

ACKNOWLEDGMENT
This work was supported in part by Hi-Tech Research and
Development Program of China (863 Project) 2006AA11Z2
08, 2006AA11Z229; the National Basic Research Program of
China (973 Project) 2006CB705506 and National Natural
Science Foundation of China 50708055.

REFERENCES
[1] The Transportation Research Boards, Highway Capacity Manual,
National Research Council, Washington, D. C. 2000.
[2] A. Katz, D. Zaidel, and A. Elgrishi, "An experimental study of driver
and pedestrian interaction during the crossing conflict," Human Factors,
vol. 17, no. 5, pp. 514-527, 1975.
[3] C. V. Zegeer, K. S. Opiela, and M. J. Cynecki, "Effect of pedestrian
signals and signal timing on pedestrian accidents," Transportation
Research Record, vol. 847, pp. 62-72, 1982.
[4] S. Khasnabis, C. V. Zegeer, and M. J. Cynecki, "Effects of pedestrian
signals on safety, operations, and pedestrian behaviour - Literature
review," Transportation Research Record, vol. 847, pp. 78-87, 1982.
[5] C. I. Howarth, "Interactions between drivers and Pedestrians: some new
approaches to pedestrian safety," Human Behaviour and Traffic Safety,
L. Evans and R. C. Schwing, eds. Plenum Press, New York, NY, USA,
pp. 171-178, 1985.
[6] Y. Kwon, S. Morichi, T. Yai, "Analysis of pedestrian behavior and
planning guidelines with mixed traffic for narrow urban streets,"
Transportation Research Record, vol. 1636, pp. 116-123, 1998.
[7] P. A. Dueer, "Dynamic right-of-way for transit vehicles: Integrated
modeling approach for optimizing signal control on mixed traffic
arterials," Transportation Research Record, vol. 1731, pp. 31-39, 2000.
[8] Z. Z. Tian, T. Urbanik, R. Engelbrecht, K. Balke, "Pedestrian timing
alternatives and impacts on coordinated signal systems under split
phasing operations," Transportation Research Record, vol. 1748, pp.
46-54, 2001.

738