Preliminary Study on Road Slipperiness Detection using Real-Time
Digital Tachograph Data
Jinhwan Jang
Dept. of Highway Res., Korea Inst. of Civil Eng. and Building Tech., 283 Goyangdae-ro, Goyang, South Korea
Keywords: Road Slipperiness, Digital Tachograph, Wheel Slip.
Abstract: Faced with the high rate of commercial vehicle-related traffic accidents, digital tachographs (DTGs) are
mandatorily installed in commercial vehicles in Korea. However, the current DTGs do not seem to be effective
for reducing accidents. One reason for this can be attributed to the absence of useful information for drivers
under dangerous road conditions such as black ice. In this study, an innovative technique to identify slippery
spots on the road using DTG data is proposed. The DTG can collect two types of vehicle speed: one is wheel
rotational speed and the other is vehicle transitional speed. The difference between the two speeds is referred
to as wheel slip, which can be exploited as a surrogate measure for detecting road slipperiness. A confidence
interval of wheel slip was established using data collected in dry road conditions; if any data point that exceeds
the predefined confidence interval is observed, a slippery road spot can be identified. The proposed method
was preliminarily tested in four types of winter road conditions and showed satisfactory results.
1 INTRODUCTION
Traffic accidents caused by slippery road conditions
are of great concern to society. According to Korean
statistics, 7,849 traffic accidents that caused 221
fatalities and 13,736 injuries occurred on slippery
roads over the last three years (KNPA, 2017). The
fatality rate on icy roads is 27% higher than on dry
roads. One study in Sweden insisted that only 14% of
drivers maintain appropriate acceleration and
deceleration under slippery road conditions and 50%
of drivers misjudge slippery road surfaces as normal
ones (Bogren, 2010). The risk of traffic accidents is
known to increase nine and twenty times under snowy
and icy road conditions, respectively (Luque, 2013).
To mitigate the tragic numbers, early detection of
slippery spots on the road followed by quick
notification to drivers and road managers seems to be
imperative. To this end, a road weather information
system (RWIS) has conventionally been deployed to
gather road weather data (Boon, 2002). However, as
RWIS can only collect spot data, its utility could be
restricted, given that road slipperiness is known to
vary even in short roadway sections (Nakatsuji,
2003). Due to this limitation coupled with the high
cost, RWIS has not been extensively deployed in
Korea. Recently, as increasing numbers of vehicles
are getting connected for various purposes, a new
cost-effective approach to detecting road slipperiness
using probe vehicles as a mobile sensing platform has
been garnering attention worldwide.
Several studies have been performed concerning
weather-related hazardous road conditions using
probe cars. A probe-based road weather data
collection system in Finland was developed by Pilli-
Sihvola (2000). The system gathered various road
weather-related data, including temperature,
humidity, friction, and video images. Probe cars for
the system were fitted with various devices—a
friction tester, a GPS antenna, an infrared sensor, a
combined air temperature and humidity sensor, a
video recorder, and a cellular modem. A real-time
tire-road friction coefficient estimation system using
probe vehicles equipped with a vehicle motion sensor
and a differential GPS module was invented by Wang
et al. (2004). A novel methodology for tire-road
friction estimation using a genetic algorithm and the
unscented Kalman filter combined with tire-models
and probe vehicle data was proposed by Nakatsuji et
al. (2005 and 2007). A probe application for the
identification of road slipperiness and coarseness
using vehicle body and wheel rotation speeds was
demonstrated by Dong et al. (2006). The potential use
of data gathered from a Vehicle Infrastructure
Integration (VII)-enabled probe vehicle in weather-
Jang, J.
Preliminary Study on Road Slipperiness Detection using Real-Time Digital Tachograph Data.
DOI: 10.5220/0006954702630267
In Proceedings of the 5th International Conference on Vehicle Technology and Intelligent Transport Systems (VEHITS 2019), pages 263-267
ISBN: 978-989-758-374-2
Copyright
c
2019 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
263
related applications and products for surface
transportation was proposed by Petty et al. (2007). A
probe-based weather-specific field study in Detroit
was conducted in 2009 and 2010 by Drobot et al.
(2010) and Chapman et al. (2010). A technique to
estimate the maximum friction while simultaneously
diagnosing the road conditions was discussed by
Koskinen (2010). Recently, Wyoming Department of
Transportation (WyDOT) has been deploying a
connected vehicle-based road weather information
system as part of a connected vehicle pilot project
(WyDOT, 2018).
As of 2011, digital tachographs (DTG) have been
obligatorily installed on every commercial vehicle in
Korea to reduce traffic accidents related to
commercial vehicles, for which the accident rate is
1.7 times higher than for general vehicles. A DTG
collects several types of vehicle data—location,
speed, brake activation, time of driving, etc. A
substantial number of DTGs installed on trucks are
connected to a truck management center via cellular
communications to transmit real-time data for
efficient logistics. In this study, a simple but novel
technique is proposed to identify slippery road spots
using the real-time DTG data. Cargo trucks are
considered to be advantageous over other types of
vehicles in that they are generally operated
throughout the nation on a 24/7 basis. Also, truck
drivers tend to choose night driving so as not to be
trapped in congestion, so information on icy spots on
the road ahead could be highly valued considering
that black ice is not easily discerned at night.
2 METHOD
2.1 Basic Concept
On a slippery road surface, vehicle wheels have a
high risk of skidding while decelerating, as well as
spinning in place while accelerating. This, called
wheel slip and expressed in (1), can be measured by
calculating the difference between the wheel
rotational and vehicle transitional speeds. In a DTG
platform, the wheel rotational speed can be measured
using wheel pulse sensors installed on each wheel of
a vehicle and the vehicle transitional speed can be
obtained using a GPS sensor installed in the DTG.
, 0≤s≤1
(1)
where s=wheel slip, w
w
=angular velocity of wheel,
r
w
=effective wheel radius, v
w
=vehicle transitional
speed, and w
w
r
w
=wheel rotational speed.
The two speeds are theoretically identical under
dry road surface conditions. However, they show
some discrepancy due to reasons such as
measurement errors of sensors, the characteristics of
vehicle tires made of rubber, and so on. Fig. 1 shows
values of wheel slip according to vehicle speed and
acceleration under dry road conditions: low and high
correlation with speed and acceleration, respectively.
The high correlation with acceleration was exploited
for measuring slippery road spots and the method will
be described in the next section.
(a)
(b)
Figure 1: Relationships between wheel slip and vehicle
dynamics: (a) speed and (b) acceleration.
2.2 Confidence Interval
The statistical confidence interval concept of a
regression line (or equation) expressed in (2)–(4) was
applied to identify road slipperiness. According to the
identical and independent distribution assumption of
the error term of the regression line, the confidence
interval of the i-th estimate can be determined. For
this study, the dependent variable corresponds to
wheel slip and the independent variable corresponds
to acceleration of a vehicle.
(2)
(3)
(4)
VEHITS 2019 - 5th International Conference on Vehicle Technology and Intelligent Transport Systems
264
where
=i-th dependent variable (estimate), x
i
=i-th
independent variable (observation), b=y-axis
intercept, e
i
=i-the error, s
e
=standard error of
regression line, y
i
=i-th observation, and n=number of
sample.
For a desired level of false alarm rate (α), a two-
sided confidence interval which includes (1-
α)*100% of the wheel slip can be defined with its
upper and lower limit values becoming thresholds
(Thr) of the regression equation for the particular
estimate of wheel slip, as expressed in (5). If any
wheel slip exceeds the thresholds, a slippery road spot
can be identified.
(5)
where N
-1
=inverse of normal cumulative distribution
function and α=significance level.
Table 1 shows the result of the regression analysis
of the relationship between wheel slip and
acceleration plotted on the right side of Fig. (1). To
determine the statistical confidence interval of the
wheel slip, a z-value corresponding to 99.7% was
employed.
Table 1: Regression analysis of the relationship between
wheel slip and acceleration.
Statistics
Value
Slope (a)
-1.2
y-axis intercept (b)
0.5
Correlation coefficient
0.8
Determination coefficient
0.6
Standard error (s
e
)
4.7
Number of sample
259
2.3 Algorithm
Based on the method stated above, the algorithm to
identify slippery road spots using real-time DTG data
can be established (see Fig. 2). The 99.7% confidence
interval (z=3) determined using historical DTG data
collected under dry road conditions is employed to
investigate whether a current wheel slip calculated
using real-time DTG data deviates from the lower or
upper limits set by the regression statistics
represented in Table 1. If any data point falls out of
the limits, the road spot is subsequently identified as
slippery. The slipperiness information is not only
displayed on the DTG installed in the commercial
vehicle for the driver, but transmitted to the traffic
management center via cellular communications for
following drivers and road managers.
Figure 2: Algorithm for identifying slippery road spot using
DTG data.
3 FIELD EXPERIMENT
To verify the proposed algorithm, field experiments
under various slippery road surface conditions were
performed in the 2017–2018 winter season on local
and arterial roads. A passenger car operated by a
skilled male driver with more than 20 years driving
experience was used for collecting real-time DTG
data including GPS coordinates, wheel speed, GPS
speed, and so on. For visual verification, a digital
video recorder time-synchronized with the DTG was
employed to gather video image data. The driver was
instructed to slightly accelerate followed by
decelerate while driving on the slippery road spots.
Fig. 3 shows the experiment results under four
types of slippery road conditions: ice, light snow,
slush, and heavy snow. Each slippery condition was
diagnosed by the proposed method, where the wheel
slips calculated by equation (1) exceeded the
confidence intervals predefined by equation (5) and
the regression statistics represented in Table 1.
Notably, the wheel slips gathered on the icy road
surface were more prominent than other surfaces in
terms of slippery road identification. This aspect can
be recognized as reasonable because the tire-road
friction coefficients on icy roads are generally known
to be much lower than on snowy or slushy roads
(Mcshane, 2003).
Preliminary Study on Road Slipperiness Detection using Real-Time Digital Tachograph Data
265
(a)
(b)
(c)
(d)
Figure 3: Field test under road conditions: (a) ice, (b) light
snow (c) slush, and (d) heavy snow.
4 DISCUSSIONS
Confronted with severe traffic congestion in
metropolitan areas in Korea, truck drivers are inclined
to operate their heavy vehicles loaded with high-
priced freight at night, when road surface conditions
are not easily discernible. Hence, road slipperiness
information gathered by mandatorily installed DTGs
on commercial vehicles for delivering the collected
information to following truck drivers, as well as road
agencies for a prompt response such as applying
chemicals, should be highly regarded. The suggested
method is simple but it was proved to reliably collect
slippery road spots with an extremely low cost
compared to conventional technologies such as
RWIS. It should be noted that the proposed method is
not applicable solely to DTG data. It can be broadly
applied to any devices that can collect vehicle body
transitional and wheel rotational speeds. Among such
devices are cooperative-intelligent transport systems
on-board units and smart-phones receiving on-board
diagnostics data of vehicles via the Bluetooth
communications.
However, there are some issues to be addressed to
improve the performance of the proposed method. In
this article, only the wheel slip method based on GPS
and wheel pulse sensors is presented to identify
slippery road spots. So, the proposed method cannot
be applied where there is interference with GPS
signals, especially in urban canyons or mountainous
areas. An alternative solution to the problem is to use
the rotational speed difference between the driving
and driven axle (or wheel) of a vehicle. However,
since current DTGs do not collect wheel rotational
speeds by the axle, the alternative method could not
be tested in this study. Furthermore, the proposed
wheel slip method can only collect slippery road spots
on the wheel track of a vehicle, not on the whole
driving lane. So, it is highly recommended to fuse the
proposed method with other techniques such as road
status recognition using vision sensors to enhance the
utilization of road slipperiness information from the
perspective of road users.
5 CONCLUSIONS
With rapid advances in wireless communications
technologies coupled with the evolving smart car
industry, increasing numbers of cars are getting
connected. In the era of a connected car environment,
more opportunities to exploit vehicle sensors to
obtain road status information become possible, such
VEHITS 2019 - 5th International Conference on Vehicle Technology and Intelligent Transport Systems
266
as the possible realization of road operations and
maintenance based on IoT, the cloud, big data, and
mobile technology. One example of the concept to
increase safety under slippery road conditions was
presented in this article. Wheel slip, defined as the
difference between vehicle’s transitional and wheel
rotational speeds, gathered by the DTG is employed
to recognize slippery road spots. Using the
confidence interval of the wheel slip under dry road
conditions, wheel slip data observed on slippery roads
can be identified. As of January 2018, around 30,000
DTGs in commercial trucks are connected and their
real-time data on a second-by-second basis are
transmitted to smart truck management centers
operated by private sectors. Considering that the total
number of commercial vehicles equipped with DTGs
is some 400,000, the size of the market for applying
the suggested method is substantial.
Due to the absence of a test facility that can
simulate various weather conditions, the field tests
could not but be limited in this study. The quantitative
measure of road slipperiness is the friction coefficient
between the vehicle tire and pavement, but the
friction is known to have a wide range in similar
adverse weather situations. The friction, due to the
experiments being carried out on real roads, could not
be measured for this study owing to safety concerns.
Fortunately, an adverse road weather simulation test
bed is under construction with the aid of the Korean
government, so the method presented herein could be
broadly tested under various friction conditions in the
near future in a quantitative manner. Techniques to
identify other anomalies on the road such as potholes
could be a study subject for subsequent research.
ACKNOWLEDGEMENTS
This work was supported by a grant from the Korea
Agency for Infrastructure Technology Advancement
(KAIA) (No. No.18TLRP-C145770-01).
REFERENCES
2017 Traffic Accident Statistics, Korean National Police
Agency (KNPA).
P. Luque, J. Wideberg, and D. Mantars, 2013. ITS to
Improve Safety and Efficiency OBD-II and Smartphone
Apps, CreateSpace Independent Publishing Platform.
J. Bogren and P. E. Caran, 2010. SRIS-Slippery Road
Information System, Intelligent Vehicle Safety
Systems.
C. B. Boon and C. Cluett, 2002. Road Weather Information
Systems: Enabling Proactive Maintenance Practices in
Washington State, Washington State Transportation
Center, University of Washington.
T. Nakatsuji and A. Kawamura, 2003. Relationship
between Winter Road Surface Conditions and
Vehicular Motion: Measurements by Probe Vehicles
Equipped with Global Positioning systems,
Transportation Research Record, No. 1824,
Transportation Research Board of the National
Academies, Washington, D.C.
T. Nakatsuji, I. Hayashi, A. Kawamura, and T. Shirakawa,
2005. Inverse Estimation of Friction Coefficients of
Winter Road Surface: New Considerations of Lateral
Movements and Angular Movements, Transportation
Research Record, No. 1911, Transportation Research
Board of the National Academies, Washington, D.C.
Y. Pilli-Sihvola, 2000. Floating Car Road Weather
Information Monitoring System, Transportation
Research Record, No. 1741, Transportation Research
Board of the National Academies, Washington, D.C.
J. Wang, L. Alexander, and R. Rajamani, 2004. GPS Based
Real-Time Tire-Road Friction Coefficient
Identification, University of Minnesota.
X. Dong, K. Li, J. Misener, P. Variyia, and W. Zhang, 2006.
Expediting Vehicle Infrastructure Integration (EVII),
California PATH Research Report, Institute of
Transportation Studies, University of California,
Berkeley.
K. R. Petty and W. P. Mahoney, 2007. Enhancing Road
Weather Information through Vehicle Infrastructure
Integration, Transportation Research Record, No.
2015, Transportation Research Board of the National
Academies, Washington, D.C.
S. Drobot, M. Chapman, E. Schuler, G. Wiener, W.
Mahoney, P. Pisano, and B. McKeever, 2010.
Improving Road Weather Hazard Products with
Vehicle Probe Data, Transportation Research Record,
No. 2169, Transportation Research Board of the
National Academies, Washington, D.C.
S. Koskinen, 2010. Sensor Data Fusion Based Estimation
of Tyre-Road Friction to Enhance Collision Avoidance,
Doctoral Dissertation, The Tampere University of
Technology.
https://wydotcvp.wyoroad.info/, Accessed in April, 2018.
W. R. Mcshane, R. P. Roess, and E. S. Prassas 2003. Traffic
Engineering (Second Edition), Prentice-Hall, New
Jersey.
Preliminary Study on Road Slipperiness Detection using Real-Time Digital Tachograph Data
267
