IoT Based Humps and Pothole Detection on Roads

and Information Sharing

1
E.RAJESH, 2T.VASANTHANILAVAN
1,2
II nd YEAR, DEPARTMENT OF ECE,
ANNAI TERESA COLLEGE OF ENGINEERING,THIRUNAVALUR,VILLUPURAM

Abstract— One of the ways to reduce road accident is to makes smooth and comfortable travel for passengers. The
identify the humps and potholes present in the path. In this major cause for irregularities of the road such as cracks,
paper, an internet of things based road monitoring system (IoT- potholes, and humps are led to damage the vehicles [2]. The
RMS) is proposed to identify the potholes and humps in the road. potholes and roughness level in the roads were monitored by
The pathway which is affected by the pothole is greatly using the smart-phones accelerometer for capturing the
influenced by the scattering signal of the ultrasonic sensor. So the vibration. The Gaussian model based mining algorithm is used
magnitude of the reflected signal decreased due to the roughness
of the surface and the signal amplitude is difficult to analyze. The
to detect the abnormalities of the road [3]. The Haar wavelet
Kirchoff''s theory basically applied for real-time analysis and it transform based on multi-scale image segmentation is used to
has certain limitations. To overcome this difficulty, an identify the changes in the damaged regions of Kullback
accelerometer has been included with the ultrasonic sensor to Leibler. This method automatically detects the damages in the
measure variation present in the signal and optimized using pathways. So, that there will be a smooth flow of
honey bee optimization (HBO) technique. The IoT-RMS transportation [4]. A track monitoring system using an
automatically updates the status of the road with location accelerometer with global positioning system has been
information in the cloud. Each road vehicles can access the introduced by Chellaswamy et al. [5]. The system
information from the server and estimate the speed according to automatically identifies the abnormalities present in the track
the potholes and humps present on the road. The simulation has
with location information.
been done and the result shows that the IoT-RMS can be
accommodated in road vehicles to reduce the accidents. The To reduce the roadbed damages ground penetrating radar-
proposed system is implemented and tested using Arduino Uno based system is used in the driven head at the regular speed of
with ESP 8266. 600 km/h. This will improve to maintain the road by
eliminating errors and pavement from blind spots [6]. Infrared
Keywords— Internet of things (IoT), road monitoring, HBO wave-based detection is used to detect the ice formation in the
optimization technique, pothole and hump detection hilly regions and this system analyzes the temperature of the
asphalt surface. Here the data are collected together and
I. INTRODUCTION evaluated to determine whether the ice is formed or not [7]. In
Nowadays the infrastructure development plays an agricultural marketing, the transportation plays the vital role.
important role in the economic growth and success of a Since the farmers transport the agricultural crops by means of
country. Development of infrastructure meets the great transportation. So it is essential to transport the cultivated
challenge and it creates a way to enter in to the outside world. crops at the right time to the market. A new approach is used
A well-structured network is used for communication, trade, to detect the reliability of the road with less transportation
industry, and social mobility of a country. The road provides time [8]. An IoT based track condition monitoring system is
maximum safety when it is maintained properly and the used for maintaining the quality of the track. The status of the
surface discontinuity may cause an increasing number of track is updated in the cloud and share the information with
accidents. The bad road condition is the main reason for all the the driver for safety measures [9].
truck accidents, according to the survey in [1]. The weakened The collision between vehicles on the road has been
road system increases the maintenance cost and the negative increased due to high speed, unawareness of the damaged
effects on the axle and suspension system of the vehicles. road, cracks, and potholes in the roads. The condition road on
Various factors which affect the performance of the road are a highway is shown in Fig. 1. The increasing number of
1) Heavy traffic which causes the repetition of the load 2) low- accidents due to humps and potholes forced us to design a
quality materials and inappropriate moisture condition at monitoring and information sharing system for reducing the
construction. Road transportation is an important system that accidents. The system automatically updates the information
about the pothole and humps in the cloud and shares the same potholes and hardness for avoiding accidents. The block
to the vehicle which is on the same road. This paper diagram of the proposed pothole detection system is shown in
introduces both the accelerometer and ultrasonic sensor based Fig. 2.
pothole detection with a global positioning system (GPS). The
system actively monitors the road and continuously updates A. Factors Affecting Pothole
the location of abnormality (latitude/longitude) to the cloud The major cause of potholes in the road is mainly due to
and processes the data which is received from the vehicles and heavy traffic and capacity to bear the load. Apart from the
send the alert information to the entire vehicles move towards construction of road material quality the capability to adapt the
the problematic area. If many vehicles cross the same location, changes and bear the influence made due to heavy load and
the accident alert will be sent to the entire vehicle from the different circumstances. This may be very low or high-
server. The proposed pothole detection and information temperature changes and traffic that affect the layers of the
sharing system has different advantages pavement. Some research says that mostly the potholes are
caused due to the attributes of materials like an excess of
 This system provides alert information in time to the asphalt content, sand particles, susceptible asphalt at low
driver to prevent accidents. temperature cause less density to the surface.
 it also provides information to the highway
maintenance department for easy rectification. B. Depth Measurement of Pothole
 it is a cost-effective method for to detect the humps The measurement of roughness and depth of potholes has
and potholes. been detected by NPRA on the network of Norwegian national
roads for so many numbers of years. That has been registered
and possible measures have been developed. Many records are
gathered annually. But, only a few years ago the ultrasonic
device has been introduced to measure the changes. This will
sense the object and also detect the movement of the vehicles.
This technology has not boomed in those years due to
improper updating of data and accuracy of measurements.
Since the potholes depth varies from one to another in meters.
So a new version of an ultrasonic sensor has been used to
sense the surface of the road and detects the potholes. Both the
accelerometer and ultrasonic sensor is fixed in a flat surface of
every vehicle and the controller (Arduino Uno with ESP 8366)
always monitors the signal and captures the location through
the GPS if the pothole is detected. The controller always
watches both the accelerometer and ultrasonic sensor signal
and verifies both the signals. If both signal variations are
Fig. 1. Highway with pothole. exceeding the threshold level then the controller confirms the
location of vibration and fetches the coordinates. Such that the
II. PROPOSED ROAD MONITORING SYSTEM information of a particular pavement will be updated in the
cloud. This information is intimated to another vehicle that
arrives by that way.
The Proposed system to monitor the surface of road uses
Ultrasonic sensors and accelerometer for sensing and
detecting. Sensing and Pre-processing operations are
performed on obtaining data and then given to honey bee
optimization algorithms that perform monitoring the surface of
the road. In this proposed road surface monitoring system
consists of 1) Obtaining information of road condition. 2)
Detection of potholes in the pavement and 3) Intimation about
the abnormal location through the server to the other vehicles
traveling to that area.
As we define the changes in the road reduces its
capability to serve as smooth path and safe riding. There are
many ways that cause damage and will intimate its
maintenance. Some of the important causes of crashes, cracks,
potholes, and rash driving cause lead to roughness. In India,
50% of the roads are damaged due to potholes. This will
probably trigger to maintain the pavement. The survey has
been done for Indian roads and focused mainly based on
Fig. 2. Proposed hump and pothole detection and information required for detection. The architecture of the proposed
sharing system pothole detection and information sharing system is shown in
Fig. 3. The data is updated in the cloud through the wireless
The Arduino Uno is used in IoT-RMS because it is flexible
to code its instructions, inbuilt ADC and extra features are not

Fig. 3. Architecture of the proposed hump and pothole monitoring system

conditions are detected by the controller and the
network, the system will navigate the location and gives alert corresponding locations are detected concurrently by it. The
to other vehicles. The data captured from the ultrasonic information related to the road are received, stored, and
sensors is processed and estimated using custom computer transmitted to the cloud server, which is done by the data
software by embedded C language. This makes easy detection gathering and translation layer. The gathered data are
of potholes that present in the pavement and prevents form translated into useful information by the conversions layer and
accident. Even there are different other ways to prevent an
accident, the proposed method guide the vehicle in pothole these are transferred to the next layer. The road data
area in an effective way. So that completely driving through management layer forms an IoT-based cyber network by
damaged road path is prohibited during real-time connecting the controller present in the vehicle. Using this
implementation. cyber network, information about all other region is easily
updated and monitored easily. There are two major advantages
C. IoT Platform due to this, they are, and first, updating the road related
The IoT platform is easily developed for monitoring road information is easily done. The algorithm is framed in such a
abnormalities. For easy access, one needs to understand the way that it adopts any changed situation, thus any information
technology of the service provider. There are various regarding the addition or deletion of the road is easily shared
international standard bodies such as IEEE Standards among the network. Secondly, the information regarding the
Association [10, 11], the Industrial Internet Consortium [12, road abnormalities is sent to the cloud server and maintenance
13], Internet Protocol for Smart Objects Alliance [14], is department, for reducing their work to locate the problem.
working on the proliferation of architecture and The Arduino Uno with ESP 8266 can be connected to a
communication protocol. The proposed architecture for IoT- cloud server, the RemoteXY allows us to manage the device
based road monitoring system consists of 4 layers: 1) first is from anywhere in the world. The ESP8266 module will be
the data gathering and translation layer 2) the railroad data configured as a client to connect to a WiFi access point. The
management layer 3) the decision-making layer and 4) the access point must have an Internet connection. The
application layer. Sensors are connected in the flat surface of RemoteXY library will ensure registration of the device on the
the vehicle, and the signal state is continuously observed and cloud server. The mobile application will connect to the cloud
stored in the memory based on the signal which is received server and not to the device directly. Thus, the device will be
from the accelerometer, these jobs are continuously performed available from anywhere on the Internet. There might be
by the data collection and conversion layer [15]. The unusual several reasons for the connection error. The main ones are: 1)
The N-IoT and the vehicle are not connected correctly, the
contacts are reversed, or not connected to the right contacts, or
not connected at all. 2) there is no power supply to the
ESP8266, when the power is supplied to the ESP8266, the red
LED must be lit 3) there is not enough 3.3 V power supply for
the ESP8266 4) the ESP8266 module is defective.
D. HBO Algorithm
HBO algorithm is used to optimize the received signal
from the accelerometer (DACC), an ultrasonic sensor (DUS) and
from the GPS receiver (DGPS). Generally, a HBO algorithm
consists of the queen bee, drone bees, and worker bees. The
initialization process is done for the queen and drone and the
queen has been selected from the individual which has best
fitness value. The steps involved in the HBO algorithm for
estimating the humps and potholes in the road are shown in
Fig. 4.

Fig. 4. Flow chart for HBO algorithm

III. SCATTERING SIGNAL FROIM THE SENSOR

A. Current Signal Amplitude
The coherent signal amplitude is observed from the
ultrasonic of every realized defect present in the highway.
This statistical state is placed above the roughness and
governed as:
∑ N þo
þo = k=1 co (1)
co N
Where, βcoσ represents the response of scattered signal
from each realized surface k, σ represents the statistical state
of defect of surface roughness, and N represents the total
number of realized states of roughness. To cumulate these
response made by statistical state, the effect leads to out-of-
phase which is positioned above the surface and the coherent scattering signals. It is difficult to assess the signal
incompatible effects around the surface is taken off. The quantitatively by using analytical methods. To overcome this
remaining in-phase coherent signal from (1) is reduced for its difficulty, a numerical method is used to analyze the signal
roughness in order to the smooth surface which consists of and it has the capability to deduce scattering from any
|þoco| |∑N þo |/N direction with any angle. In this approach, the mean signal
= k=1 co (2)
o=0|
|þin o|
|þin amplitude is obtained from the statistical distribution under
σ=0
normal incidence for the hump and potholes. To study the
Where β represents the response from smooth defect with performance of defects such as potholes and humps different
σ=0. The level of the roughness surface is proposed within
angular roughness has been considered (S=0.065ϒin and
two techniques such as 1) FE technique and 2) Kirchhoff
S=0.201 ϒin.). The comparison is performed between various
theory. The HBO method is used for high level scattering of
surface roughnesses with the smooth surface. An
signal whereas Kirchhoff theory is used for low level
accelerometer is placed in the flat surface of the vehicle and it
scattering of signal. There is some limitation in Kirchhoff
will produce signal variations when the vehicle runs over
theory since the shadowing of surface are not considered. The
humps and potholes. The controller checks both the signals
accurate solution of flat surface is obtained by comparing the
from the accelerometer and the ultrasonic sensor and decides
signal from input waves. The property of surface has a
the abnormality and picks the location immediately.
function that is expressed as radius of curvature γ. According
to the nature of Gaussian surface the parameter is defined by IV. SIMULATION RESULTS AND IMPLEMENTATION
distribution function. The minimum value to spread the wave
is important and defined by γmin. The measured signal which is received from the road
0.1µ2 monitoring unit is used to identify the humps and pothole
o
ynin = (3) present in the highway. The simulation is carried out using
o
where μ0 is the length of the surface correlation. Roughness is Matlab 2017 and the measured voltage from the sensing unit
inversely proportional to ϒmin, So, Kirchhoff theory is not (both the ultrasonic and accelerometer). For simulation,
valid for very high levels of roughness. The accuracy different roughness has been considered and the received
represented by this theory is quantitatively denoted by a rule signal is processed by the proposed system; there are twenty
based on [16] states of roughness has been taken from s=0.017 ϒin to s=0.35
ϒin. The attenuation of mean signal required to the realization
Min ynin cos3 8in ≫ 1 (4) is provided according to the state of roughness. When the
It is physically related to radius of surface, b min incident roughness is less, then the convergent solution is also
wave number qinc and incident angle θinc. By combining these decreased and the roughness is increased till the surface is
expression the condition for Kirchhoff theory is considered during the presence of convergent solution. This
0.1 Min µ2o coc38in
study of convergence is compared with Kirchhoff theory
≫1 (5) simulation which is presented and discussed by Zhang et al
o
[17].
B. Total Signal Amplitude
The amplitude of signal which is scattered from the rough
surface, humps and pothole will be decreased. The
combination of coherent and diffusion signal amplitude forms
total signal amplitude. It reflects multiple responses but, only
the values from each iteration are considered for calculating
the average amplitude. This will reduce the variation that
occurs in-phases of reflected signal. In back side scattering of
signal it contributes coherent field and diffusion field to obtain
total field. This extension is numerically expressed as:
þo = |þo | (6)
totaS co,k
Increase in roughness will result in reduction of total
signal amplitude Fig. 5. Comparison of the amplitude of reflecting wave for
o
|þo | |þ | Kirchoffs and HBO.
total
= co,k o=0 (7)
o | |þ |
|þin When there is an increase in roughness of surface then it
C. Back Scattering produces the value of 1. It shows that the HBO provides
accurate value than Kirchhoff theory at a high level is shown
Ultrasonic sensor is used to detect the potholes and the in Fig. 5. From Fig. 5 one can easily understand that the
information will be updated in the server. If any vehicle amplitude of the reflecting wave gets reduced in the rough
travels in the same location again, the potholes wherever surface compared to smooth one. Since statistical states
present in the road will be intimated to the corresponding observe coherent signal at multiple realizations and also it
vehicle from the serve. Mostly indirect incidence occurs in requires an exact count of the true coherent signal. Only a few
this case. When indirect incidence is being encountered the scattered signals are considered due to an increase of
scattering from the rough surfaces contains both diffuse and computational expense and benefit in obtaining accuracy.
 To represent the variations in coherent signal amplitude The measured acceleration is compared to the acceleration
three states of roughness is considered such as s=0.125μin, which is received during normal operation for both the MEMS
s=0.210μin and s=0.350μin (ie) lower, middle, higher levels are accelerometer. Fig. 7 shows the output which is received from
shown in Fig. 6. Fig. 6 illustrates that the amplitude variation the accelerometer fewer than two different speeds such as 40
is larger if the surface roughness increases. In case of the km and 60 km. The signal variation is present in 6.2 km
lower level, it requires only few realization around the order of distance for both speed scenarios and it is shown in Fig. 7(a)
hundred and it obtains little variation. In case of a very rough and (b) respectively. The comparison result shows that the
surface, the requirement is thousands of realization that leads proposed HBO method provides a better result than the
to the expectation of more RMS values. When the length of geometry measurement and filtering methods. Fig. 8 shows
crack is larger, then the surface realization is decreased to the set up of the proposed hardware implementation.
obtain a solution of convergent at the coherent field. This will
also increase the computational expense of every unit cell. So,
it is not a preferable trade-off.

Fig. 6. Variation of the coherent signal for three different Fig. 8. Hardware implementation set up.
roughness surfaces.
V. CONCLUSION
The acceleration signal is measured by the MEMS An IoT based road abnormality monitoring and the status
accelerometers placed on the bottom surface of the vehicle. sharing through the cloud is proposed in this paper. The
These measured accelerations are filtered and made to be ultrasonic sensor and the accelerometer are placed at the
effective using the HBO algorithm. bottom surface of the vehicle to find the irregularity in the
road. The simulation is carried out for two different speed
scenarios such as 40 km and 60 km using MATLAB 2017.
The location of abnormality is measured with the help of a
GPS receiver which is linked with IoT-RMS. Here the HBO
optimization algorithm is used to optimize the measured signal
from the sensors. The performance of IoT-RMS is compared
with the Kirchoffs method and the result shows that the
proposed system with HBO provides better performance under
all the speed scenarios. The comparison comes to a conclusion
that the IoT-RMS is efficient for pothole detection at all
levels. If the system detects humps or pothole, immediately it
stores the coordinates in the server and alert information will
be shared to the other vehicles which are passing in that
location and hence the good percentage of accidents can be
reduced.

References
[1] Jose M. Álvarez, Antonio M. López, Theo Gevers, Felipe Lumbreras,
“Combining Priors, Appearance, and Context for Road Detection”, IEEE
Transactions on Intelligent Transportation Systems, vol. 15(3), pp.
1168-1178, 2014.
Fig. 7. Measured signal from the accelerometer for two [2] Meiqing Wu, Siew-Kei Lam, Thambipillai Srikanthan, “Nonparametric
different speeds: (a) 40 km (b) 60 km. Technique Based High-Speed Road Surface Detection” IEEE
Transactions on Intelligent Transportation Systems, vol.16(2), pp. 874 –
884, 2015
[3] Yulu Luke Chen et al., “Inexpensive multimodal sensor fusion system International Conference on Power, Control, Signals & Instrumentation
for autonomous data acquisition of road surface conditions”, IEEE Engineering (ICPCSI), pp. 1-6, 2017.
Sensors Journal, vol. 16(21), pp. 7731 – 7743, 2016. [13] L. J. Bond, P. Ramuhalli, M. S. Taw_k, and N. J. Lybeck, ``Prognostics
[4] Harikrishnan P. M, Varun P. Gopi “Vehicle Vibration Signal Processing and life beyond 60 years for nuclear power plants,'' In Proc. IEEE
for Road Surface Monitoring” IEEE Sensors Journal, vol. 17(16), pp. Prognostics Health Manage. (PHM), Montreal, QC, Canada, pp. 1-7,
5192 - 5197, 2017. 2011.
[5] C. Chellaswamy, V. Akila, A. Dinesh Babu, N. Kalai Arasan, “Fuzzy [14] S. Torhorst, N. Hölzel, and V. Gollnick, ``Identi_cation and evaluation
logic based railway track condition monitoring system”, IEEE of the potentials of prognostics and health management in future civil
International Conference ON Emerging Trends in Computing, aircraft,'' In Proc. Eur. Conf. Prognostics Health Manage. Soc., Nantes,
Communication and Nanotechnology (ICECCN), pp. 250-255, 2013. France, pp. 1-5, 2014.
[6] Xianlei Xu, Suping Peng, Yunhai Xia, Wanjun Ji, "The development of [15] C. Chellaswamy, S. Arul, L. Balaji, "Design and analysis of an
a multi-channel GPR system for roadbed damage detection, intelligent collision avoidance system for locomotives", IET
Microelectronics Journal, vol. 45, pp. 1542-1555, 2014. International Conference on Sustainable Energy and Intelligent Systems
[7] Mats Riehm, Torbjorn Gustavsson, Jorgen Bogren, Per-Erik Jansson, (SEISCON 2011), pp. 833-838, 2011.
"Ice formation detection on road surface using infrared thermometry", [16] F. G. Bass and I. M. Fuks, Wave Scattering From Statistically Rough
Cold Regions Science and Technology, vol. 83-84, pp. 71-76, 2012. Surfaces. Oxford, UK: Pergamon Press, 1979.
[8] Moslem Ouled Sghaier, Richard Lepage Ecole de technologie [17] J. Zhang, B. W. Drinkwater, and P. D. Wilcox, “Longitudinal wave
sup´erieure 1100 Rue Notre-Dame Ouest, Montr´eal, Qu´ebec, Canada scattering from rough crack-like defects,” IEEE Trans. Ultrason.
“Change detection using multiscale segmentation and Kullback-Leibler Ferroelectr. Freq. Control, vol. 58, no. 10, pp. 2171–2180, 2011.
divergence: Application on road damage extraction” IEEE International [18] C. Chellaswamy, A. Vanathi, S. Duraichami, P. Glaretsubin, “Optimized
Workshop on Analysis of Multi-temporal Remote Sensing Images (Multi- Vehicle acceleration measurement for Rail Track Condition
Temp), pp. 1-4, 2015. Monitoring”, IEEE International Conference on Computing and
[9] C. Chellaswamy; L. Balaji; A. Vanathi; L. Saravanan, “IoT based rail Communication Technologies (ICCCT’17), pp. 155-160, 2017.
track health monitoring and information system”, IEEE International
conference on Microelectronic Devices, Circuits and Systems
(ICMDCS), pp. 1-6, 2017.
[10] 2015. IEEE-SA Internet of Things (IoT) Ecosystem Study. [Online].
Available: http://standards.ieee.org/innovate/iot/study.html
[11] Standard for an Architectural Framework for the Internet of Things
(IoT), IEEE Standard P2413, May 2016.
[12] C. Chellaswamy, V. Chinnammal, A. Dhanalakshmi, C. Malarvizhi, “An
IoT based frontal collision avoidance system for railways”, IEEE