Proceedings of the Third International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC 2019)

IEEE Xplore Part Number:CFP19OSV-ART; ISBN:978-1-7281-4365-1

Internet of Things based Accident Detection

System
S.Mohana Gowri1 P.Anitha2 D.Srivaishnavi3 M.Nithya4
Department of Computer Department of Computer Department of Computer Department of Computer
Science and Engineering Science and Engineering Science and Engineering Science and Engineering
Sri Krishna College of Sri Krishna College of Sri Krishna College of Sri Krishna College of
Engineering and Engineering and Engineering and Engineering and
Technology, Technology, Technology, Technology,
Coimbatore, India Coimbatore, India Coimbatore, India Coimbatore, India
[email protected] [email protected] [email protected] [email protected]

Abstract – Drowsiness and accident detection system is an turn out to be dangerous, particularly for such callings like
equipment gadget that expects to expand street wellbeing. This drivers.
framework includes the establishment of an equipment unit
that comprises of vibration sensor and heartbeat rate sensor to Accidents are most basic at late night and early morning.
identify the accident and eye flicker sensor in order to This is the body's characteristic sleep period. Drowsiness
distinguish the sleepiness of the driver. It utilizes GPS to follow can likewise in the day time too. More seasoned grown-ups
the area and GS M to send the message. The primary are bound to have a tired driving accident in the day time [2-
motivation behind this framework is to alert the unfortunate 5].
casualty's family about the incident and it can distinguish the
status of the vehicle driver as for sleepiness or weariness Driver tiredness results in diminished vehicle control
conditions progressively. The vibration sensor detects the
which is one of the significant reasons for road accidents.
accident and the beat rate sensor checks the heartbeat of the
driver. At the point when the vibration happens and the
Driving execution diminishes with expanding laziness
heartbeat level is high, it will be expected that accident has which results in accidents establishing 20%-23% of all
been happened and message revealing the driver's heartbeat vehicle accidents. The National Highway Traffic Safety
rate and his area will be sent to his relative. The eye flicker Administration (NHTSA) has gauges that 100000 revealed
sensor is fitted in a wearable glass. When it recognizes the eyes accidents are caused because of the drowsiness of drivers
are shut, the alarm bell starts to beep. On the off chance that every year in the U.S. alone [6-10].
the bell is turned off inside 20 sec, the condition is ordinary else
the condition will be regarded as unusual and message As per the global status report on road safety 2018, the
detailing something strange will be sent to his relative.
all out number of deaths caused because of road accident
Keywords— Accident detection, Drowsiness detection,
Embedded systems, GPS , GS M
has leveled out at 1.35 million per year. India meets with the
maximum number of accidents and accidental fatalities on
I. INTRO DUCTIO N the earth. The most extreme number of accidents are
accounted for in the vehicle division, that is, on road as well
Life of the general population is under hazard, a million as railroads. The 2030 agenda for s ustainable advancement
of people groups bite the dust each year as consequence of was propelled which goes for reducing the amount of deaths
street car accidents. This is a direct result of the absence of and injuries emerging because of vehicle collide.
emergency facilities in our nation.
With the developing populace the utilization of vehicles
Drowsy driving is the hazardous mix of driving and has turned out to be surplus and this has prompted to
sluggishness or weariness. This ordinarily occurs if a driver increment the traffic hazards and the road accidents, which
has not rested enough, however it might likewise occur causes immense death due to poor emergency facilities. The
because of untreated sleep disorders. No individual knows inaccessibility of the exact techniques for accident event
the exact minute when sleep comes over their body. recognition next to a reliable finding instrument with a rapid
Sleeping during driving is unmistakably risky, yet being reporting feature is the serious issue under examination.
sluggish influences the capacity to drive securely regardless Below fig.1 clarifies how the framework will function
of whether they don't fall asleep [1]. during crisis. It is assembled utilizing the microcontroller –
Arduino Uno. It comprises of three sensors – vibration
Weariness is such a psychophysical state of a man, which sensor, beat rate sensor and eye blink sensor. Battery is
does not take into consideration a full focus. It impacts the utilized as the vitality supplier and it is accessed through
human response time, in light of the fact that the worn out switch. Power supply is utilized to give a consistent 5V
individual responds much slower, contrasted with the rested power supply. The qualities from vibration sensor and
one. Appearance of the main indications of a weakness can heartbeat rate sensor are sent to the microcontroller and

978-1-7281-4365-1/19/$31.00 ©2019 IEEE 159

Authorized licensed use limited to: Auckland University of Technology. Downloaded on June 02,2020 at 07:52:19 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the Third International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC 2019)
IEEE Xplore Part Number:CFP19OSV-ART; ISBN:978-1-7281-4365-1

dependent on the qualities area removed from GPS is sent to travelers and can without much of a stretch be coordinated
GSM. The message is then send utilizing GSM. The yield with the current wellbeing frameworks in the vehicle. This
from eye blink sensor is sent to microcontroller and will likewise expand the general road safety, by revealing
dependent on its value bell will be worked. A switch will be any sort of rash heading to the nearest control room.
utilized to turn off the bell.
The front and back recognition camera embedded with
an android application is utilized. According to the reports
from the back camera they will report the android
application. Be that as it may, the inquiry is the warning
through Android Application won't observed to be a
proficient one. Android application is reliant on systems
administration, charging and numerous things as well.
In deep learning, methodology was proposed to
recognize recommendations so as to avoid the road
accidents. The need of deep learning here is that it manages
enormous measure of information. Deep learning prompts to
lot of complications when managing these sorts of
information.

According to the "Driver Drowsiness Recognition" they

are utilizing an eye detection algorithm. It cannot be
characterized of how much the calculation will be
exceptionally precise. And furthermore, it doesn't clarify
about how it accomplishes the accuracy. The adequacy of
this framework should be examined.
Fig. 1 Drowsiness and Accident detection system
III. PROPOSED SYSTEM

II. EXISTING S YSTEM The proposed framework is planned with the arduino
In existing framework, a pair of glasses is utilized as the uno microcontroller. In proposed framework the hardware
detection of drowsing system. It is fitted with a light weight device will be spot close to the dashboard of the vehicle.
eye blink sensor. The eye blink sensor is IR based. The light This device will comprise of an eye flicker sensor, vibration
waves from the sensor won't influence the retina. They are sensor, beat rate sensor, GPS, GSM, buzzer, battery, switch,
intended to recognize the unaware condition of driver [11- power supply and switch for bell
12].
The eye flicker sensor distinguishes the closure of eyes.
The yield from the eye blink sensor will be utilized to a It is implemented utilizing the infra-red light which will go
warning application. The warning framework is executed through the eyes. The IR light causes no damage to the eyes.
utilizing JAVA programming. The light goes through the white light that is the point at
which the eyes are open. It will be reflected back by the
dark light that is the point at which the eyes are shut. This
It won't caution the driver when he feels sleepy. It won't section and impression of the IR light is demonstrated by
intimate the family/companions of the driver on the off the on/off of the red LED in the eye flicker sensor. It will be
chance that he felt unconscious or met with an accident utilized to recognize unconsciousness of driver. At the point
The framework will be mounted on a vehicle, before the when the output from eye flicker sensor is low, signal will
driver's seat. It will persistently have an attention on the be enacted. On the off chance that the signal is not set to
individual driving the vehicle subsequently monitors the low physically, message will be sent.
individual's activities and give voice messages to the driver,
training him about the safety measures estimates each The vibration sensor SW-420 is utilized to identify the
opportunity he comes inside the vehicle. In the event that he vibration of the vehicle. The beat rate is then determined
is observed to be unenergetic, he will be alarmed with a utilizing the beat rate sensor. It is accepted that the vibration
voice message revealing to him that he is tired and he sensor gives a yield of high when there is an accident. In
should stop the vehicle. On the off chance that he is this way, to affirm the occurrence of accident the beat rate is
observed to be drunk while driving, the GPS directions of checked. On the off chance that the beat rate is likewise
the vehicle alongside the details of the vehicle, are sent to higher than the typical worth it is affirmed that accident has
the closest control room, with the goal that suitable move been happened. Message will be sent as accident identified.
can be made by the police. In this way, this framework has
the ability of ensuring the driver's security alongside the co-

978-1-7281-4365-1/19/$31.00 ©2019 IEEE 160

Authorized licensed use limited to: Auckland University of Technology. Downloaded on June 02,2020 at 07:52:19 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the Third International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC 2019)
IEEE Xplore Part Number:CFP19OSV-ART; ISBN:978-1-7281-4365-1

Message is sent using arduino compatible GSM. GSM is

connected with the microcontroller. The values from the IV. MODULES
sensors determine whether to send message. GPS is used to
The various modules in the drowsiness and accident
identify the latitude and longitude of the location an d the
location will be sent as google link. detection system are
• Accident and abnormal pulse detection
• Sleep detection
• Location tracking
• Message sending

V. MODULE DESCRIPTION

A. ACCIDENT AND ABNORMAL PULSE DETECTION

In the project, vibration sensor will be associated with
the arduino board. At the point when there is no vibration
distinguished, vibration sensor yield will be 0 (low voltage),
generally its yield will be 1(high voltage).

At the point when the vibration is high, it may be

expected as accident has been happened. To affirm the
circumstance, beat rate of driver must be checked. Heartbeat
rate sensor is utilized to check the beat.

Heartbeat sensor is a plug and play pulse sensor for arduino

and arduino compatibles. It tends to be utilized by under
studies who need to incorporate live heart/beat rate
information into their ventures. Heartbeat s ensor adds
intensification and noise dropping to the equipment. It is
quicker and simpler to get solid heartbeat readings.
Heartbeat sensor works with either a 3V or 5V power
supply.

The beat sensor is interfaced with arduino and the arduino is

associated with the LCD show which shows the beat
qualities. At the point when the beat worth is additionally
higher than ordinary worth, it will be considered as accident
has been happened.

At the point when the yield from vibration sensor is low, the
vehicle is under ordinary condition. However, during this
Fig. 2. Flow chart - Working of drowsiness condition, if the beat worth is lower than ordinary worth, it
and accident detection system is expected that heart is irregular.

Fig. 2 depicts the working of the framework. The T ABLE I ACCIDENT DET ECT ION VALUES
framework begins by powering ON the battery. Arduino
begins. Vibration is checked. On the off chance that it is high, Vibration Pulse value (x) Result
beat is checked. In the event that heartbeat is likewise high, High x>6 && x<60 Hear Abnormal
area is followed utilizing GPS and sent as message utilizing
High || Low 60<x<200 Heart Normal
GSM.
Low x>200 Accident
On the off chance that the vibration is low, beat is checked.
In the event that heartbeat is additionally low, area is tracked In the two conditions, the LCD show shows the conditions
and sent as message. Meanwhile, that the yield from eye flicker in like manner.
sensor is high, the condition is ordinary. If it is low, it implies
the eyes are closed. At that point the signal begins. On the off B. SLEEP DETECTION
chance that it isn't turned off physically and remains at yield The eye-blink sensor is utilized to distinguish the closure of
high, area is followed and sent as message.
eyes. It works by passing the eye and eyelid territory with

978-1-7281-4365-1/19/$31.00 ©2019 IEEE 161

Authorized licensed use limited to: Auckland University of Technology. Downloaded on June 02,2020 at 07:52:19 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the Third International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC 2019)
IEEE Xplore Part Number:CFP19OSV-ART; ISBN:978-1-7281-4365-1

infrared light, at that point examining the adjustments in the the GSM module get AT order, it react with OK. It implies
reflected light utilizing a phototransistor and differentiator GSM module is functioning effectively.
circuit. It works with 5V power supply.

VI. RESULT
Interface managed DC power supply of 5 volts to the
eye blink sensor. Dark wire is ground, next center wire is After functioning of the project, it’s time to have a
yield and red wire is certain supply. To check whether the glance at hardware.
sensor working, associate just two wires that is the 5V and
ground wire. Leave the yield wire detached. At the point Fig. 3. is the image of the project circuitry. When the
when eye is shut, LED will be off and the yield is at 0V. switch of battery is on, the entire circuit gets power and
Put eye flicker sensor glass on face inside 15mm separation starts functioning.
from eyes, and view the LED squinting on each Eye squint.
At the point when the eye is open, the yield will be high. When the condition is normal there is no vibration
The yield is offered legitimately to microcontroller for sensed in the vibration sensor and the pulse value displayed
interfacing applications. in LCD is normal.

A bell is associated with the arduino. At the point when the When the condition is normal there is no vibration
yield from eye flicker sensor is low, the signal will be set to sensed in the vibration sensor and the pulse value displayed
high. On the off chance that the driver feels sleepy, he will in LCD is normal. At the point when the yield of vibration
be cautioned by the ringer sound and he can off the signal. sensor is low, it implies that the vehicle is not having any
In the event that he didn't off the bell, it would be issue.
considered as the driver felt unconscious and something is
irregular.

C. LOCATION TRACKING
GPS represents Global Positioning System. It is utilized
to follow the latitude and longitude of any area on the earth.
It identifies the accurate UTC time (Universal Time
Coordinated). GPS module is utilized to follow the area of
accident in this paper. The device gets the directions from
satellite for consistently, with time and date. To discover
just the scope and longitude arranges, utilize the $GPGGA
string. Fig. 3 Working model of Drowsiness and Accident
Detection S ystem
GPS module sends the information identified with
following position, and it sends numerous information in At the point when the yield of vibration sensor is observed
NMEA group. NMEA arrangement comprises numerous to be high, it implies that the vehicle has been met with an
sentences, where just one sentence is required. This accident. To affirm this, beat of driver is likewise found and
sentence begins with $GPGGA and contains the directions, the worth is more noteworthy than typical. Thus the
time and other helpful data. occurence of accident is affirmed and showed in the LCD.

This GPGGA refers to Global Positioning System Fix At the point when the yield of vibration sensor is low, it
Data. Concentrate facilitate from $GPGGA string by
implies that the vehicle is not having any issue. Yet, to
including the commas in the string. In the event that the
$GPGGA string is discovered, store it in an array, at that know the state of driver his heartbeat is checked
point latitude is to be found after two commas and occasionally. In the event that his heartbeat is not exactly
Longitude is to be found after four commas. ordinary, message "Heart Abnormal" will be shown in the
LCD.
D. MESSAGE SENDING
The SIM900 is a finished quad-band GSM/GPRS At the point when the LED is ON in eye squint sensor, it
module which can be implanted effectively utilized by implies that the eye is open. Whe the LED is OFF in eye
under studies in task to send voice, SMS and information at blink sensor, it implies that the eye is closed. Along these
low power utilization. AT refers to ATTENTION. It is lines, a message "Rest distinguished" will be shown in the
utilized to control GSM module. There are a few directions LCD. A bell sounds at this condition. In the event that it
for calling and informing utilizing GSM with Arduino. For isn't set to OFF physically, it is expected that the condition
testing GSM Module, use AT direction. At the point when is something strange.

978-1-7281-4365-1/19/$31.00 ©2019 IEEE 162

Authorized licensed use limited to: Auckland University of Technology. Downloaded on June 02,2020 at 07:52:19 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the Third International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC 2019)
IEEE Xplore Part Number:CFP19OSV-ART; ISBN:978-1-7281-4365-1

2016 IEEE 5th Global Conference on Consumer Electronics

(GCCE’16), 2016, pp. 401-402
Messages will be received during the three conditions, for
[6]. J.Fenila Naomi, Robini A Hybrid Autoscaling for Cloud applications
example, sleep abnormal, heart rate abnormal and accident using an Efficient Predictive technique in Private Cloud Indian
identified by a mobile number given in the task. To know Journal of Science and Technology February 2019 Vol 12(8) pp1-7
the state of driver in every one of the three cases his [7]. Anitha P, “ Enhancing database security through encryption standards”,
heartbeat rate is additionally sent. Area of the vehicle is Journal of Advanced Research in Dynamical and Control Systems,
vol 10, 12-Special Issue, pp. 1029-1032
followed and sent as google guide connect. [8]. Srivaishnavi.D, “ E-Mushroom Farmer Society using Andriod
Application”, Journal of Advanced Research in Dynamical and
VII. CONCLUSIONS AND FUTURE WORK Control Systems, vol 10, 12-Special Issue, pp. 1105-1109
[9]. G. Victoreia, D. Yazhini, G. Parameswari, E. Gurumoorthi, and
A. CONCLUSIONS Vijayabarathy, “ Driver fatigue monitoring system using eye
closure,” International Journal of Modern Engineering Research,
This project is completely going to avoid the accidents
vol. 4, no. 11, pp. 26-31, November 2014.
because of tiredness of the driver and to intimatethe relative
[10]. K. Chang, B.-H. Oh, and K.-S. Hong, “ An implementation of
of driver if there should arise an occurrence of accident or smartphone-based driver assistance system using front and rear
irregular conditions. The objective of this project has the camera,” in Proceedings of the 2014 IEEE International Conference
upsides of minimal effort, portability and little size. on Consumer Electronics (ICCE’14), 2014, pp. 280-281.
Fundamental objective of the accident recognition [11]. Hemadri, V. B., Padmavati Gundgurti, G. Dharani Chowdary, and
Korla Deepika. "A Novel on Biometric Parameter’s Fusion on
framework project is to diminish the chances of losses in
Drowsiness Detection Using Machine Learning." in International
such accidents and a framework to caution the driver if there Conference on Computer Networks and Communication
should be an occurrence of tiredness and identify accident Technologies, pp. 1-8. Springer, Singapore, 2019.
consequently has been exhibited. This framework can be [12]. Zhang, Chao, Xiaopei Wu, Xi Zheng, and Shui Yu. "Driver
actualized adequately for a whole city or nation with huge drowsiness detection using multi-channel second order blind
identifications." IEEE Access 7 (2019): 11829-11843.
populace like India.

B. FUTURE WORKS
In future we can include greater administration applications.
It is a wonderful thing to drivers and the individuals who
travel a great deal. It is a more affordable one with the goal
that every one of the general population can utilize it and
observed in day-today life. In future we can get to the extra
data and correspondingly we can expand speed as per the
need. SMS will be made to send to three individuals. In
addition to SMS, call will be made to one among the three
individuals. This framework can have wide application
prospects and it will assume significant job in everyday life
in future.

REFERENCES
[1]. B. Reddy, Y.-H. Kim, S. Yun, C. Seo, and J. Jang, “ Real-time driver
drowsiness detection for embedded system using model compression
of deep neural networks,” in Proceedings of the 2017 IEEE
Conference on Computer Vision
[2]. J Mohan Kumar, Rishabh Mahajan, Deeksha Prabhu, Debasmita
Ghose, “ Cost effective road accident prevention system”, in the
proceedings of IEEE”, Contemporary Computing and Informatics
(IC3I), 14-17 Dec. 2016, Noida, India.
[3]. Kavitha, Dr.M.Sujaritha,” A survey of Exploratory chemometrics
analysis of spectroscopic data”, Journal of Advanced Research in
Dynamical and Control Systems, 12-Special Issue, pp 1058 -1064
[4]. M. Nithya ,N. Pooranam Secure e-payment system using cryptography
and stegnography with two-fish algorithm Journal of Advanced
Research in Dynamical and Control Systems August 2018 12-
Special Issue 1072-1075
[5]. L.-B. Chen, W.-J. Chang, J.-P. Su, J.-Y. Ciou, Y.-J. Ciou, C.-C. Kuo,
and K. S.-M. Li, “ A wearable-glasses-based drowsiness-fatigue-
detection system for improving road safety,” in Proceedings of the

978-1-7281-4365-1/19/$31.00 ©2019 IEEE 163

Authorized licensed use limited to: Auckland University of Technology. Downloaded on June 02,2020 at 07:52:19 UTC from IEEE Xplore. Restrictions apply.