Content uploaded by Anup Lal Yadav
Author content
All content in this area was uploaded by Anup Lal Yadav on Mar 22, 2024
Content may be subject to copyright.
2023 International Conference on Advances in Computation, Communication and Information Technology (ICAICCIT)
979-8-3503-4438-7/23/$31.00 ©2023 IEEE
An Intelligent System for Vehicle Collision Avoidance
System Using Internet of Things
Anup Lal Yadav
M. M Deemed to Be University,
Mullana133207, Ambala, Haryana, India.
anupsaran@gmail.com
Sandip Kumar Goyal
M. M Deemed to Be University,
Mullana133207, Ambala, Haryana, India.
skgmmec@gmail.com
Abstract—The increasing number of vehicles on the roads has
resulted in a higher risk of accidents and collisions. To address
this critical issue, this research proposes an intelligent system
for vehicle collision avoidance using the Internet of Things (IoT).
The system aims to enhance road safety by providing real-time
monitoring, analysis, and proactive collision avoidance
mechanisms. The system operates in three main stages:
perception, decision-making, and action. In the perception stage,
sensor data from vehicles and infrastructure are collected and
processed to gather relevant information about the surrounding
environment. This includes detecting the presence, position, and
velocity of nearby vehicles, pedestrians, and obstacles. In the
decision- making stage, the collected data is analyzed and
evaluated using machine learning techniques. The system
generates real-time predictions and risk assessments based on the
analyzed data to determine potential collision scenarios and
prioritize the most critical situations. This system also has an
over-speed detection feature that monitors speed and notifies the
driver when the car exceeds a certain speed restriction.
Keywords— Vehicle Collision Detection, IoT, Machine Learning,
Decision Making.
I.
INTRODUCTION
The increasing number of vehicles on the roads has be-
come a significant concern in many parts of the world.
This phenomenon, often referred to as traffic congestion, has
several negative impacts on society, the environment, and the
economy. As the number of vehicles on the roads increases,
traffic congestion becomes a common issue in urban areas.
Congestion can lead to longer travel times, increased stress
for commuters, and reduced overall efficiency of transportation
systems. The rise in vehicle numbers contributes to increased
air pollution, primarily through emissions of pollutants such
as carbon monoxide, nitrogen oxides, particulate matter, and
volatile organic compounds. Poor air quality can have
serious health implications for residents, particularly those
with respiratory conditions.
The growing number of vehicles also contributes to the
depletion of natural resources, increased energy consumption,
and the release of greenhouse gases that contribute to climate
change. The influx of vehicles can put strain on existing
transportation infrastructure, leading to increased maintenance
costs, the need for road expansions, and potential delays
in infrastructure development. Higher vehicle numbers can
lead to increased road accidents and fatalities, particularly
in areas with inadequate road design or traffic management.
Traffic congestion can have economic implications, including
increased fuel consumption, reduced productivity due to time
wasted in traffic, and potential negative effects on businesses
that rely on efficient transportation networks.
The growing number of vehicles emphasizes the importance
of effective urban planning and transportation management.
Cities may need to invest in alternative modes of
transportation, such as public transit, cycling lanes, and
pedestrian- friendly infrastructure. Advancements in
technology, such as intelligent transportation systems,
autonomous vehicles, and real-time traffic monitoring, offer
potential solutions to mitigate the impact of increasing vehicle
numbers on the roads. Encouraging behaviors such as
carpooling, ridesharing, and the use of public transportation can
help reduce the overall number of vehicles on the road and
alleviate congestion. Government policies and regulations,
such as congestion pricing, emission standards, and incentives
for electric vehicles, can play a crucial role in managing the
increasing number of vehicles and addressing related
challenges.
Vehicles typically drive quickly on the same highway.
though, if the driver is unable to react during the car’s abrupt
stop due to a driver’s attention, lengthy driving exhaustion or
a sudden decrease in speed of the vehicle in front, there may
be a considerable chance of a crash. To distinguish the cars
coming from the side or the rear, however, drivers must also
pay attention to their reflections in the mirrors. The driver
of an automobile that is abruptly approaching cannot react
right away, even while carefully checking, without honking
the horn. To prevent any conflict, it is necessary to establish
collision avoidance systems. The radar sensor prevents end-to-
end collisions. Smart techniques are used to interpret sparse
distance data in order to provide the driver with the
appropriate warning signals, as opposed to employing
sophisticated
712
2023 International Conference on Advances in Computation, Communication and Information Technology (ICAICCIT) | 979-8-3503-4438-7/23/$31.00 ©2023 IEEE | DOI: 10.1109/ICAICCIT60255.2023.10465838
Authorized licensed use limited to: Indian Institute of Technology (Ropar). Downloaded on March 21,2024 at 07:11:19 UTC from IEEE Xplore. Restrictions apply.
readings of distance data to evaluate the hazard of car
accidents. The maintenance of systems like the acceleration
skid control (ASC) system, anti-lock braking system (ABS),
and powertrain management system is made more challenging
by the presence of multiple electrical control devices in
modern automobiles [1]. Therefore, upgrading conventional
harnesses to smart and automotive networks is extremely
concerning.
A collision avoidance system is a technological solution
intended to prevent or mitigate collisions between vehicles
or between a vehicle and pedestrians. Vehicle collisions and
accidents continue to pose significant risks to both drivers and
pedestrians. To address this issue, researchers have turned to
advanced technologies such as the IoT to develop intelligent
systems for vehicle collision avoidance [2]. The intelligent
system for vehicle collision avoidance also initiates
proactive measures to prevent collisions. This includes
providing alerts and warnings to drivers, activating automatic
emergency braking systems, adjusting vehicle speed, and
coordinating with traffic management systems to optimize
traffic flow and prevent congestion.
Fig. 1 Stages of collision
Figure 1 shows the distinct stages of vehicle collision
with different conditions. For instance, the range in which
the number of fatal and injury-causing accidents increases
is between the speeds of two cars at the time of a rear-
end collision, which is between 40 and 60 km/h. There is
a considerable chance of accident when the automobile being
driven is moving at 80 km/h and the car in front is moving
at 20 km/h. There will inevitably be a collision if the driver’s
automobile is moving at 50 km/h and the car in front of it is
moving at 20 km/h. If the speed of both vehicles can change
again, no one can prevent collision. Here are several significant
advantages of collision avoidance systems:
A.
Increased safety
Safety is the principal benefit of collision avoidance
systems. These systems employ a number of sensors, like as
cameras, radar, and lidar, to identify potential collision
hazards, alert the driver, or even act autonomously to avoid an
accident. By providing timely warnings and interventions,
collision avoidance systems reduce the likelihood of collisions
and safeguard both vehicle occupants and pedestrians.
B.
Reducing Accident
Collision avoidance systems actively monitor the vehicle’s
surroundings and can detect potential hazards that the driver
may miss, such as a vehicle in the blind spot or a pedestrian
crossing the street. By issuing warnings or applying the brakes
automatically, these systems can help prevent accidents by
adding an additional layer of awareness and response.
C.
Mitigation of collision severity
Even when a collision cannot be completely avoided,
collision avoidance systems can still play a crucial role in
mitigating the impact’s severity. They can instigate actions
such as pre-tensioning seat belts, deploying airbags, and
applying maximum force to the brakes in order to reduce the
collision’s impact and minimize occupant injuries.
D.
Assistance in complex driving situations
Advanced collision avoidance systems frequently come with
tools like automated emergency braking, adaptive cruise
control, and lane departure warning. These features aid the
driver in complex driving situations, such as heavy traffic,
highway driving, and maneuvering in confined areas. Crash
avoidance systems make travelling safer and less stressful by
providing automated assistance.
E.
Improved pedestrian safety
Pedestrian detection and collision avoidance systems are de-
signed to specifically identify and safeguard pedestrians. These
systems employ sophisticated sensors to detect pedestrians in
the vehicle’s path and issue warnings or apply the brakes as
needed. Collision avoidance systems contribute to reducing
accidents and protecting vulnerable road users by emphasizing
pedestrian safety.
F.
Enhanced driving experience
Collision avoidance systems can provide additional features
that improve the driving experience. Adaptive cruise control,
for instance, can maintain a safe distance from the vehicle
in front of it, thereby reducing the need for constant pace
adjustments. Lane-keeping assistance can assist drivers in
staying in their lane, thereby reducing driver fatigue on lengthy
trips. Not only do these features increase safety, but they also
make driving more convenient and enjoyable.
G.
Insurance benefits
Some insurance companies offer discounts or incentives
for collision avoidance-equipped vehicles. It has been
demonstrated that these systems reduce the frequency and
severity of accidents. The installation of a collision avoidance
system in a vehicle may result in reduced insurance premiums
and cost savings.
The integration of IoT and machine learning enables the
intelligent system to operate in real-time. As data is
continuously collected and analyzed, the system can provide
timely alerts and warnings to drivers to prevent potential
collisions. Additionally, the system can initiate automatic
emergency
713
Authorized licensed use limited to: Indian Institute of Technology (Ropar). Downloaded on March 21,2024 at 07:11:19 UTC from IEEE Xplore. Restrictions apply.
braking systems, adjust vehicle speed, and coordinate with
traffic management systems to optimize traffic flow and pre-
vent congestion. The major contribution in this paper are as
follows:
•
To propose an intelligent system for the detection and
monitoring of collision of vehicles.
•
The proposed system is able to collect the data from
vehicles and analyze the data for decision-making.
•
The system includes detecting the presence, position, and
velocity of nearby vehicles, pedestrians, and obstacles.
•
The system generates real-time predictions and risk
assessments based on the analyzed data to determine potential
collision scenarios and prioritize the most critical situations.
The remainder of the essay is structured as follows. Section
II presents the associated work. Section III provides a brief
overview of the planned vehicle over-speed detection and
accident-avoidance system. Conclusions and suggestions for
further work are compiled in Section IV.
II.
LITERATURE REVIVEW
There are several research has been conducted for the
detection and identification of collisions of road accidents.
IoT and machine learning approaches are combined in the
smart vehicle collision prevention system that Pathik et al.
[3] presented. The technology uses roadside infrastructure and
sensors built into cars to gather data about the surroundings in
real-time. In order to analyse the data and forecast potential
collision scenarios, machine learning algorithms are used. The
system provides timely alerts and warnings to drivers, and in
critical situations, it automatically activates emergency braking
systems to prevent collisions.
Agarwal et al. [2] highlight different technologies like
LORAWAN, RFID etc. for vehicular communication. They
show comparative study between different technologies. They
highlight how such technologies help in vehicle to everything
communication. Dhaya et al. [4] proposed about collision
detection on transmission poles. To track actual collision
occurrences on transmission poles, a wireless collision
detection system has been created. The model employs a
vibration sensor to translate vibrations into digital data, which
is then read by a microcontroller unit and connected to various
peripheral devices. Alvi et al. [5] reviewed about various
technologies which are used in vehicular ad hoc network.
They highlight their strength, challenges, and mythologies for
predicting road accident. Onesimu et al. [6] proposed an
intelligent collision avoidance system that integrates IoT and
machine learning. The system employs a combination of
sensors, such as radar and lidar, to monitor the vehicle’s
surroundings. The collected data is processed using machine
learning algorithms to detect and predict potential collision
risks. The system provides real- time alerts to drivers and takes
preventive actions, such as adjusting vehicle speed or
steering, to avoid collisions.
Agarwal et al. [7] design proximal policy optimization-
based collision avoidance system for route planning. It
minimizes path loss problem and works well for better
throughput. Mohamed et al. [8] propose lane change detection
system. This
system activates automatically when the vehicle is in danger
condition. Uma et al. [9] propose radio resource allocation
scheme for effective utilization of radio resources in vehicular
communication network. The proposed system which help
emergency vehicle to reach destination on time. Their system
is also able to detect accident by sensors. Rongcai et al.
[10]
survey about road accident, collision etc. They use deep
learning model for preventing road accident. The proposed
model can help us lower the number of fatalities caused by the
absence of emergency personnel at the accident scene. Goyal
et al. [11] present deep learning based IOT model that can
detect and predict the pre-accident or pre-collision condition
and send an alarm message informing users that a collision
is likely to happen. Muslim et al. [12] provide an agent-
based collision avoidance system which is used to improve
the performance of vehicle in term of speed, direction etc.
Through this system collision can be avoided between vehicles
up to 25 percent. Shubham et al propose safety framework
for reducing road accident using time to collision measure.
They use deep learning model for increasing road safety.
Nanda et al. [13] presents an intelligent collision avoidance
system that combines IoT and deep learning techniques. The
system collects data from sensors and cameras installed in
vehicles and uses deep learning algorithms to analyze the data.
The system can accurately detect and classify objects in the
environment, such as other vehicles and pedestrians, to assess
collision risks. Real-time warnings and alerts are provided
to drivers, and emergency braking systems are activated if
necessary.
Yuan et al. [14] presented a comprehensive review article
discusses various IoT-based intelligent collision avoidance
systems for vehicles. It provides an overview of the different
sensors, communication technologies, and machine learning
algorithms employed in these systems. The review highlights
the advancements and challenges in implementing such
systems and discusses the potential benefits in terms of road
safety and accident prevention.
III.
PROPOSED SYSTEM
The system consists of a radar sensor and Arduino (Fig.2).
The sensor in the car recognizes the object or car in front of it
and sends the information to Arduino. When a particular car is
close to the one in front, the suggested method automatically
reduces the speed of the vehicle. This system also has an
over-speed detection feature that monitors speed and notifies
the driver if the car travels faster than a certain limit [15].
Following resources are used for collision avoidance system
•
Sensor – It is used to find the object in front of the vehicle.
•
Arduino – Arduino reduce the speed of vehicle for
collision avoidance.
•
Vehicle – Vehicles are equipped with sensors, Arduino,
and other resources.
The proposed Intelligent System for Vehicle Collision
Avoidance using IoT holds great promise in improving road
safety, traffic management, and overall transportation
efficiency.
714
Authorized licensed use limited to: Indian Institute of Technology (Ropar). Downloaded on March 21,2024 at 07:11:19 UTC from IEEE Xplore. Restrictions apply.
Fig. 2 Collision avoidance system
From Fig 3, we can see that sensor identifies the number of
vehicles which are in front of car such as 4 cars, 1 unknown
person, pedestrian etc. It sends information to the Arduino and
Arduino prevent accident by reducing the speed of vehicle.
Fig. 3 Real time implementation of Collision avoidance system
From Fig.4, we can see the bird eye plot graph which
shows objects in front of vehicle. It shows the obstacle under
150 meters, another vehicle under 168, 172 and 177 meters.
Red circle shows the vehicles which contain sensors. Yellow
square box shows the lane boundaries. When Arduino gets
the information about all the objects which are in front of the
vehicle. Arduino regulates the vehicle speed when a specific
vehicle is close to the vehicle in front.
TABLE I
FOLLOWING PARAMETERS ARE TAKEN FOR COLLISION AVOIDANCE SYSTEM
Fig. 4 Bird Eye Plot Graph
IV.
EXPERIMENTATION RESULTS
We have selected the SDOT (Seattle Department of
Transportation) dataset refers to a collection of data and
information related to transportation and infrastructure in the
city of Seattle, Washington. The dataset is maintained by the
Seattle Department of Transportation and includes a wide
range of data types, such as traffic volumes, collision records,
road- way characteristics, parking information, public
transportation schedules, and more. The SDOT dataset
provides valuable insights into transportation patterns,
infrastructure conditions, and traffic-related issues in Seattle.
It is used by researchers, urban planners, policymakers, and
transportation professionals to analyze and understand various
aspects of the city’s transportation system. Some common
types of data available in the SDOT dataset include: Traffic
Volume Data: This data provides information about the
volume of traffic on different roads in Seattle. It includes data
on average daily traffic counts, peak-hour traffic volumes, and
traffic patterns at various times of the day. Collision Records:
The dataset includes records of traffic collisions that have
occurred in Seattle. These records provide details about the
location, date, time, and contributing factors of each collision,
allowing for analysis of safety issues and identification of
high-risk areas.
Roadway Characteristics: Information about roadway
characteristics, such as road geometry, speed limits, pavement
conditions, and signage, is included in the dataset. This data
helps in assessing the condition of the road network and
identifying areas that require improvements. Parking Data:
The SDOT
715
Authorized licensed use limited to: Indian Institute of Technology (Ropar). Downloaded on March 21,2024 at 07:11:19 UTC from IEEE Xplore. Restrictions apply.
dataset also includes data related to parking, such as parking
zones, regulations, and occupancy rates. This information
assists in managing parking resources and studying parking
behavior in different areas of the city. Public Transportation
Schedules: Data on public transportation schedules, routes, and
stops is available in the dataset. This information is crucial
for understanding the availability and accessibility of public
transportation options in Seattle.
TABLE II
EXPERIMENTAL RESULT ANALYSIS
Table 2 shows experimental result analysis in which
collision avoidance system shows upto 7 objects which are in
front of the vehicle. Object detection time is upto 25 Seconds.
Dimensions shows the area that means 1000* 1000 is the
length and width of the route. Through collision avoidance
system, collision can be avoided upto 90 percent. Sensor
detection rate shows how much vehicles are equipped with
sensors identified by collision avoidance system. Its success
rate is upto 75 percent The correlation of attributes can be
represented in the Figure 5.
Fig. 5 Correlation of Distinct Attributes
After finding out the correlation among all the attributes,
we are evaluating the collision count during a month in the
whole year shown in Figure 6.
Fig. 6 Collision Count per Month
Fig. 7 Collision on turns left and right
Fig. 8 Collision during distinct time intervals
716
Authorized licensed use limited to: Indian Institute of Technology (Ropar). Downloaded on March 21,2024 at 07:11:19 UTC from IEEE Xplore. Restrictions apply.
Analyzing collision counts by daytime provides valuable
insights into when accidents are more likely to happen. This
information can be used for traffic management, law
enforcement, and public awareness campaigns to reduce
collisions and enhance road safety during specific time
periods.
Fig. 9 Collision count by day time
Fig. 10 Collision due to influenced driving conditions
Figure 10 shows the damages of property, injuries, collision,
and serious illness due to vehicle collisions. Figure 11 shows
the collision count by day time which depicts the around 11.05
percent maximum collision count while noon time depicts the
minimum collision time. .By harnessing the power of IoT, this
system aims to reduce accidents and save lives, making our
roads safer for everyone. Continued research and development
are essential to realize the full potential of this innovative
technology.
V.
CONCLUSION
The integration of Internet of Things (IoT) technology in
vehicle collision avoidance systems holds immense promise
for enhancing road safety and preventing accidents. By lever-
aging real-time data collection, advanced analytics, and
proactive measures, these intelligent systems have the
potential to revolutionize the way we approach collision
avoidance. With IoT, vehicles can be equipped with sensors
that collect information about their surroundings, including
the presence and movement of other vehicles, pedestrians, and
obstacles. This data is then processed and analyzed using
machine
learning algorithms, enabling the system to detect potential
collision risks and make informed decisions. One of the key
advantages of IoT-based collision avoidance systems is their
ability to operate in real-time. By continuously monitoring the
environment and analyzing incoming data, these systems can
provide timely alerts and warnings to drivers, enabling them
to take appropriate action to avoid collisions. Additionally, in
critical situations, the system can initiate automatic emergency
braking systems or adjust vehicle speed to prevent accidents.
REFERENCES
[1]
M. Elkady, A. Elmarakbi and J. MacIntyre, ”Integration of vehicle
dynamics control systems with an extendable bumper for collision
mitigation”, International Journal of Automotive and Mechanical En-
gineering, vol. 12, pp. 2893-2913, 2015.
[2]
Agarwal, V., Sharma, S., Agarwal, P. (2021). IoT based smart transport
management and vehicle-to-vehicle communication system. In Computer
Networks, Big Data and IoT: Proceedings of ICCBI 2020 (pp. 709-716).
Springer Singapore.
[3]
N. Pathik, R. K. Gupta, Y. Sahu, A. Sharma, M. Masud, and M. Baz,
“AI Enabled Accident Detection and Alert System Using IoT and Deep
Learning for Smart Cities,” Sustainability, vol. 14, no. 13. MDPI AG,
p. 7701, Jun. 24, 2022. doi: 10.3390/su14137701.
[4]
Dhaya, R., Kanthavel, R. (2020). A wireless collision detection on trans-
mission poles through IoT technology. Journal of trends in Computer
Science and Smart technology (TCSST), 2(03), 165-172.
[5]
Alvi, U., Khattak, M. A. K., Shabir, B., Malik, A. W., Muhammad, R.
(2020). A comprehensive study on IoT based accident detection systems
for smart vehicles. IEEE Access, 8, 122480-122497.
[6]
J. A. Onesimu, A. Kadam, K. M. Sagayam, and A. A. Elngar, “Internet
of things based intelligent accident avoidance system for adverse weather
and road conditions,” Journal of Reliable Intelligent Environments, vol.
no. 4. Springer Science and Business Media LLC, pp. 299–313, Jan. 25,
2021. doi: 10.1007/s40860-021-00132-7.
[7]
Agarwal, V., Sharma, S. (2021, June). IoT based smart transport man-
agement system. In Advanced Informatics for Computing Research: 4th
International Conference, ICAICR 2020, Gurugram, India, December
26–27, 2020, Revised Selected Papers, Part II (pp. 207-216). Singapore:
Springer Singapore.
[8]
A. Mohamed et al., “RoadMonitor: An Intelligent Road Surface Con-
dition Monitoring System,” Advances in Intelligent Systems and Com-
puting. Springer International Publishing, pp. 377–387, 2015.
[9]
Uma, S., Eswari, R. (2022). Accident prevention and safety assistance
using IOT and machine learning. Journal of Reliable Intelligent Envi-
ronments, 8(2), 79-103.
[10]
Rongcai, Z., Hongwei, X., Kexin, Y. (2023). Autonomous collision
avoidance system in a multi-ship environment based on proximal policy
optimization method. Ocean Engineering, 272, 113779.
[11]
Goyal, S. B., Bedi, P., Kumar, J., Ankita. (2022). Realtime accident de-
tection and alarm generation system over IoT. Multimedia Technologies
in the Internet of Things Environment, Volume 2, 105-126.
[12]
Muslim, H. (2023). Design and Evaluation of Lane-change Collision
Avoidance systems in Semi-automated Driving. IEEE Transactions on
Vehicular Technology.
[13]
Nanda, S., Joshi, H., Khairnar, S. (2018, August). An IOT based
smart system for accident prevention and detection. In 2018 Fourth
International Conference on Computing Communication Control and
Automation (ICCUBEA) (pp. 1-6). IEEE.
[14] Yuan, C., Lin, Y., Shen, J., Chen, L., Cai, Y., He, Y., ... Yu, Q. (2023).
Research on Active Collision Avoidance and Hysteresis Reduction of
Intelligent Vehicle Based on Multi-Agent Coordinated Control System.
World Electric Vehicle Journal, 14(1), 16.
[15] Patil, S., Raju, N., Arkatkar, S. S., Easa, S. (2023). Modeling vehicle
collision instincts over road midblock using deep learning. Journal of
Intelligent Transportation Systems, 27(2), 257-271.
717
Authorized licensed use limited to: Indian Institute of Technology (Ropar). Downloaded on March 21,2024 at 07:11:19 UTC from IEEE Xplore. Restrictions apply.
