International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 07 Issue: 05 | May 2020 www.irjet.net p-ISSN: 2395-0072

AUTONOMOUS CARS USING RASPBERRY Pi

Nandhini N1, Rakshana R2, Revathi L3, Siva Swarnamalya4, Nagaraj V5
1,2,3,4Under graduate, Department of Electronics and Communication Engineering,
5Assistant Professor, Department of Electronics and Communication Engineering,
Knowledge Institute of Technology, Salem, Tamil Nadu, India
---------------------------------------------------------------------***--------------------------------------------------------------------
Abstract: It is for an online comprehensive solution to The idea of autonomous cars started with
manage internet banking. Autonomous cars are the future "phantom autos" in the 1920s, where the car was
smart cars anticipated to be efficient and crash avoiding controlled through a remote control device. In
ideal urban car of the future. The purpose of the project is the 1980s, we witnessed the emergence of self-
to build up a car which can move with any driver. It just sufficient and self-managed autonomous cars. A
need a command from the user and then it will start major contributor to the autonomous car field
moving towards the destination. The project aims to build was the NavLab at Carnegie Mellon University,
a autonomous car prototype using Raspberry Pi where researchers developed the Autonomous
processor. Sensor interfacing is used to provide necessary Land Vehicle (ALV). In the same decade, the
data from the real world to the car. The car is capable of "Prometheus project," sponsored by Mercedes in
reaching the given destination safely and intelligently 1987, achieved a major result with the design of
thus avoiding the risk of human errors. Many existing their first robotic car to track lane markings and
algorithms like traffic light detection, obstacle detection other vehicles. In the 21st century, the increasing
and speed control are combined together to provide the interest in autonomous cars has been fueled
necessary control to the car. The technology that primarily by low-cost, high-performance
autonomous cars and “connected vehicles” use would technologies in various areas.
allow the vehicles to communicate with roadside
infrastructure like traffic lights and road congestion, and As the human population grows and the number
then use this information to curtail fuel consumption and of cars increases, this creates a stressful impact
emissions significantly. on our transportation infrastructure, ranging
from roads and parking spaces to fuel stations
Key Words: Connected vehicles, Navigation (for fuel engines vehicles) and charging stations
system, Obstacle detection, Path planning, (for electric and hybrid vehicles). In the past few
Sensors decades, governments have taken serious
measures for road safety, with many introducing
1. INTRODUCTION both static and dynamic technologies such as
closed-circuit television (CCTV) cameras, road
An autonomous car refers to a computer- sensors, and more.
controlled car that can guide itself, familiarize
itself with surroundings, make decisions, and However, despite these efforts, in the United
fully operate without any human interaction. The States alone, road accidents caused more than
primary drivers behind the emergence of 32,000 fatalities in 2014. The number of
autonomous cars include: the need for driver fatalities increased to more than 35,000 in 2015,
and driving safety, growth in population, demonstrating that despite the use of existing
expanding infrastructure, increase in the number technologies, human errors still occur.
of vehicles, the need for efficient time
management, and resource utilization and The traffic conditions in developing countries,
optimization. like India, are more complex owing to varied
road conditions, a heterogeneous mix of vehicles
The evolution and emergence of autonomous and chaotic traffic. Since India is a developing
cars are the result of remarkable research country there is a constant requirement for good
results coming from the fields of wireless quality transportation, infrastructure, and
communication, embedded systems, navigation, services.
sensor and ad hoc network technologies, data
acquisition and dissemination, and data Roads in India normally have speed breakers so
analytics. that the vehicle’s speed can be controlled to
avoid accidents, because of our system we also
have undefined potholes (formed due to heavy

© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 6340
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 05 | May 2020 www.irjet.net p-ISSN: 2395-0072

rains and movement of heavy vehicles) which using unified map built with various onboard
created immediately after of completion of the sensors to detect obstacles, other cars, traffic
road. It is a major reason for traumatic accidents signs, and pedestrians. The proposed map
and loss of human lives. contains not only the information on real
obstacles nearby but also traffic signs and
To minimize human errors and reduce life- pedestrians as virtual obstacles.[2]Using this
threatening situations on the road, alternative map, the path planner can efficiently find paths
technologies such as connected cars and free from collisions while obeying traffic laws.
autonomous cars are being explored. The proposed algorithms were implemented on
a commercial vehicle and successfully validated
2. LITERATURE SURVEY in various environments, including the 2012
hyundai autonomous ground vehicle
2.1 Autonomous Cars: Research Results, competition.
Issues, and Future Challenges- Rasheed
Hussain ;Sherali Zeadally [1] 2.3 Short Term Traffic Prediction for Edge
Computing Enhanced Autonomous and
Throughout the last century, the automobile Connected Cars- Shun-Ren Yang ; Yu-Ju Su ;
industry achieved remarkable milestones in Yao-Yuan Chang ; Hui-Nien Hung [3]
manufacturing reliable, safe, and affordable
vehicles. Because of significant recent advances Autonomous and connected cars (ACCs),
in computation and communication together with edge computing (EC), have been
technologies, autonomous cars are becoming a recognized as a promising solution to achieve
reality. Already autonomous car prototype green intelligent transportation for smart cities.
models have covered millions of miles in test This paper aims to address short-term traffic
driving. Leading technical companies and car prediction, a fundamental enabler for the success
manufacturers have invested a staggering of ACC applications, under the European
amount of resources in autonomous car Telecommunications Standards Institute multi-
technology, as they prepare for autonomous access EC (MEC) architecture that exhibits
cars' full commercialization in the coming years. constraints different from conventional cloud
However, to achieve this goal, several technical computing. First, a data-centric experiment
and nontechnical issues remain: software platform is designed and implemented to
complexity, real-time data analytics, and testing facilitate traffic prediction algorithm
and verification are among the greater technical development. This paper further proposes a
challenges; and consumer stimulation, insurance novel short-term traffic prediction model that
management, and ethical/moral concerns rank integrates a traffic light model and a vehicle
high among the nontechnical issues.[1] Finally, velocity model, considering limited computing
to enable cost-effective, safe, and efficient resources of MEC servers. We note that the
autonomous cars, we discuss several challenges effects of traffic lights are complicated and have
that must be addressed (and provide helpful not been rigorously examined in most, if not all,
suggestions for adoption) by designers, of the related work. This work models the
implementers, policymakers, regulatory queueing time when a driver arrives at a road
organizations, and car manufacturers. intersection and faces a red light. Moreover, to
forecast the vehicle velocity, we propose a novel
2.2 An Autonomous Driving System for low-complexity semiparametric prediction
Unknown Environments Using a Unified Map model considering periodic features and
-Inwook Shim; Jongwon Choi; Seunghak Shin spatial/temporal correlations of dynamic road
[2] events. The experiment results demonstrate that
our vehicle-velocity prediction model achieves
There have been significant advances in self- almost equivalent accuracy to the well-known
driving cars, which will play key roles in future Long Short-Term Memory Neural Network
intelligent transportation systems. In order for model, requiring much lower computational
these cars to be successfully deployed on real complexity. [3]
roads, they must be able to autonomously drive
along collision-free paths while obeying traffic
laws. In contrast to many existing approaches
that use prebuilt maps of roads and traffic
signals, we propose algorithms and systems

© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 6341
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 05 | May 2020 www.irjet.net p-ISSN: 2395-0072

2.4 Road profile recognition for autonomous space with just one backward maneuver. To
car navigation and Navstar GPS support- W. accomplish the autonomous fuzzy behavior
Holzapfel; M. Sofsky; U. Neuschaefer-Rube [4] control, the car-like-mobile-robot has trained to
park in just 2 parking dimensions based on the
Autonomous car navigation based on updating training data obtained from sensor information
dead reckoning (DR) by road profile recognition generated offline by adopting a fifth-order
(RPR). The navigation system requires sensors polynomial as the reference trajectory. The
to detect changes in altitude and driving proposed controller is an ANFIS architecture
direction which are installed in modern cars for that generates turning angle as output. As long as
different purposes (e.g. ABS sensors).[4]The the states (positions and orientations) of the
layout of the navigation system is discussed and robot are measurable at each discrete-time step
simulations are carried out over driving during the control process, this controller can
distances of approximately 150 km on the basis make the robot follow feasible trajectories by
of realistic road data and ordinary sensor just knowing the initial configuration of the
accuracies. Positioning errors of lower than 10 m robot and park successfully at the prescribed
(standard deviation) are observed. To achieve goal position.[6]The simulation results which are
this accuracy the synchronization error between based on real dimensions of a typical car
measured and mapped data must be continually demonstrate the feasibility and effectiveness of
estimated. The introduced navigation method is the proposed controller in practical car
ideal to complete present commercial car maneuvers.
navigation systems using Navstar GPS.
2.7 Design of multifunctional autonomous car
2.5 RFID-based autonomous mobile car- Jen- using ultrasonic and infrared sensors -
Hao Teng; Kuo-Yi Hsiao; Shang-Wen Luan [5] Ayesha Iqbal; Syed Shaheryar Ahmed [7]

Radio Frequency Identification (RFID) system is The daily routine problems that common man
looked upon as one of the top ten important faces on roads while commuting are becoming a
technologies in the 20 th century. Industrial serious problem with each passing day. People
automation application is one of the key issues in get late and meet accidents. The model of
developing RFID. Therefore, this paper designs autonomous car presented in this research paper
and implements a RFID-based autonomous aims to solve these issues by taking humans off
mobile car for more extensively application of the wheels, so that they do not have to drive
RFID systems. The microcontroller of Microchip anymore and the risk of accidents, getting late
PIC18F4550 is used to control the autonomous and traffic congestions can be reduced to a
mobile car and to communicate with RFID minimum. This car is able to follow the track,
reader.[5] By storing the moving control overtake other cars, detect obstacles, take sharp
commands such as turn right, turn left, speed up bends and turns, follow traffic signals and turn
and speed down etc. into the RFID tags on its lights under low light conditions. [7]Circuit
beforehand and sticking the tags on the tracks, diagrams for performing all these functions have
the autonomous mobile car can then read the been presented and the mechanical model of the
moving control commands from the tags and car has also been shown in the paper, which is
accomplish the proper actions. Due to the practically implemented and successfully run by
convenience and non-contact characteristic of the authors.
RFID systems, the proposed mobile car has great
potential to be used for industrial automation, 3. METHODOLOGY
goods transportation, data transmission, and
unmanned medical nursing etc. in the future. The four key technologies in self-driving car,
namely, car navigation system, path planning,
2.6 Autonomous parallel parking of a car-like environment perception and car control, are
mobile robot by a neuro-fuzzy behavior- addressed and surveyed. The automatic control,
based controller- M. Khoshnejad; K. Demirli architecture, artificial intelligence, computer
[6] vision and many other technologies are
integrated into the self-driving car, which is a
The concept of sensor-based behavior is used to product of the highly developed computer
design a neuro-fuzzy control system for a car- science, pattern recognition and intelligent
like-mobile-robot. The problem addressed is the control technology.
parallel parking in a rectangular constrained

© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 6342
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 05 | May 2020 www.irjet.net p-ISSN: 2395-0072

Car Navigation System-During self-driving, two WORKING PRINCIPLE

issues, which are the current location of the car
and how to go from the location to the Block diagram for Autonomous cars is shown in
destination, must be resolved. Certainly, the fig 4. The proposed system is prototype function
above two issues can be solved by a human’s of autonomous car with multiple objective, like
own knowledge in human driving. However, in potholes detection, traffic light detection and
self-driving, the car must be able to speed regulation in school zone. For the above
automatically and intelligently locate its position functions are implemented with Raspberry pi
and perform the path planning to destination. processor and various sensors. The robot
For this objective, the on-board car navigation arrangement is implemented for the car setup
system is deployed on the self-driving car. with Gear motor. Ultrasonic sensor is used to
detect the humps and potholes. Color sensor is to
Location system-The main purpose of the detect the traffic lights. For example, when it
location system is to determine the vehicle detects the red light, the car stops. This speed
location, which generally can be classified into sensor indicates the speed at which the car is
relative location, absolute location and hybrid actually travelling. It is used to control certain
location. For relative location, the current handling systems, as well as certain braking
position of self-driving car is obtained by adding systems. It is controlled by a signal generator that
the moving distance and direction to the prior is spun, and makes an electrical pulse that is then
position. sent to the vehicle's computer. Zigbee
communication is used to identify the school
Electronic map (em)-EM is used for digital map zone. So the car slows down at the schools. Gsm
information storage, which mainly includes sends the message through the mobile phone to
geographical characteristics, traffic information, concerned person.
building information, traffic signs, road facilities,
etc. It is expected that special EMs for self- 4.1 Raspberry Pi Proccessor:
driving, such as automatic road sign recognition,
car’s driving information interacting among self- The Raspberry Pi is a credit card-sized single-
driving cars, will be developed in the future. board computer. There are currently five
Raspberry Pi models in market i.e. the Model B+,
Global Path Planning-It is used to determine the Model A+, the Model B, the Model A, and the
the optimal driving path between the start point Compute Module (currently only available as
and end point. Generally, the typical path part of the Compute Module development kit).
planning algorithms, such as Dijkstra algorithm, All models use the same SoC (System on Chip -
Bellman-Ford algorithm, Floyd algorithm and combined CPU & GPU), the BCM2835, but other
heuristic algorithm (Seshan and Maitra, 2014) hardware features differ. The A and B use the
are employed to fuse the EM information and same PCB, whilst the B+ and A+ are a new design
calculate the optimal path. but of very similar form factor. The Compute
Module is an entirely different form factor and
4. BLOCK DIAGRAM cannot be used standalone. The project aims to
build a monocular vision autonomous car
prototype using Raspberry Pi as a processing
chip. In this project, we have used the model B
Rev 2. It comprises of a 512 MB RAM model with
two USB ports and a 10/100 Ethernet controller.
It is powered by USB and the video output can be
hooked up to a traditional RCA TV set, a more
modern monitor, or even a TV using the HDMI
port. This gives you all of the basic abilities of a
normal computer. It also has an extremely low
power consumption of about 3 watts.

4.2 PIC microcontrollers

Fig 4.1 Block Diagram of Autonomous Cars They are a family of specialized microcontroller
chips produced by Microchip Technology in
Chandler, Arizona. The acronym PIC stands for

© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 6343
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 05 | May 2020 www.irjet.net p-ISSN: 2395-0072

"peripheral interface controller," although that outdoor images.

term is rarely used nowadays. A typical
microcontroller includes a processor, memory, 4.6 GEAR MOTOR
and peripherals. PIC microcontrollers are the
worlds smallest microcontrollers that can be The DC geared motor is used here to control the
programmed to carry out a huge range of tasks. movement of the robot and help in carrying the
These microcontrollers are found in many robot to the desired area. The gear assembly
electronic devices such as phones, computer helps in increasing the torque and reducing the
control systems, alarm systems, embedded speed of the shaft. It can operates on 24.0 volt,
systems, etc. Various types of microcontrollers 250mA with no load condition at speed of
exist, even though the best are found in the 150rpm.
GENIE range of programmable microcontrollers.
These microcontrollers are programmed and 4.7 MOTOR DRIVER
simulated by a circuit-wizard software. Every
PIC microcontroller architecture consists of Motor drivers acts as an interface between the
some registers and stack where registers motors and the control circuits. Motor require
function as Random Access Memory(RAM) and high amount of current whereas the controller
stack saves the return addresses. circuit works on low current signals. So the
function of motor drivers is to take a low-current
4.3 ULTRASONIC SENSORS control signal and then turn it into a higher-
current signal that can drive a motor.
Ultrasonic sensors (also known as transceivers
when they both send and receive, but more 4.8 GSM/GPS
generally called transducers) evaluate attributes
of a target by interpreting the echoes from radio GSM (Global System for Mobile communication)
or sound waves respectively. In this project, they is a digital mobile network that is widely used by
are used to detect the distance of obstacles from mobile phone users in all the parts of the world.
the car. It has been used in vehicles to support GPS is a space-based satellite navigation system.
drivers with their driving tasks like parking. It provides location and time information in all
weather conditions, anywhere on or near the
4.4 SPEED SENSORS Earth. GPS receivers are popularly used for
navigation, positioning, time dissemination and
The wheel speed sensor was initially used to other research purposes. A GSM modem requires
replace the mechanical linkage from the road a SIM card to be operated and operates over a
wheels to the speedometer, eliminating cable network range subscribed by the network
breakage and simplifying the gauge construction operator. It can be connected to a computer
(elimination all moving parts except for the through serial, USB or Bluetooth connection. A
needle/spring assembly). With the advent of GSM modem can also be a standard GSM mobile
automated driving aid, such as electronic ABS, phone with the appropriate cable and software
the sensor also provided wheel speed data to the driver to connect to a serial port or USB port on
controllers to assist the operator in maintaining your computer. GSM modem is usually
control of the vehicle. The vehicle Speed sensor preferable to mobiles.
is also used for the proper shifting up of gears for
the vehicle maintenance. 4.9 POWER SUPPLY

4.5 COLOR SENSOR A power supply is an electrical device that

supplies electric power to an electrical load. The
Color can be a useful feature in autonomous primary function of a power supply is to convert
vehicle systems that are based on machine electric current from a source to the correct
vision, for tasks such as obstacle detection, voltage, current, and frequency to power the
lane/road following, and recognition of load. As a result, power supplies are sometimes
miscellaneous scene objects. Unfortunately, few referred to as electric power converters. Some
existing autonomous vehicle systems use colour power supplies are separate standalone pieces of
to its full l extent, largely because colour-based equipment, while others are built into the load
recognition in outdoor scenes is complicated, appliances that they power. Examples of the
and existing colour machine vision techniques latter include power supplies found in desktop
have not been shown to be effective in realistic computers and consumer electronics devices.

© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 6344
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 05 | May 2020 www.irjet.net p-ISSN: 2395-0072

Other functions that power supplies may 6. RESULT

perform include limiting the current drawn by
the load to safe levels, shutting off the current in
the event of an electrical fault, power
conditioning to prevent electronic noise or
voltage surges on the input from reaching the
load, power-factor correction, and storing
energy so it can continue to power the load in
the event of a temporary interruption in the
source power.

4.10 Battery

Battery is used for required power supply needed for

operating the different components in the robot. Here
we used a 12v dry cell Re-chargeable battery for the
purpose. Lithium-Polymer (Li-Po) Battery, 3 Cell (3S) 7. CONCLUSION
11.1V2200MAH, Very small in size and weight,
Connector Plug:XT60 connector, Discharge Current: In this paper, a method to make a self driving
25*2200maH=55Amp. car is presented. The different hardware
components and their assembly are clearly
5. FLOWCHART described. Driverless car revolution which aims
at the development of autonomous vehicles for
easy transportation without a driver. For the
economy, society and individual business this
autonomous technology has brought many broad
implications. Cars that drive themselves will
improve road safety, fuel efficiency, increase
productivity and accessibility; the driverless car
technology helps to minimize loss of control by
improving vehicle’s stability as they are designed
to minimize accidents by addressing one of the
main causes of collisions: Driving error,
distraction and drowsiness. The algorithm
mention in this paper has been successfully
implemented in prototype of Autonomous car.

8. FUTURE ENHANCEMENT

The work could be enhanced by improving the

algorithm by adding advanced machine learning
to it. Multi layered processors can be used for
fast processing. The present obstacle detection
Fig 5.1 Flow Diagram for Autonomous car algorithm just detects the obstacle and stops, but
in future it can be improved by the avoiding the
obstacle, and go through another way using
advanced obstacle detection algorithm.

9. REFERENCES

[1] RasheedHussain , Sherali Zeadally

(2018)" Autonomous Cars: Research Results,
Issues, and Future Challenges", IEEE
Communications Surveys & Tutorials,Vol. 21,No.
6, pp. 122-128.

© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 6345
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 05 | May 2020 www.irjet.net p-ISSN: 2395-0072

[2] Inwook Shin;Jongwon Choi;Seunghak

Shin(2015)”An Autonomous Driving System for
Unknown Environments Using a Unified Map”
IEEE Transactions on Intelligent Transportation
Systems,Vol.16 ,No.4,pp.661-692 .

[3] Shun-Ren Yang;Yu-Ju Su ; Yao-Yuan Chang

(2019)” Short Term Traffic Prediction for Edge
Computing Enhanced Autonomous and
Connected Cars” IEEE Transactions on Vehicular
Technology,Vol.68,No.4,pp. 3140 – 3153.

[4] W. Hzapfel; M. Sofsky ;U. Neuschaefer-

Rube (2003)" Road profile recognition for
autonomous car navigation and Navstar GPS
support ", IEEE Transactions on Aerospace and
Electronic Systems ,Vol.39,No.1,pp.2-12.

[5] Jen-Hao Teng ; Kuo-Yi Hsiao ; Shang-Wen

Luan (2010)" RFID-based autonomous mobile
car", 28th IEEE International Conference on
Industrial Informatics,Vol. 21,No. 6, pp. 22-128.

[6] M.Khoshnejad ; K. Demirli (2005)"

Autonomous parallel parking of a car-like mobile
robot by a neuro-fuzzy behavior-based
controller ", 2005 - 2005 Annual Meeting of the
North American Fuzzy Information Processing
Society (NAFIPS) , vol. 12, No. 4, pp. 447-465.

[7] Ayesha Iqbal ; Syed Shaheryar Ahmed

(2017)" Design of multifunctional autonomous
car using ultrasonic and infrared sensors",
International Symposium on Wireless Systems
and Networks (ISWSN)

[8] Gowdham Prabhakar;Binsu Kailath;

Sudha Natarajan (2017)" Obstacle detection
and classification using deep learning for
tracking in high-speed autonomous driving",
IEEE Region 10 Symposium (TENSYMP)

[9] Shai A. Arogeti ; Nadav

Berman(2012)"Path Following of Autonomous
Vehicles in the Presence of Sliding Effects", IEEE
Transactions on Vehicular
Technology,Vol61,No.4.pp.1481 – 1492.

[10] Myeon-gyun Cho (2019)" A Study on the

Obstacle Recognition for Autonomous Driving
RC Car Using LiDAR and Thermal Infrared
Camera", Eleventh International Conference on
Ubiquitous and Future Networks (ICUFN)

© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 6346