Abstract

Cars have become increasingly safe

and convenient since the rise of the
commercial automobile industry in
late 1980s. Recently car makers have
begun to introduce advance
driver-assistance systems such as
adaptive cruise control (which
automates accelerating and braking)
and active lane assist (which
Recent trends in automates steering). Researches in
autonomous cars has progressed
Automobile remarkably over the years and it has
been found that these systems are

Engineering capable enough that the new vehicles

can drive themselves in slow-moving
highway traffic. In 2010, four
Self driving / driverless vans traveled from Italy to
China. In August of 2012, Google

Autonomous announced that its self-driving cars

had completed over 300,000 miles of
accident-free autonomous driving.
Cars The Electric Automobile company,
Tesla Motors came up with its first
DEVANG KANTHARIA autonomous car Tesla Model X.
Although self-driving cars may still
AMAN CHHEDA seem like science fiction, many auto
suppliers, and carmakers project that
ANJALI MISHRA fully autonomous cars will be
available before 2020.
SOHAM AGARWAL This report begins by describing the
landscape and key players in the
self-driving car market. Current
DWARKADAS J. SANGHVI capabilities as well as limitations and
COLLEGE OF ENGINEERING opportunities of key enabling
technologies are reviewed, along with
a discussion on the impact of such
advances on society and the
environment. This report also reviews
legal and regulatory uncertainties.
Finally, predictions about changes in
the car-industry are made.
● To eliminate traffic congestion
Introduction and have smart parking
solutions. This will lead to
increase the fuel efficiency and
● A driverless car is an
reduce carbon emission and
autonomous vehicle that can
time wasted in traffic jams and
drive itself from one point to
for finding parking areas.
another without assistance from
● Reduce the space of parking
a driver.
and making it available for
● An autonomous vehicle is
other infrastructure
fundamentally defined as a
development.
passenger vehicle.
● Increase roadways capacity as
● An autonomous vehicle is also
well as increase the speed limit.
referred to as an
● Autonomous vehicles could
autopilot,driverless car,
alleviate or completely solve
auto-drive car, or automated
these serious problems.
guided vehicle (AGV).
● Main advantages are for the
automobile companies who are
Purpose, Goals of Self Driving trying to monetize their
autonomous vehicle industries.
Cars & Potential Advantages
● Electronics and Software
● More efficient, balanced and companies also have equal
safer transportation system . advantage opportunities by
● To eliminate traffic accidents providing their technologies
caused by human errors.
 and systems embedded on the steering, and Lane Keeping
vehicle. Assistance (LKA) Type II in
any combination.
3. Level 2​: The driver is obliged
Classification of Automated
to detect objects and events and
Systems respond if the automated
● A classification system based system fails to respond
on six different levels (ranging properly. The automated
from none to fully automated system executes accelerating,
systems) was published in 2014 braking, and steering.
by ​SAE International​, an 4. Level 3​: Within known, limited
automotive standardisation environments (such as
body, as Taxonomy and freeways), the driver can safely
Definitions for Terms Related turn their attention away from
to On-Road Motor Vehicle driving tasks, but must still be
Automated Driving Systems​. prepared to take control when
This classification system is needed.
based on the amount of driver 5. Level 4​: The automated system
intervention and attentiveness can control the vehicle in all
required, rather than the vehicle but a few environments such as
capabilities, although these are severe weather. The driver must
very closely related. The types enable the automated system
are: only when it is safe to do so.
1. ​Level 0​: Automated system 6. Level 5​: Other than setting the
has no vehicle control, but may destination and starting the
issue warnings. system, no human intervention
2. Level 1​: Driver must be ready is required. The automatic
to take control at any time. system can drive to any
Automated system may include location where it is legal to
features such as Adaptive drive and make its own
Cruise Control (ACC), Parking decisions.
Assistance with automated
 ● Till date, there are only high
end models of Level 3 vehicles
available for drive, but that too
in legal areas as these are not
yet completely safe to drive.

Implementations & Technology

❖ GPS​: Space based satellite
navigation system that provides
time and location information
anywhere.
❖ DGPS​: Enhancement to GPS to
improve location accuracy to
10 cm.
❖ DIGITAL MAPS​: Process by
which a collection of data is
● SENSORS
compiled and formatted into a
❖ SELF STEERING​: steering
virtual image.
systems that use camera that
❖ BUMPER MOUNTED
watch road markings and radar
RADAR​: 4 radars mounted on
and laser sensors that track
the car’s front and rear bumpers
other objects.
enable the car to be aware of
❖ LIDAR​: Optical remote
vehicles in front of it and
sensing technology to measure
behind it and to keep
distance to target by
passengers and other motorists
illuminating with light.
safe by avoiding bumps and
crash.
❖ ULTRASONIC SENSORS
ON REAR WHEEL​: An
ultrasonic sensor on one of the
rear wheels helps keep track of
the movements of the car and
 will alert the car about the
obstacles in the rear.

❖ Camera is basically the

system's eyes.
❖ DEVICES WITHIN THE ❖ They are superior to read signs,
CAR​: Inside the car are detect traffic, pedestrians,lane
altimeters, gyroscopes, and markings and anything else that
tachymeters that determine the might be in front of the vehicle.
very precise position of the car. ❖ Being the vision system it not
❖ PROGRAMMED TO only detects the obstacle but
INTERPRET COMMON also detects what the obstacle
ROAD SIGNS AND REAL is.
LIFE ROAD BEHAVIOUR​: ● COMPUTERS AND
The software has been PROCESSORS
programmed to rightly interpret ❖ Processor is controlled
common road behavior and basically by Driver Px2 system.
motorist signs. Several such ❖ This system will feature two
programs fed into the car’s Parker processors as well as
CPU will work simultaneously, two of Nvidia latest graphical
helping the car make safe and processing units (GPUs) called
intelligent decisions on busy Pascal.
roads. ❖ Drive PX 2 delivers 24 trillion
❖ MICROPHONE​: Can detect deep learning operations per
sirens of approaching second. Deep learning, a branch
emergency vehicles. of machine learning, will
● HIGH DEFINITION CAMS enable the car to figure out the
 world around it — like, for mapping and vehicle driving
example, if that’s a dog or decisions.”
police car with its siren on. ❖ BlueBox is already being tested
❖ GPUs have caught on in AI, in with four of the five largest
part, because of their ability to automobile companies in the
do “parallel computing,” a world.
technique that involves multiple
calculations happening
Potential Disadvantages
simultaneously.
● COMPUTER (BLUEBOX) ● Loss of driving-related jobs in
❖ Computers that process sensor the road transport industry.
data in real-time can vastly ● The automobile insurance
improve how autonomous cars industry might suffer as the
make decisions on the road, technology makes certain
whether that involves slowing aspects of these occupations
down ahead of traffic jams or obsolete.
slamming on the brakes when a ● Potential loss of privacy and
person steps into the road. risks of hacking. Sharing of
❖ In the near term, BlueBox information through V2V
could be tested to enhance (Vehicle to Vehicle) and V2I
advanced driver-assisted (Vehicle to Infrastructure)
systems that alert drivers in protocols.There is also the risk
dangerous situations, like of terrorist attacks.
automatic braking and ● Self-driving cars could
blind-spot warnings. potentially be loaded with
❖ BlueBox incorporates “the explosives and used as bombs.
embedded intelligence and ● There is also the risk that traffic
machine learning required for congestion might increase,
complete situational rather than decrease.
assessments, supporting ● Research shows that drivers in
advanced classification tasks, autonomous cars react later
object detection, localization, when they have to intervene in
 a critical situation, compared to
if they were driving manually.
Conclusion
● Autonomous cars in current
transportation environment is A recent UK study has looked at the
unsafe. Many autonomous car transformative implications of
accidents have been recorded self-driving vehicles on cities. Shared
from 2011-2016 which includes autonomous vehicles could increase
vehicles of big companies like available urban space by 15 to 20
Tesla motors (Fatal accident percent, largely through the
took place in Williston, Florida elimination of parking spaces. Today
on 7 May 2016 while a Tesla central London has about 6.8 million
Model S electric car was parking spaces and a parking
engaged in Autopilot mode. coverage of around 16%! Many large
The occupant was killed in a cities have even larger coverage ratios
crash with an 18-wheel for parking space of up to 30%.
tractor-trailer) Freeing up this space would make our
cities greener, increase quality of life
and also create the potential for
additional housing.
Autonomous vehicles will also make
the rural communities more attractive
because shared travel to nearby cities
becomes widely available, affordable
,and Google (Google's own
and does not lead to loss of productive
accident reports, their test cars
time.
have been involved in 14
Consider autonomous vehicle only
collisions, of which other
development areas and highways that
drivers were at fault 13 times. It
are limited to autonomous vehicles.
was not until 2016 that the car's
This could reduce costs as lane
software caused a crash).
markings and signage would no
longer be needed, the lanes could be
narrower and throughput per lane high numbers.
would be higher. Increased use of car-sharing for local
transport reduces the overall demand
for vehicles which in turn reduces
greenhouse gas emission for
manufacturing automobiles.

References
● http://velodynelidar.com/lidar/h
dlpressroom/How%20Google
● http://www.driverless-future.co
m/?cat=17
Autonomous vehicles could greatly ● https://www.national.co.uk/tech
decrease greenhouse gas emissions in -powers-google-car/
urban traffic because ● http://www.seminarsonly.com/
Car-sharing services could offer local Engineering-Projects/Embedde
mobility for a highly competitive d/Driverless-Car.php
price based on a fleet of smaller, ● https://www.google.co.in/amp/s
lighter cars which therefore cause /www.forbes.com/sites/aarontill
fewer emissions. ey/2016/08/22/nvidia-unveils-n
Local car sharing fleets would be ew-processor-for-its-self-drivin
ideal adopters for alternative, g-car-supercomputer/amp/
low-emission drives (electric cars, ● https://www.google.co.in/amp/
hydrogen, fuel cells). Because of their www.telegraph.co.uk/cars/featu
higher utilization levels, higher initial res/how-do-driverless-cars-wor
capital costs for the new technology k/amp/
as compared to the gasoline engine ● www.wikipedia.com
would not matter as much.
Autonomous cars used for local trips
would be an ideal application for
getting electric cars into operation in
About The Team
● Devang Kantharia
Email Id:
[email protected]
Mobile No. 9819900749

● Aman Chheda
Email Id: ​[email protected]
Mobile No: 9967025250

● Soham Agarwal
Email Id:
[email protected]
Mobile No. 9699601994

● Anjali Mishra
Email Id: ​[email protected]
Mobile No. 9930638953

( First year- Production

Engineering ,DWARKADAS J.
SANGHVI COLLEGE OF
ENGINEERING)