Presentation On Project: Presented by

You might also like

Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 26

Presentation

On
Project

Presented By: Submitted To:


Aditi Singh (17EJCCS009) Mr. Sachin Gupta
Abhishek Varshney (17EJCCS006) Project Guide
Bhavya Jain (17EJCCS040)
Anuj Godara (17EJCCS028)
Proposed Project Title

3D simulation for self-driving car using CARLA simulator


and driver warning system for path generation, path
selection, static collision checking, velocity profile
generation.
Tools & Technologies

Front-end: CARLA Simulator, Unreal Engine

Back-end: Python, Machine Learning,


OpenCV, CNN
List of Modules
1. Introduction to Self Driving Car
:
2. Sensors

3. Carla Simulator

4. Unreal Engine

5. OpenCV

6. Path Generation

7. Static Collision Checking

8. Path Selection

9. Velocity Profile Generation

10. Future Scope


How the Project is relevant to SIH Problem

In SIH there was a problem statement of Pothole detection system.

Pothole detection system (integration of a system capable of detecting


potholes on the roads using the mobility of the car being driven to ensure
driver and rider safety by detecting potholes on its own and alerting the driver
accordingly).

Screenshot of the SIH Problem Statement:


Project Area

Autonomous Vehicles/Self-Driving Vehicles:

An autonomous vehicle, or a driverless vehicle, is one that is able to operate


itself and perform necessary functions without any human intervention, through
ability to sense its surroundings.
Autonomous driving normally refers to self-driving vehicles or transport
systems that move without the intervention of a human driver. In 2014, SAE
International (Society of Automotive Engineers) published the J3016 standard
to define the various development levels up to fully autonomous vehicles. The
levels for autonomous driving range from Level 0 (no automation) up to Level 5
(full vehicle autonomy). The term self-driving is often used interchangeably with
autonomous. However, it’s a slightly different thing. A self-driving car can drive
itself in some or even all situations, but a human passenger must always be
present and ready to take control.
Self-driving cars would fall under Level 3 (conditional driving automation) or
Level 4 (high driving automation). They are subject to geofencing, unlike a
fully autonomous Level 5 car that could go anywhere.

Autonomous vehicles can-


Reduce traffic congestion (30% fewer vehicles on the road).
Cut transportation costs by 40% (in terms of vehicles, fuel, and
infrastructure).
Improve walkability and livability.
Free up parking lots for other uses (schools, parks, community centers).
Reduce urban CO2 emissions by 80% worldwide.
Objective

• 3D simulation for self driving car

• Project focuses on training the model for the preset waypoints

• Uses Carla Simulator, Unreal Engine to accomplish this objective


Abstract

Project tiled “Self-driving car using CARLA simulator” is a 3D


simulation for self-driving car, warning system for driver
focusing on training the model for the preset waypoints using
CARLA simulator and Unreal Engine to accomplish this goal
as well as to regulate the car speed. This project is mainly
written in Python programming language.
This is a model which will run on Carla simulator, our goal is
to control the vehicle to follow a racetrack by navigating
through preset waypoints.
The vehicle needs to reach these waypoints at certain desired
speeds, so both longitudinal and lateral control will be
required. The trajectory feedback will contain the car, start
and end positions, entire path/path travelled and a small
shaded region which denotes the subset of interpolated
points to be sent into the controller for control updates.
Linear interpolation will be used between waypoints to
provide a finer resolution path/speed requests for the
controller.
Introduction

A self-driving car is a vehicle capable of sensing its environment


and operating without human involvement. A human passenger
is not required to take control of the vehicle but is required to be
present in the vehicle. An autonomous car can go anywhere a
traditional car goes and do everything that an experienced
human driver does.
Background
In the 21st century, in today’s age, the need for self-driving cars is
rising exponentially as autonomous systems integrated with
warning systems for drivers provide for user comfort and
accidental safety in a better way than manual control.
Automation can help reduce the number of crashes on our roads.
Government data identifies driver behavior or error as a factor in 94
percent of crashes, and self-driving vehicles can help reduce driver
error. The greatest promise may be reducing the devastation of
impaired driving, drugged driving, unbelted vehicle occupants,
speeding and distraction.
Solution
CARLA has been developed from the ground up to support
development, training, and validation of autonomous driving
systems. In addition to open-source code and
protocols, CARLA provides open digital assets (urban layouts,
buildings, vehicles) that were created for this purpose and can be
used freely.
Unreal Engine is the world’s most open and advanced real-time 3D
creation tool. Gives creators across industries the freedom and
control to deliver cutting-edge content, interactive experiences,
and immersive virtual worlds. 
Functionality of Project

The functionality of the chatbot includes:

1. The requirements for autonomy


2. Self-Driving Hardware and Software Architectures
3. Safety Assurance for Autonomous Vehicles
4. Vehicle Dynamic Modeling
5. Vehicle Longitudinal Control
6. Vehicle Lateral Control
7. State Estimation - Linear and Nonlinear Kalman Filters
8. GNSS/INS Sensing for Pose Estimation
9. LIDAR Sensing
10.An Autonomous Vehicle State Estimator
Sensors

Vehicle that is capable of sensing its environment

Combine a variety of sensors to perceive their surroundings:


○ Radar
○ Lidar
○ Sonar
○ GPS
○ Odometry 
○ Inertial measurement units 
Benefits
Scenarios for convenience and quality-of-life improvements:
Independence of elderly and the physically disabled
Sending pets to a veterinary appointment.

But the real promise of autonomous cars is the potential for


dramatically lowering CO2 emissions.
Reduce traffic congestion (30% fewer vehicles on the road)
Cut transportation costs by 40% (in terms of vehicles, fuel, and
infrastructure)
Carla Simulator
CARLA has been developed from the ground up to support
development, training, and validation of autonomous driving
systems.

In addition to open-source code and protocols, CARLA provides


open digital assets (urban layouts, buildings, vehicles) that were
created for this purpose and can be used freely.

The simulation platform supports flexible specification


of sensor suites, environmental conditions, full control
of all static and dynamic actors, maps generation and much more.
Unreal Engine
World’s most open and advanced real-time 3D creation tool

Gives creators across industries the freedom and control to deliver


cutting-edge content, interactive experiences.

Game engine developed by Epic Games, first showcased in the


1998 first-person shooter game Unreal.

Written in C++, the Unreal Engine features a high degree of


portability, supporting a wide range of platforms.
OpenCV
● Open source computer vision and machine learning software

● Consists programming functions mainly aimed at real-time


computer vision.

● Built to provide a common infrastructure for computer vision


applications and to accelerate the use of machine perception in
the commercial products.

● Being a BSD-licensed product, OpenCV makes it easy for


businesses to utilize and modify the code.
Architectural Strategies

● Path Generation

● Static collision Checking

● Path Selection

● Velocity Profile Generation


Path Generation

●For the path generation section of the project. We will need to


compute the goal state set (the set of goal points to plan paths to
before path selection) using the get_goal_state_set() function, as
well as some of the path generation helper functions.

●The resulting spiral will be sampled to generate the path.


This functionality needs to be implemented in sample_spiral().
Static Collision Checking

For motion planner, we will be editing collision_checker.py.


In particular, we're going to be implementing circle-based
collision checking on our computed path set using the
collision_check() function. We will implement the circle location
calculation for each point on the path.
Path Selection

The path selection portion of the project involves you evaluating


an objective function over the generated path set to select the best
path. The goal of this section is to eliminate paths that are in
collision with static obstacles, and to select paths that both track
the centerline of the global path.
Velocity Profile Generation

This is all captured in the compute_velocity_profile() function in


velocity_planner.py. We have implemented the physics functions
at the end of the file which will be used for velocity planning.
Future Scope

1. Pothole detection system (integration of a system capable


of detecting potholes on the roads using the mobility of
the car being driven to ensure driver and rider safety by
detecting potholes on its own and alerting the driver
accordingly).
2. Collaboration with RTO (regional transport offices) to
install the system in each and every car to ensure rider
safety and ensure futuristic vision of the government.
References

● https://towardsdatascience.com/how-to-train-
your-self-driving-car-using-deep-learning-
ce8ff76119cb
● https://carla.readthedocs.io/en/latest/start_intro
duction/
● https://towardsdatascience.com/deeppicar-part-
1-102e03c83f2c
● https://carla.org

You might also like