An Autonomous Smart Car based on Artificial Intelligence

Automated vehicles might not be ruling the roads just yet but the big players in car manufacturing are investing heavily into what is expected to be the biggest breakthrough in motoring since model T. Research in autonomous navigation was done from as early as the 1900s with the first concep

Project Title

An Autonomous Smart Car based on Artificial Intelligence

Project Area of Specialization

Artificial Intelligence

Project Summary

Automated vehicles might not be ruling the roads just yet but the big players in car manufacturing are investing heavily into what is expected to be the biggest breakthrough in motoring since model T. Research in autonomous navigation was done from as early as the 1900s with the first concept of the automated vehicle exhibited by General Motors in 1939. However, most techniques used by early researchers proved to be less effective or costly. We have focused on one main applications of an Automated Vehicles here and designed a prototype vehicle for that. That major issue is during heavy traffic a driver has to continuously push a brake, accelerator and clutch to move to destination slowly. We have proposed a solution to relax the driver in that situation by making vehicles smart enough to make decisions automatically and move by maintaining a specified distance from vehicles, avoiding obstacles, pedestrians on the path and also follow the basic traffic rules without any human intervention. Moreover, an autonomous car must be capable of avoiding obstacles, pedestrians on the path and also follow the basic traffic rules without any human intervention. To accomplish this, a driverless car must have an artificial intelligence system that senses its surroundings, processes the visual data to determine how to avoid collisions, and operates car machinery, like steering and brake etc.

Project Objectives

The main objective is to purpose a prototype of an Artificial based Autonomous car.

• To develop a baseline model for autonomous car

• To develop an algorithm for following a basic track 

• To detect the lane, sign, and human for autonomous car

• To prioritize the selection regarding decision making

• To develop a route optimization algorithm to choose the shortest path

 

Project Implementation Method

Deep Learning

Deep learning is a class of machine learning algorithms that uses multiple layers to progressively extract higher-level features from the raw input. It is a type of neural network that contains convolutional layers. CNN's are used prevalently in image recognition deep learning models. The intuition is that CNN is especially good at extracting visual features from images from its various layers. We will use a deep-learning approach to teach our vehicle to do the same, turning it into a DeepPicar. This is analogous to how you and I learned to drive, by observing how good drivers (such as our parents or driving school coaches) drive and then start to drive by ourselves and learn from our own mistakes along the way.

The Nvidia Model

The model uses the video images, exacts information from them, and tries to predict the car’s steering angles. This is known as a supervised machine learning program, where video images (called features) and steering angles (called labels) are used in training. At the core of the NVidia model, there is a Convolutional Neural Network. 

Control

We will use OpenCV to detect color, detect edges, detect line segments.  We had to set, upper and lower bounds of the color red, many parameters to detect line segments. We used the Canny edge detection function which is a powerful command that helps to detect edges in an image. Moreover, for pedestrian detection, we used a few logos as pedestrians. We took some photos and then labeled each image.

Moreover, the hardware gadget comprised of Raspberry Pi 3 Model B+ kit with 2.5A Power Supply. This is the brain of our Car. This latest model of Raspberry Pi features a 1.4Ghz 64-bit Quad-Core processor, dual-band Wi-Fi, Bluetooth, 4 USB ports, and an HDMI port, 64 GB micro SD Card this is where your Raspberry Pi’s operating system and all of our software will be stored, car, 18650 batteries and compatible chargers, Google Edge TPU USB Accelerator, Set of Miniature Traffic Signs and a few Lego figurines, 170 degrees Wide Angle USB Camera.  Moreover, we will use google routing algorithm to ensure that our autonomous vehicles do not travel through high-density areas and do not get stuck in traffic jams, left turns or U-turns.

Benefits of the Project

1. Reduction of accidents

2. Releasing of driver time and business opportunities 

3.  More space and less congestion in the cities in fine

4. New potential market opportunities 

5. In the military autonomous vehicle can be used to reach dangerous and remote areas

6. It can also be used in the warehouse to carry goods from one place to another

Technical Details of Final Deliverable

Our autonomous car is capable of avoiding obstacles, pedestrians on the path and also follow the basic traffic rules without any human intervention. To accomplish this, our driverless car has an artificial intelligence system that senses its surroundings, processes the visual data to determine how to avoid collisions, and operates car machinery, like steering and brake, etc. Furthermore, we will use a route optimization algorithm to choose the shortest path if there are multiple paths and avoid traffic jams.

Final Deliverable of the Project

Hardware System

Core Industry

Transportation

Other Industries

Others

Core Technology

Artificial Intelligence(AI)

Other Technologies

Others

Sustainable Development Goals

Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com