Self-driving Car Using Machine Learning and Neural Networks

This self-driving car uses AI, Machine Learning & Neural Networks to drive autonomously on a track. It uses Nvidia Jetson Nano as the main onboard computer for performing all the AI, Machine Learning & Neural Networks tasks.

Setup

To get started with this project, follow these steps

1. Order parts from the list of materials below

   • 1x Nvidia jetson nano 4gb version
   • 1x Micro SD Card 64gb (buy class 10 only)
   • 1x Waveshare IMX219-160 Camera, 160c FOV camera (170° FOV will work as well)
   • 1x Wltoys A969 RC Car 1:18 Scale 2.4GHz RTR 4WD Short Course Truck (you can use any rc car/scale)
   • 1x 17gram servo for steering if using Wltoys A969 RC Car
   • 1x 20A Brushed Electronic Speed Controller or brushless depending on your car
   • 1x Universal 7 amp BEC (5.0 volt) or you can use dc to dc step down converter (min 4Aamp)
   • 1x Jetson nano acrylic case
   • 1x 30cm camera flex cable
   • TP-Link USB AC600 600 Mbps WiFi Wireless Network Adapter for Desktop PC with 2.4GHz/5GHz High Gain Dual Band 5dBi Antenna Wi-Fi dongle
   • Female to Female jumper cables
   • 1x PCA9685 PWM Servo Motor Driver
   • 1x 7.4v 5000mah li-ion battery or 11.1v 3000mah 30c lipo battery depending on your car and speed controller
   • 1x 4mm arylic sheet for mounting all the parts and making the camera mount arm
   • 4X M3 40mm standoff
   • 1x M3 8mm screw pack
   • 20x M3 nuts
   • 20x M3 10mm or 8mm standoff
   • A track to run your car on (we printed our track on a flex banner 18x10 feet)

2. Follow the hardware setup

   • Hardware Setup.

3. Follow the Software Setup.

4. Run through the Running the software.