Skip to content

A personalized AI driving assistant designed to help new drivers build confidence and improve driving skills through real-time feedback, adaptive learning, and guidance on safe driving practices

Notifications You must be signed in to change notification settings

memidhun/Roady

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

36 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

readme-banner

Roady🎯

Basic Details

Team Name: SineWave

Team Members

  • Team Lead: Sidharth V Menon - Saintgits College Engineering
  • Member 2: Midhun Mathew - Saintgits College Engineering
  • Member 3: Abin Joe Francis - Saintgits College Engineering

Project Description

Roady is an innovative project designed to assist learner drivers in honing their driving skills. Developed by the team at SineWave, Roady leverages advanced technologies to provide real-time feedback and personalized training modules. The project aims to make the learning process more engaging and effective, ensuring that new drivers gain confidence and competence behind the wheel. Whether it's mastering parallel parking or understanding road signs, Roady is your go-to companion for a safer and smarter driving experience. 🚗💨

The Problem (that doesn't exist)

The average driver does not have a reliable good driver that can help them with their driving. There are an increasing number of accidents on the road and these risk the lives of people on and off the road(I know it sounds useful but who listens to anyone).

The Solution (that nobody asked for)

We have made a Machine Learning model that can monitor the driving based on the steering input and tell the driver whether their turns were tight enough.

Technical Details

Technologies/Components Used

For Software:

  • Python,C
  • Arduino IDE, Git, VS Code
  • Tensorflow, scikit-learn, motor-driver.h
  • chatGPT, Figma,

For Hardware:

  • Arduino UNO, SG90 Servo, Motor Driver Sheild, WebCam
  • Arduino IDE, PySerial, Soldering Iron, Foam boards, Hot Glue

Implementation

For Software:

Installation

For the model we have used Python 3.11 and as such we are using Tensorflow 12.1.1 and the corresponding numpy version. Arduino IDE version 2.33 with the arduino uno motor driver sheild library is required. The software interface is entirely designed in Figma and is implemented using VS Code.

Project Documentation

For Software: UI

This shows all the UI elements designed for this project. It shows the way the UI reacts to various scenarios and what the model output and the UI output for the various cases

For Hardware:

Schematic & Circuit

image

We are using a L293D motor driver sheild to power the motors controlling our rover, one side is connected to one set of wheels and the other to another set of wheels. The polarity on the motors provided throught he sheild controls the direction of motion and the direction of movement of the rover.

Build Photos

Build side profile Build Front Facing

Project Demo

Video

Rover1-ezgif com-video-to-gif-converter (1) The video demo demonstrates the movement of a remote control rover we built as part of the project for demoing the working of the project and this shows the camera feed we pulled and the working of thr rover and its various indication lights.

Additional Demos

Predictor This is footage of a model we trained to predict steerin angles and identify sloppy turnings and to check if proper steering inputs are being provided to the vehicle.

Team Contributions

  • Sidharth V Menon: Model creation, debugging, Git management
  • Midhun Mathew: Arduino programming, circuitry, web page development
  • Abin Joe Francis: Dash and interface design and web page design

Made with ❤️ at TinkerHub Useless Projects

Static Badge Static Badge

About

A personalized AI driving assistant designed to help new drivers build confidence and improve driving skills through real-time feedback, adaptive learning, and guidance on safe driving practices

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C++ 72.6%
  • Python 27.4%