Charge & Go: EV Ride Rent Predictor
Teledemics-powered EV rental pricing with KMeans risk clustering and XGBoost regression.

About the Project
Charge & Go is a mini project that predicts EV ride rental pricing from driver behavior and telematics signals. It uses risk profiling to classify trips into Low, Medium, and High risk, then applies dynamic multipliers to estimate the final price.

The app is built around a machine learning workflow and a Streamlit dashboard. It is designed to be simple enough for a fresher to understand, but complete enough to demonstrate clustering, regression, and pricing logic.

Features
Risk classification using KMeans clustering.

Price prediction using XGBoost regression.

Dynamic pricing based on risk, time delay, and distance overage.

Streamlit dashboard for interactive input and output.

Glassmorphism-style UI and live simulation elements.

Dataset
The project uses the ADAS-EV Dataset from Kaggle. It contains telematics features such as speed, acceleration, braking intensity, obstacle distance, and lane deviation, which are used to estimate risk and pricing.

Tech Stack
Python.

Streamlit.

Pandas.

NumPy.

scikit-learn.

XGBoost.

Matplotlib.

Seaborn.

How It Works
Load the EV telematics dataset.

Preprocess and normalize the feature values.

Use KMeans to group trips into risk levels.

Train XGBoost to predict the price multiplier.

Show the result in the dashboard with a final quote.

Project Structure
text
charge-go/
├── app.py
├── config.py
├── data_loader.py
├── model.py
├── ui_tabs.py
├── utils.py
├── data/
│   └── ADAS_EV_Dataset.csv
└── README.md

Installation

bash
git clone https://github.com/hemanthv358-svg/Mini-Project
cd charge-go
pip install -r requirements.txt

Run the Project
bash
streamlit run app.py

Sample Pricing Logic
Low Risk → 1.1x.
Medium Risk → 1.5x.
High Risk → 1.9x.

Extra time and extra distance increase the final amount through additional penalties.

Model Performance
According to the report content, the XGBoost model achieved R² = 0.923 and MAE = 0.118, while KMeans clustering achieved a silhouette score of 0.672.

Future Scope
Real-time GPS and OBD-II integration.

Database support for bookings.

Payment gateway integration.

Mobile app version.

Explainable AI for price transparency.

Why This Project Matters
This project shows how telematics data can improve EV rental pricing and help make pricing fairer for safe drivers while charging more accurately for risky behavior.

Author
Hemanth V
M.Tech Data Science
JSS Science and Technology University, Mysuru