Cars are spewing out up to 306,000 extra tons of carbon dioxide per year because their tires are under-inflated. These emissions directly contribute to increased greenhouse gases in the atmosphere. Reducing pollution by keeping ones tires at the proper pressure is a crucial step towards sustaining the fuel economy.
Technology can facilitate a solution to the lack of Tire Pressure Monitoring Systems in India. Data-driven Machine Learning and Image Recognition approaches can be used to analyze historical data and estimate the tire pressures in our vehicles.
We present a new AI based mobile application that utilizes the mobile camera to capture, Open CV package to estimate the tire pressure and Machine Learning models to predict the percentage drop in fuel efficiency if the vehicle is driven with the current state of the tires. Our solution also recommends nearest gas stations and mechanics.
- The user navigates to the app and uploads an image file.
- Open CV processes the image and provides the tire radii measurement to Random Forest Classifier.
- The Classification model is deployed on IBM Watson Machine Learning platform.
- Flask API receives the predicted state of input tire image, and outputs the percentage drop in fuel efficiency.
- The user can also access the Google Maps API and locate the nearest gas station or mechanics.
- IBM Watson Studio - employs artificial intelligence and user-friendly tools to empower and streamline your data analytics.
- Watson Machine Learning - helps data scientists and developers accelerate AI and machine learning deployment on IBM Cloud Pak® for Data.
This project is licensed under the Apache 2 License - see the LICENSE file for details.