KeDo

• 

  WaterTable Prediction

• 

  Yield Prediction Features using Google Cloud AutoML Service

• 

  Chat Screen

• 

  SQL Store Table for Chemical Compostion on CloudSQL

• 

  IoT Dashboard Screen

• 

  Auto Order Screen

• 

  Marketplace Screen

Inspiration

The woes of farmers to predict good crop on the farm at suitable conditions under stressed climatic conditions inspired me for this project

What it does

This solution takes geolocation data and pulls out information of soil profile, water profile and weather profile from past datasets then plugs the data with reported yield achieved for different crops in past years, then ranks the best yield reported for those conditions. Apart from predictions, This solution is also common marketplace for different farmer needs and also provides endpoints for IoT devices to be connected.

How I built it

I built prediction system using AutoML of google cloud platform while also leveraging cloudSQL service to store all collected data and store retrieved data of soil profiles, weather and water profiles. The existing data is pulled by image on Heroku with a .tech domain hosted on .tech platform

Challenges I ran into

I ran into problem of training it, But nailed it to great extent. HTML and PHP integration with Google service helped to resolve frontend and backend integrations easily.

Accomplishments that I'm proud of

The way I have approached this problem from a different perspective of algorithmic approach and providing the solution based various different data factors

What I learned

I learned the ease of solution development using Cloud platform.

What's next for KeDo

Testing the achieved predictions with exact experimental validation which provide valid data markers to correct the predicitions and provide higher confidence rates.

Built With

• automl
• bootstrap
• cloudsql
• css
• gcp
• html
• javascript
• php
• xampp