A lot of data is collected from crop production sites and we thought it wise to use this data to make informed decision like regulation of pest and disease control based on the images taken from the site.
What it does
Alert on infestation of pest and disease, low moisture content, high temperatures, and humidity value. Based on this the foreman can make informed and timely decisions.
How we built it
Build a simple Android things IoT solution, deployed to azure via IoT Hub next we used stream analytics to sink the incoming payload from the device to blob storage and Power Bi for visualization. Once the data is added to the blob storage an azure function is triggered which takes the incoming data send it to our azure machine learning model (Published from Azure Ml Studio as a web api) then based on the severity level output from the ML Model the azure function sends an SMS notification to the foreman.
Challenges we ran into
Getting enough data to train the machine learning model Inaccuracy in the temperature sensor due to the heating effect from the NXP board
Accomplishments that we're proud of
It works! It was interesting to share all this knowledge with our fellow students at Multimedia University