Our main aim with this app is to send notifications to people when their pool is full, when ocean/sea water level start rising at an alarming rate, or when there seems to be a threat of a flash flood from a nearby river. People today waste so much water that in the coming 50 years all the water bodies would either dry up or become contaminated. This inspired us to take a step forward and try to give our children a better future, where they too can enjoy the luxury of water.

What it does

GlassHalfEmpty is an app that sends you an SMS alert when your bottle/glass is full. It also tells you the percentage of contamination of your water, if it gets dangerously high. The best part is that the use of this app is not only restricted to small glasses, bottles or cups.

How we built it

We trained our Clarifai model to detect different cases of water level and contamination. A server then receives the live video feed from the user and gives back continuous data values. These values are then interpreted and an SMS alert is sent when appropriate.

Challenges we ran into

It was hard to train our Clarifai Machine Learning Model since we were not familiar with machine learning concepts.

Accomplishments that we're proud of

We are proud that in the end, we were able to train the Clarifai Model and implement the Clarifai API. We also successfully called the Twilio SMS API.

What we learned

We learnt how to give a livestream video to a server, how to train AI models, and call the Twilio SMS API.

What's next for Glass-Half-Empty

Next we are thinking of outsourcing the model so that more data points from different sources are available. This enables more waterbodies to be accessed and also improve the accuracy of the data points.

Share this project: