FallyHelpBot

Logo of the project
Final Hardware assembly
Example of text when fall is triggered
Hardware Schematics

Inspiration

To provide an affordable solution that monitors the safety of the elderly.

What it does

It uses Machine Learning to detect a falling motion from the user which will trigger the Telegram bot to send a notification to the designated care person of the user.

How we built it

We build the hardware using Bluno Beetle, IMU Sensor, and recycled watchband. It used Bluetooth protocol and send the IMU data to a central computer. The computer uses the data for Machine Learning to detect the fall motion. We also implement a Telegram Bot that will automatically send a notification to the care person that was selected.

Challenges we ran into

We faced a slight inaccuracy in detecting the motion through machine learning initially and required a few training models to improve the accuracy

Accomplishments that we're proud of

We are able to do a meaningful hack and implement a working prototype at the end of this short hackathon period using our previous knowledge. We were also able to use Telegram Bot programming which is something new for our group as well.

What we learned

We learned more about Machine Learning and its application while utilizing hardware components that we were previously familiar with. We learn a new type of programming in the form of Telegram Bot as well.

What's next for FallyHelpBot

We can have better hardware that is not exposed to interference and is more water-proof. We can use better protocol for communication to enable longer distance detection.