Arduino x Kway project

This project is part of the Arduino x Kway challenge. The Nicla Sense ME board is a board that has several sensors such as pressure , temperature , humidity , gases. But in my opinion the most surprising sensor is the Self-learning AI smart sensor. With that in mind , I decided to make an App to be a kind of assistant that shows this information and helps the user in the detection of toxic gases.

Blue Universe

What is Blue Universe?

Blue Universe is the App I made to connect the board with the cell phone. To make the App, I used MIT App Inventor.

First , I made a splash screen :

In this screen , I just added a button to move to the next screen, the button code is this :

Now let's see the main screen:

On this screen I installed the BLE extension for MIT App Inventor that you can download from this link: https://iot.appinventor.mit.edu/#/bluetoothle/bluetoothleintro

I installed the Notifier component as well:

If you want , you can change the organization of the components , but mine was like this:

And this is how it is when you press the IAQ chart button:

The Code

I'll put the App download in the downloadable files in .apk and .aia format too , but I'll explain how it works here too.

First , the app search for bluetooth connections:

Initialize the global characteristics according to the Arduino IDE Code:

While bluetooth is connected, the app receives updates from the board:

Display in the text boxes the value of each characteristic:

This part configures what each button will do:

E

dge Im

pulse

Unfortunately my board burned out at the very end when I was doing the test part, so because of that I couldn't finish all the tests and the integration with the app and Edge Impulse. I was only able to test two gases (Butane and Chlorine), but my plans were to test more gases for a better application. So, because of this , the code part only has the functions to see weather conditions such as pressure, temperature, altitude, CO2, air quality, steps and humidity. But even so, I'm going to put the link of my public project of my application on Edge Impulse , and I'll put the .zip for the library of the deployment on the downloadable files as well.

https://studio.edgeimpulse.com/public/182727/latest

EXPANDED TECHNICAL DETAILS

Neuro-Specialized Machine Learning Assistant

This project uses Edge Impulse and the Blue Universe ecosystem to create a smart assistant for people with specialized neurological needs.

• Edge Impulse Motion Classification: The Arduino captures data from a 6-axis IMU. The firmware runs an on-device neural network that can recognize specific hand gestures or tremor patterns with high confidence.
• Blue Universe App Handshake: Connects to the MIT App Inventor dashboard. When the AI identifies a specific gesture (e.g., an SOS signal), it triggers the smartphone to place a call or send a location alert.

Optimization

• Quantized AI Inference: Uses a quantized version of the ML model to ensure the entire neural network can run on the Arduino's limited SRAM while maintaining sub-100ms reaction times.