1. Get the rumbling input
2. Clean the signal (use of a low pass or a band pass filter)
3. Understand if the user might feel hungry.
4. Maybe have other metrics like time for understanding the likelihood of the user being hungry
5. Dataset for understanding what rumbling actually feels like

Currently, we have a digital stethoscope for data acquisition. Planning on using an Arduino Nano BLE 33. It’s essential to differentiate from other noise (even conversations that the participant might have during the time of wearing the device)

How complex should the system be?

• TinyML approach
• Baseline - Bandpass filter approach
• Have time for context? Wake words as cues?

What setting are we using the system in?

• Continuous?
• Before/after food?

How can we use it on a day-to-day basis?

• Button monitor on the pant buckle?
• Adjustable belt