trachhub

raspberry pi based trachsense hub for monitoring respiratory health.

setup

first time setup

you will need git, python >= 3.9, and uv installed.

1. install system dependencies: these are required to build numeric and bluetooth packages on 32-bit arm.

   sudo apt-get update
   sudo apt-get install -y libatlas-base-dev libopenblas-dev libdbus-1-dev build-essential libffi-dev libssl-dev

2. clone and sync:

   git clone https://github.com/TrachSenseADC/trachhub.git 
   cd trachhub 
   uv sync

3. database setup: the setup.sh script handles the full installation of postgresql 13 and timescaledb. for a manual install, refer to the script logic.

quick start (using setup script)

for a fully automated installation on a fresh pi:

curl -sSL https://raw.githubusercontent.com/TrachSenseADC/trachhub/main/setup.sh | bash

running the app

if already configured, simply run:

cd trachhub
uv run app.py

to enable the internal virtual emulator for developer testing:

uv run app.py --debug

if you encounter issues with architecture compatibility on the pi, try a clean sync:

rm -rf .venv
uv sync --no-cache

using the app

once the app runs, it will print the local server URL. open that in your browser.

1. skip wifi: you can choose to skip the wifi setup and move directly to trachsense connection.
2. connect: find "trachsense" in the list of bluetooth devices (use ctrl+f if needed).
3. data stream: you might see a visual error message saying "failed to connect"—you can ignore this for now if values start printing in the terminal.
4. visualization: once connected, you will see live values and a real-time plot.

data storage

readings are stored in diff.csv. if the server fails to write this file on a pi, you can use the bt_check.py script as a fallback to verify hardware communication.

to see a complete stationary graph of your recorded data:

uv run plot.py

if you still encounter issues, please let Yashas (ybhat@umd.edu) know.