From sensor readings to a wearable system
This collaborative research project brought together embedded sensing, remote monitoring, and classification of movement and sleep-related signals. I contributed to a system that connected physical sensor acquisition to data preparation and analysis.
The system
A Raspberry Pi Zero W collected readings from inertial, PPG/SpO₂, and infrared temperature sensors over I2C. The processing workflow prepared sensor samples in windows for analysis, while Node-RED supported remote monitoring.
SVM and Random Forest models were used across the fall- and sleep-related classification work. The project connected three layers:
- Acquisition: obtaining motion and vital-sign readings from embedded sensors.
- Preparation: preprocessing samples and assembling labeled windows.
- Analysis: applying classification methods and interpreting the resulting outputs.
My contribution and the outcome
I contributed as part of the wearable-sensing research team. The work led to three conference publications covering the monitoring system, fall detection, and sleep tracking.
The project is a practical example of my interest in electronics and software together: a useful analysis depends on the sensor interface, acquisition workflow, and preparation of the data reaching the model. The linked papers describe the team’s methods and experimental results.
Research context
This was an experimental research prototype. Its published evaluations concern the stated sensing and classification tasks; clinical validation is outside that scope. Earlier papers list my name as Abdelrahman Fawaz.