medibot.
a 3D-printed patient-monitoring device that matches every reading to the right patient by face.
medibot is a portable patient-monitoring station: a 3D-printed case with a Raspberry Pi 4 and an Arduino inside, built for remote monitoring where a full bedside setup isn't practical.
it takes four measurements: SpO₂, heart rate, blood glucose, and non-contact body temperature. every reading is stored in a database, and a small camera runs facial recognition so each set of readings is filed under the right patient automatically. a Python app shows the live measurements and each patient's history.
how it's put together
- sensors SpO₂ and heart rate, blood glucose, and a non-contact temperature sensor, read by the Arduino.
- brain Raspberry Pi 4 running the database, the camera pipeline, and the app; the Arduino handles sensor capture.
- camera facial recognition ties each reading session to the right patient record, no manual entry.
- app Python interface showing live readings and per-patient history from the database.
- enclosure 3D-printed case holding the electronics, sensors, and camera as one portable unit.
