This project is designed to:
- Reduce number of calls made to nurses.
- Facilitate Patient interactions with its environment.
- Enhance nurse response for emergency situations.
The system is divided into 3:
- Triggers - the defined triggers (e.g. fall detection or sound detection) are monitored by camera and microphone.
- Target devices - Devices that are controlled by the system (e.g. homeAssitant, LAN, Telegram chat).
- WakeUp system - acts as a controller to map the triggers to the target device. Mapping is controlled from MFC GUI.
- audio_classification - classifies the audio (snaps, knocks etc), which is used to trigger the smart devices.
- beepy - This is a variant of the original beepy library beepy-v1 that is used to warn the patient that a trigger is currently reading the environment.
- fall_detection - tracks the upper body angle, and used to detect a fall, which is used to trigger emergency situation and send alert to nurse.
- fall_detection_standing - tracks the full body, and used to detect a fall, which is used to trigger emergency situation and send alert to nurse. Orignal code from here
- morse_vision - uses blink detection to convert it into morse code, which is used to trigger the smart devices.
- morse_audio - uses the amplitude of a voice to convert it into morse code.
- telegram_bot - sends the alert to a telegram channel via a bot.
- wakeonlan - wakes up the target device with the wake on lan protocol.
- python_blinking - detects the eyes blinking, which is used to trigger smart devices.
- wakeup_server - the controller, that maps the triggers to the target devices.
- whisper - real time speech transcription, to trigger emergency and non-emergency situtations.
- zeromq - custom module to update on the run the triggers from the wakeup server with zeroMQ.
