Last Updated on January 9, 2025 by pg@petergamma.org
Who uses Home Assistant for physiolgical sensors? – Home Assistant is for house sensors.
- When Retö Röllin from Switzerland developed a Raspberry Pi which sends ANT+ to MQTT:
It took not long until someone connected this code to Home Assistant. To Home Assistant? But Home Assistant is for house sensors. Why to connect ANT+ sports sensors to Home Assistant? Because HA has a large community?
If we look at the Home Assistant integrations, there are many for sports sensors and smartwatches. Smartwatches are only of limited use for doing physiology. But they are affordable, and eventually a good starting point to get to know difficult multi-sensor platforms such as home assistant.
Also Udo Berndt synchronized the Apple watch to Home Assistant and streamed the sensor data into InfuxDB and visualized that data in Grafana:
OpenBCI is a physiological multi-sensor platform with up to 16 channels. Peter Gamma from www.petergamma.org tried hard to find a solution to expand OpenBCI and add more channel, but did not find a convincing solution.
According to supporters in the OpenBCI forum, channels, slow channels with for instance temperature values can be added over BrainFlow and Python.
- But we only know of an example which connects OpenBCI over the Bluetooth dongle to ESP32 to Firebase.
- Home Assistant and Firebase is based on InfluxDB.
- We found there is a pathway which goes from OpenBCI Cyton over the Bluetooth dongle to ESP32 to the Firebase Real-time Database (RTDB) which is rather complicated:
https://ocw.cs.pub.ro/courses/iothings/proiecte/2022/open_bci_esp32
- The community of Home Assistant users seems to be much larger than the community for physiological multi-sensor devices.
- If we look at more advanced Home Assistant projects as for instance those of Adrian Spiess from Switzerland:
https://www.youtube.com/@AndreasSpiess/videos
- We can see that those advanced software projects often consists of many sub-projects.
- Some of those use Sqlite, others InfluxDB, etc.
- It seems as if these projects have grown naturally like some plants in garden.
Peter Gamma from www.petergamma.org does not want to develop physiological multi-sensor devices for the rest of his life, which can compete with those from g.tec, BIOPAC, etc. But he believes that it will become easier to build such devices, and those will become more affordable.
An interesting project is for instance to add more physiological sensors to OpenBCI. Some physiological signals are slow, as for instance respiration. And is there not a way to add such channels without the need of InfluxDB?
To work with InfluxDB is difficult. How to read, write and delete data in InfluxDB with Python is shown here:
How much easier is it to have an SQLite database and read and write values with Python?
But also the more complicated platforms become easier to handle, the more examples which are available, and the large the community for those platforms is.
Home Assistant seems not to be very different to other platforms such as Firebase Real-time Database (RTDB), Microsoft Power BI, Azure Cloud etc.
And if we have some examples as for instance the Apple watch which is already integrated into Home Assistant, is it not relatively easy to modify these examples to a software platform which is more suitable for physiology?