Last Updated on January 22, 2023 by pg@petergamma.org
Since several years, we are looking for an easy way to visualize and analyse physiological sensor data from multi-sensor devices, but failed to this date. We reviewed LSL and MQTT pathways. But according to our own little experience, neither LSL nor MQTT is necessary for our project. We have OpenBCI which is a multi-sensor device with up to 16 channels and 3 additional AUX inputs. But what to do, if we want to add more sensors, which are not EEG, ECG and MEG sensors, which do not fit to the OpenBCI architecture? More and more sensors can be attached to Home Assistant, which is a multi-sensor platform as well. BalenaHealth is not a multi-sensor platform, but many sports watches are. We have introduced a pathway from balenaHealth to Home Assistant and InfluxDB for sports sensors:
But is this the easiest path? A good starting point to put data into a database are .csv files. We looked at these instructions from InfluxDB before, and they where always too complicated for us as non-coders. Then we found the instructions from Paul Warren Eaton from Finland how to inject .csv files into InfluxDB:
These instructions are easy and well explained, and documented in several YouTube videos. If we start a new multi-sensor project, it is worth to have a closer look at these instructions, if we know that we will choose InfluxDB sooner or later anyway.
The other relatively easy instructions for InfluxDB we found are those from computer engineering student Rahul Banerjee who likes playing around with different libraries and APIs:
https://thenewstack.io/getting-started-with-python-and-influxdb/
If we will choose Python anyway sooner or later, and also InfluxDB, it is worth to study these instructions. And if we deal with sensors which offer code examples how to read sensor data, is it not to check if we can connect those to Rahul Banerjees code examples?
We can start with Scott Hardens sound card ECG, use his Windows software which shows the heart rate in real-time, and store it to .csv. We can use an OpenBCI 3-lead ECG and OpenBCI GUI, store data in .csv and analyse it with HeartPy. As a next step we can use SQlite to store our sensor data. There are many simple examples available on YouTube how to do this. But before we start scripting and code our own multi-sensor Adinstruments LabChart software, is it not worth to study Rahuls InfluxDB instructions, before we start coding? But let us do it as Paul Warren Eaton from Finland did it, more step by step and in more detail.