A Platform for Indoor Air-Quality (CO2) Sensing to combat SARS-CoV2

Dear All

I’d like to announce an initiative in the Berlin TTN community to help operators of public spaces in properly ventilating them, and to inform their patrons about the air quality.

The non-profit initiative called Clair Berlin (as in CLean AIR), you can find detailed information on our website (sorry, currently in German only). We are launching our test operations today, and would like to scale throughout the city of Berlin if funding permits.

We welcome contributions to the code base, as well as other regions around the world starting to set up their own platforms. Currently, we are four individuals doing the development and management work in our spare time and out of our own pockets; therefore, we also welcome financial backing to help the initiative scale.

Technical Infos

  • We use the Berlin TTN for sensor data communication to the Clair Platform
  • The Clair Platform is a custom software stack orchestrated via docker swarm. The entire stack, including our infrastructure as code, is available on GitHub as open source.
  • We use commercially available LoRaWAN CO2 sensor nodes during the initial test operations. In particular, the [Elsys ERS CO2 LIte](https://www.elsys.se/en/ers-co2-lite/]. However, the platform is architected such that it can connect other sensor models as well with minor modifications.
  • We also developed a DIY CO2 sensor on the basis of the Sensirion SCD30 sensor module and an Adafruit Feather M0 LoRa Board. Our prototype Arduino Software plus information on how to build your own CO2 sensor is available in a separate repository, and we added an instructions page on the Clair Berlin website.

We hope to provide a valuable service to the TTN community and are happy to receive comments and feedback!

Happy sensing



Don’t have you problems with the power usage for the ndir sensors?

Have you considered e-co² versions?

Our sensor is just a prototype to better understand the constraints of the network. In the field, we use commercially available products like the Elsys ERS CO2-Lite (which gets several years of battery life out of the Senseair LP8 NDIR sensor).

We just added a description how to build our DIY sensor on the project website.

Really a great project and sweet naming.

Due to the high demand i observed a significant higher price for SCD30 Sensors during the past weeks.

A simple local meassuremnt rating (trafficlight / Ampel) would be a real improvement.
But you stated this already on your project page.

Really a great project!

Yes CO2 concentration can be an indicator for an aerosol concentration! The simpler way is to take the number of people and the amount of air in the room and then simply determine the time to ventilate.


This is definitively another option. I would’t call it “simpler”, though. You need to know the number of people (a Pax-counter can help, which is probably more complex than a CO2-sensor), you need to know the size of the room, and you need to take into account the activity level of the attendance. Singing produces about 30 times more aerosols than being quiet. In this regard, measuring CO2 levels is less indirect.

Hi Uli,
we have a product which is designed for ventilation assistance related to SarsCov2. Have a look at https://www.corona-assist.de/air
We provide complete packages for those that dont have the technical expertise to install/operate their own hardware/infrastructure. But we are not limited to this - as long as we get an MQTT-subscription to the sensor-data we can use any sensor/infrastructure available. So maybe a talk/messaging about possible cooperation makes sense - just let me know.
Regards to Berlin from Wiesbaden