Ventilation Assistant

During the course of the project, I decided not to use window sensors, as these would have to be attached to the window, whereas the DHT11 module is placed in the room. The aim was to create a simple, intelligent system that would be suitable for everyday use. Instead of measuring the window status directly, the success of ventilation is assessed by changes in room humidity. The decisive factor is not the pure duration, but how strongly and quickly the humidity decreases. This resulted in an adaptive ventilation logic. At the start, the start time and humidity are stored, and at the end, the final humidity and duration are stored. The user specifies whether the window was fogged up when it was closed, which serves as a human sensor for condensation and supplements the data. The system calculates an effect factor from all values, which adjusts the next recommended duration. Outdoor temperature and humidity from Open-Meteo improve the estimate under realistic conditions. The web interface shows status, estimated duration, and history of past ventilation, including details on humidity, fogging, and influence. Push notifications via ntfy.sh warn of high humidity or optimal closing time. A password-protected admin area allows adjustments such as target humidity, intervals, and sensor calibration. Overall, the project evolved from a simple display to a learning assistant without additional hardware that adapts to actual conditions.

Today, I implemented AJAX into my project. Before I did that, I had to refresh the site manually when I wanted more recent data. Now the two variabels submit new data automatically every 5 seconds, which is then displayed on the webserver without refreshing. I also added a status “bar”, which shows if I have to open my window because of a too high or low humidity.

I included the two needed libraries, then created two variables for temperature and humidity. I also implemented the .html-code for the webserver.