Would definitely be an interesting longitudinal study to put a sensor in a set of classrooms, some of them with a your light system to suggest opening a window,

Would definitely be an interesting longitudinal study to put a sensor in a set of classrooms, some of them with a your light system to suggest opening a window, and some of them not being told about the sensor, which would just log data as a control. You could compare grades and long term success in people in those classes. Even in the short term it would be interesting to see if it affects any education metrics.