Using the data set from this project, supplemented by data from Sony Mobile—which linked smartphone data with tracking data from fitness armbands—DTU researchers have shown that you can monitor a person’s sleep patterns very precisely, yet in a simple, non-invasive and fully automated way, by registering whether their smartphone screen is switched on or off. This could pave the way for large-scale studies of sleeping habits etc.
The graphs show sleep patterns for two people over a month and based on the time of day. The red fields indicate that the phone’s display is switched on.
The model interprets the blue area as the person’s sleep. The person represented in the top figure has a fairly regular daily rhythm. They wake at around 5:30 am, except for a few days—presumably holidays or weekends.
The light blue sections indicate that the model is uncertain as to whether the person is sleeping, as they are not following the usual patterns.
The person represented in the bottom figure uses their phone more erratically. The model therefore assumes that their sleep rhythm is also erratic and that they check their phone at night. This person also uses their phone much more than the person in the top figure.