My own inference from this discussion is that it’s likely Apple ecosystem devices' ability to infer sleep stages improved considerably by using data from a device (the watch) physically connected to the body that can detect movement, O2 saturation (patent protection be damned), pulse rate, etc.

In fact, I haven’t a CLUE how the iPhone, sitting on its charger, could make ANY inferences about sleep, which in clinical sleep medicine typically requires actual MEASUREMENTS of extra ocular muscle activity, EEG monitoring, and body movement. Indeed, one might suspect that if someone recently deceased but interred with his iPhone in a vest pocket might be recorded by the phone as “sleeping,” at least until the battery in the phone discharged.

Does anyone here have information about what actually happened? The only information I can see the phone collecting would be motion, but that would mean the deceased person being transported back to his home town would be recorded as AWAKE because of phone GPS data recording.

The appropriate question, of course, is just what Apple DID use to infer sleep status from iPhone owners who don’t want another computer sitting on their wrists.

(answering PART of my own questions, better late than never)

My curiosity whetted, I did do a simple web search and discovered that the sensors used on the phone include the microphone (listening to the user breathe), the accelerometer (recording changes in linear motion) and the gyroscope (recording rotational motion). Apparently having the iPhone on your mattress can allow it to make some inferences regarding motion while it listens to its owner’s breathing, but it seems OBVIOUS to me that such inferences are FAR more likely to be accurate when recorded by a device (the watch) that’s physically attached to the body. I’ve not yet discovered how it “knows” that my eyes are looking around in my sleep, however.

--?
Jim Robertson