19 comments

[ 3.0 ms ] story [ 50.4 ms ] thread
Looking at the screenshots, it seems that you place your phone at one corner of your bed as you sleep. I am surprised that this would work. Does anyone here have an opinion? There are quite a few critical reviews that say it doesn't.
Yes, you place the phone on the corner of your mattress. So when you move in bed, it can detect that because your mattress shakes. And how much you move in bed is related to your sleep cycle (it's called Actigraphy). I tried it last night and i remember i woke up twice and checked the time. The graph produced by the app matches these times. The alarm simply went off at the set time, but it says that the app needs 2 nights to calibrate to your bed and your sleep cycle, so i can't say how well that works.
I use SleepCycle. Can't tell you if it is a sham or not-- it seems to work and seems to reflect my subjective assessments of my nights sleep
I've tried it for a week or so; didn't work for me. The accelerometer did pick up on my movements, though.

Also, I think putting my iPhone at the corner of my bed is a little risky. When sleepily reaching for your phone in the corner of the bed it's easy to drop it, scratch it against the wall, etc. Even when you're generally a careful person; it's going to wake you up 300 days a year, and it has to go wrong only once.

I'm hoping sleepmate does a better job.

> There are quite a few critical reviews that say it doesn't.

Can you show the reviews URL for the rest of us?

I'm curious if anyone knows how this type of solution would work if there is more than one person sleeping in your bed.

Wouldn't two people in the same bed be pretty likely to not have synchronized sleep cycles? Or would they be synchronized? Anyone know about this?

From the FAQ:

"How well does Sleep Cycle work with 2 people sharing a bed?

Our experience is that it works very well. Normally your body will "block" the movement of the other person and the iPhone will not detect that movement. You can test this yourself by using the test function in Sleep Cycle."

After about two weeks of usage: It seems to work very well, at least in the sense that my movements are registered correctly. I would estimate that it causes a noticeable difference waking up about 1/3 of the time.
I'm using this application for a few weeks now. The application needs to be calibrated for the first few days, but after that it works really well. I'm pleasantly surprised by the accuracy of the application and the ease of use. The low price ($0.99) makes it easy to try it out too, so if you care about these kind of things, I can recommend it.

One thing I've noticed is that in your sleep, you're sometimes rolling over while being "half awake": you wouldn't remember it if you would fall back asleep, but if you're inside the specific timeframe you've set your alarm to, this application will wake you up, and you indeed do feel more fresh than coming out of a deep sleep.

And the charts are nice too.

How long before mattresses with the built-in accelerometers start producing? I actually don't move that much, so for it'd be safer bet to have the accelerometers in many places.
Smart Mattresses™
It doesn't work with Tempura mattresses though, which I have. I haven't felt so cheated in a long while ...
I'm guessing, since I sleep in the same bed as my wife and my daughter's new puppy, that this would have no chance of working with me.
Aww, I was hoping you strap it to your head
What happens if I get an e-mail or sms during the night? Will the vibrator cause the app to think I am awake?
Just have a look at their FAQ to see how badly apple's closed development model hurts applications all over the place. For this application to really work, there's a huge list of precautions you have to take and most of them are based on the fact that third-party apps don't enjoy the same API as the apple based ones do.

The app doesn't work if you lock the phone, the app doesn't support the iPod Touch, the phone needs to be connected to power because there's no concept of different (hardware) sleep states.

And of course, the app would not wake you up if you were to get a phone call or SMS during the night.

Most of these limitations would go away on open hardware with an open SDK.

How did they figure out how to make this, e.g., what movements constitute different sleep states? Are there research papers on this?

I wonder if you could use changes in body temperature instead? (Obviously not using an iPhone though.)

There are some papers about it. Look at the discussion for the wakemate, a month ago or so.