Hi all

I am thinking of developing an application like the SleepCycler tool for the iPhone. (See http://www.lexwarelabs.com/sleepcycle/).

Some of the tasks are quite time consuming, because of that I would need the help of you.

Some thoughts:

1. Log the movement during your sleep
2. Develop algorithms to interpret the data
3. interpret it, figure out in which sleep phase you are
4. Develop reliable alarm functions
5. Make the app attractive with nice GUI, icons, publishing

Specially the algorithms and interpretation will be very tricky.
First of all, I am just logging my movement during the sleep. Then I will try to interpret it. It will be much easier as more data there is available. If you are willing to help, please let me know by replying to this thread.
I already have a basic app that logs the movement into a file. And I try to interpret it with excel. You could submit your data anonyme. All you would have to do is to start the app and put the phone on your bed like in the above link shown. It then will generate a log file which you would have to send me.

Are you interested in an app like that?
As a comunity, I think we should be able to develop it!

Note: I would be willing to pay more than the 1$ to get an app like that, IF it would be available. How ever as I beliefe in OpenSource, all my apps will be for free and under the GPL.
If you read the FAQ on the above listed webside, you wil lsee that there are several restrictions due Apples policy. Thats wy I have a N900 with Maemo instead an iPhone...

i would pay for an app that gives me good stocksquotes that's easy to used and apps for telling if my flight is on time or what's the next flight out of town.

@caco3: there is already a python one called sleeppy, see

http://talk.maemo.org/showthread.php?t=50985 for thread and
http://maemo.org/packages/view/sleeppy/ for the package.

maybe you can cooperate with the author to improve the app?

Blues

Wow, thank you for your tip.
Even I was searching for a while on google and this forum, I did not find anything about Sleeppy.

Saves me a lot of work
As it is also written in Python, I might be able to help on it myself

You may be interested in the S60-Application Happy Wakeup
http://www.happywakeup.com/en/How_it_works.htm

The reason why I mention it is that they use audio data from the mic rather than acceleromater sensor data (as sleeppy or the iphone apps do).

i don't know which gives the more accurate resulsts... but i just don't want to put my N900 *in* my bed. I could live with having it beside my bed or on the floor under my bed or whatever... but i wouldn't sleep well knowing each movement could throw the n900 to the floor.

The most accurate REM-phase indicator should come from eye(lids) movements. Deep sleep indicator should be the (lack of) body movements. I don't know about audio cues. It seems like a crude workaround for not having accelerometer input.

Cant see this ever working when i share a bed with the mrs!

@cheesy's member: might want to ask for support for a preliminary shagging-phase that goes by unrecorded by the system

No, this is the need for a new application: The Humpometer!

You'll be able to clock yourself (and your company), get energy graphs and with some sound input you'll know how much fun you had.