As the title hints at though, the past month has been a great example of the vast gap between a prototype (as seen in my last YouTube video) and a production ready product.
At first I thought doing all of the alarm clock stuff (with Google Calendar integration) would be easier because I was now doing it on Android, with Java and all of Androids high level features. Boy was I wrong.
On Android apps are meant to be ephemeral. Meaning that if you want your app to be a permanent fixture on the system, you really have to fight the system. I've ironed out most of these bugs over the past month, but it's been a long process because I can only do one real world test a day.
I've got it down to the Alarm working very reliably (so far). But the user facing Activity seems to want to die in the middle of the night some times, still working this bug. On the hardware side of things, the Bluetooth connection has some durability issues. Not every night, but often enough, the IOIO disconnects and won't reconnect without a power cycle. Still working this issue.
And lastly, I've determined that the sensor it's self, can't accurately measure a whole Queen sized bed. So I've ordered the parts and will use two sensors, one for each side of the bed. This will give me the added benefit of being able to figure out how many people are actually in bed. Not sure what I will do with this info yet, but more info is always better.
Once I fix the last remaining software problem, and add in some spit and polish, I'm thinking about releasing the app stand alone (sensor parts disabled in options). It will help me with wider testing, and I think the apps Google Calendar integration is a really killer feature on it's own. I'll probably release a free version with a donate micro-transaction or something.