Jawbone's company blog covered its UP activity and sleep trackers, with posts analyzing user sleep data.
HN links it for the Napa earthquake sleep data and circadian rhythm posts; one commenter says once-a-minute heart rate capture is of limited use.
Recently, I planned to use a JS interactive chart library in my blog posts instead of prerendrting the graphics using R/ggplot2, but I decided against it. I realized something: what's the point of an interactive chart on a blog post, which uses only static, unchanging data?
For example, look at the interactive chart used in Jawbone's Earthquake blog post: https://jawbone.com/blog/napa-earthquake-effect-on-sleep/
Does it gain any benefit from being interactive as opposed to just using an image? Yes, you can mouse over each point for exact times/percentages, but that should already be implied by properly-defined chart axes, and it doesn't help enhance the meaning of the chart (e.g. the spike)
Additionally, from my own testing, each charting library has its own quirks in both construction and mobile usage (e.g. if chart implements a zoom feature, you can't scroll it on mobile) so every chart I make would require additional QA to make sure it works as expected, and it would slow down blog post creation to a halt.
If I was creating an application with dynamic, unpredictable data, then interactive charts would be valuable, but otherwise interactive charts with normal data seems like a gimmick.
Also, one very important caveat with SVG chart libraries: you can't easily convert the chart to an image, with makes going viral harder. (Canvas-libraies are not affected, though. Highcharts has an export module too, albeit it's not very convenient.)
Mind dump of thoughts/sites:
http://www.fitbit.com/
http://nikeplus.nike.com/plus/what_is_fuel/
https://jawbone.com/up
Lots of hardware devices and some software integrations, but need to think outside the box.
www.dietbet.com/ - Gamification of working out is a great idea/starting point, but what about taking it further. Choosing what you eat can be more important than working out.
A food/photo sharing/tracking software of pictures of everything you eat and possibly automated/encouragement via mechanicalturk based on does it look healthy would be cool. Also could align with friends/groups who are trying to eat healthy - people could comment and say WTF/Seriously... or better yet, ask wheres the photo or have you not eatten?
Some other directions would be more in line with a chrome plugin that goes off ever 30min for you to do a pushup or something.
Just some thoughts.
Looks like Fatigue Science makes a device called the READIBAND [1], which is used to track sleeping patterns so you can adjust your schedule accordingly, but right now it is only available to military, sports, and big industrial clients. Sounds pretty interesting, in that sports teams are seeing a real benefit.
Can someone who uses one of these performance monitors weight in? Doesn't have to be the READIBAND, but what about Fitbit Flex [2], Jawbone Up [3], or the Nike Fuelband [4]. Does anyone actually notice a marked difference in alertness throughout the day?
[1] http://fatiguescience.com/readiband-2/
[2] http://www.fitbit.com/
[3] https://jawbone.com/up
[4] http://www.nike.com/ca/en_ca/c/nikeplus-fuelband
Their privacy statement is short and clear:
https://jawbone.com/legal/privacy
~1400 words.
I liked this section:
When you use the Jawbone Companion app we connect with and upload to our servers the address book and calendar on your device along with other data that we normally collect through our mobile apps described in the app usage data section below.
I wonder how calendars are used.
Capturing HR once a minute is of really limited use. More than benchmarking against resting HR, the interesting thing is to look at heart rate variability, and to do that you need to scan several times per second.
For that using purpose-built devices is just better.
I am thinking:
https://jawbone.com/store/buy/up3
or
http://www.getqardio.com/qardiocore-wearable-ecg-ekg-monitor-iphone/
One thing to note is that while Jawbone has a nice API[1] and summary data export, there isn't a way to access device/raw data.
There does appear to be some reversing of undocumented API calls that provide some better resolution, although I don't know how up to date that is: http://eric-blue.com/projects/up-api/
[1] https://jawbone.com/up/developer/endpoints