Let’s say I’ve got a great new, low cost device that can log my blinking rate - I’m not sure how this could correlate to anything right now - but what 3rd party services should I look to post my time stamped data log so that I can make it available to others to help me work on this? which are best standards to use (e.g. time - Unix time, UTC etc…)
You might be interested in looking at blink frequency and tiredness and stress. Not sure how the evidence is on the relationship, but the idea is out there anecdotally. Would love to hear more about this device.
Afraid I can’t answer your main question though.
Is google docs an option for posting the data logs?
This is a good question. I just tweeted it out to see if we could attract some answers from the broader community.
Hi Max. I saw Gary’s tweet on this and thought I’d take a stab at it.
I’m working on a project called BodyTrack which is trying, among other things, to create and/or leverage existing tools to do what you are talking about. Standards that we are using and encouraging in the community are 64-bit floating point unixtime for timestamps and JSON for text-based data exchange. We are building a pretty cool architecture and hope to soon start publishing API specs, but we’re not quite available for use yet.
Other efforts that are synergistic and may be closer to being usable for this application include:
[list]
[]Eric Blue’s TRAQS.me project
[]Pachube
[]Sensor.network
[]Locker Project
[/list]
Feel free to contact me if you want further details. I haven’t yet used these other systems, but I can at least answer questions about my current understanding of how they fit into the ecosystem.
Anne
Hi Max!
I’m getting ready to launch a new site, Giraph, that helps users keep track of anything they like. It sounds like your needs would be met well by our API. Our API documentation is not currently online, but if you’d like to get in touch with me personally I’d love to help you.
Our API is very simple to use. We accept UNIX timestamps or any other format allowed by PHP: http://www.php.net/manual/en/datetime.formats.php. Generally I would recommend UNIX timestamps, however, since they do away with any potential time-zone problems. Data is submitted to the API using JSON, which is a very simple standard to learn and work with.
Best of luck with your project!
Tom Weingarten
http://giraph.com
Are you kidding? I would kill for a blink rate sensor and thought of building one if I had time for hacking hardware!
One obvious but low hanging fruit is…dry eye/computer vision syndrome. Sign me up for a beta!
I haven’t tried it, but I think I remember someone claiming that you can pick up eye EMG signals in the raw data feed from the Zeo. You just need to put together a cable for the serial port in the back and use their python raw data library if you’re interested to try it. No hardware hacking required!
Anne
I’m intrigued. I probably won’t ever find time to do this, but if I did, I’d make a small cell phone vibrating motor buzz in the headband when my blink rate dropped too much; ie if my last blink was more than X seconds ago. I think you could retrain your eyes to blink more this way relatively easily.
[quote]I’m intrigued. I probably won’t ever find time to do this, but if I did, I’d make a small cell phone vibrating motor buzz in the headband when my blink rate dropped too much; ie if my last blink was more than X seconds ago. I think you could retrain your eyes to blink more this way relatively easily.
[/quote]I think it would be even nicer if it would be hard ware but software.
A webcam should be able to tell when you blink. It could blank the screen if you don’t blink enough.
Blink-rate is currently used for drowsy detection in cars (along with several other variables). The in-car cameras are fairly low resolution, comparable to a webcam (e.g., I think Bosch uses a 1MP camera).
In addition to dry-eye prevention, I would love to have a notification that, after detecting a low blink-rate, asks how alert I feel and based upon my response, suggests that I do one of a number of alertness-promoting activities: a game of ping pong, a cup of coffee, or running a new idea by a colleague. After my activity, having returned to my computer, I might be prompted to enter both my chosen activity and my post-activity self-reported alertness level - this could easily be consolidated into a single-click.
Over time, such a system could make better suggestions, moving them to the top of the suggestion-list in the notification. For example, coffee might be found to work best just after lunch but ping pong could be better in the late afternoon.
*Of course, vigilance or working memory tests could be substituted for self-report.
I developed a blink rate sensor which also gives a notification every time you don’t blink often enough.
My goal was to help people who spend a lot of time in front of a computer screen and have problems with dry eyes, eyes strain and computer vision syndrome.
Anyone interested in a beta version?
I would love to get feedback from the QS community, while you can enjoy a healthier blinking rate already
Peter