Feasibility of a wearable that uses a the Adafruit SI1145 UV

[lizny]

I did test with Flora, and confirm the sensor behaves correctly, sending back non-zero readings. Of course I've already sewn my Gemma into my snazzy UV measuring spring hat in a fit of optimism that the code issue would work out and I don't want to have to tear it all out again. Besides, I want the Flora for something else that actually needs its greater powers. ~ponder~

[matthew k bardoe]

I decided to go ahead and make mine with the flora... I tore one out of a project I was done with, and now have a gemma and a trinket to play around with.

CAM01058.jpg (239.25 KiB) Viewed 832 times

[lizny]

Same here, running with Flora. The hat is pretty but the inside is messy since I ripped everything out and redid it. On the upside, the Flora has an on-off switch, which is handy.

You can't really see without a video, but the display alternates between showing the current UV index and the total exposure since turning the hat on. When the exposure display fills all the pixels, I've reached the 'barely perceptible 24 hours later sunburn' level and it's time to go inside. I plan to keep the hat on my desk at work for lunchtime excursions. I don't normally wear sunscreen to work, so having the hat to keep track of my exposure is really practical.

IMG_0606.jpg (23.41 KiB) Viewed 819 times

[adafruit_support_rick]

I like it!

[lizny]

Thank you - I just wore it and it works. Am plotting version 2 with bluetooth support so I can track accumulation over time on my ipad, and change the calibration for skin type or sunscreen without having to reprogram the hat :o)

[matthew k bardoe]

@lizny

Looks great. I would love to see some of your code somewhere. I was wondering how you computed the time exposures. It looks beautiful as well. Nicely done.

Matt

[lizny]

I'll probably put it on github when I've polished it a bit more for public consumption, it's pretty raw right now.
The calculation is simple - it is

exposure = UVindex * 3 / 7 * seconds;
cumulative += exposure;

You can add an additional scale factor for skin type which I haven't done yet. Seconds is the number of seconds between readings. I have that to account for the variable time the animation takes depending on how many pixels are lit. Burn threshold is cumulativve > 3600.

There is a lot of useful information here:

http://www.davisnet.com/product_documen ... adings.pdf

[adafruit2]

These are great - We'll also try to figure out why the sensor working with gemma - didn't realize it would be so popular ! :)

[lizny]

Using Gemma drops nearly twenty bucks from the cost of the project - I'd make hats for both my kids.

[hogmartin]

Brilliant (NPI) project, Liz! I really want to make one now for my mom to wear while she's gardening. Two questions:

1. Does it have to be a hat? By that I mean, does it require direct sunlight to accurately measure the UV index? If I were to build it into an armband (like those workout mp3 players e.g.), and it was shaded by her body for some of the time while she was working, would it read inaccurately low? Or is the ambient UV light still enough to provide a safe reading?
2. Has anyone gotten this working with the Trinket? I suspect it would have the same issues as the Gemma, being ATTiny85-based, but I don't know for sure.

[lizny]

Thank you :)

The sensor datasheet says the measurements are most accurate when the light is angled straight on or nearabouts -- it gives specific angles to try to stay in. and yes, if the sensor is in her shadow but she's in the sun, it will not be accurate. It doesn't have to be a hat - and really your mom should have a gardening hat - no? I always wear one or regret it if I don't. A necklace would work with the sensor at the back of the neck where the sun hits you when you are bent over, and the display in the front.

I've tried the trinket and had the same results as Gemma. I'm sure ladyada will work out the problem, it is beyond my ken.