Japanese researchers spin up toilet paper gyroscopes for science Japanese boffins have measured the spin-speed of toilet rolls to work out who's on the loo. The idea, apparently, is that because a throne is a good place to collect health monitoring data, the researchers wanted a reasonably non-invasive ID technique. One that wouldn't get bogged down in the obvious privacy issues of trying …
This post has been deleted by its author
Assuming this can be made into a reliable algorithm that works with more than half a dozen people to choose from... exactly what do they plan to infer from this "health monitoring data"?
Personally, I'm pretty sure the amount of paper I use varies quite a bit between visits. According, y'know.
exactly what do they plan to infer from this "health monitoring data"?
Identify the 3 x 10 sheet-at-a-time wipers that keep clogging up the toilet? As they already have a Bluetooth device there they could also ID the BT signature of their phone and log how much time they spend dumping core (sorry Linux users :) ). I bet Sports Direct would pay for that sort of data.
Just a few ideas.
Management everywhere are currently waiting with bated breath for the results of a larger study. Once that's in, they can mandate how much paper you're allowed to use and how often you can use the facilities. And it will all be done in the name of "efficiency" to weed out the malingerers. Oh.. and more profit from the savings on paper and employee time wasted on the porcelain throne.
Never in a million years would I have thought you could identify people this way, but it makes sense. Bravo to these researchers.
I have no idea what use personally identifying people on the toilet is, but given that this is Japan, maybe they are developing a toilet that will check the 'contents' as you flush to see if you have colon cancer, kidney disease or other conditions that may be detectable via your #1 and #2.
I guess if it figures out you have something wrong with you, the next time you snag some toilet paper a voice will come out of the toilet "please check with your doctor, you may have colon cancer!" That surprise toilet voice would be good for people suffering from constipation!
I've seen those Lovely Ledges, most disturbing.
Also, obligatory Simpsons.
Then folding toward oneself puts the dimples on textured paper to the outside, and who wouldn't prefer dimples on their bum to pimples.
Practically, dimples also catch and hold more "material".
Also if the installer is a clutz (not looking in any mirrors), a roll with the tail to the front will snag on protruding screw heads when pulled with any vigour.
"Then, the system identifies the user with the feature values.”
followed by:
"And the authors say it's privacy-protecting, ..."
What??
Have they made sure it can deal with negative rotation values, for those people who pull too hard and then have to rewind it (or like Lae Ming above)? If not, then the software could have some kind of dump and an overflow problem so things could get wiped.
"In the laboratory environment, someone whose habit was to pull on the roll four times would be correctly identified 92.5 percent of the time."
But surely, the amount of paper used is a variable? (Maybe not for #1s, but certainly for #2s.)
Also, there would be a lot of other parameters to be taken into consideration. This week's menu in the cafeteria. Make of the toilet paper, especially when procurement keeps changing it. Office parties. Hayfever season(s). Holidays associated with typical foods. And so on. We're deep inside non-linear territory here.
And then some clever bastard adds Picolax to the canteen food (This link is not safe for work)
Yup, confirmation that I'm reading elReg again ...
So, an internet connected bog roll gyroscope - totally secure of course.
Will they be providing a couple of power outlets so I can bring in my smart toaster and fridge freezer with me to the karzy so that they can all discuss the results and alter my diet for me?
Seriously, considering the amount of smart rubbish many idiots seem to carry about with them the opportunities for data mining and mischief must be endless.
FAIL.
Good study, wrong conclusion. The identification accuracy is uselessly bad.
From my own experience, my TP use varies widely. Both in dispense speed and use. Sure, it's easily distinguished from at least one other in the household who uses about a roll per day. But I expect a bi-modal use distribution (a few squares vs a fistfull) without the ability to distinguish between members of the group.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
