Slashdot Mirror
Researcher Hacks Nine Sleep-Tracking Devices To Test Their Accuracy (brown.edu)
A determined researcher at Brown University extracted "the previously irretrievable sleep tracking data from the Hello Sense, from the Microsoft Band, and nine other popular devices," according to an anonymous reader, "by decompiling the apps and using man-in-the-middle attacks." Then they compared each device's data to that from a research-standard actigraph. Their results?
The Fitbit Alta seems to be the most accurate among the other nine in terms of sleep versus awake data... Our findings tell that these consumer-level sleep reports should be taken with a grain of salt, but regardless we're happy to see more and more people investing in improving their sleep.
I was asleep!
“He’s not deformed, he’s just drunk!”
At work. In the nap room.
Is there a kinect or similar motion sensing based applicaton for this? That's kind of the obvious way to do it instead of accelerometers on the wrist when what you really want to track is the torso - especially if people already have a kinect.
"Computer hacking" is a federal felony. Nowhere in the law does it say what that is so anything self-described as "hacking" counts. It also means that if a reporter says you've been "hacking" that's a felony accusation, no matter what you did. Something to think about.
You can thank the computer security industry for this. With their ethical hats.
Detecting REM sleep with a worn device or app is virtually impossible without polysomnography, a test that records your brain waves, blood oxygen levels, heart rate, breathing, as well as eye and leg movements during the study.
If we consider only the results from the AMI MotionLogger as the “gold standard” for our accuracy study,
Why is the AMI MotionLogger considered the "Gold Standard"? Just because it's outrageously priced?
the results from my experiment should still be taken with a grain of salt.
It's a nice undergrad experiment to learn about research methods to be used for the day when research for peer review will be done.
If you have trouble sleeping, change your habits
DON'T STAY UP LATE. i'm usually in bed a little after 10 and no later than 10:30pm
go get some exercise. being exhausted from running will do wonders to help you sleep
Don't watch TV before bed. No TV in bedroom
Stay away from electronics before bed including gaming
don't drink alcohol before bed
don't eat too close to sleep
keep your bedroom cool around 50 degrees or less
At the bottom he disclaims a few points that mean this was a pointless exercise with too many variables and not enough controls.
It probably violates tons and tons of laws. Reverse engineering, decompiling, any attempt to even understand how the process works is illegal or is believed to be illegal by law enforcement. And adding terms like man in the middle etc does not help.
sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
Article Title: Researcher hacks nine sleep-tracking devices to test their accuracy"
First sentence of summary, right under the title: A determined researcher at Brown University extracted "the previously irretrievable sleep tracking data from the Hello Sense, from the Microsoft Band, and nine other popular devices.
Note that the first says "9 devices", the second says "1+1+9 devices".
I know it's asking a lot for TFS to match TFA, but is it at least possible to match Title and TFS?
"I do not agree with what you say, but I will defend to the death your right to say it"
the Fitbit Alta seems to be the most accurate among the other 9 in terms of sleep versus awake data.
The Fitbit Alta seemed to detect asleep vs awake almost identically to the other devices. Most of the other devices could do far more and a few differed in their results.
It's like using a fitness tracker and saying well the Fitbit was the best at identifying I'm not yet dead. Whoop de do.
Anyway. $139 for the Fitbit Alta, or $3.99 for the Sleep app on the phone. Both seem to give the same Asleep vs awake results but at least the latter can tell you how asleep you are.
Scary that this is coming from a researcher.. They mention that polysomnography is the defacto standard for sleep measurement and then state that the best way to run this experiment is to use all 10 devices at the same time? How about you use each device for 1-2 nights at the same time as the polysomnography and then you can compare the accuracy of each device against an industry standard? As if you are going to strap 10 devices on and around you and have any kind of normal sleep activity..
I've had two sleep studies.
One when I was first diagnosed was on their premises, properly wired up to all kinds of gear and when it was over I received a multi-page report with charts and graphs that clearly showed the data.
The second was done by 'Snap diagnostics' at home. The gear came loose, fell off repeatedly and I spent most of the night without it attached. They then took four months to tell my doctor to increase my pressure.
No data, no report, no evidence that the test even worked.
I still don't think they had any.
Apple Watch not tested - fail.
I can only comment on the Xiaomi Mi Band 2 since it's the only device that tracks sleep that I've had. The times for going to sleep and waking up are pretty accurate. Of course I can't say anything about the validity of the "deep sleep" time.
Since most of these devices seem to use similar hardware I don't think there can be huge differences on them
Way to go, Chicken Little. I always figured you for a feathery, thoughtcrime propaganda stooge.
Data extraction methods
Most of these techniques are more akin to screen scraping than decompiling or reverse engineering.