OutRun – Open-source, privacy oriented, outdoor fitness tracker
github.com
github.com
There are so many devices, each have their own algorithm for sleep tracking and the device I pick largely depends on the accuracy of the device's sleep tracking.
However there are new apps, like Nukkuaa, an (EU only) sleep analysis app that uses any heart-rate tracker (like a bluetooth fitness chest strap) and infers sleep quality from the data.
A generic app expands what devices are available to me. Right now I use CGM tape and stick a Fitbit Charge 3 on my tricep - but I dislike the Fitbit app and if I change to another brand I cannot export my logs.
A third party app has the potential to train against more data than any individual single brand could and, with appropriate tagging, could possible offer better accuracy. Open source would be icing on the cake.
Additionally, it would be great if devices like the Fitbit Charge could be used as a bluetooth heart rate monitor that can be used on apps that consume trackers on a (presumably) standard API (like TrainerRoad, Zwift, etc).
The new charge 6 does have this capability. [0]
At such a price (more than my Apple Watch) I expect all the analysis to be done on the phone.
But small specialized shops like Oura can't fund constant R&D on hardware dev and sleep tracking algorithms, infrastructure, plus continuous SW patches & updates to existing customers and products, without subscriptions, because they lack Apple's other revenue streams and muscle to negociate steep discounts on HW and cloud infra, so the one-time sale fee of Oura's HW product is not enough to cover all those on-going dev costs and keep the company afloat.
Not an Oura customer or shareholder, just though I'd share some light on the costs and challenges of competing in this space. It just doesn't work without continuous subscription based revenue streams. Low margins on HW sales are not enough to fund such businesses.
Software app subs are one thing, but hardware is something else, IMO.
> research-grade sensors
Are these people really this shallow? What does that even mean
So, to calibrate a sleep tracking device, you have a person wear the device, while also doing the sleep study. You do this a bunch of times. You train some ML models to try and make the outputs from the sensor data, after processing, the same as the study data.
After some degree of accuracy you declare success.
Now, does it work? In broad strokes, yes. You can (easily!!) see the effect of alcohol on sleep quality. If you have a crap night vs a good night, sure, a wrist based consumer device can figure that out.
Actual details? Eh. I wouldn't trust the devices for anything but directional data.
The more sensors devices get, the better than ML model can be trained.
Now it has been awhile since I last worked on this stuff (I actually just sat next to the people doing the work), so maybe there is some revolutionary new technique out there, but if not, it is still ML models trying to correlate things and match them up to what a bunch of fancier sensors said during studies.
And even then, they can not really measure your sleep. If you want to measure sleep, you will need to have wires on your scalp that measure what happens inside your brain.
--- For primarily younger white males
But still pretty good seeing as they infer this largely from heart rate and movement.
What else works well with gadgetbridge?
Any ios equivalent?
That’s not ideal, but I’m definitely not the target market as I’m happy to buy one of the higher end Garmin models.
[1] https://www.gamesindustry.biz/take-two-reportedly-in-tradema...
Parsing GPS data is surprisingly simple.
It surprised me to learn that different apps will “smooth” out measures like elevation and distance — it’s actually quite a rich problem space with a few interesting solutions.
> I wouldn't necessarily say abandoned. I still work on it from time to time, but progress is very very slow and I cannot prioritise it over other things atm
Also, it doesn't appear to be getting a lot of activity, which suggest it's perhaps not a going concern.
https://github.com/alex-hhh/ActivityLog2/wiki/Just-The-Scree...
Incidentally, I'd love to have an open source, privacy-respecting nutrition tracker that works on open source desktops. (I find my laptop much, much faster than phone for recording this data, at least in the ad hoc way I'm doing it, including things like weights of every ingredient that goes into a dish.)
The most complete nutrient data for basic foods seems to be http://www.ncc.umn.edu/food-and-nutrient-database/ which you need to pay for, so I wonder how an open source implementation could work.
I use Cronometer daily which shows you the datasource for each food you add, and if it's not NCCDB (the one at the link above) then chances are it's not going to have much info. Yet that extra info is exactly why I'm using a nutrient tracker rather than just a calorie tracker.
In that realm, there is also GoldenCheetah:
- https://www.goldencheetah.org/ - https://github.com/GoldenCheetah/GoldenCheetah
(When I saw "Open-source, privacy oriented, outdoor fitness tracker", I thought it would be available for an open source platform, so I clicked. It being for a closed, all-your-data-are-belong-to-us platform wasn't expected. Maybe the comment saved some clicks.)
They own your device 100%. Of course they can access your data...