Hi, i Just found the AI endurance App for karoo 3 in the AppStore. I hoped for a live data of dfa in Training. Is it possible to implement? Thx for ur great work, Michael.
The idea was that the extension only collects the HRV data and then sends that data together with the power data to AI Endurance and DFA a1 as well as readiness and thresholds are all calculated in the cloud in order to not drain the battery of the Karoo.
I’m curious myself though how much it would actually drain the battery if we calculated it live/on the device. Live DFA a1 could also be optional and the user could enable it, consciously accepting the trade off for battery live. I’ll run some experiments how bad it drains the battery on my Karoo soon
If alphaHrV can do it on a garmin watch with relatively modest battery burn I’m sure the Karoo can handle it with 2GHz.
#easyformetosay
Will see how it goes… 
New version (1.0.0) of our Karoo extension is out with live DFA alpha 1. I ran a few tests and it didn’t affect battery life at all so it’s even enabled by default.
There’s new fields:
- “AIE DFA A1” and “AIE RF” (single number fields)
- “AIE DFA A1 GRAPH” (full graph of the last 20 min a1 vs power - should be used as a sole field on one screen)
This release changed our app signing key, so a normal in-app update will fail with an error. To update: uninstall the extension, then reinstall it from the Karoo app store. Afterward, re-pair your Bluetooth chest strap and reconnect your AI Endurance account. This one-time step is only needed for 1.0.0; future updates will install normally.
Wow, what a great Work, will try it tomorrow. Thx 
Hi, i did a Testride today, works very well in comparison to the AlphaHRV. But after 15min of riding at moderate endurance Power the karoo live dfa starts to drift from 1.4 to 0.8 for aprox 4 minutes at same power while the garmin stays stable at 1.4. For the rest of the 2 hour ride both was equal in a 0.1 range. Is there a kind of learning or calibration in this time? Thx, Michael.
Thanks for riding with both Karoo and Garmin! Karoo extension is using the same algorithm as AI Endurance for post ride processing and I know there are some differences to the alphaHRV algorithm (alphaHRV has to be more creative to work with Garmin underpowered hardware)
I suspect the drop happened due to higher artifact count in the beginning of the ride (sometimes there’s not enough sweat yet on the strap, always helps to wet it a bit with water at the electrodes before the ride) and alphaHRV algorithm handled it better in this case.
Glad it worked well!
Hello, today it worked flawless. Yesterday i can see 3% artifact in the Garminlog for 3 minutes, at the time of the drop.
Greetz, Michael.