Same here. Missing one workout for next Tuesday
That workout does show up in TP.
I started my subscription in the last 48 hours
Same here. Missing one workout for next Tuesday
That workout does show up in TP.
I started my subscription in the last 48 hours
It’s likely related to the free trial restriction as above. Will remove those tomorrow and also add a ‘refresh export’ button to the account page. I’ve reached out for manual fix in the mean time through the messages (need your email)
Markus, to confirm:
Does this apply to all Garmin watches, or just the ones that have overnight HRV logging?
I have a Venu 3. I do see RMSSD already, but I assume Garmin sends that to intervals. I don’t see SDNN.
SDNN is only used by Apple Watch. Or rather it’s the only thing that Apple makes available and is what’s stored in the hrvSDNN field in Intervals.icu API. Everyone else uses RMSSD
Free trial export restrictions are now removed and there’s a “Refresh Exports” button in Apps page
That did resolve the issue for me. Thanks!
Does AIEndurance allow for cadence-based exercises? Especially with submaximal strength exercises (low rpm)
We are currently cadence agnostic in our workout prescription meaning you can choose whatever cadence you like
Hello @Markus_Rummel_AIE ,
I am currently testing your AI and would like to share some observations. I am primarily a runner, averaging about 5 hours of running per week, and I do some cycling as well, between 0 and 3 hours depending on the weather conditions.
When I identify myself as a runner, the AI does not take my cycling sessions into account and does not suggest any cycling workouts. Conversely, if I classify myself as a triathlete without swimming, it requires that I have many days with double activities, which does not fit my profile.
Is there a plan to accommodate runners who occasionally cycle?
My goal is to participate in a 70 km trail at the end of June, and I would like to have a training plan that incorporates cycling sessions starting in March/April.
Another point I would like to address: I mentioned that I want to run a 10 km in February, aiming for a time of 44 minutes.
My last 10 km was on October 13, where I achieved a time of 46 minutes. However, the AI predicts a time of 49 minutes for my upcoming race.
I haven’t explored all the features yet, but it seems promising.
Thank you for your attention!
EDIT:
When analyzing the calendar, it did not record my 10KM race on October 13th, there is only the warm-up :
On Coros, my race is present:
When I try to manually upload the Coros .FIT file of my 10K, I get this error :
The AI did not record my virtual bike rides.
Is it better to connect only the watch app (Coros in my case) to get the most raw data, or to connect both Coros and Intervals? Or maybe just Intervals? Knowing that the Coros data is uploaded to Intervals.
I record my indoor rides with my Garmin head unit.
It allows for in-activity HRV recording, which is used by AIE to calculate various metrics.
It does require some setup on the head unit to enable HRV logging. It also requires that you connect AIE and Garmin directly.
Best to connect both Coros and Intervals - we automatically handle/avoid duplicates.
Triathlete with zero swimming hours would be the way to go. The hours you set for running and cycling per week are what the system tries to accommodate so maybe too many double days because of that? Also you can fine tune the plan via drag and drop after.
Update on Integration - Import from Intervals.icu now complete:
We added 2 b) yesterday after Intervals.icu made webhooks available - thanks David! - and all seems to be working well. Let me know in case there’s any issues though!
You can connect multiple platforms at the same time, we automatically try to avoid duplicates and if they do happen clean duplicates every 24h.
HI @Markus_Rummel_AIE
I continue to see issues with planned workouts not being sent from AIE to intervals.
Here’s a current example:
AIE shows a workout for next Tuesday and Thursday:
intervals only shows the Thursday ride, but omits the Tuesday ride:
Thanks!
I changed the update of the export slightly so hopefully should b fixed now. Let me know if the issue persists though please.
@Markus_Rummel - I am sorry for inconveniencing you here:
Three questions:
I am sure I will come up with more questions - so keen to subscribe if these initial issues are ironed out.
With my very best wishes
Haha yeah I have PTSD from doctors.org.uk email addresses because my responses to emails always bounce - had a few over the years where I wrote a long response to a question (including yours today) and it never got through. Seems like they just block everyone
for 2: I see 71 runs and 217 rides for your account now (import can take a while sometimes to complete). I just triggered an import from Intervals.icu again in case something’s still missing just to be sure. Also it doesn’t matter if the activities have (bad) HRV data, they still get imported and should show in the calendar no matter the HRV data situation. Once you get that zip file upload it to Intervals, they have a functionality to import the entire file I’m pretty sure, and then disconnect/reconnect to AI Endurance and all data from the last two years still missing should make it’s way in.
for 3: yes once you rename them the model will get automatically re-trained within the next 24-48 hours.
I must have missed the part where the workout name has an impact on the model.
Do I need to follow a certain naming convention? My past “race” workouts are primarily 200, 300, 400, and 600k. And next summer I plan on doing a 1550k. Should I include that in every event name past and future?
As of this week you can actually see the weight of each datapoint in the past data chart in the tooltip and via the line thickness, see attached screenshot
Update: You can now
Apologize if some of this is covered in the posts above.
Overall I’m quite happy and am likely to subscribe to AI Endurance.
I think it dovetails nicely with intervals.icu. Just picking one thing, AI Endurance’ own calendar doesn’t total and summarize weekly load. So to figure out where you might be getting an off week or particularly loaded week takes some work. But intervals.icu does that perfectly.
I have a few quibbles with the AI Endurance’ planning, but one is the one I find nearly universal with automated planning system. TrainerRoad is perhaps the most notorious for it. You get hit immediately with huge load. For context I’m at the lowest load portion of off-season, just now starting to think about structured training. Been months since I did a ride longer than 2 hours or more than about 5-6 hours per week (which the system should know because I uploaded history). With my A event set for June, I put 14 hours max and set the slider for meeting available hours to zero. I was hoping I’d get eased back into training, and letting the system know I was OK with 14 hours as I reach peak load months from now. But no, I get immediately hit with 12+ hours and 3-4 hour rides on the weekend. That’s doubles my load and hours from the past few months right in week one. Dang! I had to game the system by creating a new plan with much lower total and weekday hours so I didn’t get brutalized in December. I’ll re-adjust it later when I’m ready for those hours.
Again, overall still happy, and likely subscribing.