Would Heat Training distort AI FTP Detection?

TrainerRoad Software
1

Greetings from Germany,

I’m on a fresh, kind of lowish volume plan since January with 3 harder and 1 endurance session per Week. 3 Weeks on, 1 off.
Base 1 done. Base 2 done. Currently on a rest week.

From now & then I try to rotate in a further endurance ride, run or train now workout.

After the first month, I got my expected 4 Watt FTP bump, which was consistent or kind of in line for me over the last years, if I did my prescribed workouts (peaked 273 last fall).
Yes I cut a little bit of the rest week after Base 1, but nothing too bad.
Profile is public, if somebody wants to look it up & I saw the podcast, talking about underestimating restweeks. I will honor Restweek 2 more.

The second Base Block (of which I will attach a screenshot), went pretty smooth as well.
I finished all prescribed workouts. Rotated in a walk & added additional endurance workouts in 2 of the 3 weeks.
I only had some minor quirks, like a sweespot ride I had to mark as harder then usual, because of suboptimal nutrition after a vo2 max day.

However after this block, my ftp dropped about 2 Watts.
Surely I wasnt getting slower from that block.
I mean, I even still feel stronger than 4 weeks ago.

Maybe no big deal and maybe even correct!?, but here I realized the following:

The last workout i did was heat training related (heat ramp test), which I did for the very first time and want to incorporate regularly / purposeful from now on.

I didn’t know until now, that I could remove workouts from AI FTP detection and adaptive training. So my bad, if it had an impact.
Only after accepting the drop I then realized, that this training approach could possibly affect AI FTP detection!?

So sorry for the long story, but here are my actual questions and hopefully takeaways for everyone else:

Will Heat Training mess up AI FTP because it (for example) gets distorted power to heart rate data?

If so, should I therefore allways manually take these sessions out of the equation, by ticking off AI FTP and Adaptive Training?

If it may have had an impact on my AI FTP, could I please trigger a new detection?

How would you ensure, to tell adaptive training the truth about how stressfull (in terms of TSS for example) the heat seassion was? Because TSS will obviously be too low.

Thanks to all and have a nice day.

Base2
Base21920×939 144 KB

1 Like
2

Hey @Memento! Welcome to the TrainerRoad forum!

Great work!! :flexed_biceps:

Glad you have come to this decision. For the best results, follow the Plan as closely as you can- after all, it has been designed for you!

To address your questions:

Will Heat Training Affect AI FTP Detection?

Heat training could influence AI FTP Detection if it significantly affects your power data. Since training in heat increases cardiovascular strain, your heart rate will often be elevated relative to power output. If this results in lower-than-usual power numbers, it could slightly skew AI FTP Detection.

Should You Exclude Heat Training from AI FTP Detection & Adaptive Training?

If your power is lower than expected due to heat stress, it’s a good idea to manually exclude those workouts from AI FTP Detection and Adaptive Training, as you did.

If you want to do heat training, a good strategy is to incorporate heat training during endurance rides instead of high-intensity workouts. That way, your key sessions—the ones that drive fitness adaptations—aren’t negatively impacted by heat stress.

Can You Trigger a New AI FTP Detection?

AI FTP Detection is only available every 28 days, so a new detection can’t be triggered manually before then. However, you can ensure you’re training at the right level by continuing with your workouts and accurately filling out the post-workout surveys. If your workouts feel easier than expected, marking them as such in the post workout surveys will allow Progression Levels to adjust, ensuring that your training remains appropriately challenging. That way, you will continue to progress and will be prompted to with a new AI FTP detection result in due time.

For more details on how AI FTP Detection and Progression Levels work together, check out this article:
:link: Why is AI FTP Detection Only Available Every 28 Days?

Key Takeaway: Your FTP is just one measure of your fitness, but Progression Levels track day-to-day improvements in your ability to express that FTP across different training zones. This ensures that your workouts remain appropriately structured, even between AI FTP Detections.

Final Thoughts

:white_check_mark: If heat stress impacts your power, exclude those workouts from AI FTP Detection.
:white_check_mark: Consider keeping heat training to endurance sessions so it doesn’t affect your high-intensity work.
:white_check_mark: AI FTP Detection happens every 28 days, but Progression Levels ensure your training stays on track in between detections.

Hope this helps. Let us know how you get on and if you have any other questions for us :slight_smile: .

1 Like
3

I’m not certain this is the best approach…almost all of the training advice I have seen re: heat training has been to do it as Z2 training, at the most.

1 Like
4

Hi,

First of all thank you for the warm welcome & your detailed answer.
That pretty much clears everything up for me.

I will exclude my heat sessions from AT and AI FTP from now on.

I think there also may have been some inaccuracies in my wording regarding what test protocol I completed and what I plan to do with it.

Of course I will spend my heat training in low zone 2.
Most likely at around 45 - 50 % of FTP at the most.
And of course these will be additional sessions. Driven by core body temperature zones only.
Not any of the prescribed trainerroad workouts for that week.

@Power13
Thats the plan.

I think there’s a misunderstanding regarding the heat ramp test.
It’s not a protocol to determine any kind of power, like your typical ramp test.
It was a test protocol to ramp up your core body temperature, to determine specific heat zones, as described for example by zwift insider:

Thanks to the both of you. :slightly_smiling_face:

3 Likes