πŸŽ‰πŸŽ‰πŸŽ‰ Introducing AI FTP Detection πŸŽ‰πŸŽ‰πŸŽ‰

Hey everyone!

AI FTP Detection is now available in Early Access. We know that testing (assessing) can be stressful, and we’ve been hard at work devising a better way. This feature replaces the need to test by using AI to analyze your training progress, previous experience, and many other factors to give you an informed and data-backed FTP.

How To Enable AI FTP Detection:

Step 1. Log into the TrainerRoad website

Step 2. Access Account then Early Access from the navigation menu

Step 3. Under AI FTP Detection, select Enabled

Step 4. Proceed to the apps to use FTP Detection

How To Use It:

Step 1

Make sure you have at least 12 TrainerRoad workouts or have imported outside rides with power data. If you don’t have 12 workouts or haven’t imported rides, don’t worry. You can still enable the feature in Early Access, and the feature will automatically become available once you complete 12 workouts.

Step 2

Once you’ve enabled Early Access for AI FTP Detection, open the latest version of the mobile or desktop app.

Step 3

When a Ramp Test is your next scheduled workout, go to the Career tab in the app. Select Use FTP Detection to be presented your new FTP.

There are two reasons that may prevent us from detecting your FTP. First, your data may still be processing. If so, close and relaunch the app after a minute. Second, your FTP has changed within the last 14 days.

Step 4

Select Accept to use the new FTP. Your Progression Levels will be updated automatically, and a workout will replace the Ramp Test on your calendar.

For more info, check out our support article.

FAQ

How is this different from FTP updates on other platforms?

Most other platforms use what we call capacitive efforts to estimate your FTP using an FTP estimation model. This means that they rely on you to complete all-out efforts within your training data.

For example, if you had a very hard 20-minute climb during a ride (ridden at ~ 105% or greater of current FTP), they might use that effort to estimate a higher FTP. Our model does not rely on capacitive efforts. Instead, our model predicts your FTP based on your training history and your personal biometrics.

Will AI FTP Detection consider unstructured outside rides as well?

Yes. AI FTP Detection takes into account all rides with power or HR data.

Do I need to use a heart rate monitor?

Nope! All you need is consistent, structured training with a power meter or VirtualPower. If you’re riding without a power meter, wear a HR monitor. We recommend you wear a HR monitor on all rides for data collection for future potential features.

Can I still complete an FTP Test?

Yes, you still have the option to complete your scheduled Ramp Test.

How do I know AI FTP Detection is accurate?

The best validation for your detected FTP is whether you can successfully complete Productive workouts after you accept the new FTP. If that is not the case, we want to know about it.

What does AI FTP Detection look at to detect my FTP? What types of efforts or workouts does AI FTP Detection consider?

AI FTP Detection looks at your training history and your biometrics to detect your FTP. Because AI FTP Detection considers so much of your training data, it’s very unlikely a single type of effort or workout would substantially sway the model. You don’t have to do capacitive or all-out efforts for the feature to work. In fact, it works if you only do sweet spot or aerobic riding!

Does this use or estimate the Ramp Test result?
AI FTP Detection’s goal is to give you a training benchmark (FTP) that gives you the most productive training. Rather than looking at a single effort, AI FTP Detection uses many factors to provide an FTP that ensures that your training is scaled to your current fitness level. You do not need any FTP tests in your training history to use this feature.

Can one β€˜standout’ ride or effort impact my detected FTP?

This is what sets AI FTP Detection apart from other FTP estimation tools. Because our model looks at your training history more holistically, an outlier ride or effort will not unduly influence the model. AI FTP Detection will consider outlier rides and efforts, but will not depend on them to detect your new FTP.

How can AI FTP Detection be accurate if I’m just returning from an offseason or more than one week off the bike?

AI FTP Detection was trained on an enormous data set and can predict decreases in FTP even after long (weeks to months) training interruptions.

I use a different power meter for my outdoor rides than for my indoor rides. How will FTP Prediction handle any differences between the two power meters?

This is one of the benefits of using a model as opposed to using an estimate. We don’t simply estimate your FTP based on outdoor power data (which is often reads higher than indoors). Instead, we include your outdoor power data in addition to the various data points we already have to predict your FTP.

167 Likes

Thanks to you and the team for all your hard work on this! HUGE NEWS.

16 Likes

Congrats to the TR team! Super excited

10 Likes

Good job team!!

8 Likes

Thank you! The whole team has been working so hard on this - we are really excited to share it with everyone! :partying_face:

74 Likes

We’ll done to the team. Everything is working but need to wait a while as I’ve recently updated with ramp test. Great tool to have though

Deleted this weeks ftp update from the ramp test and can now use the estimator. It’s higher so obviously its right! Seriously though, I reckon the estimate is more accurate than the test. I’d had poor sleep in the days leading up to it and recent Zwift race had me thinking it would around 280, ramp was 274 and estimate is 279.

10 Likes

Hey @ambermalika this is NOT a critique, but just curious for the reasoning for Goddard as the Ramp replacement workout? Is it that it should be a good feeler? And the duration is short enough for most/similar to Ramp?

(I assume it’s the same for everyone, and I remember Nate saying there was a standard replacement for now)


Love, love, love the look of the new feature. I am keen to test mine. I’m a bit daunted by the number, but in AT I will trust.

3 Likes

@ambermalika Will this only work with scheduled ramp tests? or will it show up with scheduled 8 or 20 minute assessments? Will self scheduled assessments work as well, or only those scheduled via a plan/plan builder?

2 Likes

Yep, this is hardcoded for the moment, but we already have an update for this that will recommend a custom workout as the replacement. Stay tuned as that will be out soon!

14 Likes

I don’t see an option to enable FTP estimation :sleepy:

9 Likes

Deleted Cache/cookies?

1 Like

yup, same here

At the moment, this will only work with the Ramp Test, but we’ll soon have it available for 8-Minute and 20-Minute tests as well.

Currently this will work for Ramp Tests that are part of a custom plan, as well as manually added Ramp Tests.

One caveat is that you can’t use the FTP Detection feature if you’ve updated your FTP within the last 14 days. Otherwise, it should work with any Ramp Test.

3 Likes

Thanks everyone at TrainerRoad for all the work that went into this!

2 Likes

Same as @ArHu74 for me…not showing up as an option. Only Polarized Plans.

3 Likes

Got Goddard -4 as my replacement workout. I’ll be sticking with my planned workout anyway.

1 Like

I don’t see the option either

Check back in a few minutes. For some reason this was visible to some athletes before the full web deploy. Should be available shortly.

11 Likes

I tried a new browser and a private window, still doesn’t show up

2 Likes