Aerotune on-road metabolic testing

It seems Inscyd has a competitor.

Bjorn Kafka was on That Triathon Skow recently. I thought it was a great interview. As usual Mikael has great guests and does a fantastic job interviewing.

Kafka was also on the Endurance Innovation podcast:

It sounded like the software is still in beta but available. It’s inexpensive compared to Inscyd - he said 25 or 30 euros for a month of access. I’ve been wanting to try Inscyd but $200 per test is a bit much for a dabbler like myself.




I also recommend the interview in Endurance Innovation podcast (episode 99 and 100) with both founders. Looking at their offer, the protocol is very simialr to Insycd, the price is quite astonishingly low in comparision and you have the same test results and some more additional insights about your training. It is basically based on Mader’s work, like insycd so it should be the same in principle.

I really prefer the direct-to-consumer model for testing like this, as you are paying for some automated analysis, and I think the price should reflect that. I think the insycd model where it is through coaches has merit as well, since you get knowledgeable analysis with the numbers, but that makes repeated testing very expensive. If I want to track my vlamax over time and do several tests, the cost of the tests is an important factor, especially if you are primarily looking at a single number from the tests.


It would be quite nice addition to TR in the future to include this sort of profiling and AT based on the results. The algorithms behind are not very unique and it would be quite interesting USP.

Yes, it should be very similar. The model uses a ton of assumptions and “magic numbers” so those are very likely where inscyd and aerotune have some dissimilarities.

However, I find it fairly amusing that Kafka and also Weber claim that the algorithm is super complicated. It isn’t really.


It does sound like both worked through the formulas and then validated them against real world testing. Kafka’s first business was running a testing lab. Jan Olbrecht had similar software that he was using with swimmers for years.

Merging my notes from Scientific Triathlon and Endurance Innovation podcasts

Basis for relating VO2max to basic aerobic training volume:

  • need stress to promote growth
  • how much? minimum and maximum
  • theoretical basis traces back to late 1800s (Wilhem Rusch spelling?), and more recently the work of Alois Mader
  • It is theorized from protein syntheses processes within the cell and how much stimulus (kilojoules of work) these processes need to adapt as well as how big of an overload they can handle.
  • works very well in base training

Basic algorithm:

  • Mader’s formula for amount of VO2max you can use in 24 hours
  • take a % of that to get best growth
  • convert %VO2max-24hours to KJ
  • based on KJ recalculate the protein synthesis to see how fast you recover
    (implies some iteration of steps 2-4 if recovery is >24 hours or substantially less than 24 hours)

Example 1 (ignoring influence of vlamax)

  • VO2max of 70, optimal stimulus is ~12,000-14,000KJ/week
  • zone2 of 210W is ~750KJ/hour (210W * 3600 sec)
  • they said ~15-18 hours max where 12,000/750 = 16 hours; 14,000/750 = 18.6 hours
  • the numbers given don’t completely add up, so without more detail its a rough order of magnitude

Other examples:

  • VO2max of 60, average z2 power of 180W, 13-16 hours max (up to 14,000KJ/week)
  • VO2max of 45, average z2 power of 140W, 11-15 hours max (6500-8000KJ/week)

So an example closer to my numbers would be VO2max of 45, average power of 170W (612KJ/hour), and therefore MAX base endurance hours of ~10-13 hours/week.

@iamholland did that help clear up the VO2max-volume relationship?


Maybe this paper?
“A mathematical model of regulation of protein synthesis by activation feedback: some reflections on its possibilities and limits in describing muscle mass adaptations with exercise”
If anyone has access to it, I’d like to have a look.

Anyway, doesn’t seem there is any other scientific validation study for this though.

Hi Norman, I have the paper if you want it. It’s also available on the popular Science Hub website if you’re so inclined.

Mader wrote a book that was mentioned in the podcast:

The book describes with lots of practical examples metabolic models and simulations that Mader has been using.

For a while, there was a English web site by Mader himself:

but that disappeared a few years ago. Some of it can still be accessed via archive:

Unfortunately, many of the figures are missing, so a lot of the thing would only make sense if
you had the book.

The protein synthesis is covered both in the book as well as the English website above,
but I haven’t seen the part describing aerobic capacity and protein synthesis. I do remember
Jan Olbrecht (Mader’s student) also mentioned that the training volume depends on aerobic capacity. Interestingly enough, Olbrecht is a big proponent of supercompensation principle
while Mader seems to completely dismiss supercompensation.

One thing that is completely missing from any of these “models” is adaptation processes and
hormonal adaptations. There’s a great book by Viru that is not widely known or cited:


I have studied Mader’s work and papers for quite some time. It’s interesting work and I agree
with you that it’s not super complicated (at least the steady-state model isn’t, because it’s
not using any coupled differential equations).

I wanted to replicate the model and verify the results. But, as you say, the model relies on several
“magic” parameters and therefore you need to have a large database of measured lacatate (and possibly VO2).

Also note that Mader’s model is very simple. There are other much more complex models out
there that simulate cell much more accurately. And if you look at those complex models
you see that there are hundreds of “magic” parameters needed to simulate a single cell
let alone a group of cells or a muscle. For example, a recent paper describes metabolic simulation:
and they provide code as well:

This model is orders of magnitude more complicated and complex than Mader’s.
(Note: Complex/complicated doesn’t necessarily mean better).


The reference you mention is:

ROUX, W.: Gesammelte Werke über Entwicklungsmechanik der Organismen, Bd. 1: Funktionelle Anpassung . Engelmann, Leipzig 1895




1 Like

Yeah, this graphic is at first not really straightforward to understand and apart from all the other stuff I feel this one is not yet optimally explained.

But once understood it does make sense - at least for me.
So as @bbarrera already contributed to this thread, all the columns are basically the least and maximum times of optimal training stress for the respective athlete based on his VO2max and therefore potential protein synthesis capability. Also lactate clearing capability, available glycogen and a few other parameters are considered.

So then each column for each training zone or training target shows the weekly min max times.
That means: if you are doing only base training in one week, you would be best served with at least the light blue column and max the darker blue column.

For each other target zone the red columns are the intensity times and the blue columns are the low intensity times. So depending from his parameters the Sprinter Athlete 1 should do (when in a VO2max block) maximal 1:06 hours in VO2max range and then still had to ride another 8:42 h in base range (either in separate workouts as well as in part as Time in zone in the VO2 workouts of that week).



Not explicitely given. But I far as I understand this and similar CP tests you are supposed to do them on separate days. I did Test 1 on one day and Test 2 the next day.

I bet it doesn’t change anything if you wait two days or even three days in between tests (as long as you don’t tax your glycogen stores or recovery status by something extraordinary in between). Since this way you won’t be any more recovered on day 2 or day 3 after test 1 even if you just lay on the couch. On the other hand you don’t want to stay away from your normal workout routine for 3 or 4 days - so best doing it on two consecutive days.
Probably you wouldn’t introduce too much error if you would do test 1 on a convenient day in this week and do test 2 on the same weekday with the same lead in (rested, glycogen stores replenished) in the following week.

Layman me would think that you are still at your same metabolic level over the course of two weeks.

If it is feasible doing test 1 in the morning, rest up during the day and maybe doing test 2 in the evening of the same day - I don’t know.

1 Like


1 Like

I’m doing an aerotune test in the next couple of days and have just signed up for the month of testing.
There is not much information on the test protocol.
Do you need to stay seated for all 3 tests as with inscyd?
What about gear changes on the sprint test?
Do you need to lap the efforts?
Do they need to be perfect times or is there tolerance on the efforts. Eg, 20 mins can be 18-23 mins etc

If anyone have experience of doing this I would be interested in their test protocol.

1 Like