Sixty years of physiology,
finally measurable outside a lab.

The problem

The dose is invisible until it is expensive.

A coach prescribes an action and waits. Whether the dose landed shows up a training block later, as progress that stalled or an injury that arrived. Between those checkpoints an athlete trains below what they could tolerate, or above it, and nothing in the session says which. That is the good case. Everybody else has a wearable estimating effort, an app recording what they typed, or nothing at all.

What is established

Critical power: power in, duration out.

The critical power model resolves exercise tolerance into two parameters. CP, in watts, is the highest sustainable output. W′, in joules, is a finite work capacity available above it. Calibrated from a few trials, the model returns how long an effort lasts at any power above CP, and it has held across cycling, rowing, running and isolated joint contraction for six decades.^1,6,7

Why it stopped at the gym door

Lifting is intermittent by construction.

A set of eight followed by two minutes of rest is not continuous work. Capacity drains during the set and partly returns during the rest, so the model's central assumption fails at the level of construction rather than approximation. The same total work above CP can take three minutes or fifteen depending only on how the rest is arranged, and a model with no recovery term gives both sessions the same answer.

The missing term

What returns during rest is measurable.

What comes back between sets is largely phosphocreatine, and it comes back fast: recovery after exhaustive work is biphasic, with a fast component whose half-time is on the order of twenty to thirty seconds and most of that fast recovery complete inside two to three minutes.³ Those are exactly the rest intervals lifters use, which is why small changes in rest change a session so much, and why recovery cannot be treated as a second-order correction.

There is direct precedent for adding the term. Morton and Billat formulated a critical power model for intermittent exercise, and Skiba and colleagues modelled the reconstitution of W′ during recovery below CP; the resulting balance model is standard in cycling analysis.^5,8 No equivalent exists for resistance training. That literature compares two or three fixed rest intervals and reports group means,^2,4 which establishes that rest matters without producing a function a model can use. The obstacle has been measurement, not theory.

What it gives a coach

Capacity, one driver at a time.

An exercise is a combination of movement drivers. Decompose it, give each driver its share of the work, and give each driver its own critical power, recovery rate and reserve. A prescription can then be checked against capacity driver by driver, before the athlete has lifted anything. Three sets of ten at 100 kg with ninety seconds of rest, on a low bar squat:

Driver	Share	Work	Power in the set
Squat	70%	21.9 kJ	244 W
Hinge	20%	6.3 kJ	70 W
Anti-extension	10%	3.1 kJ	35 W
Total		31.3 kJ	348 W per set

Against each driver's own ceiling at that load, rep count and rest, the squat and hinge sit inside and anti-extension is over. The set fails on bracing, not on legs. Extend the rest from ninety seconds to 130 and it drops back under, before the athlete has touched a barbell.

What has to be measured first

Two terms, and only one of them is easy.

W_total = W_load + W_body

W_load is work on the external implement: measured displacement and a known mass, with no biological assumption. W_body is work on the athlete's own mass, both centre of mass elevation and segment motion, and it needs a body segment mass model. Commercial devices report the first term and not the second, and the size of that omission depends entirely on the movement.

FITNDEX derives both terms from ordinary training video. Object tracking gives displacement and timing, which validate directly against tethered equipment; bar work follows from those and a known mass. A centroid model gives body work from pose and segment masses. The segment mass model is the assumption under test, and DXA is what validates it.

The program

Ordered so nothing is built on an untested assumption.

Step one is a single short session per subject: a barbell curl measured three ways at once, against a tethered criterion device and taped distances as ground truth. Once the camera is validated against a tether, later studies do not need a laboratory. Subjects follow a written protocol at their own facility and upload video.

Known limits

What this does not claim.

Mechanics, not metabolism

The system establishes whether a relationship holds and with what parameters. Not why.

The segment mass model is an assumption

Validated against DXA in step four, and not used in steps one to three.

Curve form is not assumed

Hyperbolic, exponential and power law forms are fitted and compared.

Recovery is measured, not identified

Kinetics consistent with phosphocreatine resynthesis. Attribution to a substrate needs spectroscopy.

References

1. Monod H, Scherrer J. The work capacity of a synergic muscular group. Ergonomics. 1965;8(3):329-338.
2. de Salles BF, Simao R, Miranda F, Novaes JS, Lemos A, Willardson JM. Rest interval between sets in strength training. Sports Medicine. 2009;39(9):765-777.
3. Harris RC, Edwards RHT, Hultman E, Nordesjo LO, Nylind B, Sahlin K. The time course of phosphorylcreatine resynthesis during recovery of the quadriceps muscle in man. Pflugers Archiv. 1976;367(2):137-142.
4. Grgic J, Schoenfeld BJ, Skrepnik M, Davies TB, Mikulic P. Effects of rest interval duration in resistance training on measures of muscular strength: a systematic review. Sports Medicine. 2018;48:137-151.
5. Morton RH, Billat LV. The critical power model for intermittent exercise. European Journal of Applied Physiology. 2004;91:303-307.
6. Moritani T, Nagata A, deVries HA, Muro M. Critical power as a measure of physical work capacity and anaerobic threshold. Ergonomics. 1981;24(5):339-350.
7. Poole DC, Burnley M, Vanhatalo A, Rossiter HB, Jones AM. Critical power: an important fatigue threshold in exercise physiology. Medicine and Science in Sports and Exercise. 2016;48(11):2320-2334.
8. Skiba PF, Chidnok W, Vanhatalo A, Jones AM. Modeling the expenditure and reconstitution of work capacity above critical power. Medicine and Science in Sports and Exercise. 2012;44(8):1526-1532.

The critical power model and the phosphocreatine kinetics above are established literature. The extension to resistance training, and the driver-level capacity it produces, are implemented in the product and are the subject of the validation program described here. Study results are not yet published.