Joint angles, gait, standard tests and bar speed, measured the same way every visit and at home. Loads are modeled and clearly labeled. Every report is for your review.
The most technical answer is also the most honest one. Every Room screen and report sorts its numbers into these three buckets.
Two calibrated cameras: about 4–5° from a motion-capture lab in published research (OpenCap, 2023). Small differences are shown as ranges.
Musculoskeletal simulation from 3D motion. Best for left-versus-right and before-versus-after, not absolute values.
That's the clinician's exam and imaging. Room never claims it.
10–15 minutes in the bay. Choose a region: knee, hip, shoulder, spine, ankle, running, return to sport. Results fill in with a 3D replay.
Prescribe from the exercise library. Each exercise has measurable targets: depth, tempo, reps.
The patient's phone tracks reps, form and adherence, plus pain and effort taps. You see who's doing it.
Same bay, same tests, every 2–4 weeks. Only changes larger than measurement error are called real.
Trends, criteria met, outcomes and a plain-language summary for the patient. Exported to your EMR.
Recreational runner. Six weeks of right anterior knee pain.
| Measure | Left | Right | |
|---|---|---|---|
| Knee alignment, single-leg squat (FPPA) | 6° | 14° | Right drifts in |
| Contralateral pelvic drop | 3° | 7° | Right stance |
| Ankle dorsiflexion, knee-to-wall | 41° | 32° | Right limited |
| Running cadence | 158 steps/min | Low end | |
| Kneecap load index, +10% cadence | about −14% (modeled) | Lever | |
College soccer. Seven months after right ACL reconstruction. Ready to return?
| Measure | Result | |
|---|---|---|
| Hop tests, limb symmetry | 89–94% | Looks close |
| Knee flexion at landing | 78° L · 66° R | Stiff right landing |
| Load through the right leg | about 18% less (modeled) | Offloading |
| Single-leg jump height, symmetry | 80% | Below 90% |
| Recommendation for review | Not yet. Here's why. | |
For strength gyms and PT studios: velocity-based training, measured every rep. It's also the bridge from late-stage rehab back to full training.
Illustrative lifter, back squat. Load-velocity profile from five loads; estimated max at a 0.30 m/s minimum velocity threshold. Verified speed needs a depth camera or bar sensor.
Two synced cameras at 60–120 fps on fixed mounts, a calibration mat with hop-lane markings, and Room Core. Two cameras match lab-grade research setups; a third doesn't add accuracy.
One depth camera on a wall mount. Range of motion and single-leg tests, without modeled loads.
The patient's phone, with guided placement. Reps, form and adherence for the home program.
Estimated net value per year against about $10K of Room fees. Every input is shown and is an estimate until a pilot measures it.
Monitoring revenue assumes Medicare's 2026 remote therapeutic monitoring codes. Billing them requires a device that meets their rules, so Room is designed to support those workflows once listed. It is not billable today.
The first study compares Room with a motion-capture lab and force plates in 60 people, half of them patients. Alongside it: goniometers in clinics, sensitivity to recovery over time, and bar speed against a reference sensor. Results get published, good or bad.
We're looking for a small number of clinics and sports medicine groups to shape Room Clinic and take part in the validation study.