Help with Postural

Your chronological age plus the years contributed by three scored postural deviations: forward head (CVA), rounded shoulders (FSA) and lateral head tilt (Tilt).

Each adds between 0 and its own ceiling of 6, 5 and 2 years, for a maximum of +13. It's a trend indicator built to be trackable rather than a clinical grade. The direction it moves is the part to watch.

We measure two more, sagittal alignment and shoulder symmetry, and show you both without scoring either. Sagittal alignment because nobody has published a normal range for it in healthy adults, and we won't invent one. Shoulder symmetry because the physiotherapist reviewing our method asked us to stop pricing it without evidence for the degree being penalised.

Three are scored:

• CVA: craniovertebral angle. Forward head: how far your ear sits ahead of the base of your neck, from the side. Max +6 years.
• FSA: forward shoulder angle. Your shoulder's forward roll from the base of your neck. Max +5 years.
• Sym: shoulder symmetry. Left-to-right shoulder height difference, normalised for how tall you appear in frame. Max +3 years.
• Tilt: lateral head tilt. Your ear-to-ear line, measured off horizontal from the front. Max +2 years.

A fifth is measured and not scored:

• Sagittal alignment. Trunk lean: the angle at your hip, from shoulder through hip to ankle. Max +0 years, on purpose.

We didn't invent that last one. It is the hip angle used by SAPO, a postural assessment tool used in research, where five blinded raters agreed on it to an ICC of 0.93 on the right side and 0.86 on the left. Nobody has published a normal range for it in healthy adults, so we have nothing honest to score it against and we won't make a threshold up. We compute the angle, store it, and show it to you, and it adds nothing to your Postural Age. One thing to hold against that 0.93: it came out of a lab, from markers a clinician placed by hand on bony landmarks, which beats what a phone photograph can do.

At least two of the three scored markers have to be measurable for a valid scan, and neither unscored marker counts toward that two. Below that, Postural asks you to retake rather than filling in the gaps. The full thresholds are published on the app page.

Postural uses Apple's Vision framework for 2D body pose detection, reading 13 landmarks per view. Each view is captured across 12 frames and every joint is taken as the median of those readings, which removes most of the frame-to-frame jitter. If your shoulders or hips drifted during the capture, the app throws the set out and asks again rather than scoring it.

It cannot see depth, rotation or muscle tone, because it works from a photograph. Angles are computed in normalised image coordinates and the thresholds are calibrated for that, not read off a clinical goniometer.

Read your score as a consistent trend across time, not an absolute measurement. It is not a medical diagnosis.

Because single-photo measurement has real variance. How you happened to stand, the light, and small differences in where the phone sat all shift the computed angles by a degree or two, which can move the score by a year or so.

Postural handles this in two ways. The headline number on the reveal screen is a rolling average of your last three scans' posture penalty rather than one reading, and every change is weighed against normal scan-to-scan variation before you're told anything about it. That's why the app will tell you a week was "holding steady" instead of dressing up a small wobble as progress or regression.

A result needs at least two of the three scored markers measurable with enough confidence. The usual causes, in rough order of frequency:

• Loose clothing hiding your shoulders, hips or ears
• Dim light, or strong light coming from behind you
• Standing too close or too far, leaving part of you outside the frame
• Not standing side-on to the camera for the side view

If it keeps happening, the app will offer a manual capture button and tell you it can't get a clean read rather than guessing. Try a brighter room with a plain wall, and something more fitted.

No. Every photo is stored only on your device, in the app's own directory, and each file is flagged at the filesystem level to be excluded from iCloud backup. An automated test verifies that flag on every build, so it can't be dropped by accident in a later release. Photos are never uploaded to any server.

Deleting the app removes them permanently.

• Your scan scores: stored against your account in Firebase Firestore, so your history survives a reinstall. Scores, not photos.
• Anonymous product analytics: screen opens and actions like "scan completed", sent to PostHog. No photos, no scores, no email. Session replay is switched off, so nothing from the camera or scan screens is ever recorded.
• Crash reports via Firebase Crashlytics: device model, iOS version and a stack trace. No identity.
• YouTube traffic when you play an exercise video, your device connects to YouTube directly. We don't send your posture data there.

Your posture photos are never transmitted. Full detail in the Privacy Policy.

Yes. Postural uses Sign in with Apple or Google Sign-In to hold your account, which is what lets your scan history survive reinstalling the app or moving to a new phone. We never ask you to create a password with us and never see one.

Settings → Delete Account, inside the app. That permanently removes your account and all associated scan history from our servers. Local data, meaning photos and on-device history, goes when you uninstall.

If you're stuck, email support@srstech.in and we'll process the deletion for you.

After each scan, Postural looks at which of your markers are costing you the most years and picks the matching movements from a library of 28 curated stretches and strengthening exercises. Each one comes with a video demonstration and credits its creator.

The routine is reassembled after every scan, so it follows your posture rather than running a fixed plan.

Day 7 shows your baseline and your streak, and no improvement figure. Twenty-one minutes of mobility work over a week doesn't move your structure, so any change at that point would be measurement noise, and Postural won't present it as progress.

Day 28 is where Postural is willing to call a change real. Four weeks is a product decision rather than a research finding: long enough for a genuine shift to clear the measurement noise, short enough that you are still using the app. Consistency is the variable you control.

Unlimited scans, your Postural Age, the reveal screen and the coach note on what changed are free, and stay free. That includes the Day-28 update itself: both photos, your before and after Postural Age, and the caption on what moved. When a scan puts years on your score, your home screen names the marker adding the most of them, at no charge. You can browse the exercise library and play any single movement from it without paying.

Postural Pro adds the personalised daily routine built from your own weakest markers, your scan history with the side-by-side comparison of any two scans in it, the full five-marker breakdown, and sharing the Day-28 card or the timelapse. It is $69.99 a year with a 14-day free trial, or $7.99 a week. The trial applies to the annual plan and converts to a paid subscription unless you cancel at least 24 hours before it ends.

No. It's a consumer wellness app, not a medical device. Postural Age is a geometric trend indicator; it doesn't account for pain, strength, hypermobility or how your particular body is built, and it can't diagnose or treat anything.

If you have posture-related pain or dysfunction, see a physiotherapist or another qualified healthcare provider.

Postural is in active development, targeting late 2026 on the App Store. Join the waitlist on the app page and you'll get one email on launch day.

iPhone running iOS 18 or later. iPad support may follow in a later update.