A body-anchored volume keeps the operating experience consistent after the user moves

What it is

If the interaction volume is anchored to the body, valid input is a patch of air relative to the chest, head, or waist, and the patch travels when the person travels. After a move, the same body-relative motions still count; there is no room origin to find again. What stays consistent is the proprioceptive operating experience, not distance to a wall. A headset menu stuck in front of the face, or a near-field tool rack that translates with the user, uses this anchor.

Why it happens

A local frame is originated at the skeleton root, head, or pelvis. Hand coordinates are transformed into that origin before containment. When the person translates or turns, the transform updates with the root, so the geometry of a relative motion is unchanged: a pinch in front of the chest is still a pinch in front of the chest on the other side of the room. Spatial memory is unloaded from the room onto the body, matching the sense “my hand is in front of me.” The cost is root quality: shoulder heave and a hip center that wanders left-right make the whole volume shake like a tray hung off the torso. A body anchor also cannot outrun capture range: if the person walks where the sensor can no longer see, the chest-relative box still exists in math and the signal does not.

Studying it

Teach a chest-relative vocabulary, then repeat after translation, turning, sitting, and climbing a step if the site allows. Compare success, where the hand goes, and “it felt the same” reports against a device-fixed volume. Add jitter to the root at several levels and watch when experiential consistency collapses. Include the capture-range boundary: when the person walks to where the sensor is about to lose them, the body anchor is still mathematically right and actually unusable. Code that failure on its own so it is not blamed on the anchor.

Where it stops holding

When screen content is nailed to the room (a wall-sized image, a case that cannot turn), a body anchor lets the valid hand region walk while the thing the eyes look at stays put, splitting hand and eye space. Several people watching one public screen, each with a private body anchor, will misalign talk about “that point on the screen.” Seated with a backrest pinning the torso, the travel benefit of a body anchor is small and the jitter remains. When the root comes from an unstable monocular selfie camera, a body anchor can be harder to use than a device anchor.

Applying it

• Use a body anchor only for commands completed in front of the self. Pointing at a room-fixed object should use device or world coordinates.
• A torso center is more stable than one shoulder. If root jitter exceeds a threshold, freeze the box briefly so a shaking tray does not false-fire.
• Walk “learn it, walk elsewhere, do it again”: a body-relative motion should not require re-aiming at floor tiles. If users start hunting tiles, the anchor is still nailed to the device in practice.

Related

• Same group: C4.21.1 The coordinate anchor decides whether the volume is device-fixed or relocates with the body · C4.21.2 A device-fixed anchor requires the user to remember and return to the same physical location · C4.21.4 When the anchor is recomputed—at engagement or by continuous following—affects consistency
• Adjacent: C4.13 Full-body pose and skeleton tracking · C4.20 Interaction volume versus capture range
• Search: body-locked · egocentric · torso frame