I am blind or visually impaired: the story in one minute
Third Eye Kit

Research, Archive, Design review note

A working note, kept for the record. Back to the archive.

Design review note: where the camera sits

The founder read the design review record and raised one placement question: the phone camera, and possibly the whole sensor pod, belong on the chest rather than the belt. His reasoning: at the waist, the view is blocked by the arms, the mobility aid hand, a coat, or a bag, and a chest position covers the whole walking lane. This note records the review's answer.

What the record already says

The placement study's ninth section settled the phone's position before this question was asked: a fixed chest mount for the wearer's own phone, ultra-wide lens, landscape, no gimbal (Where the body should feel, section 9). The reasons were field of view and stability, not occlusion, but they point the same way. The chest-worn systems in the literature agree: the Mobilio phone app that cut obstacle contacts by 41 percent read a chest-mounted camera, and the biped NOA harness carries its cameras on the shoulders. No belt-mounted camera appears in any study the program cites.

So the founder is right about the camera, and the site now says so wherever the camera is mentioned: the camera rides on the chest. It was never meant for the belt; the review record's silence on the point let that impression form.

The sensor pod is a different question

The distance sensors are not a camera, and their position was chosen for geometry, not view. The down-tilted laser grid reads the ground between 0.6 and 1.25 m ahead from waist height; from the chest the same grid would look down at a steeper angle over a shorter stretch of ground, or need a wider field to keep the reach. The belt's eight motors and the pod share a housing so the buzz lane and the sensing lane stay aligned without a cable up the torso. And the belt-worn time-of-flight array in the ALVU study produced zero collisions in blindfolded trials, so a waist pod has a working precedent (the sensing study, prior art).

The founder's occlusion concern still stands for the forward grid. A swinging arm, the mobility aid hand, or a bag strap can cross a waist-level field in a way it would not cross a chest-level one. The review does not settle that from the desk.

Decisions

  1. The phone camera rides on the chest. Fixed mount, ultra-wide lens, landscape, as the placement study says. The site's pages that mention the camera now say chest, and nothing says belt.
  2. The sensor pod stays at the waist for the ground grid. The geometry of the tilted grid and the shared housing with the belt motors are the reasons.
  3. Waist against chest for the forward grid goes to the bench. Test G0 gains a test: the forward laser grid worn at the waist and at the chest on the same walk, with the arms swinging, the mobility aid in hand, and a bag on the shoulder, counting the frames in which each position loses the hazard. If the chest wins, the forward grid moves to a chest clip and the ground grid stays in the pod.
  4. The founder's question is recorded here, with the evidence for each part of the answer, in the program's standing way.