What real users of similar devices taught me

Last updated 5 July 2026

Spec sheets tell you what a device claims; users tell you why it ended up in a drawer. This write-up draws only on primary user voices — AppleVis reviews and forums, AFB AccessWorld evaluations, UK sight-loss charity reviews, blind bloggers, and a 13-participant user study. Vendor-funded findings are flagged.

What users consistently loved

Across every device, the same things earned praise: detection of head- and upper-body hazards (the one thing canes and guide dogs can’t do), graded proximity encoding that’s understood instantly, devices framed as a complement to the cane rather than a replacement, adjustable sensitivity, and — where available — direction and height information.

A Sunu Band user found it detected “walls, rails, low hanging branches” she couldn’t see in failing light. A guide-dog user could pre-plan around obstacles before her dog even stopped. The iGlasses evaluation singled out head-level branches on trails — exactly this project’s use case — as where the device shone.

What got devices abandoned, ranked

1. False positives and unpredictability — the number-one abandonment cause everywhere. Phantom obstacles while resting a wrist; alerts on nothing; well-known obstacles missed while imaginary ones buzzed. Users forgive limited range. They never forgive being lied to.
2. Feedback below the perceptual threshold, or without meaning. The BuzzClip’s single weak buzz was indistinguishable from the cane’s own vibrations for some users — and in a 13-person study it produced zero measured mobility improvement and more frustration than a cane alone.
3. Build and charging fragility. The Sunu Band died of connectors as much as false alerts — even its warmest reviewer returned his unit. “Good concept, poor quality.”
4. Clothing and mounting friction. Sleeves blocked wrist sensors; winter coats fought chest clips; anything clip-mounted gets mounted crooked.
5. Price and subscriptions against unproven reliability.
6. Weight and bulk — the top complaint even among fans of the high-end NOA.
7. Company mortality. Sunu and BuzzClip hardware became orphaned when their companies pivoted. A standalone, no-app, no-cloud device is immune to this failure mode entirely.

Where a device is worn matters more than what’s in it

The evidence sorted neatly by mounting position. Wrist: points where the arm swings, blocked by sleeves. Chest clip: weak buzz masked by cane feedback, aim-critical, fights coats. Cane: the swing generates false positives, the weight causes arm fatigue. Head: best for head-height hazards — with the caveats that scanning hurts straight-line travel and social visibility matters. The iGlasses partly succeeded because they looked like ordinary sunglasses. A baseball cap is even more invisible. And a camera-free distance sensor sidesteps the bystander-privacy objections that plague head-mounted cameras.

One genuine evidence gap: nobody has published data on sweat and multi-hour comfort of head-worn haptics. This project’s test logs will be original data.

The three rules this project takes from all of it

1. Bias hard against false positives: narrow field of view, multi-reading confirmation before any alert, and a one-touch mute.
2. A tiny but meaningful haptic vocabulary: cadence encodes distance; one unmistakable pattern reserved for “imminent”; directional feedback only if field tests prove single-point warning isn’t enough.
3. Boring-reliable beats clever: robust power connection, zero phone dependency, and it should look like a hat, not a gadget.