Market scan: why nothing on the market does this anymore
Before building anything I scanned the market for devices that warn blind and low-vision users about obstacles above cane height. The headline finding: nothing affordable serves this niche anymore. A ~€40 cap-brim device for head-height trail hazards is a real gap — and the graveyard of products that tried adjacent ideas is full of free lessons.
Throughout, confidence is flagged: [solid] = multiple or primary sources, [likely] = single decent source, [uncertain] = thin or conflicting evidence.
Sunu Band — wrist sonar, effectively dead [likely]
$299, ultrasonic sonar to ~5.5 m, distance encoded as vibration pulse rate. Discontinued around September 2024. The long-term AppleVis review didn’t recommend it: unpredictable false detections, units failing to charge, refund requests. Its core flaw was structural — the wrist points where your arm swings, not where you’re going.
Lessons: head-mounting fixes the aiming problem; the distance→pulse-rate encoding was genuinely liked; false positives killed trust.
WeWALK Smart Cane 2 — alive, $850 plus subscription [solid]
A smart cane handle with ultrasonic upper-body detection. Gen 1 was notoriously over-sensitive; gen 2 improved, but false alerts are still reported. One gen-1 user, on exactly my use case (head-height branches): it “kept beeping seemingly for no reason… after a while you just ignore it.”
Lessons: a swinging cane is a false-positive machine; a stable, head-mounted, narrow beam is a defensible alternative; users value a discreet vibration-only mode.
biped.ai NOA — the high end [solid]
A 950 g shoulder harness with cameras, AI, GPS and 3D-audio warnings, priced in the thousands. A 2026 Scientific Reports study (13 participants) found fewer collisions than cane alone; 12 of 13 would adopt it. Top complaints: bulk and learning curve.
Lessons: the market leader fixed false positives with trajectory-filtering AI — the microcontroller equivalent is persistence and hysteresis, which cost nothing. Weight is the top adoption complaint (a cap wins). The study authors note haptic feedback may beat audio outdoors, which supports this project’s core choice.
iMerciv BuzzClip — the closest analog, discontinued [solid]
A ~50 g clip-on ultrasonic device with vibration feedback, CAD $250–330. The same 2026 study documented why it failed: vibrations too weak and too frequent, no height or spatial information, feedback confused with the cane’s own vibrations, and zero measured mobility improvement. Yet users praised it as light, simple and quick to learn — the form factor was right, the execution wasn’t.
The others
Ambutech iGlasses (~$120–150, still sold): ultrasonic glasses with temple vibration. The AFB evaluation found them poor in cluttered indoor spaces but promising outdoors — and they prove head-mounted haptics have worked at low cost for a decade. Ara by Strap Tech ($2,750): dubious delivery record, no independent reviews [uncertain]. .lumen Glasses (€9,999): camera/LiDAR headset with forehead haptics — validating the forehead as a feedback site at 100× this project’s budget. Scene-description AI glasses (Envision, Meta Ray-Ban): a different job, not competitors.
The pattern
Every failed product in this space failed on false alarms and lost user trust — not on range, not on price. That’s why this project’s number-one success metric, set before a single part was ordered, is false-alarm rate. The firmware biases toward silence: multiple consecutive close readings are required before the motor speaks.