Which robot vacuums dodge cords, socks, and toys?

Independent press testing of the Dyson Spot+Scrub AI documented 20 of 24 staged obstacles avoided (TechRadar, Tom's Guide, Wired, March 2026). Dyson also reports 200-object recognition as a training-set figure. The Saros 20 Sonic and Qrevo Edge use LiDAR-plus-camera stacks; owner reports confirm routine cord and sock avoidance, but no equivalent staged test is in the public record. The Matic relies on camera-only on-device processing; no published obstacle count exists.

Dyson staged obstacle rate

20 of 24 (independent press testing)

Dyson manufacturer training claim

200-object recognition

Saros 20 Sonic sensor stack

LiDAR + camera + AI (manufacturer)

Qrevo Edge sensor stack

PreciSense LiDAR + AI (manufacturer)

Matic published obstacle benchmark

Not documented

The Dyson Spot+Scrub AI stacks dual-line DToF LiDAR with an AI HD camera and a green-spectrum LED. The Saros 20 Sonic and Qrevo Edge combine LiDAR with camera-based AI. The Matic uses camera processing only, with no LiDAR layer. Independent testing of the Dyson recorded a 31% roller hair-wrap rate versus a 26% category average, indicating contact with material the sensor array did not pre-classify as an obstacle.

Dyson sensor stack

Dual-line DToF LiDAR + AI HD camera

Saros 20 Sonic sensor stack

LiDAR + camera + AI

Qrevo Edge sensor stack

PreciSense LiDAR + camera

Matic sensor stack

Camera only (on-device)

Dyson roller hair-wrap (independent)

31% vs 26% category average

Owner reports for the Saros 20 Sonic and Qrevo Edge describe autonomous sessions in lived-in rooms without manual floor prep. The Dyson's stain-targeting loop (up to 15 repeat passes on a detected stain) extends session time in rooms with multiple soiled areas, a side effect owners need to account for. The Matic's owner forum coverage confirms adequate operation in consistent single-floor layouts; performance in heavily cluttered or changing environments is not documented.

Saros 20 Sonic unsupervised reports

Confirmed (owner reports)

Qrevo Edge session continuity

Multi-floor mapped (LiDAR)

Dyson multi-stain session length

Extended (up to 15 passes per stain)

Matic layout dependency

Routine layouts only (owner reports)

Matic heavy-clutter data

Not documented

None of the four robots require physical perimeter wires. The Dyson Spot+Scrub AI maps autonomously on first run and uses app-based virtual no-go zones. The Saros 20 Sonic and Qrevo Edge do the same via Roborock's LiDAR mapping. The Matic performs autonomous visual mapping on first run but supports no app-based zone control due to its on-device-only architecture; physical barriers are its only boundary option.

Dyson boundary hardware

None required; app-based zones

Saros 20 Sonic first-run mapping

Autonomous LiDAR

Qrevo Edge first-run mapping

Autonomous LiDAR

Matic app zone support

None (on-device processing only)

Matic boundary option

Physical barriers only

Dyson equips the Spot+Scrub AI with a green-spectrum LED specifically to support detection in low-ambient-light conditions, a design feature that implicitly acknowledges that camera performance degrades in low light. The Saros 20 Sonic and Qrevo Edge use LiDAR as the primary ranging layer, which is lighting-independent; camera-based object classification is the layer that degrades in low light. The Matic relies entirely on visual processing and no documentation of its low-light performance exists in owner or press sources.

Dyson low-light design choice

Green-spectrum LED (manufacturer)

Saros 20 Sonic LiDAR layer

Lighting-independent ranging

Qrevo Edge LiDAR layer

Lighting-independent ranging

Matic low-light performance

Not documented

Cross-platform comparative data

Not published