Robot vacuum navigation: LiDAR, camera or gyro—what actually changes

Also available in French.

Suction watts sell robots on the box. Navigation decides whether the whole room gets cleaned or the same strip gets three passes while the hall is skipped. Four technologies you will meet—and what they enable.

At a glance

Navigation	Principle	Home map	Room clean	Works in dark	Typical tier
Random	Bounce off obstacles	No	No	Yes	Budget
Gyro	Motion sensors, parallel lines	Rough or none	Rarely	Yes	Entry
Camera (vSLAM)	Visual landmarks (ceiling, furniture)	Yes	Yes	Lighting-dependent	Mid / high
LiDAR	Spinning laser, 360° ranges	Yes, accurate	Yes	Yes	Recent mid-range up

Our pick: LiDAR. Accurate map from the first run, works in the dark, room-by-room cleans—including from Home Assistant. It now appears on affordable models: under £200 / $200.

Random and gyro: avoid if you can

Random bots turn at each bump; a small room eventually gets covered, slowly and unevenly. Gyro bots run straighter lines but drift in multi-room homes and rarely keep a useful map.

Camera (vSLAM)

The robot watches the ceiling or room and tracks landmarks. Good maps and room cleans. Downsides: performance drops in very dark rooms, and an in-home camera raises privacy questions—read the vendor’s image policy.

LiDAR

The turret spins a laser and measures distances. Fast relocalisation if you move the bot; works in darkness. Trade-off: extra height—measure clearance under sofas before you buy.

Obstacle avoidance is separate

Navigation is where am I; avoidance is do not eat the cable. Entry models often only have basic IR; cables and socks still need tidying. High-end units add structured light or front cameras with object recognition.

Smart home

A reliable map is what lets Home Assistant say “clean the kitchen” instead of “clean everything”. Room cleans depend on brand integration: robot vacuums in Home Assistant. Mops and docks: mop and station guide.

All smart-home guides