LiDAR for mapping, cameras for obstacles, and the good ones use both. If I had to pick one, LiDAR — it’s the difference between a robot that cleans in neat rows and one that bounces around like a pinball.

What each one is actually doing

LiDAR is a spinning laser, usually in the raised turret on top. It measures distance in every direction many times a second and builds a fast, accurate floor plan. Because it makes its own light, it works perfectly in a pitch-black room, which matters if you run the robot overnight or while you’re at work with the blinds down.

Cameras with AI do a different job: they look forward and recognize what an object is. A cord, a sock, a shoe, a charging cable, a pet accident. That’s object identification, and no laser scanner can do it — LiDAR knows something is there but not what it is.

The lasers are eye-safe, by the way. Same class as a CD player. That question comes up a lot with pets and kids in the house.

What happens when a robot only has one

Camera-only robots struggle in low light. They navigate by recognizing visual features on the ceiling or in the room, and when the room is dark those features disappear. Performance at night is noticeably worse.

LiDAR-only robots map beautifully and then drive straight over your phone charger, because they never identified it as something to avoid. If your floors are consistently clear, that may not bother you at all. If you have kids, pets, or a home office full of cables, it will.

The premium tier is essentially both technologies working together, and that combination is most of what you’re paying for as prices climb.

Dad’s Take

LiDAR decides whether it cleans efficiently. The camera decides whether it eats your charging cable. Different jobs.