C2 Vision

Les Numériques placed leaves across the pool from the robot. In its targeted mode, the C2 Vision headed towards them and concentrated on that patch until it was clear. Standard cleaning also collected leaves, sand and small debris. The camera did not, however, make it avoid the test goggles and mask.

An owner with a fibreglass pool tried several modes and still reported missed rounded corners and loss of contact near the waterline. They wondered whether the slippery surface was responsible, but did not establish a diagnosis. This contrasts with the independent tester’s effective wall and waterline coverage; neither account should be treated as universal.

The owner reported turning away from the steps. Les Numériques likewise found that the robot treated steps as walls rather than traversing them. A camera that finds leaves therefore does not establish effective staircase or seating-shelf cleaning.

Les Numériques found the top-access filters easy to rinse and measured about two hours of runtime. Hook retrieval was necessary. Mode changes required the app, and enabling Vision required agreement to the use of some pool-floor footage for AI training. Those are meaningful operating choices, rather than automatic benefits of adding a camera.

The owner account establishes a coverage complaint, not a battery failure pattern or service history. Problems reported for the standard C2 cannot be assigned to the C2 Vision. Longer ownership evidence is still needed to judge durability.