Mammotion YUKA mini 2 1000H · Worx Landroid Vision Cloud WR320

LiDAR and vision replace the perimeter wire and the RTK base station together, but the boundary lives in a point cloud rather than on a survey grid: a yard that changes substantially, new structures or removed trees, needs re-mapping rather than a coordinate correction.

Nothing on the property needs a view of the sky: with no satellite correction in the loop, tree cover and narrow lots that constrain RTK mowers do not constrain this one, and there is no base station to mount or maintain.

Satellite-dependent autonomy: RTK precision eliminates wire-burial labor but introduces cloud and GPS failure modes not present in wire-guided systems.

AI-camera privacy cost: obstacle detection uses on-board imaging streamed to Worx cloud for model training; no local-only or opt-out mode.

Terrain above the rated 45 percent gradient

Mammotion rates the YUKA mini 2 1000H to 45 percent grade (24 degrees) on rear-wheel drive. Ground steeper than that sits outside the rating, and no behavior beyond it is published.

Thick or tall fescue lawns above 4 inches

Blade design and cutting deck geometry optimized for standard residential turf. Thick or tall-growing varieties cause bogging and reduce autonomous runtime; manual pre-mowing required.

Lawns above the rated 0.25 acre

The 1000H designation is 1,000 square meters, which Mammotion states as 0.25 acre, about 10,900 sq ft. Per-charge coverage is 2,690 sq ft, so a full lawn runs across several charge cycles. Larger properties sit outside what this model is rated to cover.

Yards with heavy mixed debris load

Obstacle avoidance handles sticks and fallen branches effectively; extreme debris (fallen limbs, dense leaf mat) causes navigation interruptions and may require pre-mow yard cleanup.

Yards over 1/2 acre

The WR320.1 is rated to half an acre. That is a lawn-size ceiling rather than a per-charge figure, so extra docking cycles do not extend it; Worx sells a one-acre Landroid Vision Cloud for larger lots.

Owners rejecting AI-camera surveillance

The WR320's obstacle avoidance relies on an on-board camera that processes imagery to the Worx cloud for AI training. Privacy policy permits image retention for model improvement. No local-only operation mode.

Terrain steeper than a 30 percent grade

Worx rates this model to 30 percent, about 17 degrees. Ground steeper than that sits outside the rating, and wet clay or loose soil can cost traction on grades below it as well.

Owners preferring predictable boundary control

Virtual boundaries rely on satellite corrections. When those drop, the mower falls back on camera V-SLAM fused with inertial data and odometry, which holds position less precisely than a corrected fix. A perimeter wire is a physical line the owner places and can see.