How to choose a robot mower: the specs that decide it

Work area: read the basis before the number

The single most misread spec is the coverage figure, because manufacturers publish it on at least three different bases: a recommended area, a tested maximum, and a rated working capacity. The same machine can carry more than one. Mammotion publishes the LUBA 2 AWD 3000HX with a recommended mowing area of 3,000 square meters (0.75 acre) and, separately, a maximum mowing area of 3,600 square meters (0.9 acre). Neither number is wrong; they answer different questions.

So compare like with like. The Segway Navimow X350 publishes 1.5 acres as a recommended area. The Ecovacs GOAT O1000 RTK publishes up to 8,600 square feet (0.25 acre) as a maximum coverage rating. The Dreame A3 AWD Pro 3500 publishes 3,500 square meters as a rated working capacity by variant, not a per-charge figure. A recommended-area number and a maximum-area number are not directly comparable, and a per-day throughput claim (Mammotion cites up to 0.4 acre per day for the YUKA mini 2 1000H) is a third thing again.

• Recommended area: what the maker suggests the machine handle comfortably.
• Maximum area: an upper coverage limit for the model.
• Rated working capacity: a per-variant capacity figure, often not per-charge.
• Per-day or per-charge throughput: a rate, not a lawn-size ceiling.

Slope: percent, degrees, and where it was measured

Slope is published in two units and against at least two locations, and both matter. A 30 percent grade is about 17 degrees; makers often print both, as Segway does for the Navimow i110N at 30 percent (17 degrees). The larger trap is the difference between the slope a mower can hold inside the working area and the lower slope it can manage at a boundary. Husqvarna publishes the Automower 420 iQ at 45 percent maximum inside the installation but only 15 percent at the boundary. Segway publishes the Navimow X350 at 50 percent (27 degrees) maximum, dropping to 25 percent (14 degrees) at the boundary.

All-wheel-drive flagships publish the steepest figures: the LUBA 3 AWD 3000 and the Dreame A3 AWD Pro 3500 both claim 80 percent, and Dreame notes its data is based on in-house or third-party lab testing with actual performance subject to variance. Treat every slope figure as a lab-derived claim, and if your slope sits near a boundary, size to the boundary limit, not the headline.

Navigation: three approaches in the wire-free market

Current wire-free mowers split into three navigation families, and the models we track illustrate each. RTK GPS with vision uses satellite positioning plus cameras: the Segway Navimow i110N (network RTK, no base antenna needed) and the Sunseeker X7 (RTK-GNSS plus VSLAM fusion) sit here. LiDAR with vision adds a spinning laser scanner: the Mammotion YUKA mini 2 1000H and the Dreame A3 AWD Pro 3500 run 360-degree LiDAR with AI vision and no RTK station. Camera-only navigation drops both wire and RTK: the Worx Landroid Vision (WR208) navigates on cameras alone, with no boundary wire, RTK antenna, or beacons.

A few machines fuse more than one method: the LUBA 3 AWD 3000 combines LiDAR, network RTK, and dual cameras. Husqvarna takes a different route on the Automower 420 iQ, using EPOS satellite virtual boundaries with onboard radar. See how two navigation philosophies compare directly in Navimow i110N vs GOAT O1000 RTK and LUBA 3 AWD vs Dreame A3 AWD Pro.

Runtime and charge time: also basis-dependent

Endurance figures carry a basis too. Segway publishes the Navimow X350 at 200 minutes full-charge mowing time and 80 minutes charging time, described as internal lab testing. Mammotion publishes the LUBA 3 AWD 3000 at 175 minutes runtime and 120 minutes charge, both from its official compare table. The Automower 420 iQ is published as typical mow time (84 minutes) and typical charging time (54 minutes), which is a different wording and therefore a different basis.

Some makers do not publish minutes at all. Dreame publishes daily coverage by power mode for the A3 AWD Pro 3500 instead of a runtime in minutes, so its runtime and charge-time fields are not published. Sunseeker lists only battery capacity for the X7, so runtime and charge time are not published either. When endurance is not published, coverage planning has to rest on whatever rate the maker does give.

What "not published" tells you

We do not fill gaps. When a manufacturer omits a figure, we record it as not published and treat that omission as information rather than a blank to be guessed. Worx does not publish runtime per charge for the Landroid Vision (WR208), so its endurance cannot be compared against mowers that do publish it. That is a real limitation of the spec sheet, and the buyer should know it.

The same discipline applies to discontinued models we cite for reference. The EcoFlow Blade and the LUBA 2 AWD 3000HX are no longer sold new, and we cite them only from official documentation, flagged as discontinued. A missing spec is never evidence of a good or bad machine; it is a boundary on what can honestly be compared.

Noise: often not comparable at all

Noise figures look simple and rarely are. Two different quantities hide behind decibels: sound pressure (measured at a distance) and sound power (the source's total output), and they do not equal each other. Mode matters too, since a mower is louder in a fast mode than a quiet one. Where a single comparable figure exists we cite it: Ecovacs publishes 57 dBA for the GOAT O1000 RTK and Segway 58 dBA for the Navimow i110N.

Where a figure is not comparable, we null it rather than mislead. The EcoFlow Blade publishes noise only as lab-measured sound power (59 dBA in Gentle mode, 65 dBA in Quick mode), a different quantity and mode basis than the noise-level figures peers publish, so no single number lines up. Sunseeker publishes 60 dBA for the X7 without stating pressure versus power or the mode tested. A decibel figure is only as useful as the basis printed next to it, and several makers do not print one.