RTK wins in open sky, LiDAR wins under trees, and cameras win only on tidy lawns

Robot-mower navigation is not a contest to find the most advanced acronym. It is a property-matching problem. RTK can hold beautifully straight lanes across an open lawn, yet struggle when trees, walls, or a roofline obstruct the sky. LiDAR turns those same trees and walls into landmarks, but may have less to read in a broad, featureless field. Camera vision removes both the antenna and rotating laser, although the lawn must remain visible enough for the mower to understand it.

The sales pitch usually stops at mapping. The lawn does not. A mower still has to cross damp turf, turn on a slope, climb a paving lip, avoid a drop, reach every zone, find its charging contacts, and leave enough daylight to complete the schedule. Precise positioning attached to weak traction merely tells the robot exactly where it became stuck.

For most compact yards with mature trees, Smart Lawn Bots would start with LiDAR. For an open rectangle, RTK remains the more economical and entirely sensible choice. Pure vision earns the narrowest recommendation because its wonderfully easy setup asks the grass itself to stay legible.

Our Top Pick Eufy E15 robot lawn mower

Read Review →

Choose the failure you can prevent

Stand at the proposed charging location and look outward. If most of the sky is visible and a powered reference station can remain fixed, RTK has a fair chance. If branches, buildings, eaves, or narrow side passages dominate the view, onboard LiDAR removes an installation variable that software cannot wish away. If the lawn looks like a clean green island surrounded by paving, fencing, or beds, camera-only mapping may be the easiest route of all.

Then walk the ground instead of looking above it. Press a heel into the dampest area, cross the steepest line sideways, inspect every rut beside the sidewalk, and measure each gate. These details decide whether rear-wheel drive is enough, whether AWD is worth paying for, and whether a supposedly multi-zone mower can reach the second lawn without somebody opening a gate or carrying it.

RTK is excellent when the sky cooperates

RTK uses satellite positioning corrected by a local reference station or a network correction service. Its advantage is repeatable position across open ground, which supports virtual boundaries, orderly stripes, efficient routes, and editable zones without burying perimeter wire. Grass color, morning shadow, and a row of fallen leaves do not determine where the mower believes the boundary lies.

The reference is also the weakness. A pole or antenna needs power, stable mounting, and a broad view of the sky. Dense canopy, close buildings, eaves, and reflected signals can reduce accuracy or dictate an absurd dock location. Vision helps an RTK mower recognize objects and bridge brief uncertainty, but the presence of cameras does not automatically make it satellite-independent.

LiDAR prefers the yards that frustrate RTK

LiDAR measures the physical surroundings from the mower. Fences, trunks, walls, beds, and buildings create a local three-dimensional scene, so the robot can navigate beneath tree cover without asking an antenna to see through branches. It can also localize in low light, although camera-assisted object recognition remains more useful in daylight.

This makes LiDAR the strongest default for an enclosed residential yard. It is not universally superior. A vast open lawn may offer fewer stable features, sensor glass must stay clean, and heavy rain or fog can interfere with ranging. More importantly, LiDAR does not supply traction, cliff detection, a wider deck, or competent map editing. Several otherwise persuasive LiDAR mowers are let down by docking errors, caster wheels, restrictive apps, or poor recovery after an obstacle stops them.

Camera vision is simple because the lawn does part of the work

A camera-only mower can remove the perimeter wire and RTK station at once. The eufy E15 and E18 use stereo cameras to identify grass, trace visible boundaries, map automatically, avoid conspicuous objects, and follow systematic lanes. On a neat lawn, setup feels closer to placing a robot vacuum than commissioning outdoor positioning equipment.

The simplicity is conditional. Yellow grass may resemble an exclusion, bare soil may resemble a continuation, flowers can become a boundary, and an unfenced sheet of grass shared with a neighbor contains no visible property line. Low light ends the working day. Damp grass, shallow depressions, and rough transitions also expose the light rear-drive chassis even when the cameras understand the scene perfectly.

That is why the eufy E15 and E18 decision is mostly about mapped area rather than greater physical ability. The E18 accepts more lawn, but it does not gain a wider deck, stronger traction, nighttime navigation, or immunity to ambiguous borders.

Sensor choice cannot rescue the wrong machine

Navigation tells a mower where to go; the chassis determines whether it gets there. The clearest example is a smooth hill versus a shallow rut. A rear-wheel-drive mower may climb the impressive-looking hill in a straight line, then beach its front caster beside a sprinkler on nearly level ground. Cross-slope travel can also make a machine slide at a gentler angle than its advertised straight-climbing maximum.

Deck width and battery behavior matter next. A seven-inch mower can maintain a quarter acre through repeated cutting and charging, but it will spend much more of the day outside than a machine with two wide discs. Acreage ratings describe how much grass the robot can manage across a schedule, not what it completes on one charge; the distinction is explained in our guide to robot mower acreage ratings.

Edges remain stubborn across all three systems. A virtual boundary can be accurate to centimeters while the cutting disc sits several inches inside the body. Flush paving lets a wheel overlap and brings the blades closer, whereas fences, walls, raised beds, curbs, and drops leave a strip for the string trimmer. ECOVACS improves the problem with a separate side trimmer, but even that mechanism cannot reach every inside corner or safely brush every fence post.

The mowers that make each system worth choosing

Several ambitious alternatives need more proof

The ANTHBOT Genie600 has the right hybrid idea, pairing full-band RTK with four cameras, but its cut quality, runtime, edge behavior, antenna requirements, parts supply, and real performance beneath trees remain unresolved. The ANTHBOT M9 adds NRTK, automatic mapping, and thirty zones, yet a weak satisfaction signal and missing ownership details make it another early-adopter purchase rather than a dependable recommendation.

The BESTMOW T100 promises 1.5-acre capacity, four cameras, RTK, and a four-inch maximum cut, but its listing conflicts over the slope limit and omits too much about the deck, battery, warranty, weather protection, and service. Yarbo reaches much farther with tracks and a modular multi-season platform, though repeated hardware failures, demanding commissioning, downtime risk, and unresolved remote-access concerns mean it should not be the only mower on a large property.

The HOOKII Neomow X2 SE demonstrates why a good sensor cannot carry an immature platform. Its LiDAR premise suits tree-covered lawns, and some owners obtain excellent large-area results, but missed lanes, failed docking, traction trouble, frequent firmware work, uncertain cellular terms, and a reported wheel-retention failure prevent a general recommendation.

The Navimow i215 is more persuasive for an enclosed lawn up to 0.37 acre, especially when trees rule out RTK and the ground is firm. It loses to the Sunseeker S4 on simplicity, to the LUBA Mini 2 on terrain, and to the X430 on throughput, but it remains a reasonable middle choice for buyers willing to refine zones and test every dock route; the i215 and X430 comparison separates those jobs directly.

Audit the yard before comparing prices

1. Measure only mowable grass. House, driveway, beds, woodland, and water do not belong in the acreage calculation.
2. Find the worst sky view. Inspect the dock, boundaries, and connecting passages rather than judging satellite visibility from the center of the lawn.
3. Walk every transition. A one-inch lip, soft rut, exposed root, or damp side slope can matter more than the advertised maximum grade.
4. Trace the route home. Every zone needs a safe, driveable connection to the charger, and a virtual path cannot open a closed gate.
5. Classify each edge. Flush paving is robot-friendly, while walls, drops, curbs, fences, and tight corners preserve manual trimming.
6. Plan for the sixth week. Confirm blades, batteries, app connectivity, subscriptions, warranty handling, major parts, and somebody able to rescue the mower.

A compatible robot mower does work. It replaces the repeated central-lawn cut with quiet, frequent maintenance and leaves tiny clippings behind as mulch. It does not collect leaves, reclaim serious overgrowth, make toys safe, repair holes, or eliminate every edge. Buyers deciding whether that remaining work is acceptable can continue with our guide to whether robotic mowing is worth the trade.

The final choice is not between three fashionable technologies. It is between three kinds of dependence. RTK depends on the sky, LiDAR depends on a readable physical scene, and camera vision depends on visible turf and boundaries. Pick the dependency your property can satisfy, then buy enough traction, capacity, and support to keep the mower moving after the first impressive map has disappeared from the phone.