- RTK is usually the first route to assess on large open lawns with reliable satellite and correction coverage; LiDAR is usually the first route to assess where canopy, buildings and narrow passages obstruct the sky.
- Local RTK, Network RTK and LiDAR-first systems carry different infrastructure, account, mapping and service dependencies, so the complete SKU and market version matter more than the headline sensor.
- Mixed sites should be commissioned as multi-zone systems: test sensor transitions, degraded operation, boundary repeatability, docking and map recovery before acceptance.

RTK and LiDAR robot lawn mowers do not have a universal winner. Start with RTK for a large, open lawn with reliable sky view and correction coverage. Start with LiDAR for a structured yard where trees, buildings or narrow passages obstruct satellites but provide stable local geometry. If the property combines both conditions, test a fusion system and verify how it behaves when one positioning layer becomes weak.
The buying decision should follow the site, not the sensor label. RTK, Network RTK, LiDAR and vision describe parts of a navigation architecture. They do not, by themselves, prove boundary accuracy, obstacle safety, automatic mapping, offline operation or serviceability.
For dealers and professional buyers, the commercial question is sharper: which system can be surveyed, installed, accepted and supported on this specific property without turning every signal problem into a return?
RTK vs LiDAR Robot Lawn Mowers at a Glance
| Decision area | RTK-first system | LiDAR-first system | Fusion system |
|---|---|---|---|
| Main positioning reference | Satellites plus correction data | Local geometric map measured by LiDAR | Satellite, LiDAR, vision and inertial layers in a model-specific combination |
| Strong starting site | Large, open lawn with clear sky | Structured yard with canopy, walls or narrow passages | Property containing both open and obstructed zones |
| Main external dependency | Local reference station or Network RTK service; sometimes internet | Onboard sensor condition, map quality, charging station and software | All included services plus correct switching and recovery logic |
| Common misunderstanding | Centimeter positioning equals obstacle detection | LiDAR equals complete navigation and safety | More sensors automatically mean more reliability |
| Installation risk | Poor sky view, multipath, correction or network loss, moved reference station | Weak or changing geometry, dirty/damaged sensor, map and calibration issues | Unclear primary sensor, unstable transitions, unresolved fault ownership |
| Dealer acceptance focus | Correction availability and repeatable virtual boundaries | Map stability, localization and boundary behavior in every structured zone | Transition zones, fallback behavior, recovery, docking and map continuity |
This table is a screening tool. A product still needs a site trial under the exact firmware, service plan and market configuration being sold.

What RTK and LiDAR Actually Do
RTK is a high-precision GNSS positioning method. The mower receives satellite signals and correction data, then uses them to estimate its position in a global coordinate system. That position can support virtual boundaries, systematic paths and repeatable work zones.
The correction source matters. A local RTK system uses a fixed reference station at or near the property. A Network RTK system obtains corrections through a service network, normally removing the need to install a local antenna. Both still require the mower to receive usable satellite signals.
LiDAR measures distance to surrounding surfaces with laser pulses. A mower can use those measurements with SLAM, inertial sensing, wheel odometry and software to build a local map and estimate its position inside it. The same LiDAR may also contribute to obstacle perception, but positioning and object recognition remain separate functions.
The distinction explains the site trade-off. RTK looks upward for satellite geometry and outward for correction data. LiDAR looks outward at the local environment. One site may offer an excellent sky view and few nearby geometric features; another may have walls, trees and passages that provide structure while obstructing satellites.
The broader wire-free robot mower guide maps RTK, LiDAR, vision and other boundary architectures. This article stays with the RTK-versus-LiDAR site decision.
The U.S. government’s GPS accuracy guidance lists satellite geometry, signal blockage and reflected signals as factors that affect positioning. NOAA’s current real-time GNSS network guidance also treats communication and multipath as field reliability issues. Those are general GNSS principles, not a failure prediction for every RTK mower.
Local RTK and Network RTK Are Different Purchases
“RTK mower” is too broad for a quotation or procurement sheet.
Local RTK
A local reference station adds physical infrastructure:
- a powered installation point;
- a suitable view of the sky;
- mounting hardware and weather exposure;
- communication between the station and mower;
- a defined relationship between station position and the saved map.
Husqvarna’s current EPOS information illustrates the consequences. Its reference station requires an open sky view, and moving the station after creating virtual boundaries can require the map to be created again.
That creates a service boundary. A customer who moves the antenna, changes its power source or relocates the charging station may have changed the positioning system, not merely moved an accessory.
Network RTK
Network RTK removes the local base station from some installations, but it does not remove infrastructure. The dependency moves into:
- geographic correction coverage;
- mobile or Wi-Fi connectivity where required;
- an activated product and regional account;
- the correction provider and its service terms;
- firmware and server compatibility;
- a procedure for outages, account transfer and end of service.
Husqvarna currently offers cloud correction in selected markets and requires stable internet across the work area for relevant cloud installations. Its EPOS support page separates that option from a local reference station and warns that model and market availability differ.
Mammotion’s current LUBA 3 page presents its iNavi Network RTK as an antenna-free service on applicable models. The same page separately discusses onboard 4G service. A buyer should record correction access and general cellular functions as separate line items instead of assuming that one “free connectivity” statement covers both.
“No antenna” means no customer-installed local RTK antenna for that configuration. It does not automatically mean no satellite, network, account, regional service or future commercial dependency.
What a LiDAR-First Mower Still Requires
A LiDAR-first mower can reduce dependence on satellite visibility, but it remains an installed system.
Dreame’s official A1 mapping procedure requires the mower to be docked, connected to the App by Bluetooth and prepared for mapping before the operator drives the boundary. Its installation page also specifies charging-station placement and includes a LiDAR cleaning brush.
Navimow’s current i2 LiDAR page takes a different route. It describes LiDAR and vision dual fusion, satellite-free positioning and automatic mapping. The page also identifies its sensor as solid-state LiDAR rather than an exposed rotating unit.
These examples show why “LiDAR mower” is not a complete specification. Dealers still need to ask:
- Is the sensor solid-state, rotating or another architecture?
- What is its field of view and protected mounting position?
- Does it map automatically or require an operator to trace the boundary?
- Which features use LiDAR, and which use cameras or other sensors?
- How is the sensor cleaned, inspected, calibrated or replaced?
- What happens after the charging station moves?
- Can the map be backed up, restored or transferred to a replacement mower?
- Which functions remain available without internet or a brand account?
LiDAR can measure in darkness without visible illumination, but the complete mower may still use cameras for boundary understanding, obstacle classification or safety. A night-capable ranging sensor does not prove that every function performs identically at night.
Which Site Conditions Favor RTK or LiDAR?
Large, open lawns
An open property with good sky view is the natural RTK screening case. Satellite-based coordinates can provide a stable global reference across long distances without requiring the mower to derive its entire position from nearby walls, trees or other local features.
Before approval, test the farthest points, low areas, edges beside buildings and any zone where trees interrupt the otherwise open sky. Confirm correction coverage across the complete work area rather than beside the charging station only.
Dense tree cover
Tree canopy can block or degrade satellite reception. Husqvarna’s current EPOS support specifically identifies dense crowns and tall buildings as possible obstructions. Some of its models use vision for a limited backup route; other systems may require a support wire or a different installation.
A LiDAR-first or fusion mower deserves the first trial under canopy because it can use local geometry. The trial still needs seasonal judgment. Leaves, trimmed hedges, parked equipment and other environmental changes can alter what the system sees. “Works under trees” is a test condition, not a permanent property of the product name.
Tall buildings and enclosed courtyards
Buildings create two RTK concerns: blocked sky and reflected satellite signals. LiDAR can benefit from nearby structural surfaces, but the dealer should test reflective materials, glass, repeated corridors, sharp corners and the transition from the courtyard into open lawn.
The correct result may be a LiDAR-first mower, a fusion mower or a redesigned work zone. A sensor comparison cannot replace a safe boundary plan near roads, water, retaining walls or public access.
Narrow passages and multiple zones
LiDAR and local perception can help a mower understand passages where satellite view is poor. Passage width, turning space, wheel slip and the return route remain mechanical and software constraints.
Commission the passage in both directions. Test entry, exit, obstacle response, route recovery and return to charge. A successful manually driven map is not proof that an autonomous task will repeat reliably.
Open but feature-poor areas
A very large, repetitive or sparsely structured area is a strong reason to test RTK first. A LiDAR-first system may still work, but the acceptance test should examine long-distance localization, loop closure, map drift and repeatable edges rather than assuming that a long advertised range solves the entire geometry problem.
Mixed properties
Many real properties combine an open front lawn, a tree-covered rear garden, a narrow side passage and a charging station beside the house. These sites are the strongest case for fusion—and the easiest place for a weak fusion implementation to fail.
Map each zone separately in the acceptance plan. Record which positioning layer is expected to lead, where transitions occur and how the mower responds if corrections, geometry, vision or network access become weak.
Why Fusion Navigation Is Becoming More Important
Current products show that the market is moving beyond a clean RTK-versus-LiDAR split.
Mammotion’s current LUBA 3 range is a useful example. The official product comparison lists LiDAR and dual-camera vision for the 1500-class model, while the 3000- and 5000-class versions add Network RTK. One product family therefore contains both LiDAR-first and LiDAR–RTK–vision configurations.
Navimow creates another distinction. The i2 LiDAR uses LiDAR and vision, while the i2 LiDAR Pro adds Network RTK to create a three-layer system.
The product name no longer tells the whole architecture. A distributor that trains staff on “LUBA 3” or “Navimow i2” without the suffix, area class and market version can misquote installation requirements and sell the wrong promise.
World Clean Biz’s LUBA versus Navimow procurement comparison covers the current model, channel and warranty context. The navigation stack still has to be checked at SKU level.
Fusion should be evaluated as behavior, not a component count:
- Which sensor establishes the primary position in each zone?
- When does the system switch or combine sources?
- Does it warn the operator when positioning quality falls?
- Does it pause safely, continue, return or wait for recovery?
- Can it relocalize without moving the virtual boundary?
- Does a network interruption affect positioning, remote control or both?
- Is the map preserved after a mower, sensor, station or account replacement?
Three sensors with unclear recovery logic can create more support paths than a simpler system matched correctly to the site.
Do Not Confuse Positioning, Mapping and Obstacle Avoidance
Marketing pages often place all three capabilities in one graphic. Procurement should separate them.
| Function | Question | Evidence to request |
|---|---|---|
| Positioning | Where does the mower believe it is? | Position-quality status, repeatable paths and behavior after interruption |
| Mapping | How are work areas, no-go zones and transport paths created and stored? | Mapping procedure, editing, backup, restore and transfer process |
| Boundary control | How closely and repeatedly does it follow the accepted safe edge? | Repeated boundary runs under relevant site conditions |
| Obstacle detection | What temporary objects can it detect? | Exact sensor set, object test, speed, distance and environmental limits |
| Safety response | What happens near water, roads, people, pets or drop-offs? | Model manual, risk assessment and site-specific protective measures |
| Recovery | What happens after signal, map, network or wheel-position loss? | Controlled fault and restart test under the approved procedure |
RTK does not identify a toy merely because it knows the mower’s coordinates. LiDAR does not prove that software can classify a pet merely because it measures a shape. Vision can add classification while creating its own lighting, contamination and data questions.
Treat every capability as a testable requirement.
Dealer Site Survey Before Quotation
A remote product recommendation based on lawn area alone is inadequate for advanced wire-free mowers.
The survey should record:
- exact address and target market;
- total area, area per zone and expected daily workload;
- sky-obstruction map by direction;
- dense canopy, buildings, eaves, walls and metal structures;
- local RTK reference-station position, power and mounting options;
- Network RTK coverage and service terms;
- Wi-Fi and mobile coverage across the route;
- narrow passages, corners, gates and hard-surface crossings;
- large repetitive or feature-poor zones;
- slopes, cross-slopes, roots, drainage and low-traction surfaces;
- water, roads, drop-offs and public boundaries;
- charging-station location, approach and drainage;
- expected seasonal and landscaping changes;
- customer account, data and map ownership requirements;
- installer, maintenance provider, repair location and warranty contact.
For remote sales, require a structured site form, annotated plan, photographs and coverage evidence. A return caused by an unsuitable property is often more expensive than the margin on the mower.
Commission the System, Not Just the Map
World Clean Biz’s robot lawn mower installation guide covers the complete installation process. An RTK-versus-LiDAR acceptance should add route-specific evidence.
Before mapping
- verify the full SKU, firmware and market version;
- record every included antenna, station, connectivity and service entitlement;
- document the intended navigation architecture for each zone;
- agree on physical safety measures that navigation cannot replace.
During mapping
- save the first accepted site plan and boundary version;
- record correction, satellite, network or map-quality indicators where available;
- mark every transition between open and obstructed areas;
- create no-go zones and transport paths under the exact manual.
Trial operation
- run the open area, canopy, building edge and narrow passage separately;
- repeat high-risk boundaries more than once;
- test docking from different zones;
- observe behavior when a permitted network or positioning interruption occurs;
- verify that the mower resumes or stops according to the documented design;
- check night or low-light operation only for functions the model claims to support;
- confirm that map edits do not shift previously accepted boundaries.
Handover
- record firmware, map version, station position and installed accessories;
- identify account owner, administrator and transfer procedure;
- document Network RTK, cellular and cloud terms separately;
- give the customer sensor-cleaning and station-movement restrictions;
- define when a site change requires recommissioning;
- obtain customer and installer acceptance.

Channel and After-Sales Risk
Wrong-site sales become “product defects”
If a dealer sells an RTK-first mower into a heavily obstructed yard without a survey, repeated positioning loss may return as a warranty claim. If a LiDAR-first mower is sold into a large site without testing map stability and coverage, a localization complaint may follow the same path.
The intake system should classify site-fit, installation, connectivity, map, sensor, traction and hardware cases separately.
Digital services are part of the product
Network RTK, cloud maps, cellular control, theft tracking and remote diagnostics may have different providers, free periods and transfer rules. Save the terms that applied when the mower was sold.
The contract should state:
- which positioning and connectivity services are included;
- the duration and renewal path;
- supported countries and roaming limits;
- who owns the account and map;
- what happens if the customer changes installer;
- how a replacement machine receives the accepted map;
- what functions remain after a service ends.
Installation changes need responsibility rules
Moving a local RTK reference station, charging station or high-risk virtual boundary can invalidate the original acceptance. Landscaping can also change canopy and local geometry.
The handover should list customer-permitted changes and changes that require the dealer to inspect or recommission the system.
Warranty does not replace site acceptance
The manufacturer, importer, seller, installer and authorized repair center may be different entities. The invoice and service agreement should identify:
- contract seller;
- local importer or regional responsible party;
- installation provider;
- warranty intake point;
- repair location;
- party paying labor, travel, freight and replacement;
- responsibility for account and map recovery.
A sensor fault, unsuitable site and incorrect installation require different remedies.
B2B Procurement Checklist
Product and navigation
- Full SKU, area class, market version and model year.
- RTK type: local base, Network RTK or both.
- LiDAR type, field of view, mounting, cleaning and calibration process.
- Vision, IMU, odometry and other positioning or safety layers.
- Primary and fallback localization behavior.
- Automatic versus manually traced mapping.
- Map backup, restore, replacement and transfer.
- Functions available offline.
Site and installation
- Sky-obstruction and local-geometry survey.
- Correction and network coverage for every work zone.
- Reference-station location, power and movement restriction.
- Charging-station location and return path.
- Narrow passages, buildings, canopy and feature-poor areas.
- Water, roads, walls, drop-offs and public access.
- Seasonal and landscaping changes.
- Documented commissioning and acceptance criteria.
Commercial and service
- Authorized territory and sales channel.
- Manufacturer, importer, seller and warranty provider.
- Network RTK, 4G, cloud and App service terms.
- Account, map and diagnostic-data ownership.
- Installer training and diagnostic access.
- Sensor, antenna, station and control-board parts.
- Warranty reimbursement for labor, travel and freight.
- Recommissioning price after site or hardware changes.
- Firmware-support and service-continuity plan.
FAQ
Is RTK better than LiDAR for a robot lawn mower?
RTK is often the better first option for a large open lawn with reliable satellite and correction coverage. LiDAR is often the better first option where tree cover, buildings and narrow passages obstruct the sky. Mixed properties should test a fusion system.
Does an RTK robot mower need an antenna?
Local RTK normally uses a reference station. Network RTK can remove the customer-installed antenna, but it may add correction coverage, internet, account and service dependencies. Check the exact model and market.
Does LiDAR work under trees?
LiDAR can localize from local geometry without depending on satellite visibility, so it is a strong candidate under canopy. Performance still depends on the exact sensor, map, software, environmental change and other onboard sensors. Test the actual property.
Does LiDAR work at night?
LiDAR ranging does not require visible daylight. The mower may still use cameras or other systems whose night behavior differs. Verify the complete product, not the LiDAR alone.
Is LiDAR used for navigation or obstacle avoidance?
It can support either or both. Some products use LiDAR mainly for positioning and mapping; others also use it for perception. Object classification and safety may depend on vision and additional sensors.
Is Network RTK the same as GPS?
Network RTK uses GNSS satellite signals plus correction data from a service network to improve positioning. It is not ordinary standalone GPS, and it still needs suitable satellite reception.
Which system is better near tall buildings?
Tall buildings can block and reflect satellite signals, so LiDAR or fusion deserves the first site trial. Test glass, repeated corridors, transitions into open space and every high-risk boundary.
Why do some new robot mowers use RTK, LiDAR and vision together?
The sensors provide different references. RTK supplies global positioning, LiDAR supplies local geometry and vision can add scene understanding. The benefit depends on how the software combines them and recovers when one layer weakens.
What should a dealer test before delivery?
Test every distinct zone, boundary repeatability, transition areas, docking, recovery after an approved interruption, map continuity, account transfer and the documented fallback behavior. Record the accepted firmware, map and infrastructure.
Final Answer
Choose an RTK-first robot mower for an open site where the complete work area has reliable satellite and correction conditions. Choose a LiDAR-first mower for a structured, satellite-obstructed site after testing its map, sensor and local-geometry limits. Choose fusion for mixed properties only after verifying transitions and fallback behavior.
For professional procurement, the sensor name is the beginning of the specification. The deliverable is a commissioned system with a verified SKU, accepted map, stable boundaries, documented digital services, recoverable accounts and a clear service owner.


