RTK delivers 1 to 3 centimetres of positioning accuracy in good conditions, against 2 to 5 metres for standard GNSS. That gap is real, but it holds only when a mower has stable corrections and a clear view of the sky. Under canopy or beside tall buildings, the same system can lose its fix and drift into decimetre-level errors, which is why boundary buffers, telemetry monitoring, and often a second sensor layer matter as much as the headline accuracy claim.
TL;DR:
- RTK achieves about 1 to 3 centimeters of positioning accuracy only under optimal conditions with stable corrections and a clear sky view.
- The accuracy can drop to decimetre-level errors near dense foliage, buildings, or reflective surfaces, requiring wider boundaries and safety buffers.
- Local base stations offer lower latency and direct control but are vulnerable to multipath errors, while network RTK depends on mobile coverage and may have higher latency.
- Incorporating sensors like visual SLAM, inertial measurement units, and LiDAR can maintain precise mowing during satellite signal outages, especially in complex environments.
- Conducting site assessments, telemetry monitoring, and staged pilots are essential before deploying RTK systems to ensure reliable, safe operation on your ground.
Table of Contents
- What RTK mowing accuracy actually means in practice
- Local base station or network RTK: which correction source suits your site?
- Why does RTK accuracy drop near trees or buildings?
- Do hybrid sensor stacks fix RTK’s blind spots?
- How do you assess a site and set safe mowing boundaries?
- How BrightView tracks RTK accuracy on real Australian sites
- When is RTK alone enough, and when do you need a hybrid stack?
- Get a site assessment before you commit to RTK mowing
- Sources
- FAQ
What RTK mowing accuracy actually means in practice
Standard GNSS works by timing radio signals from satellites and calculating position from that delay. It’s the same technology in your phone, and it’s accurate to somewhere between 2 and 5 metres on a good day. For a mower cutting a fairway or a solar farm access strip, that margin is useless. You can’t ask a machine to stay off a garden bed when its own sense of “here” wanders by the length of a small car.
RTK fixes this by comparing the raw carrier-phase signal, not just the coarser pseudorange timing, against a known reference point. The receiver resolves what’s called an integer ambiguity in the carrier wave, and once it locks that in, positioning tightens to 1 to 3 centimetres under optimal conditions. That’s the number most vendors quote, and it’s genuine, but it’s a best case, not a guarantee.
Here’s the distinction that trips people up: RTK accuracy is mostly about repeatability, not absolute truth. A mower doesn’t need to know its position relative to the centre of the Earth to a centimetre. It needs to return to the same virtual line every single pass. That relative consistency, sometimes called repeatability, is what actually keeps a cut edge straight over weeks of mowing, and it’s a more useful yardstick than the raw accuracy figure printed on a spec sheet, a point some technical analysis of mower positioning claims makes clearly.
A few variables push that 1 to 3 centimetre figure around in the real world:
- Distance to the base station or reference network. Accuracy degrades gradually the further the mower sits from its correction source.
- Satellite geometry. A tight cluster of visible satellites overhead gives a weaker fix than a spread across the sky.
- Sky view. Fewer visible satellites means fewer correction inputs and a less stable solution.
- Correction latency. Older correction data means the mower is navigating on slightly stale information.
None of this makes RTK unreliable. It means the number on the brochure is a ceiling, and your site conditions decide how close you actually get to it.
Local base station or network RTK: which correction source suits your site?
Every RTK mower needs a source of correction data, and there are two ways to get it. Each comes with a different failure profile, which matters more to an asset manager than any spec sheet claim.
- A local base station sits permanently on your property, broadcasting corrections directly to the mower over radio. You control it entirely, latency is minimal, and it doesn’t care whether the local telco has an outage. The risk sits in placement: a base station mounted near a metal shed, chain-link fence, or row of parked vehicles will pick up its own multipath errors and pass them straight to the mower.
- Network RTK, or NetRTK, pulls corrections from a regional network of reference stations over the internet, so there’s no antenna to install or maintain on site. That convenience comes with dependency on mobile coverage, network subscription uptime, and how close the nearest reference station in the network actually sits to your property.
Latency and update rate decide how robust the fix stays under stress in both models. A correction stream that updates once a second and takes half a second to arrive gives the mower a noticeably less stable solution than one updating several times a second with near-instant delivery. On large or remote sites, particularly council reserves or solar farms outside strong mobile coverage, that difference is often the deciding factor between the two architectures.
Pro Tip: Ask any vendor to show you fix rate and correction latency logs from a site with a similar footprint to yours, not a marketing demo reel. Real telemetry over a week tells you more than a spec sheet ever will.
Why does RTK accuracy drop near trees or buildings?
Sky view and multipath are the two environmental factors that most reliably wreck an otherwise solid RTK setup, and they’re both about geometry, not electronics.
Sky view refers to how much open sky a receiver’s antenna can actually see. Fewer visible satellites means a weaker geometric spread, a condition surveyors call poor dilution of precision (DOP). A mower parked under a dense canopy might see enough satellites to get a fix, just not a confident one.
Multipath is different and arguably worse. It happens when a satellite signal bounces off a nearby surface, a glass facade, a metal roof, a row of parked cars, before reaching the receiver, arriving slightly delayed and confusing the position calculation. Common multipath sources on commercial and council sites include:
- Sheet metal sheds, water tanks, and solar inverter housing
- Glass building facades and glasshouses
- Parked vehicles and shipping containers near mowing zones
- Chain-link fencing close to the mower’s working path
The observable symptom is usually the same regardless of course. The mower’s status drops from “fix” to “float,” accuracy degrades to decimetres or worse, and you’ll see it as sudden position jumps or a wobble along what should be a straight cutting line. Mitigations are mostly about geometry, too: relocate the dock or base antenna away from reflective structures, widen safety buffers around known hotspots, and treat narrow corridors between buildings as reduced-confidence zones rather than standard mowing paths.
Pro Tip: If you see repeated fix-to-float transitions in the same spot on your telemetry logs, don’t just widen the buffer there. Walk the exact location and look up. Nine times out of ten, something is blocking or bouncing the signal that wasn’t there when the site was first surveyed.
Do hybrid sensor stacks fix RTK’s blind spots?
Most commercial-grade mowers don’t rely on RTK alone anymore, and for good reason. Vendors increasingly pair it with visual SLAM (simultaneous localisation and mapping), inertial measurement units (IMU), and sometimes LiDAR, precisely because open sky can’t be guaranteed everywhere on a real property.
Each sensor covers a different gap:
- vSLAM builds a visual map of surroundings using cameras, letting the mower recognise where it is even without a satellite fix.
- IMU/VIO (visual inertial odometry) tracks short-term movement through motion sensors, holding position steady during a brief signal dropout.
- LiDAR adds obstacle and edge detection independent of GNSS entirely, useful around unpredictable hazards.
The practical question for procurement isn’t whether a system has these sensors. It’s how the mower behaves the instant RTK drops out. A well-engineered hybrid stack lets vision and inertial data carry the mower through a brief outage while it waits to re-acquire a fix, keeping the cutting line straight. A poorly integrated one either stops dead or, worse, keeps moving on a degraded and unreliable position. A combined RTK and vision approach tends to give the steadiest virtual boundaries on treed or suburban sites, and industry write-ups of hybrid positioning describe vision maintaining relative position while RTK recovers absolute alignment after an outage.
None of this is free. Hybrid stacks cost more, cameras lose reliability in low light or heavy fog, and every added sensor is another component that needs calibration and servicing. For a large open reserve with clean sky, RTK alone may be entirely adequate. For a site boxed in by buildings, fencing, and mature trees, the added complexity earns its keep.
How do you assess a site and set safe mowing boundaries?
A proper site assessment before deployment saves far more trouble than it costs. Run through this before committing to permanent virtual boundaries.
- Map the sky view. Walk the entire mowing area at the times of day mowing will actually happen, noting anywhere tree canopy, buildings, or structures cut into the horizon.
- Flag narrow corridors. Any gap between buildings, fences, or dense plantings under roughly five metres wide deserves a wider buffer and closer monitoring, not the standard margin.
- Identify multipath hotspots. Note sheds, glasshouses, parked vehicle areas, and glass facades near the planned cutting path.
- Test base placement. Position a base station or check network coverage away from reflective structures, at reasonable elevation, with clear line of sight to open sky.
- Run test passes with buffers set conservatively. Start wider than you think you need, log telemetry, and tighten only once the data proves it’s stable.
Through that testing phase, capture RTK fix state (fix versus float), satellite count, correction latency, and position residuals for every pass. That telemetry, not a single walk-through, tells you whether a boundary is safe to tighten or needs to stay conservative. Councils managing public reserves in particular should treat this data as the basis for any hazard buffer near paths, playgrounds, or waterways, a principle covered in more depth in BrightView’s robotic lawn mower safety guide.
Pro Tip: Log a full week of telemetry, including at least one overcast or wet day, before finalising any boundary near a hazard. Signal behaviour on a clear afternoon tells you almost nothing about how the same corner performs in low cloud.
How BrightView tracks RTK accuracy on real Australian sites
Numbers on a spec sheet don’t tell you how a mower behaves on your actual reserve, solar farm, or sports field. Some providers run staged pilots specifically to close that gap, because a vendor claim and a functioning boundary on your property are two different things.
Operationally, that means:
- Continuous telemetry monitoring through the pilot period, not a one-off site visit
- Conservative buffer settings from day one, tightened only once data supports it
- A preference for sensor-fusion architectures on any site with partial canopy or nearby structures
- Staged rollout rather than full-site deployment on day one
During a pilot, the metrics that matter most are boundary repeatability across repeat passes, the frequency of fix-to-float incidents, and time spent mowing per hectare. Those figures feed directly into how a contract scope gets written, and BrightView’s Australian pilot programme guidance sets out measurable ROI within 60 to 90 days as the benchmark for moving from pilot to full contract. That timeframe gives asset managers a real basis for comparison rather than a vendor’s projected figure.
When is RTK alone enough, and when do you need a hybrid stack?
Our view, after weighing the accuracy data against real deployment conditions: RTK alone is a solid, defensible choice for large open sites with genuinely clear sky, think expansive solar farms, open ovals, or unobstructed reserve land. The moment a site has mature trees, adjacent buildings, or narrow corridors, planning for a hybrid stack from the outset is the more responsible call, not an expensive afterthought.
Before signing any contract, insist on a pilot with a telemetry-based service agreement, not a verbal accuracy promise. Ask for fix rate and outage frequency data, and get the vendor’s documented failure behaviour in writing: what does the mower do the instant it loses its fix? For councils and asset managers, where a hazard near a path or waterway raises the stakes of getting this wrong, that conservative, evidence-first approach isn’t excessive caution. It’s simply the appropriate level of scrutiny for machinery operating unsupervised on public or high-value land.
— BrightView
Get a site assessment before you commit to RTK mowing
Some companies offer staged, telemetry-backed pilots before any long-term contracts are signed, rather than relying on a vendor’s best-case spec sheet.
Our service lines cover the full deployment path for asset managers and councils, from initial site assessment through to pilot programs and ongoing robotic mowing contracts once boundaries and buffers are proven stable on your site. If you’re managing a reserve, sports field, or solar asset and want real telemetry data instead of a marketing claim, get in touch through Brightview Solar & Groundskeeping’s site to arrange an assessment and see what RTK accuracy actually looks like on your ground.
Sources
- RTK GPS accuracy basics
- Rockin’ with RTK: How the tech applies to robotic mowers | Green Industry Pros
- Wire-free Robot Mower RTK accuracy guide (2026)
FAQ
Is RTK better than LiDAR for robotic lawn mowers?
They solve different problems rather than competing directly. RTK provides absolute positioning against a global reference, while LiDAR handles obstacle detection and edge sensing, which is why most reliable systems use RTK alongside vision or LiDAR rather than one instead of the other.
Does RTK need a clear line of sight to the mower?
RTK needs a clear line of sight between the mower’s antenna and the sky, not to a base station. A local base station does need line of sight to the mower for radio corrections, though NetRTK sends corrections over the internet instead.
What is RTK mowing?
RTK mowing uses Real-Time Kinematic GPS correction to guide a robotic mower along virtual boundaries with accuracy of roughly 1 to 3 centimetres under good conditions, compared with 2 to 5 metres for standard GNSS.
What are the drawbacks of robotic lawn mowers?
The main drawbacks are accuracy degradation under canopy or near buildings, dependence on stable satellite corrections or network coverage, and the need for conservative safety buffers around hazards like paths and water features. Hybrid sensor stacks reduce these issues but add cost and maintenance complexity.
Does Brightview Solar & Groundskeeping offer RTK mowing pilots?
Yes, Brightview Solar & Groundskeeping runs staged pilot programs for asset managers and councils that include telemetry monitoring and site-specific buffer testing before any ongoing contract. Current details are available through the Brightview Solar & Groundskeeping site.
