Is RTK necessary for drone surveying? Three approaches to improve accuracy

Many practitioners considering drone surveying are initially uncertain about whether RTK is truly necessary. On site, RTK-capable aircraft and high-precision positioning equipment tend to attract attention, and it can feel as if you cannot achieve the required accuracy without RTK. However, in reality whether RTK is necessary for drone surveying depends on the purpose of the site, the deliverables you need, the existing surveying system, and ultimately what you want to decide.

For example, whether you simply want to take an overview of current conditions, compare earthwork quantities, or precisely overlay existing drawings and design data makes the meaning of "required accuracy" completely different. If you don't clarify this and think, "RTK seems high‑accuracy so let's use it" or "it'll be safer to have it," you may find it doesn't produce the expected effects, or conversely that it increases unnecessary steps and costs.

On the other hand, there are certainly situations where RTK is extremely effective. When you want to align coordinates efficiently with a small crew across a large site, when you need to compare multiple measurement results against the same reference, or when you want to reduce ground-based support work while still ensuring a certain level of absolute accuracy, RTK has considerable practical importance. In other words, RTK is neither "always necessary" nor "completely unnecessary"; it is a system that should be used after properly clarifying the concept of accuracy.

In this article, to determine whether RTK is necessary for drone surveying, we first clarify the basic role of RTK and then explain three key concepts to keep in mind for improving accuracy. We also describe, from a practical standpoint, the common failures that tend to occur when relying solely on RTK and provide on-site decision-making guidelines. The content is organized so that both those considering implementation and those already operating and struggling with accuracy variability can have criteria for making decisions.

Table of Contents

• What RTK Is and Common Misconceptions in Drone Surveying

• Situations Where RTK Is Required in Drone Surveying

• Approach 1: Determine the required precision by working backwards from the deliverable

• Approach 2: Consider absolute accuracy and relative accuracy separately

• Approach 3: Achieve accuracy through the entire operation rather than relying on RTK

• Common Failures When Relying Solely on RTK

• How to determine on-site whether RTK is required

• Summary

What RTK Is and Common Misconceptions in Drone Surveying

RTK is a system that increases positional accuracy in real time while correcting errors in satellite positioning. In the context of drone surveying, it is often used to improve the coordinate accuracy of the aircraft’s position and the camera’s position during flight, which makes it easier to stabilize the alignment of photos and point clouds. As a result, it can reduce the number of reference points that need to be set on the ground and make it easier to establish coordinates efficiently over wide areas.

However, what is often misunderstood is that simply using RTK does not automatically improve the overall accuracy of a survey. The accuracy of drone surveying is not determined solely by the aircraft's positional information. Image overlap, flight altitude, shutter timing, wind effects, the shape of the target, ground surface conditions, the radio environment caused by trees and structures, and even post‑processing conditions—all of these factors interact to determine accuracy.

A common situation in the field is when you flew an RTK-capable system but didn’t get the expected accuracy. The cause is often not RTK itself but something else. For example, if satellite reception was poor near buildings, the imaging direction was biased relative to slopes or embankments, ground check points were not placed, the elevation reference was not aligned, or the coordinate settings during processing were incorrect, the benefits of RTK will not be fully realized.

Conversely, even without RTK, if control points are placed properly and the flight plan and processing conditions are arranged, there are many sites where accuracy sufficient for practical work can be achieved. In other words, RTK is a powerful means of improving accuracy, but it is not the only answer. Understanding this premise is crucial for establishing drone surveying in the field.

When RTK Is Needed in Drone Surveying

RTK is particularly effective in situations where the reliability of absolute coordinates is important. For example, when you want to overlay and compare as-built topography with existing drawings or design models, when you want to continuously compare survey results from multiple periods in the same coordinate system, or when you want to manage the progress of embankment or excavation by volume—positional discrepancies in these cases can easily lead to wrong judgments. For such uses, it is not sufficient for the shapes to merely "look plausible"; it is important that the entire site can be aligned and handled on the same reference.

Also, on large sites or sites with restricted access, the task of placing numerous ground control points is itself a burden. From the perspectives of manpower, time, and safety, there is strong demand to maintain accuracy while reducing ground work, and this is where RTK's value becomes apparent. On some sites—slopes, land under development, material storage yards, and infrastructure corridors—simply walking around to install control points can be a major effort, so systems that make flight operations and positioning as efficient as possible translate directly into practical benefits.

Furthermore, RTK is also effective when you want to make drone surveying a continuous operation rather than a one-off record. For monthly progress management, before-and-after comparisons of construction, and fixed-point monitoring for maintenance, it is important to be able to accumulate data using the same reference each time. If the positional reference varies with each flight, the shifts will mix into the differences you want to see, making them difficult to use for decision-making. RTK works to enhance that comparability.

On the other hand, there are situations where RTK is not essential. For example, if you only want to create a simple ortho image to grasp the overall view of a site, if you want to roughly look at terrain trends for an initial study, or if you prioritize understanding relative shapes over exact coordinate matching, you can often achieve your goals without RTK. The important point is not to assess the presence or absence of RTK as a question of technical superiority, but to decide based on what decisions the deliverables will be used for.

Approach 1: Determine the required accuracy by working backward from the deliverable

When considering whether RTK is necessary for drone surveying, the first thing to clarify is not "how much accuracy is required" but "what the deliverables will be used for." The term "accuracy" is convenient, but when discussed separately from the purpose on-site it can easily lead to incorrect judgments. What is important is to make clear what decisions or tasks the deliverables will ultimately lead to.

For example, if your aim is to grasp site conditions before construction and you only need to understand the site's undulations and the placement of obstacles, what you need may be overall, well-organized, easy-to-read data rather than exact coordinate matching. In that case, preventing missed captures and setting flight conditions that make targets easy to see are more important than introducing RTK. Conversely, for numerically evaluated uses such as volume calculations, checking differences from design, or verifying as-built conditions, shifts in positional or elevation references will affect the results themselves. In those cases, the value of RTK becomes greater.

A common mistake here is to assume “if we have high-precision data it will be widely usable,” and to default to RTK. In practice, however, chasing accuracy beyond what is necessary increases the burden of preparation, verification, and operation, and prevents it from becoming established on site. Accuracy is not simply something where higher is always better; being able to stably reproduce the required accuracy is more important. For practitioners, an operational approach that can deliver the same quality every time is more valuable than theoretically maximum accuracy.

Therefore, when considering whether RTK is necessary, it's easier to organize your thinking by first reviewing the deliverables from three perspectives. First: is the focus on visual assessment, or is numerical comparison required? Second: is it a single one-off record, or is ongoing comparison assumed? Third: is the data self-contained, or does it need to be connected to existing drawings or other survey results? The stronger the latter elements among these three, the greater the need for RTK.

In other words, whether to implement RTK should not begin with a discussion of the aircraft or the positioning method. First decide "what the site will determine from this deliverable," and then construct the required accuracy and operations based on that. Simply following this order will considerably improve the accuracy of the adoption decision.

Approach 2: Consider absolute accuracy and relative accuracy separately

When considering the accuracy of drone surveying, another important point is to consider absolute accuracy and relative accuracy separately. If these two are confused on site, phenomena such as "the shape matches but the coordinates are shifted" or "the overall position is correct but there is local distortion" cannot be adequately explained, and evaluations of RTK become ambiguous.

Absolute accuracy is the concept of how correctly something is positioned relative to real-world coordinates. This is important when overlaying existing design drawings, ground surveys, or measurement data from other times. For example, absolute accuracy matters in situations such as when you want to overlay a site development plan with the current terrain to determine the construction area, compare regularly acquired data month by month, or connect with other surveying results. RTK is often a system that helps ensure this absolute accuracy.

Relative accuracy, on the other hand, is the concept of how accurately shapes are reproduced within a model. For example, when looking at the trend of elevation changes across a site, the shape of slopes, how materials pile up, or the positional relationships between objects, the model can still be practically useful if the entire model is slightly offset from real-world coordinates but the internal shapes are stable. Even when comparing earthwork volumes, if measurements are taken under the same conditions each time and the relative consistency is high, it can be sufficiently useful for management.

The important point here is that, although RTK can be expected to reinforce absolute accuracy, it does not universally guarantee relative accuracy. If capture conditions are poor, shape reproduction can still break down even when using RTK. Conversely, even if absolute coordinates are somewhat loose, stable capture and processing can result in data that is relatively easy to view and use. In other words, judging data quality solely by the presence or absence of RTK can cause you to lose sight of the essential issues.

In practice, this distinction is extremely useful. For example, when data is for internal briefings or progress sharing, prioritizing relative accuracy and quickly organizing it can often be more valuable. Conversely, if you are looking ahead to quantity management, drawing reconciliation, as-built verification, and future design use, attention to absolute accuracy is indispensable. Once it is clear which should be prioritized, the priority for implementing RTK will naturally become apparent.

In drone surveying, when you say "accuracy was achieved" or "accuracy was not achieved," it is essential to explicitly state and confirm whether that accuracy refers to absolute accuracy or relative accuracy. Simply making this distinction aligns understanding on site and makes discussions about RTK implementation much more practical.

Approach 3 Achieve accuracy across the entire operation rather than RTK

When trying to improve accuracy, it's easy to focus on the equipment or positioning method, but in actual drone surveying accuracy is created by the entire operation. RTK is only part of that. If you don't understand this, you may introduce RTK yet fail to achieve the expected quality and conclude, "It's strange — we're using high-precision equipment."

First and foremost, the flight plan is crucial. Under conditions such as insufficient image overlap, biased shooting directions relative to the subject, coarse altitude settings at sites with large elevation differences, or forcing flights during windy periods, the reconstruction quality of the model will decline no matter how good the positional information is. In particular, on slopes, steps, around structures, and on developed sites with undulating terrain, flying with the same approach as over flat ground tends to produce shape errors. Before introducing RTK, you should first review whether the flight itself is appropriate for the subject.

The next important point is the verification process. You can use RTK to reduce ground control points, but that does not mean verification points become unnecessary. What is truly frightening on site is not noticing an offset. Even if the data looks visually clean, without verification points you cannot judge how much you can trust it. In practice, it is important to consciously separate the points that provide the reference from the points that check the results. RTK can streamline the work, but that does not mean you can omit quality checks.

Additionally, the handling of coordinate systems and elevations is another area that is easy to overlook. It is not uncommon for horizontal positions to match while only the elevation is off, or conversely for elevations to be close while the entire dataset is uniformly shifted. This tends to happen when the coordinate reference used in the field, the settings during processing, and the method of reconciling with existing results are not aligned. To take full advantage of RTK, acquisition, processing, and use must be connected under the same standards.

Site conditions cannot be ignored. In conditions unfavorable to satellite positioning—such as many tall trees, densely clustered structures, valley terrain with a narrow sky, or environments with many reflections—the stability of RTK can deteriorate. In such places, rather than placing too much expectation on RTK, devising flight methods and rethinking the use of control points can often lead to more stable accuracy. In other words, achieving accuracy is not about choosing a positioning method, but about assembling a measurement approach suited to the site conditions.

If you adopt this perspective, the positioning of RTK changes. RTK is not a magical tool that increases accuracy; it is an important component that supports absolute accuracy and efficiency within the overall operation. When used correctly as a component it can have a significant effect, but if you strengthen only that part while the overall design is weak, the results will not be stable. To achieve results in practice, it is more important to first establish how accuracy will be reproduced in operations than to decide whether to include RTK.

Common Failures When Relying Solely on RTK

One common mistake on sites that have introduced RTK is scaling back verification procedures because people think, "it's high-precision now, so we're fine." RTK is indeed a powerful way to improve the quality of positioning information, but that doesn't eliminate the need for quality checks. On the contrary, the more you reduce the number of control points to save labor, the greater the risk of overlooking discrepancies unless you have designed in advance where to perform the final validation checks.

The second mistake is trusting RTK numbers alone while ignoring site conditions. Satellite positioning tends to be stable in open-sky environments, but it can be easily affected by structures and trees. If you fly when the radio signal environment is unstable, positioning quality can fluctuate and those effects may be reflected in the data. Because such issues can be hard to see in the records, you need a sense for assessing the site’s spatial conditions and judging whether RTK is likely to be stable today.

The third is underestimating the importance of elevation. On site, horizontal positions tend to attract all the attention, but height is important for earthwork volumes, as-built condition, drainage gradients, and slope management. However, even if the plan coordinates are correct, if the vertical reference is not consistent, the results can be difficult to use in practice even if they do not look particularly out of place. In particular, when comparing with existing drawings or past deliverables, failing to handle the vertical reference carefully greatly reduces the reliability of the comparison.

The fourth is that the operating conditions are not consistent each time. One session may be clear with light wind, another may be windy with deep shadows, and if flight altitude and imaging conditions also differ, it becomes difficult to compare measurements of the same site. Even with RTK, if operating conditions fluctuate, the quality of time-series comparisons will not be stable. In ongoing management, what matters is not the highest accuracy of each individual run but aligning comparable conditions.

Such failures arise not because RTK is bad, but because its role was overestimated. RTK is powerful, but it is only one element of the overall system. If this is properly understood, many cases of "we added RTK but it didn't work" can be prevented in advance.

How to determine the need for RTK on site

So, how should you judge whether RTK is necessary on an actual site? One way to think about it is “whether you will overlay that data with other georeferenced data.” For example, if you plan to overlay it with design drawings, existing survey results, or future as-built management data, the value of RTK increases. If the coordinates don’t align well, you’ll need corrections and adjustments every time you compare them, which makes operations cumbersome.

Conversely, if the main purpose is simple records confined to the site or short-term progress sharing, it may not be necessary to prioritize RTK above all else. In such cases, reproducible flight conditions, standardized processing procedures, and preventing missed coverage of the target area are more important. In particular, whether field personnel can produce data of consistent quality each time is influenced more by operational design than by equipment performance.

Also, the size of the site and the burden of ground work are factors to consider. If the site is small and easy to access, it is easier to choose a method that establishes the necessary references from the ground. However, on sites that are extensive, difficult for people to walk around, contain hazardous areas, or are subject to traffic controls or access restrictions, the value of RTK, which can reduce the burden of ground work, increases. It is important to judge not only by accuracy but also from the perspectives of safety and work efficiency.

Moreover, whether the operation is continuous or one-off makes a big difference. If you collect data only once as a one-off, you can often get by with adjustments made each time. However, if repeated use is assumed—monthly, weekly, or by process—you need a mechanism to stabilize the alignment each time. In that context, RTK is meaningful not merely as a way to improve accuracy but as a means to standardize operations.

As a practitioner, rather than thinking in binary terms of "do we need RTK or not," it's far more realistic to consider "for what purpose on this site, to what extent, and how consistently do we want to achieve accuracy?" As a result, there will be sites where RTK should be deployed, and sites where you should first begin by improving flight planning and control point management. Having decision criteria makes both implementation evaluation and operational improvements less likely to waver.

Summary

Whether RTK is necessary for drone surveying is not determined by how new the aircraft is or how many features it has. What matters is the design: what you are measuring for, what decisions the deliverables will be used to make, and how you will reproduce the accuracy required for those purposes. RTK is highly effective at facilitating ties to absolute coordinates, reducing ground work, and enhancing the comparability of ongoing operations, but it alone does not guarantee survey-quality results.

There are three concepts to keep in mind to improve accuracy. The first is to determine the required accuracy by working backwards from the deliverables. The second is to consider absolute accuracy and relative accuracy separately. The third is to build accuracy across the entire operation—not just RTK alone, but including flight, verification, processing, and coordinate management. If these three are clarified, the decision to adopt RTK becomes much clearer.

What is truly needed on site is surveying data that can be used consistently every time. To achieve that, it is not enough to simply introduce a high-precision system; you need the perspective of stabilizing accuracy with methods suited to site conditions. To make drone surveying not merely a temporary trial but a system usable in everyday operations, this way of thinking is the foundation.

And if you want to link the accuracy of drone surveying to operations across the entire site, it is important to consider not only aerial measurements but also ground coordinate acquisition and simple surveying. For example, if you want to easily handle high-precision positional information on site, smoothly connect drone survey results with ground points, or apply the data to construction management and as-built verification, considering options such as LRTK—an iPhone-mounted GNSS high-precision positioning device—can make it much easier to streamline site positioning workflows. Rather than treating drone and ground positioning separately, organizing them from the perspective of how accuracy will be used across the whole site is becoming increasingly important in practice.