Accuracy and Repeatability of RTK: Which Do You Really Need on Construction Sites?

Table of Contents

• What are accuracy and repeatability?

• RTK positioning accuracy: theoretical values vs. real-world performance

• Why repeatability matters on site

• What is required on construction sites?

• Simple surveying with LRTK

• FAQ

What are accuracy and repeatability?

Accuracy indicates how close a measured value is to the “true value.” Conversely, repeatability indicates whether measurements taken repeatedly at the same point yield nearly the same result each time. For example, if repeated surveys of the same point produce almost identical results, the measurement is said to have high repeatability. Even if an instrument has high accuracy, low repeatability means its readings vary each time and cannot be trusted. On the other hand, high repeatability provides consistent results even with some error, which gives confidence on site.

A useful image is: accuracy is how close arrows hit to the bullseye’s center; repeatability is whether the arrows all cluster in the same place. Ideally, measurements should be both close to the true value and highly repeatable, but in real-world positioning these two may not always coexist due to environmental conditions. For reliable surveying on site, it is not enough to show a high accuracy number; what matters is whether that accuracy can be reproduced consistently.

RTK positioning accuracy: theoretical values vs. real-world performance

RTK (Real Time Kinematic) is a technology that corrects GNSS (Global Navigation Satellite System) position errors in real time to achieve centimeter-level (cm level accuracy (half-inch accuracy)) precision. Using dedicated RTK-GNSS equipment dramatically improves surveying efficiency, and recently smartphone-compatible RTK systems have appeared and attracted attention. When adopting RTK, many people focus on “how high an accuracy can be achieved.”

Theoretically, RTK positioning is said to yield planar position errors of several centimeters (several in) and vertical errors of several centimeters to around a dozen centimeters (several in to about a few in). Under ideal conditions, errors on the order of about 1 cm (0.4 in) can be expected. However, note that catalog accuracy is not always achievable. RTK accuracy varies on site due to factors such as:

• Satellite geometry: The configuration of satellites overhead changes with time and affects positioning accuracy. Poor geometric distribution of satellites tends to increase errors even with RTK. For example, measuring the same point in the morning and afternoon can produce differences of a few centimeters (a few in) due to different satellite geometry.

• Atmospheric effects: The ionosphere and troposphere affect satellite signals as they travel to Earth, and these effects vary with time of day and solar activity. Thus, even at the same location, slight differences in positioning error may occur depending on the time.

• Surrounding environmental factors: The environment around the survey site affects accuracy. In urban canyons or forests, satellite signals can be blocked or reflected, reducing RTK accuracy. Once a fixed solution (Fix) is lost, high accuracy cannot be maintained and the solution may revert to float (Float), increasing error.

• Differences in reference points or correction data: If you set up your own base station (reference point) for RTK, any offset in the base station coordinates will introduce an offset error in results. Network RTK (receiving corrections from a national reference network) uses a regionally unified coordinate system, but when using individual base stations, configuration can affect accuracy.

• Human error: Positioning accuracy is also impacted by human factors such as incorrect correction settings, insufficient initialization, or equipment malfunctions. Even high-precision RTK equipment cannot deliver correct accuracy if operated or configured improperly.

Thus, RTK accuracy does not always match theoretical values and varies with environment and operational conditions. The important question is whether the required accuracy for the site can be obtained stably. The next section examines this ability to “reliably obtain the same accuracy,” i.e., repeatability.

Why repeatability matters on site

In precision surveying, ensuring repeatability—that the same result is obtained whenever you measure—is extremely important. No matter how momentarily precise a positioning value is, if results vary each time you measure, the data cannot be trusted. Experienced surveyors will often re-measure important points at different times or verify with different methods, because they know that even the best equipment can occasionally display incorrect values. Satellite signal conditions or setting errors can sometimes cause aberrant results, so the habit of rechecking exists for this reason.

High repeatability gives the confidence that “the same result is obtained every time,” greatly improving the reliability of measurement data. Conversely, instruments with low repeatability produce inconsistent results, leaving you unsure which measurement to trust. Even if catalog specs indicate high accuracy, they are meaningless if the equipment cannot be trusted on site.

How can you ensure repeatability on site? Attention to the following points will help achieve highly repeatable surveys:

• Verify with multiple measurements: For important control points, if possible observe them multiple times at different times of day and confirm that the results are consistently the same. Measure the same point in the morning and afternoon or on different days and check whether differences are within acceptable limits.

• Periodic checks with known points: If there is a known control point (a point whose coordinates are known accurately) near the site, it is useful practice to measure that point before, during, or after work to confirm instrument accuracy. Consistently obtaining the correct value at a known point indicates no drift in equipment or correction data.

• Respond to anomalies: If repeated observations show large discrepancies, try restarting equipment and checking environmental factors (obstructions or radio conditions). If the problem persists, consider alternative methods such as re-measuring with a total station. Do not accept anomalous values; identify the cause.

By implementing these checks and practices, stable repeatability can be obtained. Only when repeatability is ensured does the high accuracy of RTK realize its true value.

What is required on construction sites?

So far we have explained “accuracy” and “repeatability,” but which is actually required on construction sites? In short, both are important. However, if forced to prioritize, construction sites demand stable, highly repeatable positioning more.

For construction, the most important qualities are the consistency and reliability of survey data. If measurements are stable each time, small errors are less likely to affect comparisons with design values or quality control. Conversely, if coordinates measured on different days disagree, the site can be thrown into confusion. For example, when laying out foundations using RTK, it is problematic if positions measured today shift when re-measured tomorrow. The key to smooth operations on site is always being able to reproduce the same results.

Stable operation is also essential for leveraging surveying on site. Even if theoretical accuracy is high, if RTK positioning drops or becomes unstable during work, construction will be interrupted. High-precision equipment may not perform well in urban areas surrounded by tall buildings or in mountain areas with no network coverage. Whether equipment can continue to measure stably in such environments is an important consideration.

Moreover, data consistency cannot be overlooked. If point cloud data measured on different days are misaligned, contradictions arise when integrating drawings or 3D models. Survey results must be recorded in the same reference system (for example, a geodetic datum like WGS) and be consistent so that data integration does not produce discrepancies. Only with repeatable positioning and unified survey data can highly accurate survey results be safely used in construction.

In short, on construction sites what matters more than “how high the spec accuracy is” is the ability to reproduce that accuracy stably and to keep the system usable in the field. Of course, smaller positioning errors are better, but on site, “consistently staying within a few centimeters (a few in) of error” is often more valuable than marginal improvements of a few millimeters (a few hundredths of an inch). When evaluating purchases, look beyond catalog numbers and consider repeatability, stability, and data consistency from a practical, on-site perspective.

Simple surveying with LRTK

How can you meet these “more important than raw accuracy” needs while making surveying easier? One answer is our smartphone-compatible RTK solution, LRTK, developed by our company. LRTK consists of a small high-precision GNSS receiver that attaches to a smartphone and a dedicated app. With a palm-sized receiver mounted on a smartphone and the app launched, you can start centimeter-level (cm level accuracy (half-inch accuracy)) positioning without complicated base-station setup.

Main features of LRTK:

• High repeatability and real-time verification: LRTK displays and records positioning results on the smartphone screen in real time. Re-measuring the same point and comparing results is easy, allowing you to confirm measurement stability (repeatability) on the spot. Data are automatically saved to the cloud for later review or team sharing. This reduces discrepancies when multiple people survey separately and ensures everyone on site refers to the latest unified data.

• Excellent positioning stability: The LRTK receiver is a multi-GNSS unit that supports not only GPS but GLONASS, Galileo, and QZSS (Michibiki). It is also multi-band, using multiple frequency bands to capture many satellites even in urban canyons or forests, making it easier to obtain a stable fixed solution. In mountain areas without network coverage, models that support Japan’s quasi-zenith satellite “Centimeter-class augmentation service (CLAS)” can receive correction data directly from satellites to maintain high accuracy even without an internet connection. Where communication is available, it can also connect to network RTK via Ntrip to get regional correction data from reference networks. By flexibly switching between satellite augmentation and network corrections, LRTK realizes uninterrupted positioning in various field conditions.

• Data consistency and sharing: LRTK’s smartphone app and cloud service are integrated, saving recorded coordinates, photos, and notes to the cloud. All points are handled in a unified reference coordinate system (such as a geodetic datum), and centralized cloud management prevents inconsistencies when data are later combined. Because everyone on the team can access the same latest data, data consistency is maintained even when multiple people survey. Information captured on site can be shared immediately with the office or other teams, preventing oversights and rework.

• Simplicity and versatility: LRTK minimizes specialized initial setup and is designed for intuitive use by anyone. There is no need to set up your own base station; with a smartphone and an LRTK device you can start surveying as soon as you enter the site. Measurement starts with one tap following the app guidance; no complicated operations are required. Because of this ease of use, required staking-out and as-built measurements can be performed even without a dedicated surveyor on site. Experienced surveyors also benefit from reduced time for equipment preparation and post-processing, allowing them to focus on management and decision-making. In an industry chronically short on manpower, LRTK enables anyone to perform accurate, repeatable surveying, contributing to productivity improvements and becoming a new standard for “simple surveying.”

Combining accuracy, repeatability, stability, and consistency while remaining easy to use, LRTK can significantly transform surveying styles on future construction sites. When selecting RTK equipment, consider not only catalog numbers but also repeatability, operability, and data consistency to fully realize accuracy—consider adding the next-generation simple surveying system LRTK to your options.

FAQ

Q1. What level of accuracy can RTK positioning achieve? A. Generally, RTK-GNSS surveying under favorable conditions yields planar accuracy on the order of several centimeters (several in) and vertical accuracy from several centimeters to around a dozen centimeters (several in to about a few in). While a fixed solution is maintained, errors are typically within about 1–2 cm (0.4–0.8 in). Recent smartphone-compatible RTK systems can achieve accuracy comparable to dedicated equipment. In practice, LRTK alone generally produces errors of about 1–2 cm (0.4–0.8 in), and averaging data over time has shown sub-1 cm (<0.4 in) accuracy. However, accuracy fluctuates with satellite geometry and radio conditions, so for critical measurements confirm that a stable fixed solution is maintained before proceeding.

Q2. Can RTK surveying be performed stably in urban canyons or forests? A. In urban areas surrounded by tall buildings (so-called “urban canyon” environments) or in dense forests, receiving stable satellite signals is difficult, making it hard to maintain RTK accuracy and a fixed solution. Signals can be blocked or reflected, causing position instability, larger errors, or reversion from Fix to Float. Using a multi-GNSS, multi-band receiver to increase the number of available satellites can mitigate accuracy degradation to some extent. Another approach is to move temporarily to an open area to obtain initial positioning (Fix), then continue measuring using the smartphone’s inertial sensors or AR markers to supplement positioning. However, RTK cannot be applied where satellites cannot be received at all (e.g., inside tunnels or buildings). In such cases, alternative methods—short-range radio-based ranging between local reference points and roving units, IMU (inertial measurement unit) dead reckoning, photogrammetry, or SLAM (image-based localization)—must be combined. In short, where satellite positioning is difficult, do not insist on RTK alone; flexibly choose surveying methods including conventional total stations as appropriate.

Q3. Is RTK possible at sites without internet connectivity? A. There are ways to perform RTK positioning in remote mountain or other areas without network coverage. In Japan, the quasi-zenith satellite system Michibiki provides the CLAS (Centimeter-class augmentation service); compatible receivers can receive correction information directly from satellites to achieve centimeter-level (cm level accuracy (half-inch accuracy)) positioning even without internet. LRTK offers CLAS-compatible models that can maintain high accuracy in open-sky remote sites (note: CLAS is a Japan-only service). Another method is PPK (Post-Processing Kinematic), where base station data are recorded in advance and applied later, enabling high-precision results through post-processing even without on-site communications.

Q4. How will introducing LRTK change site work? A. Traditionally, many surveying and as-built management tasks were delegated to specialist personnel or external vendors, and work often waited for survey reports before proceeding. Verifying as-built conditions against design drawings sometimes relied on staff experience or intuition, causing rework if errors occurred. With LRTK, site personnel can perform required staking-out and measurements themselves and immediately view and share results via AR on a smartphone or tablet. Waiting times and interruptions for surveying decrease, enabling real-time construction management and faster decision-making. Tasks that used to require 2–3 people can often be completed by one, easing personnel allocation. Sites that adopted LRTK report significantly reduced surveying wait times and rework, contributing to shorter schedules and improved efficiency.

Q5. Can less-experienced staff handle LRTK? A. Yes. LRTK is designed for ease of use by beginners. The dedicated smartphone app is intuitive, and positioning starts with one tap while correction acquisition and application are handled automatically. Thus, surveying can begin without troublesome configuration. However, understanding basic surveying concepts such as latitude/longitude and datums will help use results more effectively. LRTK enables personnel without deep surveying expertise to perform reasonably accurate positioning and as-built checks, expanding the range of tasks that can be handled on site and improving the speed and quality of construction management. Experienced users also gain from reduced setup and post-processing time, contributing to overall site efficiency.