RTK Reproducibility Test: A Simple Field Method to Verify Accuracy

Table of Contents

• What is RTK positioning reproducibility and why is it important?

• Simple field test methods for checking reproducibility

• Evaluating the accuracy of results: how close is close enough to pass?

• Tips and precautions for improving reproducibility

• The significance of high reproducibility (applications to design, construction, and maintenance)

• Introducing LRTK for simple, highly reproducible surveying

• FAQ

When talking about RTK positioning, the first thing people usually focus on is “positional accuracy.” RTK (Real-Time Kinematic) is a technique that applies real-time corrections to GNSS (Global Navigation Satellite System) position errors and can improve accuracy down to a few centimeters (a few inches). Dedicated RTK-GNSS equipment has dramatically increased productivity in surveying work, and recently smartphone-compatible RTK systems have attracted industry attention. Therefore, when introducing RTK, attention often goes to “how high an accuracy can be achieved.”

However, for practical field use of RTK, there is a point even more important than catalog accuracy. That is reproducibility — getting the same result whenever you measure. No matter how high the instantaneous numerical accuracy is, if results vary each time you measure, they cannot be trusted. Experienced surveyors habitually remeasure important points at different times or verify them using different methods, because even when equipment shows high-precision values, environmental factors or configuration mistakes can occasionally produce incorrect values. High reproducibility gives the reassurance of “the same result every time,” greatly improving the reliability of measurements. Conversely, low reproducibility — i.e., results that vary each time — will undermine confidence in the field even with high-performance equipment.

What is RTK positioning reproducibility and why is it important?

Reproducibility means obtaining nearly the same positioning result each time the same point is measured. In RTK, the theory allows for centimeter-level accuracy, but that accuracy may not manifest on every measurement. Major factors that affect reproducibility include the following.

• Satellite geometry: The relative positions of satellites change by time of day and affect positioning accuracy. Poor satellite geometry tends to increase errors.

• Atmospheric effects: Signal delays caused by the ionosphere and troposphere vary with time of day and solar activity and appear as small differences in error.

• Surrounding environment (multipath): Signal blockage or reflection by tall buildings or trees can introduce errors or make it difficult to obtain a fixed solution. Measurement stability varies by location.

• Reference station and coordinate system differences: Using different reference stations each time or having shifted coordinates for reference points causes offsets. Network RTK normally provides a unified coordinate system, but if you use your own reference station, be careful about the reference station’s coordinate settings.

• Operational errors: Human errors such as incorrect correction settings or insufficient initialization can occasionally cause positioning discrepancies.

For example, measuring the same point in the morning and afternoon can yield differences of several centimeters due to different satellite geometry. However, in a system with high reproducibility, such differences remain minimal (within a few centimeters), and both results can be used without issue. In a low-reproducibility system, time-of-day variations can produce large error fluctuations, leading to situations where “the coordinates measured yesterday do not match today’s coordinates.”

Ensuring reproducibility is indispensable for reliable surveying in the field. The next section presents a simple on-site test method to actually verify reproducibility.

Simple field test methods for checking reproducibility

You can easily check RTK positioning reproducibility in the field without expensive equipment or specialized experimental setups. Here is a simple procedure that even beginners can follow.

• Test point selection: Choose a point to test positioning reproducibility. If possible, use a control point with known accurate coordinates (e.g., an official survey marker), but if none exists, pick a spot on site with good visibility. Drive a nail or place a stake as a marker so you can relocate the exact same point later. Selecting an open location without tall buildings or obstacles will minimize error sources.

• Multiple measurements at different times: Measure the point with RTK multiple times. For example, measure twice, once in the morning and once in the afternoon, and preferably again on the following day. Each time, set the antenna back on the point and start positioning. For each measurement, wait until the RTK solution becomes a fixed solution (Fix) and stabilizes before recording the measurement. Taking data for several dozen seconds per observation and averaging is even better (averaging over a short period reduces the influence of instantaneous noise). The important thing is to obtain multiple results measured at different times. If you use your own mobile reference station, make sure to set the same reference station coordinates each time or use the same network RTK service (different reference station settings will introduce offsets that prevent comparison).

• Recording and comparing measurement results: Record the coordinates obtained at each measurement (latitude/longitude or planar coordinates X, Y, Z) and compare them. When you measure the same point two or more times, compute the differences between the results. Use the first measurement as the reference and calculate the differences for subsequent measurements (ΔX, ΔY, ΔZ). Determine the horizontal and vertical differences. If you measure three or more times, check the differences for all combinations. Writing them down or entering them into a spreadsheet makes difference calculations easy.

• Confirming results and additional verification: Check whether the differences between measurements are very small. For example, if you measure the same point three times and all mutual differences are within a few centimeters (a few inches), reproducibility can be considered good. If the differences are large (exceeding the criteria described below), consider possible causes: was satellite reception poor during one measurement period, did a fixed solution degrade to a float solution (Float), or was there an error in the reference station settings? If necessary, take one or two additional measurements or retest under different environmental conditions. If you used a point with known accurate coordinates, compare the measured results to the known coordinates to check absolute error. If you clearly observe errors of several centimeters or more, there may be issues with equipment settings or usage.

With the above procedure, you can quickly and simply verify RTK reproducibility. It only requires measuring the same point at different times, so you can perform it in the field without additional cost.

Evaluating the accuracy of results: how close is close enough to pass?

How consistent must multiple measurements from the test be to judge that reproducibility is good? Specific standards depend on the purpose of the survey, but here are general guidelines.

• Horizontal position differences: Under good conditions, RTK-GNSS surveying is expected to achieve horizontal errors within about 2–3 cm (0.8–1.2 in). Therefore, when repeatedly measuring the same point, planar position differences of a few centimeters (generally 3 cm (1.2 in) or less) can be considered passing. Stability in the 1 cm (0.4 in) range is ideal.

• Vertical (elevation) differences: Vertical errors tend to be larger than horizontal and, depending on conditions, differences approaching 10 cm (3.9 in) can occur. Generally, vertical differences within 5 cm (2.0 in) are considered good. It is not uncommon to see cases where the horizontal difference is 2 cm (0.8 in) while the vertical difference is 4 cm (1.6 in). The important thing is to confirm that differences, including vertical, remain within ranges acceptable for practical use.

For example, if as-built control for civil engineering requires an accuracy of about ±5 cm (±2.0 in), then reproducibility tests yielding measurement differences of 2–3 cm (0.8–1.2 in) between repeat measurements would be more than sufficient. Conversely, if horizontal differences frequently exceed 5 cm (2.0 in) or vertical differences vary by 10 cm (3.9 in) or more, this is a red flag. Such large differences suggest that RTK is not performing to its expected capability; root-cause investigation is necessary, and the method may not be suitable for high-precision surveying as-is.

To evaluate absolute accuracy, compare measurements at known reference points when possible. Compare RTK results to the control point’s accurate coordinates and check whether errors consistently remain within a few centimeters. Large systematic offsets may indicate calibration issues or incorrect reference station coordinates.

In summary, a practical guideline for good reproducibility is that repeated measurements show only very small differences (on the order of a few centimeters (a few inches)). Differences of tens of centimeters indicate poor reproducibility and necessitate review of equipment settings and measurement procedures.

Tips and precautions for improving reproducibility

If reproducibility test results are unsatisfactory, there are many practical measures you can take to improve measurements. The following points help increase RTK reproducibility and yield stable results.

• Use multi-GNSS and dual-frequency receivers: Use GNSS receivers that support multiple satellite constellations (not just GPS but also GLONASS, Galileo, QZSS, etc.) and multiple frequency bands. Increasing the number of available satellites makes it easier to maintain a fixed solution even in urban or mountainous areas, improving reproducibility.

• Pay attention to satellite geometry: Consider satellite geometry at the measurement time. Choosing times with low PDOP (Position Dilution of Precision) values, which indicate less degradation of accuracy, results in smaller and more stable errors. Use apps or websites that predict satellite geometry and, if possible, survey during times with favorable PDOP.

• Choose locations with good visibility: Survey in open areas with a clear view of the sky rather than under buildings or trees. If obstacles are unavoidable, you can first obtain a fixed solution in an open location and then move to the measurement point (getting a good initial condition can help stability). Reducing the risk of signal interruption makes consistent results easier to obtain.

• Maintain fixed solutions: During measurement, continuously confirm that RTK maintains a fixed solution (Fix). Check the survey app to ensure Fix is maintained or that position coordinates are not fluctuating. If the solution reverts to Float, do not use that interval’s data; wait until Fix is restored before recording. Comparing only data obtained under fixed-solution conditions improves reproducibility assessment accuracy.

• Handling reference stations: If you set up your own reference station, place it on the same known point each time or at least set precisely the same coordinates. If reference station coordinates change by day, relative positioning may be accurate but absolute coordinates will shift, preventing reproducibility evaluation. Also, the shorter the baseline length between the reference and rover, the better the accuracy and stability; place the reference station as close to the survey area as practical.

• Unify coordinate systems: When comparing data across multiple days, ensure all records use the same coordinate system and datum. For example, if you record one day in a geodetic datum (JGD2011) but the next day in a local arbitrary coordinate system, the numbers cannot be directly compared. Confirm coordinate system settings on equipment and apps before surveying.

• Multiple measurements and averaging: Improving reproducibility can be achieved by measuring the same point multiple times and averaging. This reduces the effect of transient errors and decreases result variability. For example, measuring a point continuously for one minute and recording the averaged coordinates will yield a more stable value than a single instantaneous measurement. Averaging measurement results also helps reduce differences when comparing across days.

• Cross-check with other methods: If you have any doubts about RTK results, cross-check using other measurement methods. For example, use a total station to compute coordinates from distances and angles, or verify elevation with leveling. Verifying important control points using different methods provides additional confidence.

Following these points will substantially improve RTK reproducibility and yield stable field positioning. In short, the keys to ensuring reproducibility are “measures to maintain centimeter-level stability from start to finish” and “the habit of verifying whether measured values are plausible by repeating measurements.”

The significance of high reproducibility (applications to design, construction, and maintenance)

Ensuring RTK reproducibility provides benefits beyond simply having confidence in a surveying method; it brings major advantages at each project stage — design, construction, and maintenance.

In the design phase, topographic surveys and field investigations provide the data to create design drawings. If survey data are reproducible, additional surveys or measurements by different teams later will produce consistent coordinates in the same system. For example, if terrain data acquired one day aligns perfectly with data taken the next day, designers can use them without worry. Low reproducibility can cause daily inconsistencies in survey deliverables and lead to contradictions on design drawings. High reproducibility ensures data consistency and supports smooth design work with fewer rework cycles.

In the construction phase, RTK-GNSS is used for staking out positions (setting stakes or layout marks) and as-built control. When reproducibility is ensured, the positions set by surveyors in the morning and rechecked in the afternoon coincide, allowing confident progress in construction. In ICT-equipped machine control, consistent reference points ensure no control drift. Also, when checking installation positions for structural elements during construction, stable survey values reduce ambiguity over whether a discrepancy is an error or a construction mistake. As a result, rework is prevented and schedules and quality are improved. Low-reproducibility measurements tend to trigger unnecessary remeasurements and corrections; high reproducibility enables reliable construction control with minimal effort.

In maintenance, reproducible survey data are crucial. For infrastructure inspections and ground-subsidence monitoring, detecting small changes over time is essential, but only if the survey error itself is small and stable. For example, if bridge pier elevations measured by RTK vary by several centimeters each time, you cannot tell whether changes are real or just measurement noise. Only when reproducible measurements around 1 cm (0.4 in) or better are achieved can you reliably state that “this measurement shows a settlement of X mm since the previous check.” High reproducibility increases detection sensitivity for long-term changes and improves the certainty of maintenance decisions. It also ensures data continuity and trust when multiple people perform maintenance measurements at different times.

Thus, improving RTK reproducibility adds value across design, construction, and maintenance. Consistent survey data underpin digital construction and infrastructure management. In ICT-driven construction initiatives such as the Ministry of Land, Infrastructure, Transport and Tourism’s i-Construction, RTK-GNSS use is fundamental, and reproducibility is a key element of accuracy management.

Introducing LRTK for simple, highly reproducible surveying

How can you ensure reproducibility while making surveying itself easy? One solution is the smartphone-compatible RTK solution “LRTK.” LRTK consists of a palm-sized high-precision GNSS receiver that attaches to a smartphone and a dedicated app, and its standout feature is far greater ease of use compared with traditional fixed RTK equipment.

Using LRTK in the field makes it even easier to secure reproducibility. LRTK displays and records positioning results in real time on the smartphone as maps and numerical values, so you can immediately confirm onsite whether repeated measurements of the same point are stable. In practice, repeated measurements at the same point frequently remain within a few centimeters (a few inches). The app can also average positioning data over a set time, and using this feature can improve precision to millimeter-level accuracy (1 mm = 0.04 in), making it possible to enhance accuracy substantially. Because verifying important control points is easy, LRTK enables the collection of highly reproducible field data.

Additionally, LRTK is easy for anyone to use. It minimizes the complex configuration typical of traditional RTK devices; following a few taps in the smartphone app is enough to proceed with surveying. No cumbersome base station setup is required — you can start work on site with just a smartphone and an LRTK receiver. This ease of use allows sites without dedicated surveying staff to perform required positioning tasks. For experienced surveyors, too, LRTK reduces time spent on equipment setup and post-survey data processing, allowing more focus on decision-making. In an era of labor shortages, LRTK — enabling anyone to perform accurate, reproducible surveying — is becoming a new standard for “easy surveying.” Combining high accuracy, reproducibility, stability, and consistency with ease of use, LRTK has the potential to significantly transform future field surveying workflows.

When considering RTK equipment, look beyond catalog accuracy to reproducibility and ease of operation that allow you to reliably realize that accuracy. Including LRTK as an option among next-generation simple-surveying solutions can dramatically improve the reliability and efficiency of field surveying.

FAQ

Q1. How much accuracy can be expected from RTK surveying? A. Generally, RTK-GNSS surveying can achieve planar accuracy of a few centimeters and vertical accuracy of a few centimeters to a dozen centimeters under good conditions. Smartphone-based RTK can achieve nearly the same levels of accuracy if correction information and settings are properly configured. In practice, LRTK receiver-alone results typically fall within about 1–2 cm (0.4–0.8 in) of error, and averaging data over a set time can yield sub-1 cm (<0.4 in) accuracy. However, note that the highest accuracy is not guaranteed at all times. Accuracy varies with satellite geometry, radio conditions, baseline length, etc., so for critical measurements confirm that RTK is stably in a fixed solution before proceeding.

Q2. Can RTK surveying be performed in urban canyons or forests? A. In urban canyons surrounded by tall buildings or in dense forests, satellite reception degrades, making it difficult to maintain RTK accuracy and a fixed solution. Signals can be blocked or reflected by buildings or foliage, causing instability and increasing the chance of a Fix reverting to a Float. However, modern receivers supporting multi-GNSS and multi-band can increase the number of usable satellites, so accuracy degradation in urban or forested environments is better controlled than in the past. For example, you can temporarily move to an open area to perform initial positioning (acquiring Fix), then proceed while supplementing missing data with the smartphone’s inertial sensors or AR features. That said, RTK is not applicable where satellites cannot be received at all (e.g., tunnels or indoors). In such cases, switching to other methods — total station surveys, IMUs (inertial measurement units), photogrammetry, or SLAM (simultaneous localization and mapping) — is necessary. In short, don’t stick to RTK in environments where satellite positioning is difficult; flexibly combine other surveying methods as appropriate.

Q3. Does LRTK support QZSS “Michibiki” CLAS? A. Yes. There are LRTK receiver models that support Japan’s Quasi-Zenith Satellite System (QZSS) centimeter-level augmentation service (CLAS). CLAS-capable units can directly receive augmentation signals from QZSS satellites and perform real-time centimeter-level positioning even at sites where access to the Geospatial Information Authority of Japan’s continuous operating reference station network or Ntrip internet connections is not available. This is very reassuring for mountain or offshore locations with cellular reception gaps, provided the sky is sufficiently open. Note that the CLAS service area is generally limited to within Japan.

Q4. How will site operations change after introducing LRTK? A. Traditionally, layout work and as-built control on site required specialist survey staff, and sometimes external survey contractors were engaged. This process often introduced time lags because construction proceeded only after survey reports were delivered. Verification against design drawings sometimes relied on judgment or experience, increasing the risk of misreading or oversight and consequent rework. After introducing LRTK, site staff can perform layout and measurements themselves and immediately verify and share results via AR displays. Real-time construction control speeds decision-making and enables early error detection and immediate correction. Furthermore, tasks that previously required two or three people can often be completed by one person, increasing flexibility in staffing. Reports from actual sites say that “waiting times and rework have been greatly reduced, improving overall project efficiency.”

Q5. Can less experienced personnel use LRTK? A. Yes. LRTK is designed for intuitive operation, so it is accessible to less experienced users. By tapping the measurement start button in the smartphone app, correction reception and calculations are performed automatically, so complex settings are unnecessary. Understanding basic GNSS and datum concepts helps, but specialized knowledge is not strictly required to get started. LRTK allows non-specialists to perform position measurements and as-built checks with a reasonable degree of accuracy, broadening the range of tasks that site staff can handle. Of course, proficiency increases the scope of use, but even beginners gain significant benefit from being able to achieve reproducible centimeter-level positioning easily.