What is on-site localization? Why it’s necessary for ICT construction and five basic procedures

When advancing ICT construction, attention tends to focus on equipment selection and the creation of 3D design data, but what commonly trips people up on actual sites is how coordinates are aligned. If the points taken in the initial survey, the positions in the design data, the current position shown by construction machinery or positioning terminals, and the measurements used for as-built management are not aligned to the same reference, they may look correct on a screen but still leave discrepancies on site. The Ministry of Land, Infrastructure, Transport and Tourism’s approach to ICT-utilized construction assumes a flow that uses 3D data from the initial survey through construction, as-built management, inspection, and delivery, and the key to maintaining that continuity on site is on-site localization.

The term on-site localization may sound difficult at first, but essentially it is the work of matching the “coordinates used on site” with the “coordinates held by machines and GNSS.” Especially on sites using RTK-GNSS or network RTK, machine guidance, or mobile positioning, the quality of this initial setup greatly affects subsequent construction accuracy and operational efficiency. Conversely, if this is left vague and work continues, what was intended to replace batter boards can become a redundant double-check, and extra re-measurements may occur when comparing as-built results, reducing the benefits of ICT construction.

Table of contents

• What on-site localization is

• Why it’s necessary for ICT construction

• Situations that require on-site localization

• Five basic procedures for on-site localization

• Common failures on sites that don’t go well

• Tips for stabilizing accuracy in operations

• Summary

What on-site localization is

In practice, on-site localization is also called localization, coordinate transformation, or site coordinate alignment. In documents from the Ministry of Land, Infrastructure, Transport and Tourism, localization is organized as converting GNSS coordinate systems to site coordinate systems, and by performing that conversion, the idea is that thereafter GNSS measurements can automatically yield site coordinate measurements. In other words, the essence of on-site localization is not to use satellite-derived position information as-is, but to adjust it to the coordinate system of the control points and design data that will be treated as correct for that particular construction.

What is important here is that on-site localization is not simply an initial machine setup. Basic design data used in ICT construction includes not only the shape of the construction object but also construction control point information and the coordinate system information to be used. As-built measurement data are managed in comparison to that basic design data. Put differently, if the coordinate assumptions are not aligned, design, construction, and inspection cannot be compared on the same footing. On-site localization is the foundation-building work to align those assumptions.

Another common misconception is that ICT construction refers only to construction using ICT-equipped machinery. In fact, Ministry documents indicate that ICT construction includes both work using ICT construction machinery and work that does not use ICT machinery but implements 3D initial surveys, 3D design data creation, 3D as-built management, and so on. That is why on-site localization must be considered not just for machine operators but as a common foundation including surveyors, construction managers, as-built managers, and data creators.

Why it’s necessary for ICT construction

The first reason on-site localization is necessary is to correctly overlay design data and on-site positions. In machine guidance and as-built verification, real-time awareness of the difference between the current position and the design is valuable. However, if the underlying coordinate system is misaligned, the displayed differences themselves cannot be trusted. Even an apparent discrepancy of several centimeters to several tens of centimeters (several in to several tens of in) can affect slopes, excavation bottoms, spreading, and structure placement, causing rework or unnecessary checks. On-site localization is the work that treats the design’s “here” and the site’s “here” as the same point.

The second reason is to connect the entire construction process. In ICT-utilized projects, a framework is adopted that consistently uses 3D data from the initial survey, design document review, construction, as-built management, inspection, to as-built drawings and record documents. If any one stage in this flow uses a different coordinate assumption, inconsistencies such as “the survey was correct but the construction is off” or “only the as-built has different values” are likely to occur on site. On-site localization is necessary not only to make each task correct but also to seamlessly connect preceding and following processes.

The third reason is to make the labor-saving aspects of ICT construction actually work. Verification documents from the Kanto Regional Development Bureau show effects such as reductions in batter board installation and as-built measurement, and reductions in work related to the initial survey, with examples where ICT methods lead to decreased personnel and man-hours. Guides for MC and MG also organize how having electronic batter boards in the machine can omit batter board installation and checking, reduce workers near heavy machinery, and improve safety. Such labor savings are only realized when the design and machine positions are correctly matched.

The fourth reason is that the nature of positioning technologies changes how preparation should be considered. RTK-GNSS obtains precise positions by receiving correction data from a base station and is well suited to large sites and multi-machine operations because correction data can be distributed to multiple machines. On the other hand, documents indicate that GNSS-based sites require localization for each site. For small-site automatic-tracking total station (TS) examples, technologies are introduced that can determine coordinates by back-sighting without localization. In short, understanding the necessity of on-site localization is indispensable when choosing which positioning method to use.

Situations that require on-site localization

The most typical situation is overlaying the positions of machinery or a rover obtained by RTK-GNSS or network RTK onto design data. For operations that use the current position in real time—bulldozer or excavator machine guidance, on-site as-built confirmation, staking out, point recording—on-site localization is a prerequisite. Especially on sites that deploy multiple moving stations on the machinery side, while sharing correction data is easy, if the initial coordinate alignment is fuzzy it spreads the same offset across the whole operation, making the initial setup particularly important.

The next common case is when multiple surveying methods coexist. Even if the acquisition methods differ—drone, TLS, TS, GNSS, mobile-device point-cloud capture—ultimately comparisons must be made on the same design surface or as-built reference to be meaningful. The Ministry’s documents also present ICT construction as a framework connecting 3D initial surveys, 3D design data, 3D as-built management, and delivery, and they state that regardless of the different measurement techniques used, the reference that must ultimately match is the coordinate baseline. On-site localization serves to translate those heterogeneous data into the same site “language.”

Furthermore, on-site localization is important when the on-site control points are not used as public coordinates directly but are operated as site coordinates adjusted to existing structures or construction conditions. Official materials explain that there can be offsets between the site coordinate system and the GNSS coordinate system, that rotational correction is necessary when control points are rotated, and that on site the control points—shown in the site coordinate system that includes survey errors—are operated as the correct reference, so GNSS must be adjusted to match the site. In other words, offsets are not anomalies but conditions to be absorbed for practical on-site handling.

Once on-site localization is done, the general idea is to continue using that setting for the duration of the project, but that does not mean a one-time setup is sufficient forever. If the work area expands, control point handling changes, equipment used changes, or provisional works or communication conditions change, the appropriateness of the setting needs to be rechecked. Rather than blindly continuing to use the initial transformation, it is important to maintain an attitude of verifying “is this still correct under current site conditions?”

Five basic procedures for on-site localization

1\. Align the coordinate assumptions. Before starting on-site localization, confirm which coordinate system is adopted among the design documents, 3D design data, as-built management methods, and delivery conditions. Ensure the zone number of the plane rectangular coordinate system is correct, identify the vertical datum, determine which points will be treated as construction control points, and check for any site-specific coordinate practices. If these are not clarified at the outset, later observations can be correct yet the settings will be wrong. The fact that basic design data include coordinate system information and construction control point information shows that this confirmation is not a mere clerical task but the core of construction data.

2\. Decide which control points and check points to use. On-site localization associates known site coordinates with coordinates obtained from GNSS or positioning instruments, so the base points must be stable. Do not rely only on easily moved temporary stakes or locations susceptible to construction influence; choose points that can be used continuously on site. Also, separate the points used for localization from those used later for verification so that the quality of the setting can be judged. To avoid overlooking site-wide bias or rotation, the selection of control points is where accuracy starts.

3\. Observe corresponding points and determine transformation parameters. Observe points that are known in site coordinates with GNSS or the equipment in use, and derive translation and rotation corrections from the differences. Official documents indicate that when site control points are rotated, rotational correction is necessary, so simply translating coordinates may be insufficient at some sites. It is also important to confirm basic conditions affecting coordinate calculation—instrument height, antenna height, prism constant, observation posture—at this stage. Treat localization not as pressing a software button but as correctly modeling the site’s geometric conditions to reduce the chance of failure.

4\. Apply the derived transformation settings to all equipment used on site. If construction machinery, rovers, data collectors, as-built verification terminals, photo management, and point-cloud inspection apps operate under different coordinate assumptions, you may find “the machinery matches but the inspection terminal is off” on site. ICT construction produces effects because data are continuous from surveying through construction, as-built management, and delivery. The localization setting must therefore be consistent across all devices used for decision-making on site, not just on a single machine. For multi-unit operations, distinguish whether what is being shared is only correction data or whether coordinate transformation settings are also consistent.

5\. Always perform residual checks and routine inspections. The true value of localization is judged not at the moment of setting but in continued on-site use afterwards. Ministry documents state that the measurement accuracy at the time of localization affects overall subsequent measurement accuracy and present ideas for residual checks and system checks via repeated positioning. In practice, checking discrepancies at verification points not used for the setting, and performing simple checks at known points before starting work—to confirm “can we use this setting today?”—is an effective routine. Treat on-site localization as part of operational management, not just an initial setup.

Common failures on sites that don’t go well

The most common failure in on-site localization is starting work with ambiguous coordinate assumptions. Mistakes such as taking the wrong zone number for the plane rectangular coordinate system, confusing public coordinates with site coordinates, or omitting the vertical datum may seem simple but cause troublesome offsets that are hard to detect until the end. That is because machines and terminals will operate consistently internally after localization, so abnormalities only surface when comparing to external known points or as-built comparisons. If the initial organization is lax, inconsistencies tend to appear after construction has progressed, so do not underestimate pre-work coordinate checks.

Another frequent problem is weak selection and verification of control points. If settings are based on points clustered on one side of the site or on temporary points vulnerable to construction influence, the results may seem correct locally but show large offsets farther away. On sites requiring rotational correction, applying only a translation tends to produce larger differences at the extremities. Treat localization as a site-wide consistency check rather than a point-by-point alignment, and separate setting points from verification points to avoid failure.

Insufficient confirmation of equipment conditions is also common. Elements that affect accuracy differ by positioning technology: antenna height, prism height, reception status of correction information, TS line-of-sight, satellite tracking conditions, and communication status. Automatic-tracking TS is less affected by satellite conditions and is easy to use on small sites, urban areas, or mountainous regions, but typically operates one TS per machine. RTK-GNSS is easier to deploy on multiple units but the quality of correction data and localization affects the whole system. If you operate different technologies with the same assumptions without understanding their characteristics, you will be unable to distinguish whether a problem is due to localization or the positioning method.

One more point to be careful of is not rechecking after a single setup. There are many factors on site that can change coordinate assumptions—work area expansion, temporary relocation, damage to known points, communication changes, or personnel changes leading to setting mistakes. Even if the initial localization was correct, that condition may not be maintained. Continuing short pre-work checks without skipping them is the quickest way to reduce re-measurement and rework.

Tips for stabilizing accuracy in operations

To stabilize on-site localization, first consolidate the “coordinates considered correct on site” on one sheet, in paper or data form, so anyone can understand them the same way. Sharing the construction control points, coordinate system in use, check points, key equipment settings, and pre-work check procedures reduces decision variance when personnel change. ICT construction differentiates quality by data linkage rather than individual skill, so making coordinate rules the site’s common language improves operational efficiency.

Next, use positioning technologies appropriately according to site conditions. On open sites deploying multiple machines, RTK-GNSS is often efficient; in urban or mountainous areas with poor satellite conditions or on small sites, automatic-tracking TS systems may be easier to operate. The important thing is not to rigidly decide which is superior but to determine what the site prioritizes. Comparing multi-unit deployment, initial preparation, line-of-sight, communication, required accuracy, and staffing when selecting a method will design the localization burden appropriately.

Also, instead of performing complex reconfigurations each time, adopt quick verification routines that can be done repeatedly. Simple checks at known points, quick comparisons with the design surface, adding coordinates to photo records, and visual comparisons with as-built verification terminals—stacking up small checks prevents large offsets from growing. When on-site localization is stable, staking out, as-built checks, and information sharing among stakeholders operate with the same coordinate language, speeding up on-site decision-making.

Summary

On-site localization is the work of aligning the coordinates held by GNSS and positioning devices to the coordinates that will be treated as correct on that site. In ICT construction, data are connected from initial survey, 3D design data, construction, as-built management, inspection, to delivery, so if coordinate assumptions shift once, the effect can spread to subsequent processes. That is why carefully following the basic procedures—confirming coordinate assumptions, selecting control points, observing corresponding points, applying settings, and checking residuals—not only secures accuracy but also leads to labor reductions such as fewer batter boards, less re-measurement, and faster information sharing.

If you want to make on-site verification work lighter, it is effective to combine not only stationary equipment but also mobile positioning means with high mobility. For example, options such as LRTK Phone allow RTK positioning simply by preparing a smartphone, or high-precision positioning devices that attach to an iPhone or iPad. Even if not used as the primary unit for large-scale ICT construction, incorporating them as tools for site verification, position recording, high-precision geotagging of photos, and faster stakeholder sharing can make coordinate operations including on-site localization more accessible.