How to Achieve Centimeter Accuracy (cm level accuracy (half-inch accuracy)) for Indoor Warehouse Positioning and Use Cases

Table of contents

• Technologies that enable centimeter accuracy (RTK-GNSS・IMU・LiDAR・SLAM)

• How to achieve seamless positioning in mixed indoor-outdoor environments

• Use cases for centimeter positioning in warehouses - Warehouse inspections - AR applications - Stake positioning guidance - 3D surveying - Tracking

• Benefits of introducing positioning (efficiency, safety, cost)

• Use of LRTK

• FAQ

Technologies that enable centimeter accuracy (RTK-GNSS・IMU・LiDAR・SLAM)

To obtain centimeter-class (cm level accuracy (half-inch accuracy)) positioning accuracy in warehouses or construction sites, several advanced technologies must be combined. The principles and characteristics of each technology are explained in an easy-to-understand way.

• RTK-GNSS (Real-Time Kinematic): This method obtains highly accurate position information by correcting satellite positioning (GPS, GLONASS, etc.) errors in real time. A reference station (base station) is installed at a known point, and the error information of the satellite signals received at that point is sent to a rover. The rover (the receiver used by the worker) applies that correction information to determine its position, reducing errors that were typically about 5–10 m (16.4–32.8 ft) with standalone GPS to a few centimeters (cm level accuracy (half-inch accuracy)). Using RTK-GNSS allows surveying over large outdoor areas with design-level precision. However, it requires a communication environment to send and receive correction information, and performance can degrade when out of communication coverage.

• Position estimation with IMU (Inertial Measurement Unit): An IMU combines accelerometers and gyroscopes to estimate travel distance and orientation from the device’s acceleration and angular velocity. By integrating IMU data from a recently known position, the current relative position can be determined. IMUs can autonomously update position without relying on external signals or markers, allowing continuous self-positioning immediately after leaving GNSS coverage. However, small measurement errors in IMUs accumulate and cause drift, so standalone long-duration or long-distance use gradually reduces accuracy. Therefore, periodic correction (reset) using external information such as GNSS is necessary. For short durations, however, IMUs can maintain high accuracy, making it feasible to “use the high-precision position obtained outdoors as a starting point, then bridge indoors with an IMU.”

• Distance measurement and environment scanning with LiDAR: LiDAR emits laser light and measures the reflections to obtain distance measurements at high frequency. By collecting many distance points, the 3D shapes of surrounding structures such as walls and shelves can be captured in detail. Because LiDAR can perform millions of distance measurements in millisecond intervals, it can acquire target distances with centimeter accuracy (cm level accuracy (half-inch accuracy)). Scanning the environment with LiDAR while moving enables creation of high-precision point cloud maps, and one can estimate one’s position on that map (self-localization). However, LiDAR alone requires map matching and needs initial position information or combination with other sensors to tie the map to absolute coordinates (such as latitude and longitude).

• Self-localization with SLAM: SLAM stands for Simultaneous Localization and Mapping. It uses feature points from camera images or LiDAR to estimate the device’s position while simultaneously building a map of the surroundings. SLAM is widely used in autonomous vehicles, AMRs (autonomous mobile robots) operating inside warehouses, and AR-capable smart glasses. Its advantage is that it can determine self-position in unknown environments without relying on GNSS or other external signals. However, when the initial absolute position is unknown, the generated map’s coordinate frame is undefined, so later alignment to a known coordinate system is required. Over time, map and position errors accumulate, but techniques such as loop closing—returning to previously visited areas to cancel accumulated errors—can maintain accuracy. SLAM’s accuracy and stability improve when combined with other sensors like IMUs or known position markers.

• UWB (Ultra-Wideband) positioning: UWB is a wireless technology using wide frequency bands at several GHz, transmitting very short pulses at the nanosecond level and calculating distance by measuring time-of-arrival differences at multiple antennas. This method can achieve real-time positioning with about 10 cm (3.9 in) accuracy in indoor environments such as factories and warehouses. By installing several fixed transmitters (anchors) in the warehouse and equipping workers or vehicles with small tags, absolute coordinates within the work area can be obtained. UWB causes little interference with other wireless systems (such as Wi-Fi) and delivers stable accuracy, but the need to install dedicated equipment and perform initial calibration incurs cost.

By combining the above technologies, centimeter-class (cm level accuracy (half-inch accuracy)) positioning can be realized both indoors and outdoors. Because each technology has its strengths and weaknesses, fusing multiple methods complements their shortcomings and enables stable, high-precision positioning.

How to achieve seamless positioning in mixed indoor-outdoor environments

In environments like warehouses where outdoor yards and indoor buildings are continuous, there are times when one switches positioning methods—“GNSS outdoors, other methods indoors.” The keys are unifying coordinate systems and automatic switching of positioning methods. Specific methods for achieving seamless positioning are as follows.

First, obtain an accurate absolute position outdoors with RTK-GNSS, then switch to relative positioning using IMU or smartphone AR functions once inside the building. For example, if you calibrate your current position to millimeter-level outdoors at the warehouse entrance using RTK and then proceed indoors, IMU and camera tracking can follow your movement from that starting point and maintain a few centimeters (cm level accuracy (half-inch accuracy)) of accuracy for a while without satellite signals. By returning to an open outdoor area to receive GNSS and reapply corrections before drift becomes large, long-term accuracy can be preserved.

Next, the indoor positioning system’s coordinates can be aligned in advance with the outdoor reference. When installing UWB anchors inside a warehouse, measure and register the anchors’ coordinates in a public coordinate system (for example, the Japanese plane rectangular coordinate system) beforehand. Then, when a forklift moves from outdoors (GNSS-tracked) into the building, tracking can continue in the same coordinate system. Devices automatically switch to available positioning methods, providing continuous location data across the entire coverage area.

Another method is to use a GNSS repeater (pseudo-satellite). This technology amplifies and rebroadcasts GPS signals received outdoors so that indoor environments mimic direct satellite reception. Simple repeaters can make the entire indoor area appear to be the same single location, but advanced systems that generate different pseudo-satellite signals per area have been developed. With such systems, even conventional GNSS receivers can determine position indoors; if RTK correction information is also received, centimeter-accuracy RTK positioning (cm level accuracy (half-inch accuracy)) can be possible indoors.

In short, in mixed indoor-outdoor sites, multi-sensor fusion and flexible use of appropriate positioning methods are essential. Use GNSS where satellite signals reach, bridge gaps with IMU or SLAM, and integrate all available radios and sensors to estimate the best position. Such hybrid positioning provides uninterrupted location services across an entire warehouse site.

Use cases for centimeter positioning in warehouses

When high-precision positioning becomes available, various operations in warehouses, logistics, and construction sites can be streamlined and enhanced. Here are concrete use cases centered on warehouses and their benefits.

Warehouse inspections

In large warehouse inspections, even if an inspector finds an abnormality, accurately recording “where” it was found can be difficult. With centimeter accuracy (cm level accuracy (half-inch accuracy)), photos taken by inspectors can automatically receive coordinates and orientation metadata. For example, a report like “looseness at the ceiling beam joint of shelf No. X” can be tagged with precise position data so the location is visible at a glance on a warehouse map. Drones and robots are increasingly used for inspecting high or narrow areas inaccessible to people, and high-precision position information supports their autonomous operation. Technologies enabling stable indoor drone flight without relying on GNSS are emerging, and real-time point cloud data can generate high-resolution digital twins (3D models), enabling next-generation facility management. All of these solutions depend on accurate self-positioning.

AR applications

Combining AR (augmented reality) with centimeter positioning (cm level accuracy (half-inch accuracy)) can intuitively assist warehouse tasks. For example, when a worker picks items, imagine the smart glasses they wear overlaying arrows and distances to the shelf containing the item. If current and target shelf positions are known within a few centimeters (cm level accuracy (half-inch accuracy)) and an optimal route is guided, workers can efficiently pick items in large warehouses without getting lost. Similarly, a forklift operator’s tablet can display real-time position and a warehouse map to navigate to the next load or the shortest route. AR visual guidance is easy for even new staff to understand, reducing human errors and speeding up tasks. In the future, AR could highlight hazardous or restricted areas in a worker’s field of view, improving safety management.

Stake positioning guidance

When building a new logistics warehouse or rearranging layouts, stake positioning (locating where equipment or shelves should be installed) is required. Traditionally, this layout marking (“墨出し”) used tape measures and chalk based on drawings, relying heavily on experience and judgment, and eliminating centimeter-level discrepancies took time even for experts. With centimeter positioning and AR, stake positioning becomes much easier. For example, a tablet could display planned installation points as virtual markers over the live view of the warehouse floor and guide workers with arrows to “move here.” Workers simply follow the instructions and mark the specified points, allowing staff without surveying expertise to perform high-precision layout marking. This prevents construction errors, reduces rework, shortens schedules, and ensures quality.

3D surveying

High-precision positioning is powerful in 3D surveying that records and understands site conditions in three dimensions. Mobile LiDAR scanners and LiDAR sensors built into modern smart devices make acquiring point cloud data easy. Combining the initial position obtained with RTK-GNSS lets you tie the collected point clouds to geographic coordinates (global coordinates). In other words, simply walking through a warehouse to scan can construct a 3D model with real geodetic coordinates. Tasks that previously required control points or post-processing alignment can now be completed in near real time. The resulting high-precision 3D models (digital twins) can be used for equipment layout planning, floor flatness checks, and long-term change detection (e.g., subsidence or equipment deterioration).

Tracking

Accurately tracking people, vehicles, and goods is important for increasing warehouse and factory productivity and safety. Centimeter-level positioning enables continuous monitoring of forklift routes and visualization of movement. Analyzing accumulated route data can identify unnecessary trips and congestion points, allowing quantitative evaluation of layout changes for efficiency gains. Equipping workers with position tags allows safety measures such as slowing forklifts when a person enters a hazardous area. Consistent high-precision tracking from outdoor yards to indoor warehouses enables strict control over loading/unloading operations, preventing loss and improving inventory accuracy.

Benefits of introducing positioning (efficiency, safety, cost)

Introducing high-precision positioning systems on-site can yield the following benefits.

• Operational efficiency: Automation and labor-saving in positioning tasks shorten work that previously required manpower and time. For example, if internal staff can handle layout changes in a day instead of outsourcing surveying to external companies, decision-making speed dramatically increases. Real-time position data used for route optimization and navigation reduces daily work time. As a result, productivity improves and more is achieved with less effort.

• Safety: Accurate position information is a powerful tool for safety management. Knowing the relative positions of people and vehicles enables preemptive warnings or automatic stops before danger occurs. Managing with objective data instead of relying on expert intuition reduces accident risk due to human error. High-precision guidance reduces operational mistakes and improves on-site safety quality. Eliminating the need for people to enter dangerous heights or confined spaces is another major advantage.

• Cost: Although introducing positioning systems requires upfront investment, they contribute to cost reduction in the long term. Centimeter accuracy dramatically reduces construction mistakes and rework costs. Performing surveying in-house instead of outsourcing reduces external fees. Recent high-precision devices are becoming smaller and cheaper, cutting deployment costs compared to traditional approaches that required extensive dedicated infrastructure. Furthermore, using Japan’s quasi-zenith satellite “Michibiki” correction service (CLAS) free of charge means no communication or service fees for correction information, making operational costs almost zero. The efficiency gains and mistake prevention lead to high ROI, making the investment worthwhile overall.

Use of LRTK

One concrete solution for achieving centimeter-class positioning both indoors and outdoors is LRTK. LRTK is a next-generation RTK-GNSS system designed for easy field use and incorporates various measures to address the challenges described above.

First, LRTK supports Japan’s quasi-zenith satellite “Michibiki” centimeter-class augmentation service (CLAS). This allows RTK positioning by receiving correction data directly from satellites even where internet communication from a base station is not available. In mountainous or out-of-coverage areas, LRTK can independently secure centimeter accuracy (cm level accuracy (half-inch accuracy)) as long as the sky is visible. If a known control point exists on site, one LRTK unit can be set up at that point as a local base station and the rover can perform relative positioning to maintain high accuracy even in enclosed indoor spaces. The flexibility to switch between standalone positioning (using CLAS) and relative positioning (using an on-site base) is powerful in environments such as tunnels.

Next, LRTK is designed to work with smartphones and tablets. A palm-sized, about 150 g integrated GNSS receiver connects wirelessly (no cable) to existing smart devices. For example, attaching LRTK to a commercially available tablet enables anyone to perform position measurements with accuracy comparable to professional surveying equipment. On arrival at a site, turning on LRTK and performing a tens-of-seconds initialization in an open area is enough to obtain a high-precision reference point. Thereafter, the smartphone’s AR functions and IMU sensors track self-position even where GNSS cannot reach. LRTK provides the high-precision starting point, and the smartphone supplements relative pedestrian navigation to seamlessly connect indoor and outdoor positioning.

LRTK also offers a cloud service for centralized data management. Position data transmitted from LRTK devices, and in the future trajectories from UWB tags and IMUs, can be managed in the cloud. This makes it easy to overlay outdoor heavy equipment movement with indoor robot trajectories on the same map. Positioning results, photos, and point cloud data recorded on site can be synchronized to the cloud in real time and shared within the organization. Without special software, data can be viewed as 2D/3D maps in a web browser, enabling remote offices to monitor site conditions.

In this way, LRTK advances centimeter positioning in terms of ease of use and integration. It packages advanced technology so that anyone on site can operate it, letting high-precision positioning become part of everyday workflow. If you face problems like “we need to measure but can’t” or “we’re troubled by position errors,” consider LRTK. Reliable position information everywhere indoors and outdoors will elevate warehouse management and construction sites to the next stage.

FAQ

• Q. What is needed to achieve centimeter accuracy (cm level accuracy (half-inch accuracy))? A. Basically, a high-precision GNSS receiver (RTK-capable) and a smartphone or tablet to display and record position data are required. For example, when using an LRTK device, simply install the companion app on a supported smartphone and connect via Bluetooth to be ready. Start positioning from a location where satellite signals can be received (outside or near a window), then move indoors and press the record button at desired points. The cumbersome base station setup and pre-survey planning traditionally required are unnecessary—measurements can begin immediately upon arrival.

• Q. Can you really get centimeter accuracy? Doesn’t accuracy deteriorate indoors? A. Yes, with proper equipment you can expect horizontal a few centimeters and vertical a few centimeters (cm level accuracy (half-inch accuracy)). If an RTK “Fix” solution with sub-centimeter errors is obtained in an open outdoor area, moving indoors immediately afterward will not cause a sudden large accuracy drop. Smartphone sensors and AR technologies maintain self-position accuracy for a short time, so positions can be recorded indoors at nearly centimeter-class accuracy (cm level accuracy (half-inch accuracy)). However, prolonged absence of any satellite reception allows drift to accumulate. For large facilities and long-distance movement, it is advisable to periodically move to a place with sky visibility and reapply RTK corrections. In validation tests where about 10 point measurements were taken and error variation evaluated, all points were within a few centimeters.

• Q. Can it be used in warehouses or mountainous areas with no network coverage? A. Yes. Receivers that can use Michibiki (QZSS) CLAS signals, like LRTK, enable high-precision positioning even outside cellular coverage. Because correction signals are received directly from satellites, the system does not depend on communication infrastructure. For indoor positioning, if Michibiki signals can be received near rooftops or windows, RTK-level accuracy can be maintained without internet. However, in environments completely without sky view, such as deep underground, satellite signals cannot reach; in those cases, complement with relative surveying from ground-measured control points or correlate with existing drawings (tie to known points).

• Q. How can surveying data be managed and shared? A. Data acquired with high-precision positioning devices can be automatically synchronized to the cloud for use. For example, LRTK allows uploading measured coordinates, photos, and scanned point cloud data to a dedicated cloud on site. Uploaded data can be viewed via a web browser as 2D maps or 3D views and shared with remote team members via shareable links. Coordinate systems such as Japan’s plane rectangular coordinate system (and other selectable systems) are supported, making overlay with CAD drawings and GIS data simple. Survey results obtained on site can be directly integrated into design drawings and management systems, dramatically accelerating data utilization.

• Q. Can staff who aren’t technically inclined use it? A. Modern high-precision positioning systems are designed with user-friendly interfaces for non-experts. In the case of LRTK, an intuitive smartphone app visualizes the current positioning state (accuracy rank, number of tracked satellites), and following on-screen guidance allows anyone to perform surveys. Device setup is as simple as mounting the unit on a pole tip and holding it vertically; no difficult adjustments are required. New users can learn the basics in a short training session, so facility managers and staff without surveying experience can use the system practically. Comprehensive support is also available if users encounter difficulties.