Comparison of Five Technologies for Achieving Centimeter-Level Positioning Indoors (cm level accuracy (half-inch accuracy))｜How Field Personnel Should Choose

The demand for "obtaining position information indoors with cm level accuracy (half-inch accuracy)" is rapidly increasing across many industrial sites such as manufacturing, construction, logistics, and infrastructure management. However, even under the broad term "indoor positioning," precision, cost, and implementation difficulty vary greatly depending on the technology used. This article compares five technologies commonly considered in the field and explains the key points for choosing the one that suits your company's environment.

Table of Contents

• Points to clarify before choosing an indoor positioning technology

• Technology 1: UWB (Ultra-Wideband radio)

• Technology 2: LiDAR and SLAM

• Technology 3: High-precision Wi-Fi positioning (Wi-Fi RTT)

• Technology 4: Magnetic mapping

• Technology 5: Vision (camera)-based positioning

• Summary of accuracy, cost, and deployment difficulty of the five technologies

• Hybrid approaches that combine technologies

• LRTK as a bridge to seamless high-precision indoor-outdoor positioning

• Practical steps for selecting technologies

• Regulations and laws to check when implementing indoor positioning

• Future trends of each technology

• Key success points learned from implementation case studies

• Summary

Things to Clarify Before Choosing an Indoor Positioning Technology

Before getting into a technical comparison, it is important to first clarify what is required at your company's site.

The first criterion is how much accuracy is required. The applicable technologies are narrowed down depending on the required accuracy level, such as "cm level accuracy (half-inch accuracy) (±3 cm (±1.2 in) or less)", "accuracy of several cm to several tens of cm (several in to several tens of in)", "sub-meter accuracy (approximately ±50 cm (±19.7 in))", and "accuracy of several meters (zone recognition) (several ft)". Setting a concrete numerical value for how much accuracy is needed in actual operations is the starting point for selecting technology.

Next is the size and shape of the positioning area. The scale and cost of the required infrastructure differ between a small area of tens of square meters (tens to hundreds of sq ft) and a vast factory floor of thousands of square meters (tens of thousands of sq ft). If the area spans multiple floors or multiple buildings, you also need to consider the design for coordination between floors and between buildings.

It is also important whether the positioning target is a person, an object, or a robot. The requirements for real-time performance change depending on whether you are tracking moving targets such as people or forklifts, or measuring fixed targets such as equipment. When tracking moving targets, the update rate (how many times per second the position is acquired) becomes an important specification.

Compatibility with existing infrastructure must also be considered. Whether Wi-Fi or Ethernet can be used, the building’s structure (the prevalence of metal structures), and how power is supplied also influence technology selection. Making maximum use of existing equipment can, in some cases, reduce deployment costs.

It is important to clarify budget and schedule constraints. Estimate the total cost—including equipment costs, installation costs, software license fees, and maintenance fees—and evaluate cost-effectiveness. It is also an important practical issue whether installation work can be carried out without halting factory or warehouse operations.

Having organized these, let's compare the five technologies.

Technology 1: UWB (Ultra-Wideband Wireless)

UWB is currently one of the most promising technologies that can stably achieve cm to several tens of cm accuracy indoors (cm level accuracy (half-inch accuracy) to several tens of centimeters). It transmits short pulses over a very wide frequency band (typically 3.1–10.6 GHz) and calculates position by measuring the time of flight between multiple anchors.

The accuracy is approximately 10-30 cm (3.9-11.8 in), and with careful design and calibration it can be improved further. The high real-time update rate enables 10-100 or more position fixes per second, making it suitable for tracking moving people and objects.

Initial deployment costs are relatively high, requiring anchor installation work and power and network cabling. However, in recent years chip prices have fallen, and integration into smartphones and wearable devices has increased. Equipment costs are expected to decline further in the future.

Suitable environments include open floors with few obstacles where radio waves propagate easily, and manufacturing lines where the placement of precision machinery is stable. In warehouses with densely packed metal shelving, radio reflections can be a challenge, but there are an increasing number of cases that can be handled by algorithmic NLoS correction.

UWB's greatest strength is its extensive track record in industrial applications. There are deployment examples across diverse fields such as manufacturing, logistics, healthcare, and construction, creating an environment where it is easy to choose highly reliable systems.

Technology 2: LiDAR and SLAM

LiDAR acquires point clouds of the surroundings using a 3D laser scanner and determines position by matching them with map data (matching). By combining it with SLAM (Simultaneous Localization and Mapping) technology, it can localize autonomously while generating a map even without a prior map.

Accuracy depends on the performance of the LiDAR used and the processing algorithms, but in suitable environments it can achieve accuracies of several centimeters to several tens of centimeters (several in to several tens of in). In particular, in environments with many distinctive features (pillars, walls, shelves, etc.), matching accuracy is high and stable positioning is possible.

LiDAR-based SLAM is commonly mounted on AGVs (automated guided vehicles) and inspection robots. A key advantage is that it does not require fixed infrastructure (anchors, etc.) and can operate autonomously on the device alone. In addition, because it produces a 3D map of the surrounding environment as a byproduct, it can be used for facility management and digital twin construction.

On the other hand, in simple, homogeneous environments (such as clean rooms with nothing but white walls), distinctive features can be lacking, which may reduce positioning accuracy. Also, because a processing computer must be mounted, the device itself tends to become larger and more expensive. Thanks to advances in algorithms that leverage AI, robustness in environments with few distinctive features has been improving year by year.

Technology 3: High-precision Wi-Fi Positioning (Wi-Fi RTT)

The IEEE 802.11mc (Wi‑Fi RTT) standard measures the TWR (round-trip time of flight) between a smartphone and a Wi‑Fi access point to calculate distance. With access points and devices that support this standard, it is possible to perform positioning with an accuracy of about 1-2 m (3.3-6.6 ft) while leveraging existing Wi‑Fi infrastructure.

Furthermore, the next-generation IEEE 802.11az standard is expected to achieve sub-meter accuracy through more precise ranging using multiple antennas. By using the wide bandwidth of the 5 GHz band, distance measurement accuracy approaching that of UWB is being targeted, and in the future it could approach cm level accuracy (half-inch accuracy).

The advantages of being able to reuse existing Wi‑Fi infrastructure are significant, but at present products that can stably achieve cm level accuracy (half-inch accuracy) are limited. While we can expect higher precision in the future, it is currently suited to applications with relaxed accuracy requirements (zone management, route tracking, etc.). Also, some cases can be addressed by firmware updates to access points, so it is important to confirm in advance whether existing Wi‑Fi equipment can be utilized.

Technique 4: Magnetic Mapping

The indoor geomagnetic field is influenced by columns, beams, rebar, and other structural elements, resulting in a unique pattern at each location. This pattern is measured in advance to create a fingerprint map, and during positioning the values measured by a magnetic sensor are matched against the map to estimate the position.

It has the major advantage that no additional infrastructure is required because it can use the magnetic sensors built into smartphones. However, accuracy is generally about 1-3 m (3.3-9.8 ft), and it is not suitable for applications requiring cm level accuracy (half-inch accuracy).

Another issue is that building renovations (such as the addition of metal components) require map updates, and accuracy degrades in environments with high magnetic noise. It is often used not as the primary system for high-precision positioning but as an auxiliary position reference combined with other technologies. In particular, it is effective as a complementary positioning method in locations where the radio signals of high-precision technologies such as UWB cannot reach.

Technology 5: Vision (Camera)-Based Positioning

Position is estimated from images captured by a camera by extracting feature points, matching them, and performing triangulation. Methods that use AR markers or known features as landmarks can achieve sub-mm (sub-0.04 in) to cm-level (0.4 in) accuracy under favorable conditions.

Advantages include the ability to operate under normal ambient light and the requirement for little additional infrastructure. On the other hand, accuracy fluctuates with changes in lighting, occlusion, and camera orientation, so careful operation is necessary to maintain stable cm level accuracy (half-inch accuracy).

Recently, Visual SLAM that leverages AI has advanced, making relatively robust self-localization possible using only a camera. Integration into smartphones and AR glasses is increasing, and further development is expected. Combining it with depth sensors (depth cameras) improves both positioning accuracy and environmental recognition.

Summary of Accuracy, Cost, and Implementation Difficulty of Five Technologies

When technologies are organized along the axes of accuracy, cost, and ease of implementation, each exhibits distinct characteristics.

UWB can achieve high accuracy from centimeter-level (half-inch accuracy) to several tens of centimeters, but it requires initial costs and installation work for anchors. It is ideal for applications that prioritize accuracy and has a strong track record in manufacturing and logistics. It features high maintainability, making it easier to maintain stable accuracy even during long-term operation.

LiDAR+SLAM requires no fixed infrastructure and can operate autonomously, but device costs are high and accuracy can be unstable in environments with few feature points. It is suitable for mounting on robots and AGVs, and can also be used together with map-building functionality.

High-precision Wi-Fi is economical in that it can leverage existing infrastructure, but the current stability of cm level accuracy (half-inch accuracy) is limited. It is a technology that holds promise for increased precision in the future, and as support for new standards advances, its appeal as an option will grow.

Magnetic mapping is inexpensive and requires no additional infrastructure, but its accuracy is on the order of 1-3 m (3.3-9.8 ft), making it unsuitable for applications that require cm level accuracy (half-inch accuracy). It can be effectively used as an auxiliary position reference.

Camera-based systems are expected to see accuracy improvements thanks to advances in AI, but they are susceptible to the effects of lighting and occlusion, and ensuring stable cm level accuracy (half-inch accuracy) requires environmental setup. They are low-cost and easy to deploy, and can achieve high accuracy when conditions are properly controlled.

Hybrid approach combining technologies

In real-world settings, it is often difficult to meet all requirements with a single technology. Therefore, a hybrid approach that combines multiple technologies is effective.

For example, with a combination of UWB and IMU, the IMU can interpolate position during moments when UWB cannot perform positioning due to radio interference, enabling continuous high-precision positioning. With a combination of LiDAR and UWB, LiDAR can handle detailed environmental perception and ensure local accuracy, while UWB can provide a reference for absolute position.

Designing such sensor fusion systems requires specialized knowledge, but the available middleware and SDKs have become more comprehensive, lowering the barrier to adoption.

LRTK as a bridge to seamless high-precision positioning indoors and outdoors

Many on-site tasks are not confined to indoor environments. In construction sites, infrastructure inspections, factory management, and so on, there is a strong need to manage positioning data while moving between outdoor and indoor environments.

One effective approach in such environments is a hybrid method that combines outdoor high-precision GNSS positioning with indoor positioning. Outdoors, LRTK, an iPhone-mounted high-precision GNSS positioning device, is used to obtain absolute coordinates with cm level accuracy (half-inch accuracy) using RTK. Even after entering indoors, it is possible to manage location information for construction management and inspection records based on the acquired reference point coordinates.

LRTK is notable for its ease of use: it can be used simply by attaching it to an iPhone. No complex installation work or specialized calibration is required, allowing field personnel to start using it immediately. An approach that begins with outdoor cm level accuracy (half-inch accuracy) positioning and gradually progresses to integration with indoor positioning is a practical first step for deployment.

By integrating with the LRTK cloud platform, positioning data, point cloud data, and photos captured outdoors can be managed in the cloud as is. By combining these with indoor inspection and construction records, digital management of the entire building can be achieved.

Practical Steps for Technology Selection

This document outlines practical procedures for selecting indoor positioning technologies.

First, we conduct on-site interviews and define requirements. We clarify accuracy, area, targets, update frequency, budget, and existing infrastructure. It is important for field personnel, managers, and the IT department to gather together and share the requirements.

Next, we will carry out a PoC (proof of concept). We will trial the shortlisted candidate technologies in the actual field environment to verify whether they achieve the expected accuracy. It is not uncommon for catalog specifications to differ from real-world performance. Conducting a PoC in advance is especially essential in factory environments with a lot of metal.

Based on the results of the PoC, we carry out system design. We plan anchor placement, network design, software integration, and maintenance planning. At this stage, it is important to also consider future scalability (expansion of the positioning area and an increase in the number of tags).

After deployment, conduct regular accuracy evaluations and adjust the system in response to environmental changes (equipment additions and modifications). Indoor positioning systems are not "install and forget"; ongoing maintenance is essential to maintain accuracy.

Regulations and Laws to Check When Deploying Indoor Positioning

When introducing an indoor positioning system, it is important to check relevant regulations such as the Radio Act and the Building Standards Act. Introducing a system while ignoring regulations not only constitutes a legal violation but also carries the risk of being required to forcibly shut down the system, so verification during the planning stage is indispensable. Check for the presence of the technical conformity (Giteki) mark on the device’s specification sheet, and if unclear, request written confirmation from the vendor as a safety measure. Because the Radio Act’s regulations may be revised, it is recommended to confirm the latest information on the Ministry of Internal Affairs and Communications website or by consulting experts.

In Japan, UWB may only be used with devices that have obtained the Ministry of Internal Affairs and Communications' technical conformity certification (Giteki). Because using imported or uncertified devices may constitute a violation of the Radio Law, verify in advance the legality of devices to be introduced.

Also, when installing anchors in the ceilings or columns of factories and warehouses, installation must be carried out in a manner that does not affect the building’s structure. For rental properties, prior confirmation with and permission from the owner or management company are required.

When collecting workers' location information, from the perspective of personal data protection it is required to clarify the purpose of collection, obtain consent, and establish data management policies. If a labor union is present, it is important to provide sufficient explanation and build consensus before implementation.

Future Trends of Each Technology

The five technologies are each expected to evolve going forward. Selecting them with future technological trends in mind leads to building systems that remain effective over the long term. To make choices that are less likely to become obsolete, it is effective to select products that comply with standard specifications, to review vendors' roadmaps, and to choose systems with open APIs.

UWB chip costs continue to fall, and standard integration into smartphones and wearable devices is progressing. With advancements in AI-powered NLoS detection and correction techniques, stable cm-level accuracy (half-inch accuracy) can now be achieved even in complex, metal-rich environments.

LiDAR + SLAM is benefiting from the miniaturization and cost reduction of LiDAR sensors, making it easier to deploy them on a wider range of devices. Advances in AI point cloud processing technologies have greatly enhanced robustness in dynamic environments.

High-precision Wi‑Fi is expected to deliver significant accuracy improvements using existing infrastructure as the next-generation standard (IEEE 802.11az) becomes widespread. High-precision positioning using 5G (5G-NR Positioning) is also attracting attention as a technology to achieve seamless high-precision positioning both indoors and outdoors.

Camera-based positioning has improved robustness to lighting changes and occlusions through the combination of depth cameras and AI technologies. With the proliferation of AR glasses and spatial computing devices, the importance of high-precision camera-based positioning is increasing.

Based on these technological trends, UWB is currently the strongest candidate in terms of the balance of accuracy, track record, and cost, but it is highly likely that hybrid systems integrating multiple technologies will become mainstream in the future.

Key Points for Success Learned from Implementation Case Studies

We summarize the key success points revealed by real-world implementation case studies.

A common feature of successful implementations is that they conduct PoCs (proofs of concept) carefully. Through PoCs in actual field environments, they can detect differences from catalog specifications and unexpected problems early, and reflect those findings in the design of the full-scale deployment.

Also, involving on-site staff in requirements definition is a key to success. Rather than proceeding solely within the IT department, incorporating the voices of the staff who will actually use the system into the requirements results in a more usable and effective system.

A continuous improvement framework after deployment is also important. Simply deploying the system is not the end; by continuing accuracy monitoring and making regular improvements, the return on investment can be sustained over the long term.

Summary

The five technologies for achieving cm level accuracy (half-inch accuracy) in indoor positioning each have their own strengths and limitations. Choosing the optimal technology based on the site's environment, budget, and accuracy requirements is the key to success.

At present, the core technologies capable of stably achieving cm level accuracy (half-inch accuracy) indoor positioning are UWB and LiDAR + SLAM, but advances in Wi‑Fi and vision technologies should not be overlooked. Also, iPhone-mounted GNSS-based high-precision positioning devices such as LRTK, which can start outdoors and seamlessly extend indoors, are a strong option for realizing high-precision position information management at a realistic cost for sites that move between outdoor and indoor environments.

Carefully comparing technologies and organizing on-site requirements, and establishing a phased implementation plan, is the shortest route to successfully achieving indoor positioning with cm level accuracy (half-inch accuracy) on site.