Rail Point Cloud Measurement with LRTK: High-Precision Diagnosis of Displacement and Settlement

In track (rail) maintenance that supports the safe operation of railways, it is critically important to accurately diagnose rail and trackbed displacement and settlement and to address them early. Repeated train passages and ground movement gradually cause rails to shift or sink. If such track deformations are overlooked, they can lead to degraded ride comfort or, in the worst case, derailments, so routine monitoring and timely repairs are indispensable. This article explains the importance of diagnosing track displacement and settlement, the limitations of conventional methods, and a new approach using point cloud data measurement combined with GNSS that has attracted attention in recent years. In particular, we focus on how a rail point cloud measurement solution using smartphones combined with RTK technology—LRTK—contributes to manpower reduction and higher accuracy in the field.

Importance of diagnosing displacement and settlement in railway track maintenance

Railway tracks constantly bear heavy loads, and over time the rails, sleepers, and the ballast (crushed stone) beneath the trackbed compact and settle, causing changes in track height and inclination. External factors such as vibration from passing trains, temperature changes, earthquakes, and heavy rainfall also cause slight movements of the rails from their design positions—track displacement (alignment deviations). When such settlement and displacement accumulate, they manifest as height differences between rails (lateral twist), widening of the gauge (distance between rails), and track distortion (alignment irregularities). If left unaddressed, these issues lead to degraded ride comfort and damage to rolling stock and track components.

Early detection and correction of track displacement and settlement are essential to maintain safety and ride comfort. For example, in sections where settlement is progressing, regular monitoring should be conducted, and when certain thresholds are exceeded, maintenance such as tamping the ballast (trackbed compaction) or jacking up the rails must be performed in a timely manner to restore height. If settlement is left undiagnosed and unaddressed, speed restrictions or suspension of train operations may ultimately be required, disrupting transport services. Therefore, for railway operators, diagnosing track displacement and settlement is at the core of routine maintenance operations.

Limitations of conventional inspection methods (track inspection cars and manual inspections)

Traditional methods for understanding track condition include track inspection cars (specialized measurement vehicles) and manual measurements by workers using levels and rules. Track inspection cars run on the tracks ahead of revenue trains and continuously measure track irregularities and distortions using lasers and accelerometers. They have the advantage of measuring large areas in a short time, but operating such specialized vehicles is costly, and measurement frequency is limited, making it difficult to capture signs of abnormalities in real time. Also, inspection car data are managed based on kilometer posts, and measured positions are shown as relative distance information. For this reason, discrepancies can occur when comparing different measurement datasets or when matching to plan coordinates, and identifying the precise location of defects in the field can be time-consuming.

On the other hand, patrol inspections have long relied on workers using levels and track gauges to measure rail height and inclination point by point. For example, placing a straightedge (scale) on the railhead to measure left-right rail height differences (line/lateral level) or setting up levels at regular intervals to measure settlement amounts are common practices. These manual inspections are simple but require significant time and manpower to measure long sections. When measurements must be completed within limited night-time windows when trains are not running, measurement ranges may be constrained and work can become rushed. Moreover, manual inspections produce sparse measurement points, so there is a risk of overlooking issues. For instance, even if heights are measured every 5 m (16.4 ft), localized settlement or distortion between those points might not be detected. Furthermore, work on the tracks involves the hazard of contact with trains, so it is desirable to complete tasks in as short a time and with as few people as possible. Conventional methods thus have limitations in efficiency, accuracy, and safety, and the field has been seeking better measurement means.

Reproducing rail geometry and using it for diagnosis via point cloud data

Recently, 3D point cloud data-based track measurement has attracted attention in the railway sector. Point cloud data are a collection of countless measured points acquired by LiDAR (laser ranging) or photogrammetry, creating a high-precision 3D model of surrounding structures and terrain. Recording rails, sleepers, and trackbed surfaces as point clouds yields detailed models that are like digital copies of the track geometry. Using these models, track conditions that used to be discernible only on site can be analyzed from multiple perspectives at a desk.

Various diagnostic indicators can be calculated from point-clouded rail data. Examples of applications include:

• Assessing rail cross-section deterioration: By slicing point cloud data in the cross-sectional direction, you can visualize wear and deformation of the rail head. Measuring the wear amount of the rail profile provides a basis for decisions on replacement or grinding.

• Settlement amount and trend analysis: By comparing point cloud datasets from different times, you can determine how many millimeters a rail height at a given location has dropped compared to the past. For example, by overlaying annual point clouds from regular inspections, you can graph settlement progress and predict future maintenance timing.

• Evaluation of track twist and vertical displacement: Extracting left-right rail height data from the point cloud allows calculation of height differences (twist amounts) over arbitrary section lengths. Checking for abrupt twists over short base lengths (e.g., 3 m (9.8 ft)) can detect ride irregularities that pose a high derailment risk. Drawing the vertical profile of the rail also makes local depressions or uplift immediately apparent.

• Measurement of gauge and alignment: From point clouds you can easily measure left-right rail spacing (gauge) and straightness (alignment irregularity) over straight sections. Changes in gauge over long distances that are difficult to measure manually can be computed continuously from point clouds, avoiding missed abnormal spots.

By reconstructing rail geometry with point cloud data, the track’s geometric condition can be understood in detail and quantitatively. Micro-changes that cannot be fully measured in the field and the relative prominence of anomalies within the whole are both visualized, enabling more accurate maintenance decisions. Point clouds function as a kind of “3D ledger,” allowing the digital recording of track transitions from past to present, which is a significant advantage.

Challenges in ensuring positional accuracy for ground-based point cloud measurement and the role of GNSS

Point cloud measurement has major advantages, but ensuring positional accuracy is a challenge when using it in railway settings. When acquiring point clouds from handheld scanners or smartphones at ground level, the resulting point clouds are recorded in a relative coordinate system based on the device’s internal sensor measurements. A single LiDAR scan does not reveal where on a map the captured point cloud corresponds to (absolute coordinates), necessitating alignment when comparing multiple measurement datasets. Furthermore, when scanning over long distances while walking, scan drift due to accumulated sensor error cannot be ignored. For example, when continuously scanning over long distances of over 100 m (328.1 ft), positional offsets of several centimeters or more between the start and end points can occur. Such errors cause inconsistencies when matching high-precision track point clouds with past data or other survey results, making it difficult to calculate precise settlement amounts or pinpoint defect locations.

The key to solving this issue is integration with GNSS (Global Navigation Satellite Systems). In particular, using RTK (Real Time Kinematic), a high-precision satellite positioning technique, allows the position of the scanning device to be obtained in real time with centimeter-level accuracy. Specifically, by receiving correction information from a reference station installed on the ground or from a networked reference station into the smartphone during measurement, the scanning device’s position can be continuously captured in a high-precision global coordinate system. As a result, each point in the acquired point cloud can be assigned absolute coordinate information. In other words, the point cloud functions as a “map with survey coordinates,” enabling direct comparison with arbitrary reference points on maps or other GIS data. This eliminates much of the effort required to precisely align point clouds from different times and greatly facilitates precise comparison of settlement amounts and accurate localization of anomalies.

Labor-saving rail point cloud measurement realized by smartphone + LRTK

A practical on-site solution that realizes RTK-enhanced measurement simply is LRTK, which combines smartphones with high-precision GNSS. LRTK equips a smartphone with a small RTK-GNSS receiver and uses a dedicated app for LiDAR scanning or photogrammetry to allow anyone to easily acquire point cloud data with absolute coordinates. Precise surveying that once required specialized instruments and skilled personnel can now be completed with a smartphone in the palm of your hand, attracting attention in construction and civil engineering.

Features and benefits of rail point cloud measurement using LRTK include:

• Ease and speed: Carry only a pocket-sized smartphone and a small GNSS device to the field, launch the app, and start scanning. Like shooting a video, you can walk along the track holding the smartphone and acquire point clouds of the rail surroundings for several hundred meters within minutes. Results are immediately viewable on the smartphone screen, making retakes or additional measurements easy on site.

• High-precision positioning: With RTK corrections, the smartphone’s self-position is continuously computed with cm level accuracy (half-inch accuracy), so acquired point clouds are tagged with accurate real-world coordinates. This reduces position drift and scale errors that plagued traditional scans, yielding low-distortion data even over long sections.

• One-person operation and labor savings: The intuitive smartphone app makes it possible for a single person without specialized surveying knowledge to perform measurements. Tasks that previously required two-person teams can be completed by one person with smartphone + LRTK. There is no need to carry heavy surveying equipment, so long-distance movement along tracks is less burdensome. This contributes significantly to reductions in personnel and work time.

• Multifunctionality and data utilization: The LRTK app integrates not only point cloud scanning but also distance and area measurement on acquired data, geotagging of photos, and AR display among other functions. Point clouds and photos captured on site can be shared and stored via the cloud instantly, allowing detailed analysis and report preparation to begin as soon as you return to the office.

By combining a smartphone with LRTK, three-dimensional measurement of tracks becomes dramatically simpler. Understanding rail geometry, which previously required large surveying equipment and specialized knowledge, can now be carried out quickly by the track maintenance staff in the field. This accuracy-and-efficiency-balanced solution is expected to be a labor-saving tool that addresses common challenges in railway sites such as labor shortages and limited work windows.

Rapid and safe operations during night work and short closures

Rail maintenance is typically carried out at night or during gaps between trains to minimize impact on operations. With limited work time to complete surveying and repairs, the speed of point cloud measurement using LRTK proves highly effective. For example, consider nighttime work after the last train where you need to check settlement in a section. Traditionally, workers conducting leveling measurements on the rails and recording heights every tens of meters would take considerable time. With LRTK, you can walk along the track holding a smartphone and obtain height data for the entire section at once. Field measurement time is drastically reduced, allowing more time for repairs and safety checks, thereby improving overall work efficiency.

LRTK’s simple measurement is also effective for short track closures (for example, blocks of several tens of minutes for daytime inspections). You can quickly acquire point clouds only at the points you need and carry back the required data, enabling completion within short windows between trains. This eliminates the need to halt trains over long sections to run inspection cars or to mobilize large numbers of staff.

Moreover, LRTK-driven labor savings contribute to improved safety. Shorter task durations reduce the time workers are exposed on the tracks, lowering the risk of contact with trains. Compared with conventional methods that involve heavy machinery and many staff, small teams operating nimbly can maintain better situational awareness and easier safety management. Field reports even note that “finishing surveying quickly gives workers peace of mind, so safety checks are done more thoroughly.” Balancing labor savings and safety improvements is a key requirement for modern railway maintenance, and LRTK is a new technology that helps meet that need.

Monitoring use of point cloud data and advanced reporting through AR/BIM integration

Using point cloud data for track management also enables more advanced monitoring methods. By regularly acquiring point clouds of the same section, you can accurately track the progression of displacement and settlement over time. For instance, conducting annual point cloud measurements for an approach section to a bridge allows quantitative evaluation of yearly settlement rates by comparing year-to-year settlement amounts. What used to be tracked with paper records or tables can now be understood intuitively through direct comparison of 3D models. Creating difference heatmaps between point clouds instantly shows which locations have settled and by how much, aiding prioritization in maintenance planning.

Furthermore, acquired point cloud data can transform reporting and sharing through integration with AR (augmented reality) and BIM/CIM models. With high-precision point clouds, you can overlay 3D data onto the real scene via a tablet or smartphone screen. For example, on site you can use AR to display a previously measured point cloud model over the current view from the smartphone camera and check deviations between the current track and past data. In areas with settlement, an AR view comparing past rail position point clouds with the current rail makes the degree of drop easy to understand intuitively. It is also possible to integrate point cloud measurement results into pre-made BIM models (which include designed rail heights and 3D placements of structures) and visualize differences from design values using color-coding. Such visualization solutions facilitate smooth sharing of track condition information not only with on-site supervisors but also with office managers and other departments.

In report preparation, using 3D views and AR images enabled by point clouds and BIM allows realistic reproduction of site conditions that are hard to convey with text or 2D drawings alone. For example, a regular inspection report might state, “Annual settlement of 5 mm (0.20 in) confirmed near the exit of the XX tunnel. See the displacement diagram on the point cloud model (Figure ○),” and include a difference heatmap or AR composite photo as the figure. This helps readers immediately grasp the seriousness and exact location of the problem. Adding objective point cloud data to the intuition and experience cultivated by field personnel dramatically enhances the persuasiveness and reproducibility of track maintenance.

Conclusion: A new era of rail inspection with simple surveying using LRTK

From diagnosing track displacement and settlement to planning maintenance, digital approaches are gradually spreading in the railway industry. Among them, simple surveying and point cloud measurement using LRTK—a combination of smartphone and GNSS—is a revolutionary method that significantly lowers the threshold for field work. If your in-house track maintenance staff can acquire high-precision rail point cloud data themselves without relying on specialized equipment and use it for continuous monitoring, the PDCA cycle of maintenance will become more refined. Labor and efficiency improvements in surveying and inspection are urgent issues for railway sites struggling with workforce shortages, and tools like LRTK can help address these challenges.

Maintenance practices that once depended on masterful craftsmen are shifting toward preventive maintenance based on quantitative data. DX (digital transformation) is coming to railway infrastructure maintenance as well. Adopting technologies like LRTK that enable field-born data accumulation and analysis will not only balance safety and efficiency but also lay the groundwork for future preventive-maintenance-oriented management. Consider leveraging LRTK, a next-generation solution, to bring new innovations to your track management.