5 Steps to Visualize Buried Pipes with Point Clouds｜Prevent Positional Misalignment Before Excavation

First, let us consider the issues and background surrounding buried underground pipes. During roadworks and site excavations, accidents often occur in which existing underground water and sewer pipes, gas pipes, power cables, and the like are accidentally damaged. Damaging an aging water pipe can lead to a large-scale leakage incident, and damaging a gas pipe can cause a gas leak and, in the worst case, pose an explosion risk. Cutting buried power lines or communication cables can cause power outages or communication disruptions in the surrounding area, seriously affecting daily life. In fact, many such incidents of damage to buried utilities are reported nationwide every year, and many of their causes stem from "not accurately knowing what is buried underground and where."

Pipes buried underground are invisible, so they have long been managed on site with the utmost care. When burying pipes, surveys are conducted before backfilling to record the piping’s position and depth, and information is preserved through photography and the creation of as-built drawings (completion drawings). At excavation sites, workers rely on those drawings and markings on the ground, and proceed with the work guided by veteran workers’ experience and intuition while estimating, "There should be a certain pipe buried around here." In some cases, the positions of underground buried objects are investigated using ground-penetrating radar equipment, or test excavations are carried out—digging small holes—to directly confirm the pipes. However, management that depends on paper drawings and human intuition has its limits, and especially in places like urban areas where complex networks of piping are laid out, it is not uncommon for the information on the drawings to disagree with the actual positions. There are countless examples of "the drawings showed nothing should be here, but when we dug we found a pipe coming from an unexpected location and had a scare," and on site there is always the risk of buried objects being mislocated.

Thus, "how to make the unseen visible" has become a major challenge in underground infrastructure management. If underground structures can be intuitively understood, it would not only help prevent accidents during excavation work but also greatly contribute to improving the efficiency of inspection and replacement planning for aging pipes. In recent years, the use of point cloud data and AR (augmented reality) has attracted attention as technologies that hold the key. A point cloud is a collection of many three-dimensional measurement points acquired by laser scanners or photogrammetry, and it is a method that can digitally record a site's terrain and structures with high accuracy as they are. Below, we explain five concrete steps to "visualize" buried pipes using point cloud data to prevent positional errors before excavation, incorporating the latest technologies such as mobile devices and drones.

Step 1: Preliminary Preparations and Collection of Underground Pipe Information

The very first step is to gather sufficient information about buried pipes before excavation work and to plan the on-site investigation. Review past as-built drawings and records to identify which types of pipes are buried where. Many public infrastructures such as water and gas are documented by municipalities or utility operators, but older buried installations may exist only on paper drawings or the records may be out of date and not match the actual layout. For that reason, cross-check multiple information sources and collect the most up-to-date and reliable data possible. Also check related above-ground features. For example, by identifying the locations of fire hydrants, valve covers, manholes, and the bases of utility poles—which connect to underground pipes—you can form an approximate image of the buried routes. In some cases, it is also effective to use equipment such as metal detectors or ground-penetrating radar (GPR) to investigate subsurface responses and confirm whether there are any unknown buried objects. This kind of advance preparation makes it clear which areas should be prioritized for investigation and scanning.

Next, develop a plan for on-site point cloud measurements. Select appropriate measurement methods and equipment according to the size of the target area, environmental conditions, and the required level of accuracy. For confined areas or locations requiring fine detail, laser scanners or smartphone LiDAR scans that yield high-density point clouds are effective. On the other hand, if you want to capture an entire long-distance piping route, aerial photogrammetry with drones is suitable for building a wide-area terrain model. As noted later, each method has its strengths, so plan to combine and use them as needed. Additionally, preparations for accurate position alignment (georeferencing) are important. Check in advance whether known control points or survey markers are present on site, and if not, consider installing temporary reference markers. This will help minimize positional discrepancies when integrating separately acquired datasets later.

Step 2: Mobile Point Cloud Scanning Using a Smartphone or Tablet

When preparations are complete, first acquire point cloud data of the site using a mobile scan with a smartphone or tablet. In recent years some smartphones are equipped with small LiDAR (LIDAR) sensors, and by using dedicated apps anyone can easily perform a 3D scan of their surroundings. If you walk around the planned excavation area and its surroundings holding a smartphone while scanning with the camera and LiDAR, the shapes of the ground and structures are reproduced on the screen in real time as a collection of countless points (a point cloud). For example, on a road the pavement, curbs, surface irregularities, and buried-pipe-related structures such as manhole covers and gas valve boxes are also accurately recorded in the point cloud. Because you can digitally copy the site itself in a short time, you can later review in detail "what was where."

Point-cloud scanning with a smartphone offers excellent mobility, and its advantage is that a single person can easily take measurements in tight spaces and both indoors and outdoors. Until now, using a 3D laser scanner required transporting heavy equipment and specialized setup, but with a smartphone, field workers themselves can take it out and measure whenever needed. Acquired point-cloud data can be immediately previewed on the device, and if anything is missing or was missed, additional scans can be performed on the spot. Also, using a smartphone makes it possible to capture data from a viewpoint overlooking the ground from roughly human height, so for shallowly buried pipelines it may be possible to detect subtle rises or depressions that appear on the ground surface. In terms of this convenience and the detailed on-site perspective recording, mobile scanning is highly effective for preliminary investigations of buried pipes.

Furthermore, by combining a high-precision GNSS receiver that can be attached to a smartphone, you can directly assign absolute coordinates in the geodetic reference frame to the acquired point cloud. Typically, the GPS accuracy of a smartphone alone has errors on the order of several meters (several ft), which can cause positional misalignment when overlaying the captured point cloud onto drawings or other survey results later. However, if you perform scanning while mounting a GNSS device that supports the RTK (Real-Time Kinematic) method to the smartphone, it becomes possible to link coordinates in the world geodetic system to each point in the point cloud with centimeter-level accuracy (cm level accuracy, half-inch accuracy). This allows the point cloud data acquired on-site to serve directly as a 3D record with accurate positional information, greatly reducing the need for time-consuming post-processing adjustments to align data to reference points.

Step 3: Wide-area point cloud surveying using drones and ground-based laser scanners

In addition to mobile scanning, measurement methods such as drones (unmanned aerial vehicles) and ground-based laser scanners are also used to cover wide areas and high elevations. Photogrammetry using drones involves taking numerous photographs of the ground from the air and analyzing them to generate point cloud data and orthophotos (composite images viewed directly from above). The advantage is that, by surveying from above, you can grasp at once wide-ranging terrain and the layout of structures that cannot be seen from the ground. To record the surroundings of buried pipelines from an overhead perspective, photographing the entire construction area with a drone and creating a 3D model allows easy confirmation of surface undulations and positional relationships with other facilities. Especially for long-distance pipelines, aerial surveying enables efficient capture of the entire route.

Recently, there has been an increasing number of cases where drones are equipped with small laser scanners (airborne LiDAR). Airborne LiDAR is effective for nighttime surveys that are difficult with photogrammetry and for acquiring ground surface morphology in areas densely covered with trees. Because it offers high point-cloud density and accuracy, it is a powerful option when you need to scan large areas quickly and in detail. However, drones are susceptible to weather conditions such as wind and rain, and attention must also be paid to flight restrictions under aviation law. Since flight permission applications may be required in urban areas or around airports, be sure to confirm this during the planning stage.

Meanwhile, ground-based 3D laser scanners (terrestrial LiDAR) are also indispensable for acquiring high-accuracy point clouds. Large laser scanners mounted on tripods can measure very dense point clouds of millions of points per second and record terrain and structures with millimeter-level accuracy (a few millimeters (about 0.04–0.35 in)). For example, when there are complex structures (bridges or plant equipment, etc.) around planned excavation sites, or when you need to measure areas that are difficult for drones to reach, such as wall surfaces or the interior of underground pits, ground-based scanners are the right choice. By scanning multiple times from different positions as needed and later integrating those scans, you can build a detailed 3D model without blind spots. Ground scanners require specialized operators and time, but because they provide precise data, they remain a reliable method around critical facilities and for applications that prioritize accuracy.

By combining different methods according to the situation—from easy smartphone point-cloud capture, to drone-based wide-area surveying, to high-precision measurements with terrestrial scanners—you can comprehensively record all information around buried pipes in 3D. On site, it is effective to first cover the main areas with quick smartphone scans, acquire an aerial overview model with a drone in parallel, and then supplement only the parts that need detailed information with a terrestrial scanner. The multiple-source point cloud data obtained this way can be integrated in post-processing to produce a single, consistent 3D model.

Step 4: Integration and processing of point cloud data and generation of buried-pipe models

Next, we enter the step of integrating the various point cloud datasets acquired on site to generate a buried-pipe location model. Point clouds obtained by different instruments or methods may have differing axis orientations and scales, so they are converted to a common coordinate system and overlaid. As described in Step 2, if the data were positioned with RTK-capable equipment, they are already referenced to the public coordinate system and can be integrated relatively smoothly. If point cloud data were recorded in a local coordinate system, an appropriate transformation is performed by matching markers placed on site or the positions of known points. When overlaying multiple point clouds, slight misalignments may occur, but if there are artificial features captured in common (for example, road signs, building corners, or manhole covers), they are fine-tuned so that those features align exactly to improve integration accuracy.

Once point cloud data have been registered into a common 3D space, the data are cleaned up by removing unwanted noise points and interpolating missing areas. If vegetation on the ground causes disturbances in the point cloud, it is removed by filtering, and if important structures are captured only sparsely as points they are reinforced by cross-referencing surrounding photographs. Recently there are point-cloud processing services that run in the cloud, and some can automatically remove noise and perform meshing (conversion from points to surface models) at the push of a button. For example, environments are becoming available that allow non-experts to instantly generate contour lines and cross-sections from acquired terrain point clouds or to automatically calculate excavation volumes.

For visualization of the buried pipe itself, it is necessary to extract the pipe’s position and shape from point cloud data and model them. In the case of new installation work, if you directly scan the pipe and acquire point cloud data before backfilling, you can extract the pipe’s surface point cloud from that data and create a pipe-shaped 3D model. Even with a smartphone scan, if the piping is exposed its circular contour will appear in the point cloud, so you can trace it and generate a pipe model about 200 mm (7.87 in) in diameter. Specialized software also has functions to fit and extract cylindrical shapes from point clouds. The resulting buried pipe model will reflect the exact route (horizontal position and depth in the vertical direction). Even after backfilling and the ground surface has been restored, having this 3D model alone allows you to accurately reproduce the pipe’s route at any time.

Even if an existing buried pipe cannot be scanned, there is a method to create a 3D model of the pipe from coordinate information in existing records and place it onto the point cloud. For example, by deriving on-site the coordinates of the pipe’s start and end points and intermediate bend points identified from old drawings, and drawing a three-dimensional line that connects those points, you can generate an approximate buried-route model. Depth information should be reflected if it is recorded on the drawings, and unknown sections can be filled in by mixing estimates and measured values. Place this against the surrounding terrain point cloud and check for any obvious inconsistencies as you position it (for example, if it clearly appears above the ground surface, there may be a depth input error). Adjusting it in point-cloud visualization software with transparency enabled makes it easier to grasp the distance from the ground.

The completed 3D model data is uploaded to cloud storage or a GIS (Geographic Information System) and shared and managed among all stakeholders. Whereas traditionally as-built drawings and photo logs were stored as paper files for each project, centralizing them as digital 3D data dramatically improves information searchability and prevents deterioration. If another construction project arises in the future, referring to this data allows instantaneous assessment of interference with other buried assets, enabling design changes or adjustments to construction methods in advance. Because pipe diameters and burial depths can be measured directly from the model, precise preliminary studies can be conducted without excavating the site to verify. Terrain data acquired as point clouds can be imported into CAD or BIM software for use in design or reviewed together by stakeholders in a VR environment, producing broad value that goes beyond mere record-keeping.

Step 5: Visualization of Buried Pipes Using Point Clouds and Leveraging Them for Safe Excavation Planning

Finally, you reach the stage of using the completed 3D point-cloud model of the buried pipes to support actual excavation work. If you have accurately positioned pipe data in three dimensions, it first helps when planning the work at your desk. For example, you can create materials to inform machine operators and workers in advance by overlaying the underground pipe model onto photographs of the surface, or output cross-sectional views in a point cloud viewer that show the positional relationship between the ground surface and the pipes. This makes points such as "at which depths and from which positions caution is required" and "where nearby buried utilities are located" immediately clear, making it easy to share safety measures. Especially on sites where multiple lifelines intersect, spatial visualization using point clouds is highly effective for preliminary risk assessment.

Furthermore, point cloud data is a powerful ally even during on-site excavation work. Recent smartphones and tablets have rich AR capabilities, and when combined with the aforementioned high-precision alignment technologies, they can even make it possible to see through the ground to the pipes beneath. Specifically, if you hold a smartphone camera over the planned excavation spot on the ground, a computer-generated model of the buried pipes is overlaid on the screen. As if the ground had become transparent, the direction in which the pipes run beneath the pavement in front of you is shown in real time. For example, information such as “exactly 1.5 m (4.9 ft) below your feet a single water pipe with a diameter of 150 mm (5.91 in) crosses” can be displayed on the screen and shared with all workers. Whereas in the past buried utilities were found by relying on planar information on drawings and on-site intuition—probing around with “it should be around here”—anyone can now intuitively “see” and confirm them, greatly reducing the risk of unnecessary trial excavations and accidental digs.

To reliably perform this AR-based visualization, it is necessary to capture the smartphone’s position and orientation with high precision. As mentioned above, general GPS accuracy has errors of several meters (several ft), so without any countermeasures the displayed position of the virtual model can be significantly offset from the actual buried location. However, by using RTK-GNSS the smartphone’s position can be determined to within a few centimeters (a few in), and when combined with LiDAR-acquired ground surface point clouds it becomes possible to align the virtual pipe model almost perfectly with real space. If “markerless high-precision AR” that does not drift even when freely walking around without placing special markers on the ground can be realized, safety checks at excavation sites will become dramatically more efficient. Workers can proceed with their tasks while always being aware of underground structures, allowing them to make appropriate judgments—such as noticing in advance when a shovel is about to hit a pipe and avoiding it.

Above, we explained five steps to visualize buried pipes with point cloud data and prevent positional misalignment before excavation. Thanks to the recent emergence of RTK-GNSS receivers for mobile devices and advances in 3D scanning technology, an era is approaching in which anyone on site can easily perform centimeter-level point cloud measurement (cm level accuracy (half-inch accuracy)) and visualize buried utilities. For example, LRTK is a high-precision GNSS device that can be attached to an iPhone, enabling a single smartphone to serve as a surveying instrument, a point cloud scanner, and an AR display terminal. By leveraging these cutting-edge technologies, accidents caused by positional misalignment of buried pipes can be prevented in advance, realizing safe and efficient infrastructure construction and maintenance.