Can Point Clouds Be Used to Create Distress Maps? 4 Points to Know Before Implementation

Creating distress maps is an indispensable task in the maintenance of infrastructure such as structures, slopes, tunnels, retaining walls, bridges, and stone masonry. Accurately identifying damage such as cracks, peeling, uplift/voids, missing sections, settlement, displacement, water flow paths, and deformation, and organizing them into drawings, determines the quality of repair plans and inspection records. At the same time, the field is constantly beset by issues such as labor shortages, time constraints, inconsistent records, and difficulty in re-verification. Against this backdrop, in recent years an increasing number of practitioners have been considering whether point cloud data with three-dimensional positional information, not just photographs, can be utilized for creating distress maps.

In practice, point clouds can be effectively utilized for creating deformation maps. However, simply acquiring point clouds does not by itself produce a high-quality deformation map. Before implementation, you need to clarify what point clouds are good at and what they are not, the kinds of representation required in the deformation map, on-site acquisition conditions, coordinate management, and operational rules. Overreliance on point clouds can actually increase rework. Conversely, if you properly design the intended use and procedures, you can improve the reproducibility of records, reduce oversights, and make it easier for multiple people to perform checks.

This article organizes, for practitioners searching for "deformation map point clouds", a practical approach to using point clouds to create deformation maps. Centered on four key points you should know before adoption, it provides a detailed, step-by-step explanation of when this approach is effective, where caution is required, and how to operate it on-site to achieve results.

Table of Contents

• Why point clouds are attracting attention in the creation of deformation maps

• Point 1: How to interpret recording accuracy and reproducibility

• Point 2 Understand information that point clouds alone cannot fully represent

• Point 3: Coordinate Management and Acquisition Conditions Essential for Deformation Comparison

• Point 4: Operational design to ensure full utilization on site

• Approach to creating deformation maps leveraging point clouds

• Situations in which LRTK is effective at sites seeking to streamline the creation of deformation maps

Why point clouds are attracting attention in the creation of deformation maps

The main reason point clouds are attracting attention for creating deformation maps is that they can capture site conditions widely and objectively in three dimensions. In conventional deformation-map creation, the typical workflow has been to verify the object on site while taking measurements, photographs, and notes, and then convert those into drawings after returning to the office. This method is a solid approach backed by experience, but the level of detail in the records tends to vary depending on the person in charge, which can leave insufficient information when someone else wants to check later. Particularly at locations that are difficult to revisit or at sites where a large area must be recorded in a short time, that weakness tends to become apparent.

Point clouds retain an object's surface shape as a collection of many points, which has the advantage of making it easy to cut cross-sections at arbitrary positions later, check distances and heights, and review from different angles. In deformation maps, it is important not only to identify the location of deformation but also to record the surrounding shape and junctions, the extent of the affected area, the tilt of slopes and walls, and the presence of steps or subsidence. Using point clouds makes it easier to supplement information that can be easily overlooked in simple planar records.

Point clouds are also effective from a sharing perspective. With only photographs and handwritten notes, it can be difficult for all stakeholders to understand the shooting direction, sense of scale, and the spatial relationships of deterioration at the same level of resolution. On the other hand, because point clouds are stored with positional information, they are easier to treat as a common foundation among clients, inspection personnel, design personnel, and repair-planning personnel. The fact that they make it easier to reduce on-site misunderstandings provides value far beyond mere drawing efficiency.

Furthermore, in recent years at maintenance sites, the importance of comparisons over time—not just single-year records—has been increasing. Distress is not something to check once and be done with; it is necessary to continuously monitor whether it is progressing, stable, or recurring after repairs. In this context, point clouds are effective as baseline data for comparison. If records can be captured in the same coordinate system, covering the same target area, and under similar acquisition conditions, it becomes easier to verify differences from previous years and to track positions. If distress/condition maps are regarded not as one-off deliverables but as the entry point to long-term management, the value of point clouds increases even further.

However, it must not be misunderstood that point clouds are merely one means of recording and do not replace the purpose of the deformation map itself. A deformation map is a deliverable that interprets observed phenomena and organizes them into a form that is easy to manage. Point clouds are a supporting source of information and do not automatically perform tasks such as assigning meaning to deformations, interpreting them, applying rules for drawing representation, or organizing data according to inspection standards. Understanding and applying this premise is what determines the success or failure of implementation.

Point 1 How to Interpret Recording Accuracy and Reproducibility

Before implementation, the first thing to consider is to clarify what you expect from point clouds. In many sites, point clouds are expected to achieve both improved accuracy and greater efficiency, but these two do not always mean the same thing. The accuracies required for creating deformation maps include several types: positional accuracy, dimensional accuracy, shape reproducibility, accuracy in determining the extent, and consistency of comparisons. If you introduce point clouds, you need to decide in advance which type of accuracy to prioritize.

For example, whether you want to precisely evaluate the crack width itself or to comprehensively document the cracks' locations and distribution on a drawing will greatly change the required data quality. The interpretation of fine defects is strongly influenced by acquisition density, distance to the target, lighting conditions, surface condition, and the amount of noise. If, in an effort to capture a wide area in a short time, you acquire a coarse point cloud from a position far from the object, you may grasp the general shape but lack the information necessary to translate it into a damage map.

On the other hand, a major strength of point clouds is that they can preserve highly reproducible records. The fact that you don't have to rely solely on an inspector's intuition or notes, and that another inspector can later review the same data, is extremely important in practice. When creating deterioration maps, leaving information at a consistent level that might be needed later—rather than only recording the parts that seemed noteworthy on site—contributes to the reliability of the deliverable. Point clouds are advantageous in this respect. If the site has been fully captured, it becomes easier after returning to the office to review the data, check cross-sections, verify dimensions, and re-evaluate the extent of deterioration.

However, high reproducibility does not mean that all deformations can be represented accurately. Point clouds can have blind spots, and acquisition quality may decline in deeply uneven areas, complex junctions, shaded regions, surfaces with strong light reflection, or wet surfaces. Also, when the target is small or surface changes are very slight, it can be difficult to judge based on point clouds alone. Therefore, accuracy in the creation of damage drawings should be considered not merely as mechanical numerical precision but from the perspective of whether the information necessary for drafting has been preserved in a state that can be reverified.

From this perspective, the success or failure of introducing point clouds is not determined by the sophistication of the measurement functions alone. Only when the type of object, the required deliverables, the site conditions, the data collector's understanding, and the post-processing procedures align will accuracy and reproducibility have practical significance. If point clouds are to be introduced for deformation mapping, it is important to design the process to include questions such as "how much needs to be recorded to be sufficient," "which deformations should be supplemented by other methods," and "who will check which screen."

Point 2 Understand information that point clouds alone cannot fully represent

The second point is not to assume that a deformation map can be completed using only point clouds. While point clouds are powerful for understanding shape and positional relationships, they do not necessarily always convey the nature of the deformation, its appearance, or the condition of the surface. A deformation map must organize not only the fact that an anomaly exists, but what kind of deformation it is, how advanced it is, which direction it is extending, and what effects it has on the surrounding area. At that time, there will certainly be information that is difficult to judge from point clouds alone.

For example, changes in the color of cracks, wet staining associated with leaks, the occurrence of efflorescence, the flow of rust stains, the weathering condition of materials, suspected delamination, and the way the surface is peeling may not be fully captured by shape alone. Since a defect map ultimately serves as documentation that informs management decisions, auxiliary information necessary for interpreting these phenomena is indispensable in addition to shape records. In practice, it is common to combine point clouds with photographs, on-site inspection notes, and results from tapping tests and visual inspections to assign meaning to observed defects.

If you do not understand this point and assume that on-site verification can be minimized simply because point clouds have been acquired, you are more likely to fail. In particular, at stages that lead to estimating causes of deterioration or considering repairs, it is necessary to organize visual information together with positional information. In other words, point clouds form the foundation of a deterioration map, but combining them with other recording methods is a prerequisite for producing the final deliverable.

Also, deformation drawings serve the purpose of organizing information for clarity. Although point clouds themselves contain a large amount of information, left as-is they can be difficult to interpret as deliverables. Even if they are convenient for field personnel, for submission deliverables you must select and omit information as necessary and translate it into drawing representations that organize deformation symbols, annotations, dimensions, and spatial relationships. What is required here is the perspective of creating drawings based on point clouds. Decisions are needed about which cross-sections to adopt, which position to use as the reference, which lines or symbols to use to indicate the extent of deformation, and how much to include in the deformation drawing.

Furthermore, at some sites the representation rules and delivery formats required for deformation drawings may already be established. In such cases, the issue becomes how to connect point-cloud-derived information with the existing rules. If you do not consider how the deliverables will be produced before introducing point clouds, you may end up having acquired the data but requiring extra work to produce drawings, increasing the verification process compared with before, and causing interpretations to vary among personnel.

Before implementation, the mindset should not be “do everything with point clouds” but rather “decide what to reliably capture with point clouds and what to supplement with other information.” In workplaces where this is clearly organized, point clouds are extremely effective. Conversely, if you introduce them while this organization is still unclear, you will end up with more records but inconsistent quality of deliverables.

Point 3 Essential coordinate management and acquisition conditions for deformation comparison

The third point is that if you want to make use of deformation maps for future comparisons, standardizing coordinate management and acquisition conditions is indispensable. Point clouds really show their strength in deformation maps not only for single-year plotting but when comparing with past data, tracking conditions before and after repairs, and grasping trends in displacement or settlement. However, as a prerequisite, if the methods of positioning and the acquisition conditions are not consistent, the reliability of such comparisons is greatly reduced.

For example, if one year’s records use an arbitrary on-site reference point and another year’s records use a different reference, the positions may look similar on a map but precise comparison becomes difficult. To judge the progression or expansion of deformations, it is important to handle data based on the same coordinate system or on aligned reference points. This is not just an issue for large-scale surveys but a basic principle that applies to all deformation maps intended for comparison.

Also, unifying acquisition conditions must not be overlooked. Point clouds can look different depending on the distance to the object, the angle, maintaining visibility, acquisition density, ambient light, surface wetness, and the effects of traffic and vegetation. If conditions differ greatly by year or by operator, areas that have not actually changed may appear different, or conversely real changes may be obscured. If you plan to use condition maps as ongoing management records, it is important to establish acquisition procedures and standards.

What is particularly important on site is to make clear what will be used as the reference for aligning positions. If you set references that can be used consistently—reference points, control points, known points, or clearly defined corners of structures—the alignment when re-acquiring data will be stable. Conversely, if the practice is to just roughly photograph the same spot each time, reproducibility will decline as soon as the person in charge changes. Before the introduction of point clouds, you should establish reference standards so that whoever is responsible for updating the deformation maps in the future will not be left uncertain.

Furthermore, coordinate management involves not only on-site work but also sharing information both inside and outside the company. When multiple parties—such as designers, maintenance teams, and contractors—handle the same positional information, having consistent coordinates speeds up discussions. Explanations of defect/damage locations can be made not as “around here in this photo” but by being commonly pointed to on drawings or point clouds. This is a major advantage when considering the scope of repairs and directing additional investigations.

If you want to make use of point clouds in deformation maps, don't judge solely by the ease of measurement. It's important to establish a registration system with future comparisons, re-surveys, and stakeholder sharing in mind. Sites that design this level of setup at the time of implementation are more likely to appreciate the value of point clouds a few years later.

Point 4: Operational design to ensure full utilization on-site

The fourth point is not simply acquiring point clouds, but creating an operational workflow that can be fully used within the process of producing deformation maps. In practice, it is often the post-measurement steps—organizing, naming, sharing, verifying, and drafting—that cause more problems than the measurement itself. To successfully implement this, you need to clearly define who, when, at which stage, which data they will look at, and what decisions they will make.

First, what’s important is how you define the target area. Deformation maps need to be organized so they are easy to read as deliverables, but if point clouds are captured too broadly they become heavy to handle, and if captured too narrowly you lose the context with the surrounding areas. Therefore, at the acquisition stage it is effective to establish rules such as “include not only the target deformation but also reference ranges in front, behind, left and right” and “do not expand unnecessary ranges too much.” This makes it easier to crop and compare data once back at the office.

Next, unifying naming conventions and management units is necessary. If site names, structure names, spans, surfaces, positions, acquisition dates, personnel, and so on are not managed according to consistent rules, you will be unable to find the data you need later. Because distress maps are often referenced in ongoing management, tidying them up sloppily as a one-off task becomes a major liability in subsequent years. In particular, when dealing with multiple locations on the same day, confusing positions can lead to serious mistakes, so rules for recording are essential.

On-site verification procedures are also important. Just because point clouds can be post-processed, operating without checking for missing data or blind spots on-site is risky. You need to perform a minimum check immediately after acquisition and have a system to add data on the spot if necessary. If insufficiencies are noticed after leaving the site, the effort required for a return visit increases significantly. When creating deformation maps, a defect in a single drawing can reduce the explanatory power of the whole, so the immediate post-acquisition verification step should not be omitted.

Furthermore, when the person handling the point cloud and the person drawing the deformation map are different, the handover rules must also be clarified. If information such as which coordinates are being used as the reference, which surface should be prioritized for inspection, what points were of concern on site, and which photos correspond to which locations is lacking, decision-making will stall at the drafting stage. In other words, not only the point cloud data itself but also a system for providing the surrounding information necessary for creating the deformation map is required.

If these considerations are made before implementation, point clouds become more than just a new recording technology; they become a tool for stabilizing the entire process of creating deformation maps. Conversely, if data acquisition is prioritized before operational procedures are in place, methods will vary by person and, as a result, you will end up with double management alongside conventional methods. To fully utilize them on site, operational design should be considered more important than measurement capabilities.

How to Create Deformation Maps Using Point Clouds

So, in practice, how should you proceed to make use of point clouds when creating deformation diagrams? Fundamentally, you should keep a consistent workflow of organizing site conditions, acquisition planning, on-site recording, post-return verification, drafting, and comparison management. The important thing is to clarify the role of the point cloud at each stage.

In the initial stage, organize the structure to be inspected and the types of defects. Depending on whether you want to see deformation of the entire wall surface, map the distribution of cracks, compile the locations of delamination, or check for steps and settlement, the required acquisition range and density will vary. If this organization is vague, failures can occur: you may capture a wide area but lack detail, or focus so narrowly on details that you cannot determine the overall positions.

Next, decide how to establish the reference. If you plan to make future comparisons, determine which coordinate system to use and what will serve as the reference for alignment. If the site is large, it will be easier to manage if you also decide here how to divide sections or surfaces. Because deformation maps ultimately need to be easy to read on paper or on screen, it is effective to make the acquisition unit match the output unit.

On site, photos and notes are recorded together with point cloud acquisition. The point cloud captures the geometry, photos supplement the surface condition, and notes record on-site observations, so that when back at the office you have the information needed for decision-making. At this stage, it is important that each photo is linked to its location information. If, when reviewing them later, you cannot tell which damage corresponded to which location, the point cloud data cannot be fully utilized.

After returning to the office, first check the data for omissions, duplicates, blind spots, and noise, and clean it up as necessary. Then create the cross sections, elevations, and developed views required for the deformation map, and translate the deformation locations and extents into drawing representations. Here, rather than making the point cloud the deliverable itself, it is important to use it as supporting evidence for the deformation map and organize it into a representation that is easy to manage. This is because those who view the deformation map are not necessarily familiar with point cloud manipulation.

Furthermore, it is important to manage the completed deterioration map by linking it to the original data. If only the drawings remain and you cannot later verify their supporting evidence, ongoing management will be weakened. Conversely, if only the original data exist and they have not been organized into drawings, they become difficult to use in routine maintenance management. Connecting the two will prove extremely valuable for future re-surveys and repair planning.

Thus, using point clouds to create damage maps is not simply about capturing three-dimensional data; it is more akin to redesigning the entire workflow of recording, verification, drafting, and comparison. Sites where implementation tends to be most effective include those where you need to record large areas quickly, where revisits are difficult, where comparisons over time are important, and where multiple people need to inspect the results. Conversely, in situations where detailed surface observation is the primary concern and visual assessment outweighs positional information, point clouds may serve only as an auxiliary tool. The important thing is not to treat point clouds as a panacea, but to assign them roles appropriate to the objective.

Situations Where LRTK Excels on Sites Aiming to Streamline Deformation-Map Creation

When using point clouds to create deterioration maps, a common practical challenge is how to stably provide positional information to the data acquired on site. This is especially valuable when surveys span multiple days, when a different person needs to recheck the site, or when you want to handle surrounding photos, point clouds, and repair locations according to the same reference. No matter how easy the point cloud is to view, if the positional alignment is ambiguous, comparisons and sharing can easily become confusing.

In such situations, a system that makes on-site coordinate verification and location recording easy determines the overall efficiency of producing deformation maps. In particular, for practical tasks such as recording deformation locations, positioning supplementary photographs, obtaining coordinates of reference points, and confirming the exact same point on revisits, it is important to be able to incorporate high-precision position information into daily work without difficulty. The deformation map itself is the final deliverable, but by putting the preceding recording processes in order, the reliability and reproducibility of the drawings is greatly increased.

In that respect, LRTK is a method well suited to field work. LRTK is a smartphone-mounted GNSS high-precision positioning device, and its characteristic is that it makes it easy to carry out recording tasks while confirming on-site positions at centimeter-level (half-inch accuracy). When you want to streamline the verification of control points and the acquisition of on-site coordinates in the preparatory stages before creating deformation maps, when you want to attach position information to supplementary photos or inspection points, or when you want to align the positional relationship between point clouds or drawings and the field, it delivers practical benefits. A major advantage is that it is easy to use in routine maintenance and inspection records without having to organize a large-scale arrangement.

If you are considering introducing point clouds into deformation map creation, it is important to think not only about point cloud processing itself but also about how to stabilize position recording at the site. For sites that want to improve the reproducibility of records and create deformation maps that can be reviewed later, the accuracy of positional information becomes the foundation of the deliverable. Using LRTK makes it easier to organize on-site the locations where deformations were confirmed and related records, and helps reduce rework when creating deformation maps. If you want to make point-cloud-based deformation map creation implementable in the field rather than leaving it as a desk study, there is considerable value in adopting high-precision position-recording methods such as LRTK.