A New Era of Infrared Inspection: Smart Inspections with Point Cloud Data Integration for Zero Blind Spots

As infrastructure facilities and buildings age, advanced inspection technologies are required for their maintenance and management. Among these, infrared inspection has become an indispensable method because it can visualize abnormalities that are not visible to the eye as thermal images, enabling early detection of exterior wall deterioration and abnormal heating in electrical equipment. However, field maintenance technicians sometimes raise concerns such as “Will we miss abnormalities in blind spots?” and “It’s difficult to accurately record the location of where images were taken.” Conventional infrared inspections have required considerable effort for thermal image analysis and reporting, and reproducibility and comparative evaluation of results have remained problematic.

For example, imagine photographing every corner of a large building’s exterior wall with an infrared camera at night. Even if you find several cold areas on the thermal images that appear abnormal, it’s not easy to organize which positions on the building correspond to those spots afterward. In the end, you might need to mark positions on drawings by hand for each photo or leave tape marks on site. The more abnormal locations there are, the more complicated the management becomes, and the greater the risk of omissions. Thus, even though infrared makes issues “visible,” more ingenuity is needed to make full use of that information.

Limitations and Challenges of Conventional Infrared Inspection

Inspections using infrared thermography cameras can be performed safely and without contact, but their operation has several limitations and challenges. The main points are as follows.

• Location identification tends to be ambiguous: Even if thermal images show abnormal areas, it is not easy to later determine exactly where those locations correspond on the structure. With photos alone it becomes unclear “which part was photographed,” and additional work such as manually writing positions onto drawings when preparing reports was necessary.

• Difficult to compare with past data: In periodic inspections, deterioration progression is judged by comparing with previous results, but quantitatively comparing thermal images taken at different angles or distances each time is difficult. If the location mapping remains ambiguous, you cannot accurately track “whether the same spot had an abnormality last time” or “whether it has spread.”

• Records tend to become person-dependent: Infrared inspection results are reported with photos and text, but judgment of abnormalities and methods of recording often rely on the experience of the person in charge. Differences in reporting accuracy can occur between veterans and newcomers, and information may accumulate with individuals rather than being shared within the organization. If a veteran technician transfers or retires, past records may not be sufficiently handed over, and successors may have to search for abnormal locations from scratch.

• Risk of missed detections due to blind spots: There is also the limitation that abnormalities hidden in blind spots that the camera could not face cannot be detected. For example, parts behind equipment, shadows at height, or areas hidden by piping or obstacles that are not captured in thermal images can lead to misses. Even experienced technicians needed significant effort to photograph extensive areas without omission.

Thus, conventional infrared inspection methods had challenges such as “ambiguous data location,” “difficulty in longitudinal comparison,” “variability depending on the person,” and “remaining blind spots.” On site, measures such as recording shooting positions in detail and ensuring information sharing within teams were taken to supplement these issues. However, to fundamentally improve accuracy and efficiency, a new system is needed that can centrally manage and utilize infrared images as digital data rather than relying on human experience and manual work. Fortunately, in recent years high-precision positioning and 3D measurement technologies that leverage smartphones have emerged, and new inspection methods that solve these problems at once are becoming a reality.

Smart Inspection Method Integrating Infrared Images and Point Cloud Data

A trump card to solve these issues is a smart inspection method that integrates infrared images with 3D point cloud data. Using newly developed systems that support LRTK (high-precision positioning technology), you can record the abnormal spots photographed with an infrared camera as position information in three-dimensional space and link them to a 3D model of the entire structure for management. Let’s look at the key points in order.

• Adding high-precision position information with LRTK: LRTK is a system based on real-time kinematic positioning technology that enables positioning with several-centimeter accuracy (cm level accuracy, half-inch accuracy) on smartphones and similar devices. The equipment is compact and lightweight, highly portable, and easy to handle in the field, so operational burden is kept low. By using this, you can attach positioning tags to each point photographed with an infrared camera with accuracy that conventional GPS could not achieve. Latitude, longitude, and height information are recorded for every detected abnormal spot, and later you can precisely identify “which position the abnormality is at” on a map or drawing. This mechanism allows precise plotting of abnormal locations on maps even for equipment scattered over wide areas, dramatically improving traceability.

• Digitizing structures by 3D point cloud scanning: With LRTK-compatible 3D scanning functions, you can capture buildings and equipment as point cloud data (a three-dimensional model represented by countless points) using a laser scanner or smartphone-integrated LiDAR. By digitizing the entire subject without omission, you can create a site “digital twin” that includes parts not visible to the naked eye. This 3D point cloud becomes the foundation for linking thermal images of abnormal areas and allows intuitive understanding of the spatial relationships of abnormalities within the entire site. For example, in plant equipment with complex piping, if everything is point-cloud-scanned to the corners, you can grasp abnormalities on hidden backsides in the digital space that are difficult to confirm visually. Also, the once-acquired 3D model can be reviewed virtually off-site, helping check for missed shots and enabling additional analyses.

• Integrated management of infrared images and 3D models: By overlaying high-precision position-tagged infrared images onto a 3D point cloud model, you obtain a three-dimensional model with thermal distribution, a so-called “thermographic point cloud.” This allows each abnormality to be marked on the model so you can see at a glance whether any inspections were missed. Data from different dates are integrated into the same coordinate system, making it easy to compare chronological changes at the same exact locations. When preparing reports, you can record accurate positions and conditions while viewing the abnormal spots displayed on the model, eliminating the need to manually number photos and organize correspondence. Also, because individual thermal images are organized not as standalone photos but as parts of the whole structure, inspection results can be database-ized so anyone can recognize abnormalities by the same standards. This visualizes records that were previously person-dependent and contributes to organizational knowledge accumulation.

• Intuitive visualization and sharing with AR technology: The integrated 3D model can be shared on the cloud or confirmed on site with AR (augmented reality) display on a tablet. For example, during repair work, a person on site can simply hold up a tablet to overlay where on a wall or equipment an abnormality was found. High-precision alignment in AR display eliminates worries like “It’s here on the drawing, but where is it on the actual object?” By sharing 3D data between the field and the office, discussions can be held remotely based on the same information and the expertise of seasoned personnel can be utilized remotely. AR’s intuitive visualization is also useful for training newcomers, enabling less experienced technicians to confirm and address abnormal spots accurately on site. Information that was difficult to grasp from video or photos alone becomes instantly clear when overlaid on the real object, allowing all stakeholders to proceed with work with a shared image.

This smart inspection method solves conventional problems such as ambiguous locations and missed detections, enabling infrared inspection results to be truly datafied and utilized. The main benefits are summarized as follows.

• Prevention of missed abnormalities: Because all data can be checked on a 3D model, missed detections due to blind spots are reduced.

• Clarification of location: Abnormalities are recorded with coordinates, so you can identify them on site later without confusion.

• Easier longitudinal comparison: Because data are integrated under unified management, you can accurately compare the same spots with past inspection results.

• Information sharing and succession: Database-ized records allow accumulated knowledge to be shared even when personnel change.

• Improved inspection efficiency: You can acquire a 3D model and inspection results simultaneously in a single site visit, reducing time spent on report creation and post-processing.

Because abnormal locations are recorded with geographic coordinates, inspection results are accumulated as asset information and are useful for the next inspection or repair planning. If data are retained, future AI analysis becomes possible, enabling predictive maintenance such as anomaly forecasting and deterioration prediction.

Benefits of “Zero Blind Spots and Datafication” Seen in Use Cases

So in what kinds of sites does this infrared × point cloud data integration smart inspection produce noticeable effects? Below are three representative use cases: building exterior walls, electrical equipment, and solar panel inspections.

Smart Infrared Inspection of Building Exteriors

Infrared cameras are used in surveys of exterior walls of buildings and condominiums to detect tile or mortar delamination and peeling. Recently, revisions to the Building Standards Act have made periodic exterior wall inspections mandatory for buildings older than 10 years, and safe, efficient infrared surveys have become mainstream, replacing tapping surveys. Traditionally, when photographing large exterior wall surfaces from the ground, blind spots occurred at heights and building sides, posing a risk of misses. With smart inspection, the entire building is 3D-scanned to obtain a digital model of the exterior wall, and infrared shooting is conducted for each surface and integrated into the model. This enables mapping of thermal anomalies across all four facades, achieving an inspection with zero blind spots. Also, because the position coordinates of abnormal tiles are recorded, planning for scaffold placement and specifying repair areas for later repair work becomes smooth. Being able to accurately check against past inspection data to see “whether any new abnormalities have occurred this time” is also a major advantage for long-term building maintenance. In fact, one commercial building that introduced smart inspection detected several tiny tile delaminations that conventional methods had missed, and early repairs prevented a falling exterior material accident. As a result, resident safety improved and repair costs were reduced through early intervention. Stakeholders commented that “we can manage the building more confidently than before.” Furthermore, analyzing temperature distribution and spread of anomalies on the integrated data helps prioritize repairs and consider repair methods.

Smart Infrared Inspection of Electrical Equipment

In regular inspections of electrical equipment at factories and substations, it is important to use infrared to detect abnormal heating early in distribution panels, transformer connections, and cable deterioration points. Traditionally, thermal images taken for each device were used to organize reports indicating “which breaker was hot” and “which connection had an abnormality,” but as equipment became more complex the location explanations became cumbersome and on-site confirmation could be time-consuming. With the smart inspection system, the entire facility is recorded in 3D point cloud data and thermal abnormalities can be spatially plotted against components inside panels. For example, even in a large distribution panel, “abnormal heating spots” are marked in red on the 3D model so you can instantly identify problem areas. Maintenance personnel can overlay that data in AR onto the actual equipment and immediately pinpoint the relevant part — for example, “the cable connection on the second position from the left in the △ stage of device 〇〇” — without hesitation and take corrective action. For critical infrastructure, detecting thermal anomalies in advance and performing part replacement or repair helps prevent sudden power outages or equipment failures, supporting safe operation. This prevents oversights in electrical equipment inspections and delays in recovery work, enabling reliable maintenance. In practice, some factories avoided sudden production line stoppages by planning replacements for degraded components identified and located through smart inspection. Visualizing abnormalities as data helps maintain stable production operation.

Smart Infrared Inspection of Solar Panels

The fusion of infrared and point cloud data is also powerful for inspecting solar panels at solar power plants or on building rooftops. Because solar panels are numerous and distributed over wide areas, conventional methods made it difficult to manage abnormalities for each panel down to the location. After photographing hotspots (abnormally heated areas) with drones or handheld infrared cameras, identifying those locations required estimating the area on a map or marking on site, which was time-consuming. With the smart inspection method, you either 3D-scan the entire plant from the air or build a coordinate system using existing drawings and LRTK positioning, and then associate infrared images with each panel. Abnormal panels are marked on the 3D model, so you can immediately see which row and which panel had a problem even later. Managers can check the data on an office PC and give accurate location instructions to field staff, and field staff can quickly find the relevant panel using AR navigation on a tablet for inspection or replacement. Data integration also streamlines the inspection work itself. Vast facilities that previously took days to check manually can be diagnosed thoroughly in a short time, reducing downtime and labor costs. Even at large-scale solar power plants, integrated data enables comprehensive zero-blind-spot inspections, reducing generation loss and fire risk.

Note that in civil infrastructure inspections such as bridges and tunnels, methods that combine infrared and other sensor data with 3D models are also attracting attention. Although the targets differ, the common idea of “datafying the whole site to visualize abnormalities” will greatly contribute to the advancement of inspection work. For example, in one large-scale solar power plant, an infrared inspection that used to take several days was completed in half a day, achieving zero misses of defective panels. Site managers evaluated that “the burden of inspection has been greatly reduced and we can operate with peace of mind.”

Conclusion

The smart inspection method, which can be called a new era of infrared inspection, dramatically improves the accuracy and efficiency of field inspections and takes infrastructure and equipment maintenance to the next stage. Moreover, technologies that leverage smartphones are cost-advantageous compared to dedicated equipment, making them attractive for small-scale inspection businesses to adopt. In the age of on-site DX (digital transformation), introducing smart inspection is not merely efficiency improvement but has significant meaning in enhancing quality control and promoting skill succession. With high-precision positioning using LRTK and utilization of 3D point cloud data, inspections can shift dramatically from “relying on intuition and experience” to “data-driven inspections.” Missed abnormalities will decrease, records will be accumulated and asset-ized, and information will be easier to share among stakeholders, ultimately enabling optimized maintenance planning and cost reduction.

Furthermore, LRTK-compatible systems support field work beyond infrared inspections. For example, if simple dimensional measurements or positioning are needed during inspection, the LRTK function allows you to handle them with just a smartphone without bringing specialized surveying equipment. This is very useful for peripheral tasks of infrared inspection, such as measuring the area of an abnormal spot to calculate the quantity of repair materials or recording the positions of equipment not shown on drawings. As LRTK plays an active role beyond infrared inspections, site digitalization will advance further and overall work will become smarter.

The approach of “zero blind spots by point cloud data integration”, which breaks away from conventional wisdom, brings new value to inspection work in fields such as electrical, architectural, and civil engineering. Experience next-generation smart inspections in the field, where infrared inspection and advanced technologies are fused to deliver both safety and efficiency.

Going forward, such smart inspections are expected to become the standard for maintenance management and to greatly transform the way inspections are conducted. This is no longer a futuristic fantasy but a real technology that is beginning to permeate our sites.

Why not take this opportunity to review conventional methods and consider introducing smart inspections?