The New Standard for Solar Panel Inspections: Pinpoint Faulty Panels with a 2 cm (0.8 in) Error Margin

Challenges with Conventional Solar Panel Inspections

In operations and maintenance (O&M) of solar power plants, solar panel inspection is indispensable. However, conventional inspection methods have many issues that create inefficiencies in the field. For example, in large-scale megasolar sites thousands of panels are arranged in regular arrays. Traditionally, inspections relied on visual checks and handheld infrared thermography cameras, requiring inspectors to check panels one by one. This approach takes enormous time and manpower, and long hours in the blazing sun impose heavy burdens on workers. If abnormal spots are missed and left unattended, generation losses accumulate and may develop into serious failures. Although discovering and addressing defects early through periodic inspections is important, inefficient methods prevent teams from fully leveraging inspection opportunities. There were also risks of oversights, and even when abnormalities were found, records were analog (paper notes or photos), making it difficult to accurately share locations and conditions later.

Recently, panel anomaly detection via infrared imaging using drones has emerged, enabling wide-area inspections in a short time. For example, airborne drone imaging can check panels representing several MW in about 30 minutes and identify hotspots (high-temperature anomalies). Note that infrared hotspot detection is limited to times of sufficient sunlight, so efficiently identifying anomalies in a short period is required. However, even if a drone identifies a suspicious panel, the on-site task of finding and repairing or replacing that faulty panel remains. Ultimately, someone still has to identify “which specific panel is faulty” within the power plant, and this step continued to consume time and effort. Remote monitoring by string and electrical diagnosis using IV curve tracers can also detect anomalies, but those methods still require measuring each panel in the string one by one on site to determine the culprit, which likewise takes time and labor.

On site, staff search for faults using layout diagrams and number lists, but finding a specific single panel among hundreds of rows is not easy. Discrepancies between positions on paper drawings and the real objects, inadequate signage or markings, and faded management numbers due to aging all cause mismatches between physical panels and data. Also, confirming a specific panel requires locating the serial number label on the back, but searching for the relevant number among a vast number of panels is unrealistic. Staff end up repeatedly backtracking within the target area to confirm, leading to on-site inefficiency.

Why “Positioning Error” Becomes a Problem in Identifying Faulty Panels

Even if an anomaly is detected during solar panel inspection, a slight error in position identification can greatly complicate on-site work. This is because panels are densely installed within the site and each looks almost identical. For example, if GPS-derived position information has an error of several meters (several feet), multiple panels will be located around the indicated coordinate. An error of 5 m (16.4 ft) could even point to a completely different row of panels. Under such conditions, searching by guessing “it’s probably around here” relies on intuition and experience and is inefficient. Especially in multi-MW-scale plants where thousands of similar panels exist, even a deviation of several tens of centimeters (several dozen inches) can lead to misidentification.

Large positioning errors also carry the risk of repairing or replacing the wrong panel. Removing a normal panel that should not be serviced, or overlooking and leaving the true faulty panel untreated, directly results in generation loss and unnecessary costs. Moreover, taking time to identify a faulty panel increases the power plant’s downtime. To perform rapid and accurate maintenance, the key is how precisely and quickly position identification can be performed on site.

Field staff have reported that “identifying the anomaly can sometimes take more time than the actual repair work.” On large sites, miscommunication among staff has led to mistakenly swapping target panels. Solving these issues requires a system that allows anyone on site to unerringly locate the faulty panel to the exact spot.

The World of 2 cm (0.8 in) Error with RTK Positioning and the Potential of Smartphones

In recent years, RTK positioning has gained attention as a technology to solve this problem. RTK (Real Time Kinematic) is a method that corrects satellite positioning (GPS) errors in real time, dramatically improving positioning accuracy. Normally, smartphone-integrated GPS has errors of about 5–10 m (16–33 ft), but using RTK reduces that to within a few centimeters (within a few inches). In fact, RTK now enables current positions to be determined with an astonishing error of about 2 cm (0.8 in). Two centimeters is only a few percent of a typical solar panel’s width (1 m (3.3 ft) or more), and is a level of precision at which the human eye hardly detects any misalignment. This “2 cm (0.8 in) error world” is poised to become a new norm in solar panel inspection.

What is particularly groundbreaking is that such high-precision positioning can now be achieved easily with a smartphone. Recently, compact RTK-GNSS receivers that connect to smartphones have appeared, and with dedicated apps, palm-sized devices can provide centimeter-class positioning (half-inch accuracy). One example is LRTK, a proprietary smartphone RTK system: a thin, lightweight device attachable to an iPhone plus an app enables anyone to easily enhance real-time positioning accuracy. Tasks that once required surveying tripods, expensive GNSS receivers, and two specialized technicians can now be completed with a smartphone in hand.

In Japan, the use of the high-precision signal (CLAS) from the Quasi-Zenith Satellite System (QZSS: Michibiki) is making RTK high-precision positioning possible even in mountainous areas outside smartphone communication coverage. With such mechanisms, large rural power plants can achieve positioning errors of just a few centimeters (a few inches) without installing base stations, further improving the technology’s practicality. The advantage of smartphone-based high-precision positioning is not only device miniaturization but also the ease of data utilization. Accurate coordinate data acquired in a smartphone app can be immediately saved to the cloud and shared with other team members in real time. With GPS positioning accuracy dramatically improved, smartphones have become worthy of being called “high-precision surveying tools,” beginning to transform field workflows.

How AR Guidance, Point Cloud Data, and Cloud Integration Change Solar Panel Inspection Workflows

The combination of smartphones and high-precision RTK positioning (for example, LRTK) dramatically changes field workflows for solar panel inspections. At the core are AR-guided navigation to faulty panels and the 3D point cloud and cloud sharing of recorded data.

① Intuitively identify faulty panels with AR: By using augmented reality (AR) technology to overlay digital information on live camera views of the site, the location of a faulty panel can be directly marked in real space. With high-precision coordinates, an app can display icons or arrows like “the target panel is here” on the actual panel. For example, when a coordinate of a faulty panel is specified in an RTK-enabled smartphone app, a virtual tag indicating that panel will align precisely in the camera view. Workers simply walk in the direction shown on the smartphone screen to arrive at the target panel without confusion. There is no need to count row numbers while holding drawings or to search for markings as before. Because the AR display does not drift even after walking several meters (several feet), the target panel can be found at a glance even across a wide area. AR’s visual guidance eliminates the hassle of matching map positions to physical objects and enables anyone to intuitively locate the target.

② Record the entire field situation with point cloud data: Combining high-precision position information with smartphone-mounted LiDAR (laser sensors) or cameras makes it easy to record the site as 3D point cloud data. For example, scanning the area around a faulty panel and the mounting structure with a smartphone can save a detailed 3D model composed of countless points. Previously, 3D as-built records required specialized laser scanners or drone imaging, but the smartphone + RTK combination enables anyone to easily perform 3D scans. By viewing the acquired point cloud data, installation angles, surrounding terrain, and shading conditions can all be visualized in three dimensions. Inspections require understanding not only the faulty panel itself but also the surrounding environment and aging effects. Recording the entire site as a point cloud makes it easy to analyze details in the office later or compare data from multiple time points to detect changes. You can also measure dimensions on the point cloud or calculate distances to obstacles, so it is useful for ancillary tasks such as assessing shading risk due to vegetation growth. Moreover, point cloud models uploaded to the cloud can be viewed remotely, allowing specialist technicians to grasp conditions and provide advice without traveling to the site.

③ Cloud integration for information sharing and record management: A key strength of smartphone DX is the ability to immediately upload field-acquired data to the cloud. When a photo with high-precision coordinates is taken, it is automatically tagged with “when, where, and which direction,” and saved to the cloud. This allows information such as “Hotspot found on panel No. X in row Y of the west area of the plant on YYYY/MM/DD” to be accumulated in a digital ledger. Tasks that previously involved handwritten notes or pasting photos into Excel reports can now be automatically aggregated and organized in the cloud. Managers in the office and remote insurance companies can share information in real time via the cloud, enabling faster judgments and instructions. Inspection reports can be generated semi-automatically by fitting data into cloud templates, greatly reducing the workload of field personnel. In addition, data management without paper forms prevents document loss and communication errors. Since histories are accumulated in the cloud, it is easy to check for past inspection omissions or duplicates.

In this way, the fusion of smartphone + RTK + AR is enabling the solar panel inspection workflow to be digitized end-to-end from “anomaly detection” to “on-site pinpointing” to “recording and sharing.” This is not merely efficiency improvement: accumulated data strengthens preventive maintenance and turns tacit knowledge into shared assets.

Typical Operational Example and Benefits: Time Savings, Improved Accuracy, and Addressing Labor Shortages

So how does field operation change when these technologies are actually adopted? Below is an example inspection scenario at a typical solar power plant and the benefits of implementation.

Operation example: At a megasolar site of about 10 MW with approximately 30,000 panels, monthly inspections adopted a combination of drone infrared diagnostics and smartphone RTK. First, automated drone flights captured thermal images of all panels in about one hour. AI analysis mapped several dozen candidate abnormal panels, and the coordinate data were sent to the field staff’s smartphones. Staff moved through the plant with smartphones in hand and followed the app’s AR navigation to each indicated point in sequence. The screen displayed messages like “Row X, Panel Y: high-temperature anomaly,” and a marker floated above the corresponding panel so the target was instantly recognizable. For panels suspected of actual failure upon on-site confirmation, marking stickers were applied and high-precision coordinate-tagged photos were taken and recorded to the cloud. Finally, a manager reviewed the photos and data in the cloud and made same-day decisions on parts replacement and warranty claims.

Effects: Where this series of tasks previously required, for example, a total of three people over two days, after implementing the technology one person can respond in one day. Screening by drones and AI dramatically reduced oversights, and AR navigation drastically cut the time spent searching for target panels on site. The ability to pinpoint the target without getting lost across large areas was particularly impactful, enabling even inexperienced staff to accurately identify faulty panels. As a result, work time was reduced to less than half, and personnel requirements dropped to less than one-third of the traditional level to perform the same inspection tasks. This translated directly into reduced operating costs and freed time for other important tasks. Maintenance of generation output and reduced O&M costs provided significant managerial benefits, making the investment in advanced technology worthwhile. In addition, reduced time working long hours outdoors contributed to lower heatstroke risk and improved work safety.

Accuracy also saw major improvements. Accurate photo records tied to coordinates eliminated omissions and recording errors. Later, when checking “which panels have been replaced” or “where previous anomalies occurred,” the relevant locations can be displayed with one click from the accumulated cloud data, making historical comparisons easy. This greatly improved the reproducibility of inspection and repair work, enabling consistent equipment health maintenance without relying on veteran intuition. Improved efficiency made it possible to increase inspection frequency as well—for example, converting detailed inspections that used to be possible only once a year into monthly checks—strengthening preventive maintenance. The approach was also effective as a countermeasure to chronic labor shortages. Using smartphone RTK and AR guidance allowed inspections that previously required two to three people to be covered by one person. Because even non-experts can follow digital guides to perform tasks, the burden of tacit on-site skills was reduced and overall team productivity improved. As a result, field teams have praised the system: “Even with limited staff, we can perform more thorough inspections than before.”

Conclusion: LRTK’s Potential to Expand into Other O&M Tasks and Simple Surveying

The smartphone RTK and AR linkage technology that can pinpoint faulty panels with a 2 cm (0.8 in) error margin is not only establishing a new standard for solar panel inspection, but also holds broad potential for future O&M tasks. For example, inspections of fences and transmission equipment within solar power plants could use high-precision coordinate-tagged photos and AR displays to manage anomalies, and substation patrols could reference past inspection histories on the spot. Smartphone RTK also performs well for site surveying and monitoring terrain changes. For simple surveying tasks, staff can obtain sufficiently accurate measurement data in the field using the LRTK system without arranging specialized surveying equipment. This is useful for planning new installations and for assessing conditions after disasters.

Moreover, after typhoons and other natural disasters, high-precision photo data can serve as evidence for insurance claims and contribute to rapid recovery decisions. Beyond solar panel inspections, the fusion of digitalization and precise geolocation is becoming the key in equipment management and infrastructure inspections. Centimeter-level accuracy (half-inch accuracy) achievable with the familiar combination of smartphone and RTK, together with information use via AR and the cloud, is expected to become the new field standard. Environments where anyone on site can use the system intuitively and data are centrally managed will not only improve efficiency and quality but also enhance work safety and comfort.

As AI analysis and robotics integration progress further, inspection DX will evolve to even higher levels. This solution, which could be called the “new standard for solar panel inspection,” is a next-generation O&M style that addresses labor shortages and skill transfer challenges. Smartphone high-precision positioning technologies such as LRTK are expanding their usefulness beyond solar generation inspection and maintenance into construction, civil engineering simple surveying, and infrastructure upkeep. Why not seize this new wave of technology and accelerate the DX (digital transformation) of your field operations? Let’s open up the future of on-site work together!