Study of a New Inspection Method in Building Inspection Operations - Verifying Efficiency with Drone x Image Analysis
Asahi Technology and JR West conducted a verification test using automatic drone navigation and image analysis for station facility inspections, confirming operational efficiency and high reproducibility.
📋 Article Processing Timeline
- 📰 Published: April 7, 2026 at 19:10
- 🔍 Collected: April 7, 2026 at 10:30
- 🤖 AI Analyzed: April 20, 2026 at 23:49 (325h 19m after Collected)
Asahi Technology Co., Ltd. (Headquarters: Himeji City, Hyogo Prefecture, Representative Director: Yasuyuki Yukinaga, hereinafter "ATCL") has conducted a study to streamline inspection operations of station facilities with West Japan Railway Company (Headquarters: Osaka City, Osaka Prefecture, President and Representative Director: Shoji Kurasaka). In this study, targeting station building roofs, outer walls, and overpasses, while strictly complying with railway-specific rules, we verified the effectiveness of highly reproducible photography that allows repeated shooting from the same route and angle through automatic drone navigation, as well as the automatic extraction of degraded areas using image analysis.
1. Background
The maintenance and management of station facilities, such as station buildings, overpasses, platform roofs, outer walls, and roofs, has traditionally centered on visual inspections by human labor. In addition, because building inspection work is carried out over a wide area, it takes an enormous amount of time to acquire data. There have been variations in the assessment of the degree of degradation depending on human knowledge and experience, and there were also challenges in the effort required to organize data and link systems together.
On the other hand, with the declining working-age population due to the declining birthrate and aging population leading to personnel shortages and the aging of facilities, there is an urgent need to create mechanisms that improve productivity by utilizing new technologies.
2. Overview of the Initiative
In this verification, based on the inspection items and judgment criteria for each station facility, and after organizing the shooting conditions, we implemented a series of procedures to extract degraded areas by taking pictures of targets such as roofs and walls from above the building using automatic drone navigation and performing image analysis. At the same time, we verified whether data collection by automatic navigation is effective for streamlining inspection operations and ensuring the reproducibility of photography from the same route and same angle, and furthermore, whether images that meet the quality required for image analysis were obtained.
3. Verification Results
We confirmed that the method combining automatic drone flight and image analysis is effective in streamlining the inspection work of station facilities. Even under operations based on site conditions, the continuous flow from photography to image analysis functioned without interruption.
■ Effectiveness of automatic drone navigation
We confirmed the clearance distance from overhead wires and obstacles around the building in advance, and based on the results, we set an automatic flight route that could secure an appropriate clearance. As a result, we confirmed that automatic navigation maintaining a predetermined safety distance throughout the flight is possible. Because automatic navigation allows shooting from almost the same route and position every time, it is possible to continuously acquire images under the same conditions and composition without relying on the pilot's skill or experience. This is thought to contribute to streamlining inspection operations by improving the accuracy of chronological comparisons and reducing the need for retakes.
1. Background
The maintenance and management of station facilities, such as station buildings, overpasses, platform roofs, outer walls, and roofs, has traditionally centered on visual inspections by human labor. In addition, because building inspection work is carried out over a wide area, it takes an enormous amount of time to acquire data. There have been variations in the assessment of the degree of degradation depending on human knowledge and experience, and there were also challenges in the effort required to organize data and link systems together.
On the other hand, with the declining working-age population due to the declining birthrate and aging population leading to personnel shortages and the aging of facilities, there is an urgent need to create mechanisms that improve productivity by utilizing new technologies.
2. Overview of the Initiative
In this verification, based on the inspection items and judgment criteria for each station facility, and after organizing the shooting conditions, we implemented a series of procedures to extract degraded areas by taking pictures of targets such as roofs and walls from above the building using automatic drone navigation and performing image analysis. At the same time, we verified whether data collection by automatic navigation is effective for streamlining inspection operations and ensuring the reproducibility of photography from the same route and same angle, and furthermore, whether images that meet the quality required for image analysis were obtained.
3. Verification Results
We confirmed that the method combining automatic drone flight and image analysis is effective in streamlining the inspection work of station facilities. Even under operations based on site conditions, the continuous flow from photography to image analysis functioned without interruption.
■ Effectiveness of automatic drone navigation
We confirmed the clearance distance from overhead wires and obstacles around the building in advance, and based on the results, we set an automatic flight route that could secure an appropriate clearance. As a result, we confirmed that automatic navigation maintaining a predetermined safety distance throughout the flight is possible. Because automatic navigation allows shooting from almost the same route and position every time, it is possible to continuously acquire images under the same conditions and composition without relying on the pilot's skill or experience. This is thought to contribute to streamlining inspection operations by improving the accuracy of chronological comparisons and reducing the need for retakes.