Developing a Software & Hardware Package with Unmanned Aerial Vehicles for Technical Diagnostics of Industrial Buildings and Facilities

The relevance of developing a technology for diagnosing the technical condition of buildings and facilities using a software & hardware package based on unmanned aerial vehicles stems from the need to enhance the safety, performance, and accuracy of inspecting industrial objects, particularly tall, hard-to-reach ones such as chimneys and ventilation stacks. Traditional methods, including industrial climbing, scaffolding, and aerial lifts, are labor-intensive, time-consuming, and pose greater risks to personnel. The goal of the study is to develop, test, and verify the new technology designed to overcome these restrictions. During the study, modular design methods, software solution optimization, pilot tests, and a set of remote-control methods, such as high-resolution visual inspection, thermographic imaging, and laser scanning using unmanned aerial vehicles, were used. The technology was tested on a real-world structure: a chimney over 150 m high. As a result, the technology that helped perform a comprehensive inspection of the object, build its detailed 3D model using photogrammetry, and identify and localize defects such as traces of corrosion and compromised masonry at the joint filler was developed and successfully tested. Thermographic and lidar inspections confirmed its relevance when diagnosing industrial facilities, including those in operation. The comparative analysis of efficiency has shown that the use of the developed technology significantly reduces the total time required for work, decreases the number of experts involved, prevents decommissioning, and minimizes professional risks. The practical significance of the study is the development of a tool for prompt, safe, and highly accurate remote monitoring that provides a comprehensive understanding of a facility's technical condition and can be integrated into an enterprise's planned diagnostics system. The results of the study provide a basis for transitioning to systematic monitoring and for further development of algorithms for automated defect recognition and classification.