en.Wedoany.com Reported - On June 10, Russia's United Engine Corporation (UEC) deployed an automated inspection system based on artificial intelligence at its UEC-Saturn plant in Rybinsk for quality checks of compressor blades in the production of PD-8 turbofan engines. After a trial run, the system has entered industrial operation, capable of identifying surface defects on blades with a detection resolution of 40 microns.
Aircraft engine blades are typical high-precision, high-reliability components. Compressor blades endure high-speed rotation, airflow impact, vibration, and temperature changes during engine operation. If minor surface defects are not detected in time, they can affect component lifespan, assembly quality, and overall engine reliability. Traditional manual visual inspection relies on operator experience and lighting conditions, leading to variability in identifying subtle scratches, dents, edge anomalies, and surface imperfections. By integrating the AI visual inspection system into the PD-8 production line, UEC aims to detect quality deviations early in the batch manufacturing process and reduce bottlenecks associated with manual inspection.
The system includes two robotic inspection stations designed for automated checks of polished compressor blades. The equipment uses machine vision to capture surface images of parts, and a neural network identifies defect types and locations. It currently handles six types of polished compressor blades for gas turbine engines. Inspection results are recorded by the system and used for subsequent quality assessments, making the status of individual components more traceable throughout the production process.
The PD-8 is a key engine model for Russia's development of a domestic civil aviation power system, with applications targeting regional jets and other aviation platforms. Engine localization requires not only design and testing but also the establishment of stable capabilities in component processing, inspection, assembly, and quality control. Blades are among the core engine components. As production tempo increases, relying solely on manual methods makes it difficult to balance efficiency and consistency. With the AI inspection system in industrial operation, it can maintain inspection accuracy while improving throughput capacity, providing quality control support for the subsequent ramp-up of PD-8 production.
Such applications also demonstrate that aviation manufacturing is accelerating the adoption of industrial artificial intelligence. By combining machine vision, neural networks, robotic material handling, defect databases, and quality management systems, the inspection process—once dependent on operator experience—can be transformed into a more standardized data-driven workflow. For engine manufacturers, AI quality inspection not only detects defects but also helps process engineering departments analyze defect sources, optimizing polishing, machining, handling, and assembly stages. As inspection data accumulates, the model's ability to recognize different surface conditions and defect characteristics will continue to improve.
The impact on the industrial chain will focus on areas such as aircraft engine manufacturing, industrial vision, robotic inspection stations, quality traceability software, and smart factory upgrades. Subsequent milestones include the system's expansion to more blade types, the extent of inspection efficiency gains, data integration with factory quality management systems, and whether it can further cover other critical PD-8 components. If operation proves stable, the UEC-Saturn Rybinsk plant will leverage AI inspection capabilities to improve production consistency of aircraft engine components, providing a new benchmark for expanding intelligent applications in key processes within Russia's aviation manufacturing industry.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
