In a major leap for the aviation industry, a research team at the Polytechnic University of Turin has developed an innovative automated inspection system for airplane engine components, specifically honeycomb structures. This development is set to reshape the way manufacturers approach quality control in line with Industry 4.0 standards, significantly improving efficiency and reducing human error.
The inspection of small and irregularly shaped engine parts, such as honeycomb components critical for engine compression, has traditionally relied on time-consuming manual processes. However, the complex nature of these inspections, combined with the limitations of human analysis, often leads to missed defects and inconsistent quality assurance. Addressing these challenges, the Polytechnic University of Turin, as part of the EIT Manufacturing AVISPA-2 project, has engineered a two-phase non-destructive testing (NDT) system that automates the process, ensuring more accurate and reliable inspections.
Automated Solution for Enhanced Precision
At the heart of this new system is the SVS-Vistek hr455CXGE 61-megapixel 10GigE Vision camera, which works in tandem with a Kuka KR16-2 robot and AI algorithms to perform thorough inspections of honeycomb engine parts. The advanced technology enables the system to detect both external and internal defects with high precision, something that traditional methods have struggled to achieve.
In the first phase, the system captures high-resolution images of the honeycomb structure using a 3D lens system, generating a "kaleidoscopic" multi-view of the part. The AI then analyzes these images, detecting external defects and determining whether the part should proceed to the next phase. If no external defects are found, the second phase involves using an optical fiber sensor to inspect the interior of the honeycomb cells, where AI segmentation techniques identify any faulty regions.
Eliminating Human Error, Enhancing Efficiency
The traditional manual inspection process is not only labor-intensive but also prone to human error due to the subjective nature of visual assessments. Variability in working conditions and the experience level of operators can further impact the consistency of inspections. By automating the process, the new system addresses these shortcomings, delivering more reliable results and reducing the need for human intervention in dangerous or complex environments.
"Our system automates the inspection of honeycomb engine parts, making it more objective, faster, and less prone to human error," said a lead researcher from the Polytechnic University of Turin. "By leveraging AI and advanced camera technology, we can significantly improve the quality control process for aviation manufacturers."
Real-World Validation
The system was tested in an aviation plant, where it demonstrated exceptional accuracy in identifying defects that could lead to part rejection. The automated process reduced inspection times to between 40-60 seconds per part, with future improvements aimed at halving this duration. Moreover, the automated approach offers full traceability and documentation at every stage, something manual inspections struggle to achieve.
As the aviation industry continues to adopt Industry 4.0 technologies, this breakthrough in automated inspections is set to become a critical component of next-generation manufacturing practices. With further optimization, the Polytechnic University of Turin's system holds the potential to significantly reduce costs and inspection times, while ensuring the highest safety and quality standards are met.
Reference: https://www.azom.com/news.aspx?newsID=63744
