Quantum Machine Learning for Defect Detection

In collaboration with Fraunhofer ITWM and a major aircraft manufacturing organization, this project explored the application of quantum machine learning techniques for automated anomaly detection in aerospace manufacturing processes. Ensuring structural integrity in aircraft components requires highly reliable inspection systems capable of identifying subtle defects in materials and assembled structures.

The WISER team developed novel Quantum Neural Network (QNN) models for predictive maintenance and quality control. By leveraging quantum-enhanced learning frameworks, the approach aims to improve the detection of manufacturing defects that may be difficult to identify using conventional machine learning techniques.

Quantum models were developed to analyze inspection and sensor data across both synthetic and real-world datasets generated during aircraft construction, enabling the identification of anomalous patterns associated with potential structural flaws. These models were evaluated for their ability to enhance detection sensitivity while maintaining computational efficiency. The work also included studying the effect of encoding strategies and architectural configurations like the shape, area and sequential or parallel ansatz in quantum algorithms. Our experiments show that while broader frequency spectra can theoretically enhance expressivity, practical trainability is strongly influenced by architectural factors such as qubit count and circuit depth. Notably, we find that QNNs perform best when the frequency spectrum is tailored to the target function’s complexity but remains as compact as possible.

The work contributes to the development of advanced predictive maintenance systems for aerospace manufacturing, demonstrating how emerging methods from Quantum Machine Learning can support more reliable defect detection and improved quality assurance in complex industrial production environments. It also provides guidelines for selecting QNN ansatze and offer new insights into the interplay between expressivity and trainability in quantum machine learning.

WISER Research Fellows: Hein Zay Yar Oo, Martyna Czuba