An unprecedented collaboration between EPFL and the University of Glasgow has yielded a cutting-edge machine-learning algorithm designed to identify concealed manufacturing defects within wind turbine composite blades—an innovation poised to revolutionize the wind energy industry.
The detrimental impact of faulty wind turbine blades on operational efficiency and financial costs is well-documented, underscoring the critical importance of stringent quality assurance measures for wind turbine manufacturers worldwide. Traditional quality inspections, limited to surface scrutiny, often fail to detect hidden defects lurking beneath the composite surface during the production phase.
Enter a groundbreaking approach, jointly developed by EPFL and University of Glasgow researchers, which integrates a patented radar technology with an AI-driven assistant to unveil potential anomalies concealed within the blade's structure. This pioneering method offers a plethora of advantages: non-destructive and non-contact in nature, it facilitates agile and swift data acquisition and analysis while consuming minimal power. The research breakthrough has recently been disseminated in Mechanical Systems and Signal Processing, marking a significant milestone in the quest for enhanced quality control measures.
At the helm of this interdisciplinary endeavor is Olga Fink, a tenure-track assistant professor of civil engineering at EPFL and the visionary leader of the Intelligent Maintenance and Operations Systems Laboratory (IMOS). Leveraging her expertise in signal processing, Fink's previous research endeavors have encompassed anomaly detection in machinery, background noise suppression in audio recordings, and classification of bird songs using innovative AI-driven methodologies.
Reflecting on the burgeoning challenges posed by the evolving landscape of wind turbine manufacturing, Fink emphasizes the imperative of adapting inspection methodologies to accommodate larger and more intricate turbine designs. As wind turbines evolve with augmented scale and complexity, the probability of manufacturing defects escalates, necessitating proactive measures to preemptively identify and rectify potential anomalies.
Complementing EPFL's contributions, the University of Glasgow team, spearheaded by Prof. David Flynn from the James Watt School of Engineering, has spearheaded advancements in prognostics and health management within the realm of autonomous systems and connectivity. Pioneering the fusion of robotics and artificial intelligence, the Glasgow researchers have harnessed a patented Frequency Modulated Continuous Wave radar, augmented by a robotic arm, to scrutinize wind turbine blade samples at varying distances. Through sophisticated signal processing techniques, they have successfully isolated features indicative of impending failures within these intricate composite structures.
As the wind energy sector embarks on a trajectory towards sustainability and efficiency, the integration of state-of-the-art measurement technologies and AI-driven analytics promises to redefine quality assurance standards. The collaborative efforts of EPFL and the University of Glasgow underscore a paradigm shift towards proactive defect detection and mitigation strategies, heralding a new era of reliability and performance optimization within the renewable energy landscape.
Source: Ecole Polytechnique Federale de Lausanne