Powerful AI augments fault detection on wind turbines

An artificial intelligence (AI) engine has been developed to automatically detect rotor blade damages on wind turbines, and aims to improve the productivity and consistency of blade inspection processes by recognizing damages based on inspection image material.

The leapfrog technology is the result of a partnership between Swiss company Sulzer Schmid, specialising in UAV technology for rotor blade inspections, and NNAISENSE, an artificial intelligence specialist.

The autonomously flying drones of the 3DX Inspection Platform of Sulzer Schmid assure high-definition quality and consistent image acquisition time, as well as 100 per cent blade coverage while minimizing human errors and operational risks.

The cutting-edge image assessment tools of the platform ensure detailed and efficient damage assessment. With the support of an AI-enabled inspection software, the review work of blade experts will be greatly simplified. Instead of having to review the entire surface of the blades, they will simply need to focus on the pre-selected areas of concern. This technology progress will not only boost the productivity of the reviewing teams but will also improve the quality of damage annotation processes.

“Maintaining the structural integrity of rotor blades is critical to maximizing energy output and ensuring the safe operation of wind turbines. We are convinced that we will be able to transfer our extensive expertise in surface defect recognition from other industries to the wind industry and are looking forward to our cooperation with Sulzer Schmid, an innovator in its own space”, commented Faustino Gomez, CEO of NNAISENSE.

Christof Schmid, chief operating officer and co-founder of Sulzer Schmid said: “Maximising end-to-end productivity is a key success factor in the highly competitive market of wind turbine inspection solutions. Thanks to our collaboration with NNAISENSE, we will be able to push the envelope in this area and significantly advance the automation capabilities of our inspection platform”.

The initial version will be able to flag all areas of concern on any given damaged blade. Future upgrades will add other capabilities such as the ability to establish damage categories and severity levels.

No posts to display