A collaboration between Sensing360 and NaN Tech Solutions
The energy transition requires wind turbines that not only run more, but also continue to function reliably and efficiently. Maintenance is a crucial factor in this. In practice, maintenance is often still reactive: malfunctions are only addressed after they occur. This leads to unplanned downtime, higher costs, and loss of performance.
Predictive maintenance with AI
As part of the MIT AI project AIRFLOW, which was awarded funding in 2025, Sensing360 and NaN Tech Solutions are developing an AI solution for predictive maintenance of wind turbines. By continuously collecting and analyzing data on the load and condition of turbine components, deviations and wear can be identified at an earlier stage. This makes maintenance easier to plan and allows turbines to continue to perform optimally for longer.
Edge cloud architecture
AIRFLOW uses an edge-cloud approach. Some of the analyses take place directly on or near the turbine, while the cloud is used for further processing and optimization. This architecture reduces the amount of data that needs to be sent continuously and enables faster analyses, even in locations with limited connectivity.
Distribution of expertise
Within the project, Sensing360 focuses on sensor technology and measuring and interpreting turbine load and condition under operational conditions. NaN Tech Solutions develops the embedded systems and AI implementations that enable efficient analysis on hardware. This combination makes it possible to achieve intelligent decision-making close to the source.
Effect on performance and sustainability
AIRFLOW contributes to higher availability of wind turbines, lower maintenance costs, and extended service life of installations. By better aligning maintenance with the actual condition of turbines, unnecessary downtime is prevented and CO₂ emissions over the life cycle of wind farms are reduced.
