Tiny On-Device Structural Health Monitoring for Wind Turbines Using MEMS Pressure Sensors

Wind turbines are extensively used as a renewable source of energy. These turbines are exposed to turbulence and highly variable weather conditions such as wind, rain, and ice, which can cause significant damage and degraded power generation performance if not detected in time. The emergence of the Industry 4.0 paradigm enables a reduction in maintenance costs and increased efficiency. Recent research has demonstrated the possibility of monitoring the aerodynamic performance of the wind turbine directly from the blades with today's state-of-the-art wireless monitoring systems. However, these need external support to process and analyze the collected data. This paper proposes an effective tiny machine learning approach to monitor the blade structural integrity directly onboard the sensor node, requiring only 3 kB of RAM and 35 kB of Flash running on a low-power ARM Cortex-M4F at 48MHz. We experimentally evaluate the proposed model with a dataset acquired in a wind tunnel with 40 MEMS barometers placed around the airfoil in 6 different damage conditions. The X G Boost classification model is purely based and trained on the blade aerodynamics, featuring a classification accuracy above 83 % over 6 labeled classes of increasing damage. With a total execution time of 579 ms while it only consumes 9 mJ, it enables real-time analysis for a sustainable battery - based monitoring system.