Menu
The solution is based on analytics of sound intensity of the wind turbine components. All signals are digitally recorded with directional indication for better sound identification inside the nacelle. In case of deviations detected in the intensity of the identified acoustic pattern, the MLOps process triggers steps leading to model re-training and better Predictive Maintenance decisions.
Every year, thousands of massive wind turbines, generating 1,400,000 GWh of clean energy worldwide, are exposed to strong winds and harsh environmental conditions. And every day, accidents happen or are narrowly avoided. In 2023 alone, more than 3,800 recorded safety events, including total nacelle damage accidents and near-damage events occurred. Total net worth of wind parks worldwide is estimated to be 1,000 billion USD.
The maintenance of wind turbines is safe and has economical sense when it is scheduled for windless days. When the turbine is not working it’s impossible to detect audible signals that could be indicating upcoming hardware malfunctions. Acoustic Probe Technologies provides “ears” for wind turbines allowing for real time sound monitoring and sound analytics ML pipeline with data inference on premises. The sound intensity probe is embedded into compact size industrial AI Edge High Performance Computer.
