Turbit AI Can Detect Rotor Bearing Damage Months in Advance

Detecting rotor bearing damage in wind turbines is crucial for their efficient and safe operation. Traditional methods of monitoring main bearing temperature can be time-consuming, expensive, and inefficient. However, with advances in artificial intelligence (AI), it is now possible to detect potential failures with precision and efficiency.

This article introduces the AI monitoring system developed by Turbit, a company specializing in AI-based monitoring systems for wind turbines. Michael, the CEO and founder of Turbit, explains how their AI monitoring system works and how it can detect main bearing failures with an accuracy of up to 0.5 degrees temperature difference and months in advance.

https://video.wixstatic.com/video/4ab135_bf79081a879641b18777d496beb44f32/720p/mp4/file.mp4

The AI monitoring system developed by Turbit uses SCADA data from wind turbines, which includes outdoor temperature, wind speed, rotor speed, and other external factors that influence the temperature of the main bearing. This data is used to train and learn the normal behavior of the main bearing temperature. The system can simulate how the bearing temperature should behave over time, even in extreme weather conditions.

As shown in the example, the AI monitoring system can detect abnormal behavior of the main bearing temperature. The system compares the actual main bearing temperature data with the simulated data and sends an alarm to the operator when there is a significant deviation. The alarm is escalated if the problem persists and, in some cases, triggers a service event.

In the example mentioned by Michael, the AI monitoring system detected a problem with the main bearing in January. The temperature of the main bearing had increased over time, indicating a possible problem. The system sent an alarm, which was escalated when the problem continued. A service event was triggered, and the problem was eventually solved by repairing the lubrication mechanism.

The AI monitoring system developed by Turbit can detect potential failures with an accuracy of up to 0.5 degrees temperature difference and months in advance. This gives operators sufficient time to plan and carry out maintenance work, minimize downtime, and prevent worst-case scenarios.

In summary, AI-based monitoring systems like the one developed by Turbit can provide a more efficient and cost-effective solution for detecting potential failures in wind turbines. With the ability to detect potential problems with a high degree of accuracy and well in advance, operators can minimize downtime and prevent costly repairs.