Machine learning algorithms have become prevalent in the data processing industry over the last decades and are now widely employed in research on fault diagnosis for wind turbines. These methods often achieve high classification accuracy when trained and evaluated on labeled datasets. However, their deployment in real-world applications remains challenging due to several factors, including the scarcity of appropriate public datasets for model training, the difficulty in generalizing models across different turbines and operating conditions, and the limited interpretability of many machine learning methods. Addressing these challenges is essential to improve the reliability and trust of such models, which will be key to their successful adoption in industrial environments.