Quality Analytics

Exercise:
Using the example of wind turbines, we took on the challenge of estimating the lifetime of installed components and thus the time till expected failure. This methodology is applicable to all types of machines and serves to avoid longer downtimes. Even if businesses comply with maintenance guidelines, the failure of individual components is inevitable.

Procedure:
The algorithms were trained with wind turbine data of wind power stations. Based on an algorithm for anomaly detection, statistical models were created to predict the remaining useful lifetime.

Goal:
Estimation of a Weibull distribution via a neural network with the goal of predicting the lifetime of the components at any given point in time.

Conclusion:
Machine learning helps avoiding shutdowns and reducing maintenance costs: The replacement of marginal components is more cost-effective than the renewal of entire component segments.

This sounds useful? Contact us anytime.

Applied Frameworks