Quiz 13 - Predictive Maintenance

Which of the following best describes predictive maintenance?

• Performing maintenance only after equipment breaks down

• Performing maintenance at regular, fixed intervals

• Using data and analytics to predict and prevent failures

• Ignoring maintenance until a safety inspection occurs

Which of the following models is most suitable for time-series forecasting of equipment failure?

• Random Forest

• Logistic Regression

• LSTM

• K-Means

In the GE Aviation case study, which of the following was a key part of their PdM system?

• Manual Excel-based tracking

• Visual inspections only

• LSTM models and digital twins

• Monthly scheduled maintenance

Which of the following is a challenge specific to implementing PdM in wind turbines?

• Too much labelled failure data

• High-speed internet access everywhere

• Harsh environmental conditions and connectivity limitations

• Lack of vibration sensors

Which of these is an unsupervised learning model commonly used for anomaly detection?

• XGBoost

• Isolation Forest

• Logistic Regression

• Random Forest

Predictive maintenance is typically more expensive in the long run than reactive maintenance.

• True

• False

Autoencoders can be used to detect unusual machine behaviour by learning from normal operational data.

• True

• False

Labelling historical failure data is usually easy in industrial environments.

• True

• False