Predictive maintenance model

The brief

A machine shop replaces tools on a fixed schedule and still has surprise breakdowns. Predict failures from sensor readings, handle how rare they are, and explain which conditions lead to each failure type, so maintenance can be planned. Good looks like: metrics for rare failures, a threshold based on cost, and explanations an engineer agrees with.

What you will build

1. Explore the sensor data and failure types
2. Engineer features such as power and temperature difference
3. Train gradient boosting with imbalance handling
4. Evaluate recall and false alarms per failure type
5. Explain predictions with SHAP

Free, open data to build it with

• AI4I 2020 predictive maintenance ↗ UCI Machine Learning Repository

Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.