Python notebook created on October 2018.

• Edited by Karim Choukri. ([email protected])
• Use case data extracted from Mary Wahl (https://gallery.azure.ai/Notebook/Predictive-Maintenance-Modelling-Guide-Python-Notebook-1)

The business problem for this example is about predicting problems caused by component failures such that the question "What is the probability that a machine will fail in the near future due to a failure of a certain component?" can be answered. The problem is formatted as a multi-class classification problem and a machine learning algorithm is used to create the predictive model that learns from historical data collected from machines. In the following sections, we go through the steps of implementing such a model which are feature engineering, label construction, training and evaluation.

Common data sources for predictive maintenance problems are :

• Failure history: The failure history of a machine or component within the machine.
• Maintenance history: The repair history of a machine, e.g. error codes, previous maintenance activities or component replacements.
• Machine conditions and usage: The operating conditions of a machine e.g. data collected from sensors.
• Machine features: The features of a machine, e.g. engine size, make and model, location.
• Operator features: The features of the operator, e.g. gender, past experience

The data for this example comes from 4 different sources which are real-time telemetry data collected from machines, error messages, historical maintenance records that include failures and machine information such as type and age. Datasets are grouped as following:

• Fault history: The fault history of a machine or a component inside the machine.
• Maintenance history: The repair history of a machine, for example, error codes, previous maintenance activities or replacement of components.
• Conditions and use of the machines: The operating conditions of a machine, for example, data collected from the sensors.
• Characteristics of the machines: The characteristics of a machine, for example, engine size, make and model, age.

• pandas
• numpy
• statsmodels
• scikit-learn
• scikitplot
• matplotlib
• seaborn
• xgboost
• itertools