predictive-maintenance-best-model

Predictive Maintenance Model

Project Overview

This model predicts whether an engine requires maintenance using historical engine sensor data.

Business Context

Unexpected engine failures can result in costly repairs, vehicle downtime, reduced fleet availability, and safety risks. This predictive maintenance model helps identify failure risk early so maintenance can be planned proactively.

Dataset Features

The model uses the following sensor parameters:

• Engine rpm
• Lub oil pressure
• Fuel pressure
• Coolant pressure
• lub oil temp
• Coolant temp

Target variable:

• Engine Condition

Best Model

Best selected model: Random Forest

Performance

Accuracy: 0.6596 Precision: 0.6728 Recall: 0.8957 F1 Score: 0.7684 ROC AUC: 0.6992

Best Hyperparameters

{ "n_estimators": 100, "min_samples_split": 2, "min_samples_leaf": 1, "max_depth": 5 }

Example Usage

import joblib import pandas as pd

model = joblib.load("best_predictive_maintenance_model.pkl")

sample = pd.DataFrame( [[700, 3.5, 5.0, 2.5, 80, 85]], columns=[ "Engine rpm", "Lub oil pressure", "Fuel pressure", "Coolant pressure", "lub oil temp", "Coolant temp" ] )

prediction = model.predict(sample) print(prediction)

Business Value

• Reduces unexpected breakdowns
• Improves maintenance planning
• Helps optimize service schedules
• Supports data-driven fleet and engine reliability decisions