HF-Sum/predictive-maintenance-engine-data

Predictive Maintenance Engine Dataset Dataset Overview

This dataset contains engine sensor readings collected for predictive maintenance analysis. The objective is to develop machine learning models capable of identifying whether an engine is operating normally or requires maintenance intervention based on operational and thermal sensor measurements.

The dataset is designed to support predictive maintenance applications for various engine-driven systems including automobiles, portable generators, lawnmowers, and compact industrial machinery.

The sensor values represent realistic operational behavior across both small and large engine environments.

Business Problem

Unexpected engine failures can lead to:

Expensive repairs Operational downtime Reduced equipment lifespan Safety risks Productivity losses for fleet operators and manufacturers

Traditional maintenance strategies are often reactive or based on fixed schedules, which may either:

miss early warning signs of failure, or lead to unnecessary servicing costs.

This dataset enables the development of machine learning solutions that support proactive maintenance scheduling using engine telemetry and sensor analytics.

Objective

The primary objective is to predict engine condition using sensor data and classify whether an engine:

is operating normally, or requires maintenance attention.

The dataset supports supervised machine learning classification tasks for predictive maintenance systems.

Dataset Features Feature Name Description Engine rpm Engine rotational speed measured in revolutions per minute (RPM) Lub oil pressure Lubricating oil pressure responsible for reducing engine friction Fuel pressure Fuel delivery pressure influencing combustion efficiency Coolant pressure Cooling system pressure used for thermal regulation lub oil temp Lubricating oil temperature affecting lubrication quality Coolant temp Engine coolant temperature used to monitor overheating conditions Engine Condition Target variable representing engine health condition (0 = Normal, 1 = Maintenance Required/Faulty) Target Variable

The target variable is:

Engine Condition

Classification labels:

0 → Normal engine operation 1 → Engine requires maintenance / faulty condition Intended Use

This dataset is intended for:

Predictive maintenance modeling Binary classification tasks Sensor analytics research Machine learning experimentation Automotive maintenance optimization Fleet management analytics

Possible algorithms include:

Decision Trees Random Forest Gradient Boosting AdaBoost XGBoost Deep Learning models Machine Learning Applications

Potential applications include:

Real-time engine monitoring systems Maintenance scheduling optimization Fleet reliability analysis Failure risk prediction Intelligent maintenance alerts Data Characteristics Structured tabular dataset Numerical sensor readings Binary classification target Suitable for supervised learning Appropriate for ensemble learning techniques Expected Insights

The dataset can help identify:

overheating patterns, lubrication-related anomalies, pressure instability, abnormal operational conditions, and sensor relationships associated with engine degradation. Limitations The dataset does not include timestamp-based sequential behavior. External environmental variables are not included. Failure severity levels beyond binary classification are not provided. Ethical Considerations

This dataset does not contain:

personally identifiable information, user-sensitive data, or confidential operational records.

The dataset is intended purely for educational, research, and predictive maintenance modeling purposes.

Citation

If using this dataset for academic or educational purposes, please cite the corresponding project repository or submission.

Maintainer

Dataset maintained as part of a Predictive Maintenance Machine Learning Capstone Project.