AI4I 2020 Predictive Maintenance Dataset

This dataset contains 10,000 records of machine operations, simulating an industrial milling machine. It includes environmental and operational variables like air temperature, process temperature, rotational speed, torque, and tool wear. The dataset is designed for classification tasks, where the target is to predict 'Machine failure' and identify the specific failure mode: Tool Wear Failure (TWF), Heat Dissipation Failure (HDF), Power Failure (PWF), Overstrain Failure (OSF), or Random Failures (RNF).

Data Quality

The dataset is of high quality with no missing values (0 nulls across all columns). It follows a consistent format and represents a balanced set of features for technical machine performance analysis.

Use Cases

Binary classification of machine failures, multi-label classification of failure modes, sensor-based anomaly detection, and optimization of maintenance schedules in Industry 4.0 contexts.

Limitations

Being a synthetic dataset, it utilizes mathematical models to simulate wear and failure rather than capturing every physical variation found in organic sensor noise from aging machinery.

Methodology

Generated using simulation models based on actual industrial machine parameters, designed to provide a realistic benchmark for predictive maintenance algorithms.