Industrial Electric Motor Thermography Dataset
This data product consists of a structured dataset of real thermographic inspections of electric motors operating in an industrial plant. It is designed to support condition monitoring, fault detection, and predictive maintenance applications. The dataset provides real-world industrial variability, including operational noise and naturally occurring faults, making it suitable for robust AI model development.
Dataset Metrics
Data Format
Dataset Schema (6 features)
| Name | Type | Nullable | Description |
|---|---|---|---|
| thermography_file | Image, radiometric JPEG | No | Thermal image containing pixel-wise temperature information. |
| visible_image | Image (JPEG) | No | Visible spectrum image corresponding to the thermal capture. |
| temperature_matrix | Array (float, 256x192) | No | Pixel-wise temperature values (°C) corresponding to the thermographic image. |
| condition_label | Categorical | No | Health status of the bearing. Values: healty, faulty. |
| acquisition_date | Datetime | No | Date and time of data acquisition. |
| inspection_batch | String | No | Identifier of acquisition campaign (YYY_MM) |
License & Compliance
About This Dataset
The dataset contains thermographic inspections captured under real operating conditions. Each acquisition includes multimodal information: - Radiometric thermographic images (infrared, per-pixel temperature information) - Visible spectrum RGB images captured simultaneously - Pixel-wise temperature matrices in CSV format - Condition labels (healthy / faulty) derived from acquisition structure - Acquisition timestamp and inspection batch identifier Data is organized chronologically and by equipment condition, enabling supervised learning and temporal analysis.
Data Quality
The dataset is based on real industrial measurements acquired using a Hikmicro B20 thermal camera, ensuring consistent acquisition parameters across all records. Key quality characteristics: - Real-world industrial variability (load changes, ambient conditions, installation differences) - Synchronized multimodal acquisition (thermal + visible + temperature matrix) - Relevant environmental information included in temperature matrices - One-to-one alignment between all modalities per inspection - Labeling derived at acquisition time (folder-based, reducing annotation noise) - Chronological organization by acquisition campaigns (YYYY_MM structure)
Use Cases
- Fault detection in electric motors using thermographic analysis - Predictive maintenance systems based on thermal pattern recognition - Multimodal AI combining infrared and RGB data for industrial diagnostics - Benchmarking anomaly detection and classification algorithms - Research in condition monitoring and industrial AI systems
Limitations
- The dataset is not a continuous time series (data is collected in discrete campaigns) - Fault types are not explicitly categorized beyond “faulty” label - Data is limited to specific industrial environments and equipment setups - Load variations are not explicitly annotated - Scalability depends on number of campaigns included (currently ~990 MB total)
Methodology
Data was collected in industrial manufacturing environments from bearings operating under normal working conditions. Thermographic inspections were performed in situ using a Hikmicro B20 thermal camera during routine inspection activities. Each acquisition includes: - A radiometric thermal image - A corresponding visible image - A temperature matrix exported as CSV Data is organized by acquisition campaigns (YYYY_MM), with manual or operational labeling of equipment condition (healthy or faulty) based on inspection context. No artificial manipulation or laboratory conditions were introduced, ensuring that the dataset reflects real-world industrial behavior.
Geographic Coverage
Source Information
Governance & Review
Usage & Restrictions
License: Proprietary License — This dataset is licensed for use by the purchasing entity. Refer to the specific license terms for permitted use.
Access Control: Download links are time-limited and single-use. All access is logged for audit purposes.