ABB IRB 1100 Robot Synthetic Predictive Maintenance Data with 2,000 Instances

A synthetic dataset of 2,000 instances and 215 features simulating sensor data from an ABB IRB 1100 industrial robot. It was created by Mohamed Lamine Soudani via a simulation model incorporating six robot axes and sensors for electrical, mechanical, thermal, and vibration parameters. The dataset, last updated in May 2026, is designed for fault detection and severity-level validation.

Use Cases

• Training fault classification models based on simulated electrical, mechanical, thermal, and vibration sensor signals.
• Benchmarking feature selection methods using the 215 explanatory characteristics to identify a reduced feature subset.
• Developing severity-level prediction models for maintenance prioritization based on three severity levels per defect type.

Strengths

• Contains 2,000 simulated instances with 215 explanatory features, providing a structured testbed for model development.
• Simulation model incorporates six robot axes and multiple sensor types (electrical, mechanical, thermal, vibration), suggesting detailed parameter coverage.
• Dataset is associated with published research results, including specific model performance metrics (e.g., LightGBM F1-macro = 0.7625).

Limitations

• Data is entirely synthetic, generated via simulation, and may not fully capture real-world noise and failure modes.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is known (2,000), but the dataset size is very small (1010.0 B), indicating limited scope.

Provenance

Source

figshare, authored by Mohamed Lamine Soudani.

Collection Method

Generated via a sophisticated simulation model of an ABB IRB 1100 industrial robot.

Freshness

Last updated 2026-05-22 12:48:59; freshness should be verified.