Triaxial Bearing Vibration Dataset of Induction Motor under Varying Load Conditions

Overview

The Triaxial Bearing Vibration Dataset of Induction Motor under Varying Load Conditions provides high‑resolution vibration data from a three‑phase induction motor used to study bearing fault diagnosis in industrial IoT settings.[page:2][web:165]

Data Collection

• Vibration signals were recorded using a MEMS‑based triaxial accelerometer attached to the motor housing near the drive‑end bearing, connected to an NI myRIO data acquisition system.[page:2]
• The motor was operated with a healthy bearing and with bearings having inner‑race and outer‑race faults of different severities (e.g., 0.7 mm, 0.9 mm, 1.1 mm, 1.3 mm, 1.5 mm, 1.7 mm).[page:2]
• Each bearing condition was tested at three load levels (100 W, 200 W, 300 W), resulting in 38 CSV files covering healthy and faulty states under varying loads.[page:2][web:165]

Signals and Format

• Data were sampled at 10 kHz with 1000 samples per channel acquisition blocks.[page:2]
• Each CSV file contains four columns: Time Stamp (seconds), X‑axis, Y‑axis, and Z‑axis vibration in units of g (1 g = 9.80665 m/s²).[page:2]
• File names encode the bearing condition and load (e.g., Healthy-with-pulley.csv, 0.7inner-100 watt.csv, 1.7outer-300 watt.csv), enabling straightforward filtering by fault and load.[page:2]

Use Cases

• Developing and benchmarking fault diagnosis algorithms for rotating machinery using supervised or unsupervised learning.[page:2][web:165]
• Comparing time‑domain, frequency‑domain, and time–frequency (e.g., wavelet) features for bearing condition monitoring.[page:2]
• Testing deep learning and domain adaptation approaches for industrial IoT condition monitoring under varying loads and fault severities.[page:2][web:165]

License

The article states that the datasets are stored on the Mendeley Data platform under DOI 10.17632/fm6xzxnf36.2; users should follow the reuse conditions and any license information given on the Mendeley dataset page and cite both the Data in Brief article and dataset DOI.[page:2][web:165]