datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
nln-emp-compressor-bearing
NLN-EMP Compressor-Analogue Bearing Benchmark
An evaluation benchmark for compressor bearing-fault detection, derived from the NLN-EMP motor-current and vibration dataset (Royal Netherlands Navy / University of Twente, CC0). Vibration is converted to the quantity a predictive-maintenance product actually reads — ISO 10816-3 band-limited velocity RMS in mm/s — and motor phase-current waveforms are kept alongside for motor current signature analysis (MCSA).
The headline result of… See the full description on the dataset page: https://huggingface.co/datasets/EitaNis/nln-emp-compressor-bearing.pdm-synth-compressor-bearing
PdM Synthetic Compressor Bearing-Fault Dataset
Physics-composed synthetic data for compressor bearing-fault detection, generated by the synth package of the PdM predictive-maintenance system. 2000 randomised compressor instances are each simulated over a degradation trajectory, producing both an ISO 10816-3 vibration-velocity RMS feature with fully consistent derived labels, and panel-level aggregate current waveforms composed by summing the compressor's own motor current with a… See the full description on the dataset page: https://huggingface.co/datasets/EitaNis/pdm-synth-compressor-bearing.iitk-compressor-bearing
IITK Reciprocating Air Compressor — Bearing Benchmark
The equipment-axis benchmark for compressor bearing-fault detection: acoustic recordings from a real single-stage reciprocating air compressor, belt-driven by a 5 HP / 415 V / 50 Hz / 1440 rpm direct-on-line induction motor under pressure-switch duty cycling. Eight labelled states — healthy, a bearing fault, and six compressor-native confounders (flywheel, piston ring, rider belt, leaking inlet valve, leaking outlet valve… See the full description on the dataset page: https://huggingface.co/datasets/EitaNis/iitk-compressor-bearing.Air_Compressor_Mannual_QNA