DRMHB - Dynamic Rotating Machinery Health Benchmark

DRMHB is a synthetic benchmark for machine-learning research in rotating machinery condition monitoring. It represents 180 calendar days for a fleet of 12 nominally equivalent variable-speed motor-pump units, combining:

• multiple healthy operating regimes;
• explicit startup, shutdown and internal regime transitions;
• machine-to-machine individuality;
• rare healthy out-of-distribution (OOD) operation;
• realistic non-fault disturbances and sensor ageing;
• progressive single and interacting mechanical faults;
• synchronized triaxial spectra, waveforms and tachometer edges;
• a causal calibration/holdout/future evaluation protocol.

All measurements are synthetic. DRMHB is not field data and is not a substitute for validation on real machinery.

The current fixture was generated with VibSynth orchestration script 2.5.0, serialized as VibFrame 0.2.0. The conversion examples now use vibframe-anndata==0.3.0 with complete evaluation preservation. The raw ZIP is unchanged: this is the published 2.5.0 fixture, not a newly generated compact-controls campaign.

Start here: dataset report, conversion examples and complete ground-truth guide.

The challenge

The task is not merely to separate healthy and faulty rows. A useful model must learn fleet normality while avoiding false alarms caused by legitimate process changes. In particular:

• R2 and R4 have similar rotational speed but different hydraulic load;
• forward and reverse transitions follow different paths because state has lag and memory;
• adaptive acquisition heavily oversamples transitions relative to elapsed time;
• equivalent machines retain stable individual vibration fingerprints;
• healthy OOD conditions are statistically unusual but not damaged;
• observable vibration need not increase monotonically with latent severity;
• some machines contain two coupled degradation mechanisms.

Industrial scenario

The simulated plant contains 12 variable-frequency-drive motor-pump sets, identified as M01 through M12. Vibration is measured at the pump drive-end bearing housing in three directions:

• pump_DE_H: radial horizontal;
• pump_DE_V: radial vertical;
• pump_DE_A: axial.

Five active regimes are used:

ID	Name	Nominal meaning
R1	ECO	Reduced speed and demand
R2	NOMINAL	Standard production duty
R3	PEAK	Temporary high-demand duty
R4	THROTTLED	R2-like speed with different hydraulic load
R5	RECIRCULATION	Low-flow recirculating operation

OFF is also an explicit state. Machines operate Monday-Thursday from 08:00 to 22:30 UTC and Friday from 08:00 to 18:00 UTC; weekends remain stopped.

Time horizon, splits and adaptive acquisition

The fixture covers:

2026-01-01 00:00:00 UTC -> 2026-06-29 22:30:00 UTC

The 46,907 observations are divided as follows:

Split	Rows	Intended use
calibration	7,984	Learn healthy normality
normal_holdout	3,657	Calibrate thresholds and false-positive behaviour
future_stream	35,266	Causal evaluation with faults, transitions and OOD

Acquisition is deliberately non-uniform:

Context	Policy	Realized rows
Startup transition	Every 5 minutes	10,752
Shutdown transition	Every 5 minutes	10,752
Internal regime transition	Every 10-15 minutes	18,571
Stable running	Every 2-4 hours	3,424
OFF	Every 8 hours	3,408

There are 40,075 transition observations because changing regions are sampled densely. Observation count is therefore not proportional to elapsed time. Temporal plots and persistence metrics must use timestamps rather than row index.

Signal acquisition and storage

Every observation begins from a synchronized triaxial acceleration realization of 1.6 seconds at 51.2 kHz. Both stored spectral products are calculated from that complete source realization:

• VEL_1K: velocity spectrum from 0 to 1,000 Hz;
• ACC_10K: acceleration spectrum from 0 to 10,000 Hz.

Each spectrum has 1,601 bins. With three directions and two processing modes, the dataset contains 281,442 spectral signals.

The persisted waveform branch is anti-alias filtered and decimated to 25.6 kHz. Its length is time-based rather than revolution-normalized:

Stored window	Samples/channel	Role
0.25 s	6,400	Ordinary OFF observation
0.50 s	12,800	Ordinary active or transition observation
1.60 s	40,960	Stratified anchor or localized event

The current fixture contains 3,240 short OFF windows, 40,551 ordinary windows and 3,116 full-length windows. Each observation has H/V/A waveforms, giving 140,721 waveform signals and 2,002,260,480 persisted samples in total.

Waveforms are quantized to an int16-equivalent grid and reconstructed as physical-unit float32 values to remain compatible with VibFrame 0.2.

Legitimate novelty and faults

Healthy active observations may contain low-probability disturbances such as RPM drift, load transients, electrical interference, sensor-noise bursts and transient impacts. The current fixture contains 1,756 such rows. Three healthy machines also receive short operational-OOD windows, totaling 88 observations.

Fault distribution:

Machine	Fault 1	Fault 2
M01	Imbalance	-
M02	Healthy	-
M03	Angular misalignment	-
M04	Bearing outer-race	-
M05	Imbalance	Bearing outer-race
M06	Mechanical looseness	-
M07	Healthy	-
M08	Angular misalignment	Mechanical looseness
M09	Bearing inner-race	-
M10	Bearing inner-race	Resonance
M11	Healthy	-
M12	Resonance	-

No fault begins before day 60. Fault severity is derived from mechanism-specific physical state, and observable response remains dependent on speed, load, thermal history and structural transfer.

Ground truth and leakage prevention

The source VibFrame keeps evaluation information in separate sidecars. The 0.3.0 workflow preserves both supported observation-aligned views and the original files, without automatically exposing targets to a model:

AnnData location	Information
obsm["ground_truth"]	Snapshot truth, aligned with obs_names
obsm["waveform_ground_truth"]	Per-waveform annotations with explicit raw-channel bindings
uns["vibframe_evaluation"]	Byte-exact original annotation files and source/path/size/SHA-256 manifest
uns["vibframe_anndata"]	Conversion, raw-channel and feature provenance
obsm["raw_*_capture_t"] and *_capture_t_known	Actual capture times and explicit/fallback flags

This includes construction labels, crop offsets and durations, quantization and hashes, machine parameters, scenario configuration and DiagGT documents/tables. All regular files under evaluation/, ground-truth/ and ground_truth/ are preserved, plus root JSON/YAML and machine JSON context. Unknown annotation fields are retained; the original Parquet schema can be recovered.

Snapshot joins use source, machine and integer UTC-microsecond snap_t, not row order. Per-waveform joins also identify the raw channel. Capture t can be later than snap_t because of cropping; it is not a different snapshot.

The reference snapshot view had 42 columns. Code must inspect actual fields instead of hard-coding a fixed schema. Do not use split, fault_*, severity, physical states, ood_operational, nuisance labels or DiagGT as unsupervised model inputs. Operational-state knowledge must be an explicitly declared experimental condition. The package never adds these fields to X implicitly.

Preservation is not automatic labelling: DiagGT intervals, consolidated observations and findings retain their original semantics. No nearest-time join is invented, and absence of a diagnosis does not mean healthy. The original archive remains dataset-wide when AnnData observations are sliced.

Distribution and local AnnData workflow

The single canonical public signal artifact is data/DRMHB.vibframe.zip. H5ADs are local derivatives, not a second public source of truth.

python -m pip install -r examples/requirements.txt

python examples/01_create_featureless_anndata.py \
    data/DRMHB.vibframe.zip derived/DRMHB.v0.3.0.raw.h5ad

python examples/02_inspect_anndata.py \
    derived/DRMHB.v0.3.0.raw.h5ad --waveform-details

python examples/03_add_features.py \
    derived/DRMHB.v0.3.0.raw.h5ad derived/DRMHB.v0.3.0.features.h5ad

python examples/05_read_evaluation.py derived/DRMHB.v0.3.0.features.h5ad

Example 01 explicitly enables ground_truth.scope="all", stores numerical values as float32, and ingests in raw-data blocks of 8 MiB. The featureless base has X.shape == (46907, 0) for the published fixture. The unchanged reference feature requests materialized 468 machine/source-scoped columns in the reference EDA, with structural missing values for other machines. They are not 468 dense fleet-wide descriptors; the notebook explains the logical-column alignment.

Example 03 calculates features only from the raw H5AD and checks that original annotations and raw coverage are preserved. Example 02 reads HDF5 metadata and public annotation APIs selectively; it does not assume AnnData backed mode keeps obsm or the evaluation archive out of RAM.

For an existing 0.2.x H5AD, preserve the original source identities and run:

python examples/04_add_complete_ground_truth.py \
    derived/DRMHB.float32.eda-features.h5ad data/DRMHB.vibframe.zip \
    derived/DRMHB.v0.3.0.features.h5ad

This adds annotations without recalculating X, var or raw samples. A transactional temporary copy needs roughly one additional H5AD of free disk. Legacy capture-time companion arrays are not retroactively recovered by this operation; new raw imports populate them. Crop information is still retained in waveform truth.

Storage and memory

The canonical ZIP is 8,972,153,939 bytes (8.356 GiB). The reference 0.2.2 run measured about 9.35 GiB for its featureless H5AD and 9.44 GiB after reference features. Those are historical snapshot-only sizes, not measured 0.3.0 sizes. Full annotations add original sidecar bytes, aligned views and capture-time matrices. Examples report their actual output size rather than extrapolating.

The default 512 MiB sidecar budget covers original encoded annotation files, not decoded tables or total process RAM. Inspect exact archived sizes with list_evaluation_files(). Raw-array compression remains none, as in the existing examples; no compression benchmark is rerun. Ground-truth conversion does not modify the VibFrame ZIP.

AnnData layout

obs   : snapshot identity, snap_t, machine, source, speed and availability
X     : n_snapshots × n_materialized_features (initially zero columns)
var   : feature descriptors, units, channel identity and provenance
obsm  : ragged raw spectra/waveforms + lengths/positions/speed
        tachometer edges + capture_t/known companions
        ground_truth + waveform_ground_truth (evaluation only)
uns   : vibframe_anndata (package/raw/feature provenance)
        vibframe_evaluation (original sidecar bytes and manifest)

See examples/README.md for precise commands and docs/GROUND_TRUTH.md for target selection, source identity, original DiagGT access and missing-data semantics.

Recommended evaluation protocol

1. Fit preprocessing and models only on calibration.
2. Use normal_holdout to select thresholds.
3. Evaluate once on future_stream.
4. Treat healthy transitions and healthy OOD as negative examples unless the research question explicitly targets novelty.
5. Use timestamps for detection persistence and delay.
6. Report per-machine and per-fault results alongside aggregate metrics.

Recommended metrics include AUROC, AUPRC, F1, ordinary healthy FPR, transition FPR, OOD FPR, wall-clock detection delay and the fraction of faulty machines with a confirmed detection.

The reference EDA confirms an alarm after six continuously positive wall-clock hours; an observation gap longer than 4.5 hours breaks the run.

Compact reference baseline

The values below come from the immutable published EDA run for the 2.5.0 fixture with converter 0.2.2. Updating the annotation API does not constitute rerunning those experiments.

The reference notebook fits four deliberately small unsupervised baselines. Thresholds are the 99th percentile of normal-holdout scores.

Model	AUROC	AUPRC	F1	Healthy FPR	Transition FPR	OOD FPR	Median delay
Robust distance	0.739	0.758	0.245	1.08%	1.16%	4.55%	35.78 d
PCA reconstruction	0.693	0.778	0.531	1.07%	1.11%	4.55%	1.75 d
Isolation Forest	0.814	0.843	0.485	1.23%	1.40%	3.41%	7.77 d
One-Class SVM	0.665	0.765	0.544	1.15%	1.29%	1.14%	0.88 d

These values are sanity checks, not leaderboard targets. Several machines are not detected by any compact baseline, and OOD false positives remain materially higher than ordinary healthy FPR for three models.

Healthy calibration also remains non-trivial in the reference features: PCA requires 7, 10 and 14 components for 80%, 90% and 95% cumulative variance, respectively, and the TwoNN intrinsic-dimension estimate is approximately 8.80.

Repository contents

DRMHB/
|- data/DRMHB.vibframe.zip         canonical published 2.5.0 signal dataset
|- derived/                      local H5ADs (not versioned)
|- examples/                     import, inspect, features, retrofit, evaluation readers
|- notebooks/DRMHB_EDA.ipynb       editable 0.3.0 EDA workflow, outputs cleared
|- notebooks/reference/          immutable executed source of published EDA results
|- docs/DRMHB.tex                 editable report source
|- docs/DRMHB.pdf                 compiled dataset report
|- docs/figures/                  EDA images and editable vector diagrams
|- docs/provenance/               figure/table and source identities
|- tools/                        report build and reference-asset recovery
`- tests/                        small, synthetic example-integration checks

The report explains the installation, daily/weekly state schedule, acquisitions, degradation, evaluation and complete AnnData workflow. Build it with python tools/build_report.py once a LaTeX distribution with latexmk and TikZ is installed. The report's 37 EDA figures and two baseline tables are recovered from the original executed notebook, not redrawn from invented data. Their manifest records source cells and hashes. The updated notebook exports new plots/tables to derived/eda_v0.3.0/, separately from the report reference assets. Review and promote a complete new run together with its numerical narrative and provenance; partial execution cannot silently replace published figures.

Limitations

• The plant, signals and labels are entirely synthetic.
• The hydraulic and mechanical dynamics are reduced models, not CFD/FEM.
• The persisted 25.6 kHz waveform discards content above 12.8 kHz.
• Spectra contain magnitude, not absolute complex phase.
• Persisted waveform length depends on acquisition policy.
• Adaptive sampling biases count-based summaries unless elapsed time is used.
• Evaluation truth is cleaner and more complete than real maintenance labels.
• Performance on DRMHB is not evidence of production readiness.

Results should ideally be validated on independent field datasets.

Citation

@dataset{drmhb_2026,
  title     = {DRMHB: Dynamic Rotating Machinery Health Benchmark},
  author    = {Gonzalez Zapico, Alejandro},
  year      = {2026},
  publisher = {Hugging Face},
  note      = {Synthetic rotating-machinery health-monitoring benchmark generated with VibSynth by TWave}
}

License

DRMHB is released under the Creative Commons Attribution 4.0 International license (CC-BY-4.0). See LICENSE.md.

Reporting results

Please state:

• the temporal split and calibration procedure;
• which raw modalities or derived features were used;
• whether RPM, regime or other operating metadata were supplied;
• how transitions and healthy OOD were scored;
• the wall-clock alarm persistence rule;
• per-machine detection coverage and delay;
• whether evaluation truth was used only after prediction.