Dark Mode
Zalingo Synthetic Manufacturing — Predictive Maintenance & Downtime RC
Synthetic Tabular Data
Tags and Keywords
Trusted By
"No reviews yet"
£749
About
Zalingo Synthetic Manufacturing — Predictive Maintenance & Downtime RCA — 100k Focused Sample (Parquet)
A 100,000-row synthetic PdM/RCA sample with labelled failure windows and downtime events. It blends high-frequency sensor telemetry with event logs (cycles, maintenance, downtime) and precomputed rolling features so teams can go straight to RUL/survival, anomaly, and root-cause modelling—no proprietary plant data, no PII.
Need larger volumes or scheduled refresh? After purchase, ask about enterprise bundles and monthly/weekly/daily subscriptions (PdM/RCA only or mixed).
Dataset Features (tight, buyer-friendly)
- site, line_id, machine_id, asset_class — Factory context.
- ts_utc — UTC timestamp (ISO-8601). granularity: 1–60s ticks (synthetic).
- event_type — sensor_reading | cycle_complete | downtime | maintenance | quality_event.
- sensor_type, reading_value, reading_unit — vibration (mm/s), temp (°C), pressure (bar), rpm, current (A), power (kW), flow (L/min).
- op_state, setpoint, shift — Run state, target, and shift label (optional).
- failure_label (0/1), failure_mode — bearing | overheating | misalignment | lubrication | electrical | other.
- time_to_failure_hours (TTF) — For survival/RUL.
- rolling_mean_vibration / rolling_std_temp / delta_current — Windowed features.
- anomaly_score_0_1 — Unsupervised early-warning signal.
- downtime_start_utc / downtime_end_utc / downtime_minutes, downtime_cause — Mechanical | electrical | material | planned | changeover.
- maintenance_type, work_order_id — preventive | corrective | condition-based.
- mtbf_hours, mttr_hours — Reliability summaries (synthetic).
- oee_availability / oee_performance / oee_quality / oee_overall — KPIs (0–1).
- throughput_units, scrap_count, scrap_reason, energy_kwh — Production context.
- country, city — ISO-2 + synthetic city.
Distribution
- Format: ZIP with Parquet shards (Snappy) + README.
- Volume: 100k rows, ~22–35 cols, 1–5 parts.
- Partitioning (full versions): by date / site / line / machine.
Usage
- Failure prediction & RUL (hazard/TTF).
- Anomaly detection with rolling windows.
- Downtime RCA (Pareto, cause trees).
- OEE improvement and energy analytics.
- MLOps QA (schema/drift checks, dashboards).
Notes/Disclaimers: Not real plant data; not for safety-critical decisions. Rates and KPIs are synthetic and not tied to any manufacturer.
QUALITY