adilabbasii/turbine-sensor-data

Turbine Sensor Streams Dataset

Production-ready gas turbine telemetry data for anomaly detection. Designed to simulate real enterprise RAG documents for the IoT Anomaly Agent Space.

Overview

This dataset provides realistic sensor streams that mirror actual SCADA exports from power plant DCS systems (Emerson Ovation, ABB Symphony). Use these files to test anomaly detection or upload your own telemetry data.

Files

File	Description	Status
gt_unit1_baseline_jan2025.csv	GE 7FA Unit 1 - Normal operation baseline	Normal
gt_unit2_bearing_event_jan2025.csv	GE 7FA Unit 2 - Developing bearing issue	Anomaly

Data Schema

Matches standard SCADA/historian CSV exports:

timestamp          - ISO 8601 timestamp (5-minute intervals)
exhaust_temp       - Turbine exhaust temperature (F)
vibration_x        - X-axis vibration velocity (in/s)
vibration_y        - Y-axis vibration velocity (in/s)
bearing_temp       - Journal bearing temperature (F)
inlet_pressure     - Compressor inlet pressure (psi)
lube_oil_pressure  - Lubrication system pressure (psi)
fuel_flow          - Natural gas flow rate (MSCF/hr)

Upload Your Own Data

The IoT Anomaly Agent Space accepts CSV uploads with this schema. Export from your historian (PI, PHD, Wonderware) and upload for instant anomaly detection and root cause analysis.

RAG Integration

These documents are indexed for retrieval-augmented generation:

• Asset identification by unit number
• Temporal context for event correlation
• Failure mode pattern matching
• Maintenance history cross-reference