Quick Start

This guide walks you through connecting SUPi to your industrial data sources and getting your first predictions running.

Prerequisites

• Access to at least one data source (SCADA, OPC-UA, MQTT, or historian)
• Network access from the SUPi instance to your data endpoints
• A SUPi license key (contact sales for a trial)

Step 1: Configure Data Sources

Navigate to Settings → Data Sources and add your first connection. SUPi supports:

# Example OPC-UA connection
{
  "type": "opc-ua",
  "endpoint": "opc.tcp://your-scada-server:4840",
  "security_mode": "SignAndEncrypt",
  "credentials": {
    "username": "supi-reader",
    "password": "****"
  },
  "sampling_interval_ms": 1000,
  "tags": ["compressor.*", "turbine.*", "pump.*"]
}

Step 2: Define Your Assets

Map your physical equipment to SUPi's asset model. Each asset gets:

• Asset type — compressor, turbine, pump, reactor, etc.
• Sensor mappings — which tags correspond to vibration, temperature, pressure
• Operating envelope — normal operating ranges and design limits
• Criticality rating — business impact if this asset fails

Step 3: Build Digital Twins

SUPi automatically generates physics-based models using your asset definitions and historical data. The system needs at least 30 days of operational data to build an initial model.

Step 4: Enable Predictions

Once models are trained, enable real-time prediction from the Monitoring → Assets dashboard. You'll start seeing:

• Remaining useful life (RUL) estimates
• Risk scores updated every cycle
• Anomaly flags with root cause indicators