Compose - Data Modeling & Contextualization

Compose is where raw data becomes meaningful information. Build digital representations of your factory by defining data models and creating instances that map to your real-world assets, processes, and products.

Why Compose?

Data without context is just noise. A temperature reading means nothing until you know which machine it came from, what product was running, and whether it's within normal operating range. Compose solves this by:

• Structuring your data – Define reusable schemas that ensure consistency across your organization
• Adding context – Link data to physical assets, production lines, and business processes
• Building digital twins – Create living representations of machines, lines, and entire facilities
• Supporting ISA-95-aligned hierarchies – Map your data to industry-standard structures for interoperability
• Accelerating development – Reuse models across teams and projects to avoid reinventing the wheel

Core Concepts

Compose operates through the following building blocks:

Models

Models are the blueprints for your data. Think of them as templates that define:

• Structure – What fields exist (temperature, pressure, status, etc.)
• Data types – Whether fields are numbers, strings, timestamps, or complex objects
• Validation rules – Min/max values, required fields, format constraints
• Relationships – How models connect to each other (machine → line → factory)
• Metadata – Documentation, privacy levels, and versioning

Example use cases:

• Define a "Production Line" model with fields for line speed, downtime, and OEE
• Create a "Product" model that tracks batch numbers, quality metrics, and timestamps
• Build a "Sensor" model with standardized fields for IoT device data

Note: The previous Instances page and UI have been removed in favor of newer function- and model-based flows. Existing references to instances in older screenshots or drafts may be deprecated.

Common Use Cases

Digital Twin Foundation

Define models for every asset type (machines, sensors, lines) and create instances for each physical asset. Your digital twin becomes a queryable, real-time representation of your factory.

ISA-95 Hierarchy

Build models that map to the ISA-95 standard (Enterprise → Site → Area → Line → Cell → Equipment). Instances automatically inherit the hierarchy, making cross-plant analytics trivial.

Product Genealogy

Create models for products, batches, and quality tests. Instances link together to provide end-to-end traceability from raw materials to finished goods.

Predictive Maintenance

Model your equipment with health indicators, maintenance schedules, and failure modes. Instances aggregate sensor data and trigger alerts when anomalies occur.

Energy Management

Define models for meters, consumption zones, and production areas. Instances calculate energy per unit and identify optimization opportunities.

Getting Started

Ready to structure your factory data?

1. Start with Models – Define the data structures you need
2. Use in Orchestrate – Build pipelines that leverage your contextualized data and function-driven logic

Each section provides detailed configuration guides, expression syntax, and best practices to help you build robust, maintainable data models.