---
title: System Architecture
slug: concepts/system-architecture
docTags: 
createdAt: 2025-08-20T12:59:32.589Z
---

This page gives you a high-level overview of how **Factry Historian** fits into your industrial data landscape. It outlines the main layers of the system, how data flows from equipment into the Historian, how it is processed, and how it is exposed to external systems.

![Overall Factry Historian Architecture](https://api.archbee.com/api/optimize/QT0Op5BmASO-D5eI7lmoL-31vLMxI7PcPtB3CR97FG9-20250820-130445.png)

## Data Collection Layer

Factry Historian connects to various types of equipment and control systems in your production environment. The most common protocols used for collecting real-time data include:

- **OPC-UA**
- **OPC-DA**
- **MQTT** (including Sparkplug B)
- **MODBUS TCP**
- **File or network based protocols** (e.g. REST APIs, SQL databases)
- **Other protocols** for proprietary or legacy equipment

Collectors running within the Historian ecosystem use these protocols to acquire data from PLCs, sensors and SCADA systems, either by polling the source directly (OPC-DA, Modbus TCP, Siemens S7, SQL, REST) or by subscribing to pushed data (MQTT, OPC-UA subscriptions).

## Core Platform (Factry Historian)

At the heart of the architecture is the Historian itself. It provides the runtime environment, APIs, configuration tooling, and internal logic to process, store, and enrich incoming data.

Key building blocks include:

- **Asset Model**: Logical structure to map your equipment and group measurements
- **Engineering specifications**: Technical details such as units of measurement and ranges
- **Event Detection**: Configurable logic to detect stops, alarms, batches, and more
- **Calculation Engine**: Used to generate derived values, aggregations, or KPIs
- **Manual Entry**: Forms for structured operator or lab input
- **Prototypes**: Templates for consistent deployment across similar assets or use cases

Data is stored in two backend databases:

- **InfluxDB** for time-series data such as measurements and calculations
- **PostgreSQL** for configuration, metadata, and events

## Outputs & Integrations

Once data is collected, processed, and enriched, Factry Historian can serve it to other systems through built-in connectors or custom integrations:

### Built-in Outputs

- **FactryOS / OEE**: Integration with our MES platform
- **Seeq Connector**: For advanced analytics and process mining
- **Grafana Datasource**: Visualize time-series data in Grafana dashboards
- **Excel Add-in**: Retrieve Historian data directly in spreadsheets
- **MQTT Output (spB)**: Publish data to external MQTT brokers (typical in UNS architectures)
- **Parquet Export**: Periodic export to files for archiving or data lakes

### Connected Systems

Using the Factry Historian (Swagger-spec) REST API, users can furthermore connect to:

- **ERP** systems for planning or production reconciliation
- **Cloud platforms (Azure, AWS)** for central data storage or analytics
- **Business Intelligence** tools for reporting or dashboarding
- **ML / AI pipelines** for anomaly detection, forecasting, or optimization

## Summary

Factry Historian sits at the center of your process data infrastructure, acting as the single source of truth for high-resolution process data, and data derived from it. It connects to industrial equipment, organizes and enriches the data, and makes it available to people and systems that need it.

This modular architecture allows for flexibility in scaling, integration, and customization depending on your production environment and digitalization goals.