Integrations allow you to take the real-time data collected by Proxus and export it to the external systems where it is needed most. Data flows from devices through the Unified Namespace to your target systems. Whether you are building a data lake, an AI model, or an alerting system, Proxus provides the connectors to make it happen.
Integration Architectures
Proxus supports two main categories of data integration: Outbound Connectors and Notifications.
If you are looking for device connectivity (OPC UA, Modbus, Siemens S7, etc.), see Protocols.
How it Works
Integrations typically operate as actions triggered by the Rule Engine.
- Event Detection: The Rule Engine monitors incoming data.
- Target Selection: When a condition matches (e.g.,
Temperature > 80), it triggers an integration action. - Data Transmission: The platform formats the payload and sends it to your target system (e.g., POSTing to a Webhook or publishing to an MQTT topic).
Resilience & Buffering
Industrial environments are prone to network instability. Proxus integrations are built for reliability:
- Reactive Pipelines: Specialized connectors (like MQTT and Azure IoT) use an in-memory buffer to handle spikes in data.
- Guaranteed Delivery: Combined with the internal buffering system, Proxus ensures that your data reaches the integration target even if the connection is temporarily lost.
- Zero-Allocation Design: High-throughput integrations are optimized to minimize memory overhead, allowing for high-frequency data streams.
Choosing the Right Connector
- For Analytics: Use Azure Data Lake, ClickHouse or InfluxDB connectors for time-series analysis.
By default, Proxus stores entities/config in PostgreSQL and telemetry in ClickHouse. Outbound Connectors are optional exports that send copies of telemetry to external targets.
- For Cloud Integration: Connect via AWS IoT Core or Azure IoT Hub for cloud-native processing.
- For General Messaging: Use MQTT or Kafka to bridge data between different IT/OT systems.