Overview

The Edge versions of the software include the ability to act as a Data Collector, getting field data using anyone of the 70+ industrial protocols available, then forward that data to a Historian System, or to a MQTT broker, 

The system can use the Store & Forward protection, so when the server that needs to receive the data acquisition is not available, the system store the data locally, and when the server becomes available the data is forwarded. 

On this page:


Canary and MQTT

We added the Store-and-Forward functionality when our software platform publishes data to the Canary Historian or an MQTT Broker, which means we temporarily store the data of a message for transmission to its destination at a later time in case the network is not accessible for any reason.

This feature is crucial in data collection projects, where the data is sent to a remote repository.


Architecture Requirements for Edge to Cloud Applications

Open

  • Open technologies → Interchangeability of components

  • Not tied into any specific vendor solution

  • High Flexibility

  • Enables self-service approach

Distributed and Secure

  • Allows processing at Edge Devices

  • Allows Hybrid storage (edge and on-premise historians)

  • High scalability and responsiveness

  • Comply with NERC-CIP network security standards

Reliable

  • Avoids duplication of data definition

  • Avoids hard-coded data mappings

  • Resilient and secure architecture

  • Reduces maintenance efforts

Distributed Data Steward

The data governance role, ensuring the collection, quality, and fitness of the data assets, including its metadata definition, is distributed across multiple systems or sites, but abides by the Universal Namespace conventions defined within the corporation.



In this section:

  • No labels