By: Kurt Hochanadel, Director of Sales and Marketing at Cirrus Link Solutions 

KANSAS CITY, February 18th, 2019 – The progression of Digital Transformation has had one primary issue to overcome: access to consumable Operational Data (OT) for Information Technology (IT) enterprise services. This is important in order to access services for data analytics and Artificial Intelligence (AI). In a legacy approach, data is stranded due to use of proprietary protocols, which makes it impossible to bridge the OT/IT gap in a cost-effective and open-standards approach. Typically, OT device data is not accessible or in a format that is consumable for IT services to enact on for analysis or other functions. Here is a simple but common example:

Example: A Modbus PLC is controlling and monitoring a compressor. It has registers of values used in ladder logic to control the compressor for operations which is typical example for industrial applications. These registers have data that when AI is applied it offers predictive maintenance to extend the life of the compressor and giving insight to schedule maintenance for serviceability and uptime.

So where is the problem? Let’s look at the Modbus PLC data from register 40,027 shown below:

When data comes to IT, there is no standard on what to do with the data. IT needs to know that the 40,027 Modbus register represents a specific measurement. For example, the compressor temperature scaled from 0 to 4095 representing 0-100 in degrees C. With these details, AI can consume the data over time coming from the PLC and derive insight on its operation. This is where the adoption of Sparkplug B bridges the OT/IT gap with MQTT.

MQTT has become the one of the most used IoT protocols. MQTT provides a one-to-many approach of publishing data to multiple subscribers on an enterprise service-bus. The MQTT client is easy to implement, even on the most constrained devices. Until now, for MQTT, there has not been a standard on how data is represented or how the topic notation is specified. Sparkplug B defines the specification of the data payload, and the topic notation. Sparkplug provides a common interface for clients (i.e. device or application) to publish and/or subscribe to data. Sparkplug payloads include the Tag name and meta information on scaling, engineering units, etc.
Below is a simple Sparkplug B payload represented in JSON (note that this is a representation of the binary encoded payload). The payload provides all the context required with the tag name and meta information that is used for Digital Transformation and OT.

{
“timestamp”: 1486144502122,
“metrics”: [{
“name”: “Compressor Temperature”,
“timestamp”: 1479123452194,
“dataType”: “Float”,
“engUnit”: “Celsius”
“scaledLow”: 0.0
“scaledHigh”: 100.0
“value”: 95.4
}],
“seq”: 1
}

Not only does Sparkplug B provide the payload and topic notation, it also defines how to manage the state of the connection between MQTT clients. This is required for operations as they need to know that the device is always online and operating. This enables its use for real-time control where a SCADA host application can utilize MQTT and Sparkplug B to operate a pipeline, factory floor or any control application.

The Eclipse TAHU Sparkplug B specification provides the standards that have been missing from MQTT. It provides all the context needed to bridge the OT IT gap and truly accomplish Digital Transformation for data in industrial applications.

Please reach out to sales@cirrus-link.com with any questions or comments.