Edge Computing – Making the Right Choices at the Edge
When considering edge computing IoT architectures, many individuals allow expedience and technology to drive their decisions. This can lead to disaster. A successful edge computing solution requires careful consideration of end goals. Each component of your architecture must perform and deliver according to your needs. Since the amount of IoT-generated data rapidly increases, selecting the optimal architecture for edge analytics becomes a critical operational and technical challenge. Mistakes can be detrimental to your business.
The first step in successful edge computing implementation is to thoroughly understand your end goals. Learn from the lessons of the past, create a clear picture of your desired outcomes and connect with seasoned technology professionals to assist in achieving your optimal infrastructure.
Cirrus Link Solutions, experts in MQTT and edge computing architecture, understands that full implementation of IoT analytics requires hundreds of decisions; however, the three most vital are time, volume and data standardization. Addressing these will result in a cohesive alignment of your business and technical needs while minimizing wasted effort and increasing success.
Time – How quickly must the data be analyzed?
A fraction of a second can make all the difference in the choice and cost of an edge computing technology implementation. Industrial control systems might need to respond in tens of microseconds to avoid damage and ensure safety. Whereas other devices, such as climate and temperature sensors, might only need to collect data once every few minutes.
When it comes to real-time data gathering and analyzing, the SCADA (supervisory control and data acquisition) system you choose is based on the data turn-a-round. SCADA systems used to monitor and control a plant or equipment in industries such as telecommunications, water, and waste control, energy, oil and gas refining, and transportation. Utilizing MQTT as data transport will maximize response times and minimize bandwidth requirements. These considerations will determine the edge computing infrastructure especially if adding edge analytics to be performed close to the source of the data. This may include within the connected devices themselves or within a local gateway device. In order to properly architect an analytics system, designers must understand the use case and the level of responsiveness required.
Volume – Where should the data go, and by using MQTT can the volume of data transmitted be reduced?
For data analysis not needed within a few seconds, edge computing processes can and should move further away from the sources of data. The distance the data can be transferred or moved is determined primarily by the volume of data and the bandwidth available. MQTT utilizing report by exception allows a larger volume of data to be monitored while reducing the bandwidth by only sending data points that have changed by a specified threshold.
Most MQTT infrastructures and IoT devices are optimized for low power and minimal cost, storing from only hours to weeks of data. In addition to determining how quickly the data must be analyzed, a decision must be made on what data will be retained and what data will be transferred for additional analysis and storage.
Data Standardization – Who are the data consumers throughout the enterprise?
It is crucial to be able to translate the Brownfield data into data that is easily consumable by IT. Operational Technology (OT) data is proprietary and cryptic in nature and typically has no context as to naming, engineering units, or scaling. The SCADA/HMI control application currently polls devices in their proprietary protocol of which any context of the data is hand entered not retrieved by the protocol.
Using MQTT and the Sparkplug B specification provides the definition to transform data into today’s IT standards providing full context. Sparkplug B is an open standard that is license free to use under the Eclipse Foundations TAHU project and can be found HERE
Moving to a publish/subscribe model with MQTT enables this transition from a one-to-one into a one-to-many approach, encouraging innovations while making it easy to adopt new technologies. Data producers publish the data in Sparkplug B format to an MQTT server.
Cirrus Link Solutions’ SCADA partner, Inductive Automation, hosts the Ignition Community Conference annually in September. This year, the Cirrus Link Solutions team will give a keynote presentation and conduct two workshops on MQTT and MQTT servers. We will be discussing how edge computing analytics at the device edge, infrastructure edge and in the centralized cloud can complement each other. We will step through the concepts of an MQTT-with-Sparkplug solution and highlight where to use it. We will also show the benefits of using MQTT with Ignition 8 and the new Ignition Perspective Module. This workshop will also contain an interactive project which will further emphasize the distinct requirements and fundamental influence time, volume and data standardization play in building a proper architecture for analytics.