How IIoT Enables AI and Big Data

Summary

As industrial companies work to make better use of their data to improve operations and increase efficiency, there are many challenges in the convergence of operational technology (OT) and information technology (IT) systems. Bridging the gap between OT and IT is essential to unlocking the insights that can be achieved through artificial intelligence (AI) and big data analytics.

At the heart of the OT-IT disparity lies the fundamental variance in data requirements. OT data consists of proprietary protocols, multiple data formats, and no contextual information. OT data is of course designed for operations, is directly coupled to applications, often exists over isolated networks, and typically relies on application programming interface (API) endpoints for data retrieval.

On the other hand, IT has very different needs to feed cloud-based AI and big data applications with data. IT needs data objects and modeling, standard data formats, and contextualized information. IT needs data to be decoupled so it can be shared, easily integrated with enterprise systems, and conform to a publish/subscribe methodology.

The key is finding a superior solution that works for OT while being able to share contextualized data to IT.
 

Bridging the divide: Enabling data connectivity

OT and IT do have some needs in common. Both are looking for secure data movement, both want to put data to use to power use cases that improve the business, and both are typically interested in saving resources (time and/or money). Addressing both the common and disparate needs necessitates robust data connectivity solutions tailored for the plant floor and designed for interoperability. The first step is to choose a protocol—a method for moving data—that is based on open standards and allows for self-discovery of the data.

Message queuing telemetry transport (MQTT) is a lightweight messaging protocol designed for efficient communication between devices in low-bandwidth, high-latency, or unreliable networks. MQTT is secure, ensuring encrypted data transmission to safeguard sensitive data.

MQTT Sparkplug is a specification that builds on the MQTT protocol to define how to use the protocol in a mission-critical, real-time environment. Sparkplug standardizes the format and structure of messages exchanged between industrial devices, sensors, and applications in Industrial Internet of Things (IIoT) environments. It provides the data model needed to define a tag value for use with OT, also providing data to IT, making it 100% self-discoverable and easy to consume.

Sparkplug gives users the ability to publish under a well-known OT-centric topic namespace to be able to publish model definitions from the edge, populate them with the contextual data for those process variables, and do all of that with report-by-exception. It’s simply a great solution for this OT to IT challenge.

Sparkplug also decouples applications from the underlying data sources, so cloud platforms can efficiently ingest data models and assets, enabling advanced analytics and AI-driven insights. Leveraging open standards like MQTT Sparkplug facilitates data modeling and self-discovery, establishing a source of truth at the edge.
 

Edge-to-cloud data flow

Decoupling applications from data sources is the real strength of an architecture built around MQTT Sparkplug. Applications can decouple and populate cloud platforms with the data models and assets seamlessly. There are many applications on the market—many vendors call them integrations or bridges—designed to facilitate this movement from edge to cloud.

These bridges exist for the major cloud platforms, making it possible to easily connect OT data models into Azure Digital Twins and AWS IoT SiteWise. Consider sending data to Snowflake, for example, to see how Sparkplug really aids in this type of edge-to-cloud data flow.

• Self-learning auto-creation of data models and assets. Using the Sparkplug standard from the Eclipse Foundation via the bridge Snowflake can atomically learn data assets requiring zero configuration and deliver the data into Snowflake.
• Efficient ingest of real-time tag data. Data is only ingested on report-by-exception, meaning values are only updated as it changes, vastly minimizing data storage requirements.
• No programming or code required. MQTT Sparkplug provides the open standard protocol that enables edge platforms and devices the native capability to represent OT data requiring only configuration for a zero-code solution.

IIoT serves a crucial role in bridging the gap between OT and IT systems, enabling the application of AI and big data analytics on cloud platforms within industrial settings. Through the adoption of protocols like MQTT Sparkplug, which standardizes message formats and facilitates self-discovery of data, industrial companies can efficiently communicate between devices and cloud platforms. This streamlined data connectivity enables real-time analytics, AI-driven insights, and efficient data flow from the edge to the cloud. By leveraging open standards, such as MQTT Sparkplug, companies can achieve interoperability and integration without extensive programming, ultimately empowering them to optimize operations, increase productivity, and remain competitive in today’s data-driven landscape.

_{This feature originally appeared in the June 2024 issue of InTech digital magazine.}