New architecture—MQTT—in new transformations

March 18, 2020

Don't connect devices to applications. Do connect them to infrastructure.

While smart data collection/analysis is a primary objective with digital transformations, many organizations still use clipboards and paper to gather data and information. Meanwhile, industrial applications are still closed, proprietary, and limiting in functionality, connectivity and licensing.

Inductive Automation's Travis Cox

And yet, I am optimistic about the future.

Organizations see the promise of digital transformation: the ability to get their operational data into cloud IT applications and do all kinds of amazing analysis, business intelligence and machine learning. But most organizations are trying to drive digital transformation from the top down. If you approach it in a way that doesn't work for the operator on the plant floor, it won’t succeed. Digital transformation really must be implemented from the bottom up, with OT on board first. 

Business leaders must create a migration strategy that implements digital transformation but also meets all OT requirements. It boils down to a single, crucial concept: an architecture change. We need to stop connecting devices to applications with protocols. Instead, we need to connect devices to infrastructure.

This new OT architecture uses MQTT, a lightweight publish/subscribe protocol that enables message-oriented middleware architectures. With MQTT, device data is published by exception to an MQTT server, either on the premises or in the cloud. Applications subscribe to the MQTT server to get data; there’s no need to connect to the end device itself. MQTT provides numerous benefits, including needing little bandwidth while having open standards, interoperability, auto-discovery of tags, and stateful awareness.

Modern challenges, modern tools

In addition to MQTT, you have web browsers, HTML5, smartphones and tablets, federated identity, and integration with the cloud. Modern web browsers are everywhere today, from PCs to smartphones and tablets, to smart TVs. They provide easy access to applications and data. Browsers are arguably the most used, developed on, and tested application medium in the world. The advancement of web technologies has accelerated from the use of HTML5 and CSS3 to mobile-responsive design to high levels of security.

Those technologies are stable and mature enough to use in the industrial sector. Modern SCADA platforms can provide access to anyone using a web browser—bringing tremendous value in delivering data across an enterprise. These platforms must also support a licensing model that is fundamentally unlimited, allowing access to data without barriers. We shouldn’t have to walk up to an HMI station as the only way to get access to data, right?

Smartphones and tablets are incredible pieces of technology. They have an amazing number of sensors built in, such as GPS, camera, accelerometer, Bluetooth and more. Applications can easily be made available to web browsers and smartphones, as native apps. With smartphones, applications can have access to these sensors, which can unlock great opportunities.

Most organizations already use several cloud applications such as email, document sharing, and customer tracking. These applications use single sign-on (SSO) with existing corporate credentials and allow access to all your applications in one place. This is possible using trusted federated identity technologies such as SAML and OpenID Connect. It’s possible to use those technologies with SCADA. Instead of having shared accounts, each person can easily log in to gain access to the data they need, securely. These systems can verify identity through two-factor authentication (2FA). This allows users to seamlessly get data from different applications easily, effectively and securely. 

And it’s easier than ever to leverage data with cloud platforms, machine learning, analytics, and artificial intelligence. The cloud offers many opportunities. It provides numerous databases, deep storage, and data lakes that make it easy and cost-effective to store all your data. The cloud provides instant availability, reliability, and scalability without the headaches of maintenance or local infrastructure. You can leverage machine-learning algorithms and tools to tune processes, predict machine failures, do forecasting and more.

Ideal Industry 4.0 from a brownfield starting point

There are hundreds of different polling protocols that require mapping in terms of name, engineering units, and other metadata. Currently, our SCADA systems are connected directly to these devices with mappings defined in SCADA. To get to a new architecture, we need edge computing, protocol conversion and platforms. 

First, we must decouple devices from SCADA. Let’s say you have 10 Modbus devices connected to SCADA. You can deploy a single edge gateway, hardware, or software solution with support of the Modbus protocol and MQTT that pushes polling closer to the PLC. You can poll more information, potentially at faster rates, and publish the values as they change to a central MQTT server.

As you acquire new sensors or upgrade equipment that supports MQTT, SCADA will immediately get access to that data without having to know about the end device. With this new architecture, data is openly shared throughout the enterprise.

And building tools on top of platforms is ideal. Platforms are environments, or ecosystems, that connect different groups and derive benefits from others participating in the platform. Platforms drive innovation with lower risk and have a large community to draw from. They typically have open access and interfaces, cross-platform compatibility, modular architecture, and the ability to scale and extend. 

It’s really great to see more and more device manufacturers embedding SCADA software in their products. You can buy a device with the software you need already installed, configured and licensed. It saves organizations time and money, and it provides the best hardware and software together. It’s much easier for users to deploy, because the vendor has done almost all of the work for them.

Travis Cox is co-director of sales engineering at Inductive Automation.