So you’re tired of all the hype and buzzwords around this Industrial Internet of Things (IIoT) stuff? You want to actually start using predictive analytics, machine learning and other IIoT technology to do things like improve your overall equipment effectiveness (OEE) or monitor your production KPIs in real time?
Hey, me too! So buckle up, friend, and let’s get started.
But wait. Where and how do we get started, and what tools do we need to do that? And how do we know they’re the right tools for what we’re trying to do?
As operations technology (OT) and information technology (IT) professionals, we face challenges in adopting and implementing IIoT technology. A systemic lack of OT/IT interoperability increases our development time and costs. Risks associated with new technology adoption and cybersecurity leave us skeptical. And traditional automation vendors try to lock customers into proprietary technology, only exacerbating our interoperability problem. It is at this point where most of us have stopped our IIoT adoption—we’re excited about the possibilities, but paralyzed with fear and skepticism of the unknown.
However, as long as you know what a control tag is, then this article is for you. Armed with that knowledge, you can build an IIoT application with little to no monetary investment or opportunity cost; and you can do it in a couple of hours instead of weeks or months. Here’s how.
Open interoperability
Like Ethernet, today’s standards-based Internet technologies—like RESTful APIs—have infiltrated the traditionally proprietary and closed-off world of automation and process control. This shift is allowing IT software and OT hardware to communicate directly. Open-source edge computing and IIoT application development platforms like Node-RED have bridged the gap between OT and IT silos, giving our valuable legacy industrial machines and equipment a path toward digital transformation.
RESTful APIs (Representational State Transfer Application Program Interface) are the software tools stitching together the Internet as we know it today. They document how software programs should request information and share resources using the HTTP/S protocol to request and deliver data in JavaScript Object Notation (JSON) format. Today, there are more than 17,000 RESTful APIs to everything from predictive analytics software hosted in the cloud to industrial automation controllers on our plant and factory floors.
RESTful APIs allow everything from cloud applications to local databases and spreadsheets to securely interface directly to industrial equipment—like Opto 22 SNAP PAC controllers—and exchange real-time production data. But perhaps more importantly, RESTful APIs enable legacy industrial equipment to access external digital data and resources to do things like autonomously improve their OEE and schedule their own service calls. RESTful APIs to an automation controller form the foundation of IIoT applications. They’re also one of the technologies that power Node-RED—the free, open-source software development platform for building Internet of Things (IoT) applications.
Node-RED provides an easy-to-use graphical interface displayed in a web browser, where prebuilt blocks of JavaScript code, called nodes, are dragged, dropped and wired together to build what’s called a Node-RED flow. Node-RED flows are what power your IIoT application. These nodes make IIoT application development simpler, easier to repeat and faster to scale. With a vast and continuously growing library of prebuilt code, Node-RED provides IIoT engineers with an easy way to connect edge computing systems—like automation controllers and I/O systems—to local or cloud-based machine learning and predictive maintenance applications.
For example, let’s say you want to create an application to poll Modbus TCP data from a device, log it to a SQL database and move it into IBM’s Watson IoT Platform. You’ll find nodes for all of those functions already developed and ready to deploy without having to write, debug or support software code, thereby reducing software development risk and accelerating time to market.
Drag, drop, wire together, deploy. Node-RED is that easy. And it’s recently become a standard feature on the secure, industrially hardened groov IIoT appliance from Opto 22.
Less buzz, more build
So how do you use these tools to build an IIoT application for your specific factory, process or plant?
Well, let’s start with something you’re probably already familiar with—a control tag. Now, you could be a ladder logic programmer, a flowchart programmer or a structured text programmer; frankly, the IIoT really doesn’t care what language or environment you use to write your control programs, as long as it can get access to the data in your control tags. The standard control tag that we all know and love is really where we first begin bridging the gap between our legacy industrial assets in the physical world and our local or cloud-based digital applications that handle predictive maintenance and machine learning. Control tags are where our Big Data originates.
For example, the screenshot above of an Opto 22 PAC control program displays a control tag that’s holding the value of a digital input point on a rack of I/O.
The controller running the control program and interfacing to the rack of I/O has a RESTful API built into it; so, with the proper authentication and permissions, just about any application—including basic spreadsheet applications, web browsers like Google Chrome and, of course, powerful cloud-based predictive maintenance and machine learning applications like IBM’s Watson IoT Platform—can directly read and even write to the control tag value in my controller. For example, in Chrome, I can make an HTTP/S request to the controller’s web server by inputting the URL required to request my digital input values from my I/O rack, specified in the controller’s RESTful API. Then the controller will respond with the tag values I’ve asked for in JSON format, which looks like the screen at the end of the previous page.
But what does a web browser have to do with predictive maintenance and machine learning? The answer is: Not a lot. It’s actually the technology operating behind the scenes that has the real power. Because I can easily pull data out of my control system and view it in applications like Chrome, it means I can use other digital applications like Node-RED to pull data out of the controller, contextualize it with edge processing and move it into cloud applications like Watson—all without having to write a single line of software code. For example, I simply drag the SNAP PAC read node into my Node-RED flow, add a node to convert Boolean data values from my I/O rack into human readable text, add a Watson node to push the data up to the cloud, provide a few configuration parameters to each node and my application is built as shown to the left.
Then the data from my control system shows up in the Watson IoT Platform like as shown below.
Leveraging open-source technology and Internet standards lets engineers focus on identifying opportunities to improve processes and rapidly develop solutions. With the majority of the code already developed and accessible through Node-RED, you can wire together an IIoT application to pull data from your industrial machinery and equipment and then move it into whichever predictive analytics application you like. Most of them know how to natively communicate using RESTful APIs, and quite a few already have Node-RED nodes developed for them.
Find out more about developing your IIoT application with open-source interoperable software tools and free code samples at http://developer.opto22.com. We’ve also posted a glossary of IoT terms at http://info.opto22.com/internet-of-things-glossary-of-terms.
For more information, visit Opto 22 at www.opto22.com.
