Blog

Why IT and OT Convergence is Critical to IoT Solutions | CBT

Written by Maggie Chang | May 3, 2022 8:00:34 AM

Why IT and OT Convergence is Critical to IoT Solutions

CBT Presents at IoT Day Slam 2022

Abstract: Why is IT/OT convergence a “thing”? We’re part of the same company, shouldn’t we be able to work together?

Get an understanding of:

  • Why IT and OT are oil and water
  • How to bridge the gap and converge technologies to increase yield, efficiency, and productivity
  • What this means for IoT and IIoT in practice

Speaker: Steve Jordahl, Chief Technologist, Hybrid IT

 

 

Transcript:

Hello, everybody. So we’re gonna get into well, we’ve heard about a lot about digital transformation, this convergence of IT, OT, or just, you know how those things blend. And a lot of people think, Well, why haven’t they already blended? Why are we having to do this paradigm shift? And so that’s what we’re going to talk about today, we’re gonna dig into that, figure out why there is such a discrepancy and how we go from point A to point B, to get a converged architecture on both sides.

So we’ll be talking about both sides individually, and then we’ll talk about convergence. And then we’ll layer on the IoT stuff.

So we’ll first talk about IT modality really, we want to get in the headspace of the IT person or the IT departments and figure out how they make decisions, why they make decisions, and how that relates, then in contrast to OT. So here is here’s how IT kind of things, right? First of all, IT people play in the data center, right, they’re not out in a refinery, they’re not in a utility site. They’re not on a barge, they’re not, you know, doing anything outside, really, they’re typically in a data center, or in an office controlling infrastructure that lives in a data center. And their function is to support the business processes, right? Email, help desk, and things of that nature. And, you know, they’re, they’re typically business-facing apps, and they all have purposes, they all need to be up. But they’re very different than what you find on the OT side of the business. When you talk networking, it’s usually Ethernet. It’s usually TCP IP, it’s usually switched and routed, it’s usually Cisco, or Aruba, or, you know, something like that. And that’s all, that’s all fine and good. That helps make the world go round that gets your email from point to point all over the globe. But then, when you kind of boil it down, the target of their work is systems and software, right, a server, or a virtual machine, or a container, or network equipment, potentially. And the software that runs on top of that, that makes the business go. Over on the right side is a picture of the OSI model, open system, and Open Systems Interconnection. And it’s it basically outlines how data gets from node to node from computer to computer, across a physical wire, or physical sets of wires, and then how it’s interpreted up the stack to the point where you see Google in your web browser. Okay. And that’s how IT kind of how IT thinks they think very, very statically in this kind of mode from both a development perspective, as well as you know, IT infrastructure management perspective, okay.

So when we look at the mindset of IT, right, failure happens, they expect failure, they don’t expect to put up an email system and have it never go down. And the result of email going down. Yeah, some people don’t get their mail for a little bit. But it’s typically not the end of the world, you might miss it might miss a deadline, or might have something that gets somewhere a little bit late. But by and large, at least 80% of the time, nothing, nothing really crazy happens as a result of an email server. Now, if you’ve been in the IT space for any amount of time, or dealt with infrastructure and IT space, it’s all about the night, right? It’s all about, hey, does this storage platform have four nine or five, nine availability? Does this other architecture do this, does this server with this particular software have the ability to be up reliably? And, you know, IT wants to get everything up as close it as close to 100% of the time as possible. But at the end of it, it ends up coming down to management decisions. Is there a budget for Is there a priority on this particular project? What’s the impact of failure if something goes down? Is the sky going up, all those things go into determining how much reliability IT can put into a system. And even if you get components that are 99%, reliable, you still have their interconnects, or things that are above and beyond the platform you bought that has those, those, you know, Nines credentials. And so at the end of it when you talk, a multi-tier system, right, where you have app servers and database servers and clients and you know, maybe some web services out in the middle, all those interconnects, as a whole, do not have 100%, or even 99% reliability. They try, but it’s typically just, it falls short. Another aspect of it is IT people think, oh, three to five years, this, this equipments gonna be obsolete, I’ll get new equipment, right, and we’ll go to the next level. And that’s the kind of expected they don’t think in terms of decades or multiple decades, right. And it has just, you know, grown so fast, right? The evolution of things has just gone really fast, everybody’s clamoring to keep up much less, you know, stay on top of things. And on top of the things they’ve already deployed, they want to go to next, and the next thing, and the next. So that’s kind of the mindset of IT. And that’s how IT goes.

Now we’ll talk about the OT side. So the playground for OT, is industrial operations, right? It’s those plants, those refineries, those utility facilities, things like that, there, there are many people that are out in the physical environment, you know, out in the weather, working on something physical, replacing pipes, or repairing electrical lines, right, something like that. And their focus is on physical processes. In oil and gas, there’s a process that takes the raw material in runs through the process, and the resulting chemical comes out, right, that process is what they’re focused on. They’re their job is to connect initially right to build these environments up. So they can do said process, to monitor them to see how they’re doing. And then, you know, to manage and control and secure and do all those things to make sure that it runs reliably. The network, though, has evolved much, much slower than the IT side, right, they think this will work. So the bottom three layers are now kind of starting to blend into the OT side of things. When you look at the equipment and software platforms that are coming from the ABB, Rockwell, Siemens, and those types of companies, they have now all pretty much adopted some sort of Ethernet, with some sort of IP above it in order to communicate across those different devices into the control system. So they are adopting the technologies that we evolved in it, and now putting them into the OT content, which is great. But when you look at anything that’s existing or has been existing for a number of years, you’re not going to find Ethernet, typically, right, you’re going to find Profit Loss, Modbus and OPC, UA, and other things like that, that might run over serial lines, right? It’s not going to be what we know from the IT side in the existing architecture. As we move forward, we are seeing a good merging of IT networking capabilities into the OT side of the business. And then as you look at the top, the top part of this graph at the Blue graphic, these three lower layers, the network, data link, and physical all kind of marry up into this OT-related graphic. Remember the OSI model is more IT-related. Those relating to this bottom, bottom aspect, and then you work your way up into production in inventory and business IoT business staff and things. So that’s how that kind of plays.

Now when we look at the OT mindset, their mindset is failure can’t happen. When failure happens, money is lost. When a chemical plant stops producing because a pump fails, they are losing money for every second that that pump is out of operation. They put in equipment expected that it’s going to be there for potentially decades, right, they spend more money on robust hardware to make sure or to do whatever they can to make sure that it will be a viable solution indefinitely going forward. And they’ve kind of focused on these deterministic reliable protocols, instead of doing the technologies that we’ve used in IT, where we have had the concept of being able to resend packages that don’t make it. But these guys don’t want to not have a packet that doesn’t make it, they want to have packets that make it every single time. That’s what they’re known. Special Purpose stuff, right? It’s not generic stuff that you customize to suit. It’s a very specific purpose, is very purpose-built, and does its job. We already talked about that. Show slow evolution, we mentioned that IT has been fast, and OT is very slow. And it’s because of the put it in, let it run for decades, type of process, that cadence that they’ve been accustomed to.

So what’s missing in OT, is a lot of the things that we are now doing more in the IT side of the business with analytics and things of that nature. Or in the case of like a connected worker play, where you want to have your workers out in the plant, be able to communicate or get help or enter your desire of choice here, while they’re out there, and not have to go back in talk to somebody to come back out and do what they were told to do. They want to be connected. And a lot of this is enabled by the network connecting those types of environments, right. But really, what they what they’re focused on is those industrial transformation leaders, right, that’s what this graphic on the bottom is, those people who are looking at how to blend these environments, they are looking to increase revenue, reduce costs, increase operating margin, and as a result of blending these two, they are able to increase those or get better results from those capabilities. And that’s, that’s what we’re really trying to write, we want to make people safe, we want to make things efficient. All that kind of rolls in and gets better when you blend OT and IT together.

So now we’ll talk about how they converge and what people are doing today. So the first or the way we have traditionally done it is we had the two systems completely isolated, right? It does its own thing, it doesn’t align its own deck because the mind just doesn’t know. Right? They don’t trust each other, they don’t have the same methodologies and approaches. And so they don’t benefit from each other as a result. So we need something different. Now the first toe in the water is really about taking OT information, the data, right the information from all the hardware, all the sensors, everything bringing some of that specific data over into the IT space where they can do some analytics jobs or do some correlations, do something to try to optimize and better utilize the equipment that is on that is in that plant. Right? That’s what they’re trying to do. As you do that trust just begins to grow. Right? They start understanding because they’re now talking, they start understanding how each side thinks what makes them tick. Why they’ll get better results if they start working together, things like that.

Now, when you go to the next stage, you start to blend, right. So there’s now bi-directional communication between the two sides. So you might have automated data push from the OT side analytics happen Automatic Data push back from the IT side. Now, that doesn’t mean that the IT data that’s pushed in is actually controlling that OT, the OT heart, right, it’s not necessarily making decisions to turn something on or off, or increase throughput or, or decrease throughput based on that information. But it can be staged, it can be put in a place where the people that are making those decisions to the controls, can look at that, and then act accordingly. But it could also be that added to the beginning of automation, where they do the analysis, the analysis comes back and something is automatically changed. On the OT side, once that trust is strong enough, once they understand how things are playing together. But you still got to kind of think about it from the perspective of IT people and IT systems may fail. So if they don’t trust that that system is going to be up and operational, when they need to do this analysis, then they can’t really trust it to give that advice back. At which point, they can’t trust it to make operational changes. So it’s, it’s not quite there. But this is a better scenario, this is got that whole leg in the water.

So now what we really want to get to is something more like this, where we have a convergence of the two. Now, at this point, they trust each other, or at least the people that are working on this project request feature. What I kind of look at this, as is the concept that we run things like high-performance computing environments, where there is an air gap between the normal IT systems or the or the desktops and things like that. And this high-performance computing system. Or another analogy is black programs, top-secret programs that you know, our defense contractors, right they have, they have to air gap things, they have to make things separate, because of the security, criticality, whatever, there are many reasons to do it. But in this case, you don’t necessarily have to say all of it, trust all of OT, you have to pick those, those very good people from both sides who get it and bring them together onto the same team where they have the same goals. And they’re needing to do things together. And with that, then you can get the best out, then you can say, I want this OT data to feed into this analytics. And I want the output of those analytics to come back. And it makes changes to the actual operating environment. And that’s when things will change. It’s not going to be for everybody in it to go do this stuff it’s going to be for select people to participate in this kind of thing, and put those OT and IT systems together on the same network with a common goal. And that’s how we know.

When we talk about IoT and IIoT, there’s not really a difference. I mean, IoT is all about industrial things. Right? It’s a subset of IoT. IoT is huge. But IoT typically talks over TCP IP, your internet-connected refrigerator, or your Alexa or you know any of those newfangled things that are coming out. They’re all talking on IP, TCP IP, when you look at IIoT devices, they may or may not, or they may gateway between a profit bus or Modbus, to TCP IP. So there’s there are elements in there that are different than IoT things many times, not always, but many times. Very specific functionalities that you don’t get out of the general-purpose IoT stuff. And gateway is a way to get them to talk on the Ethernet or, or Wi-Fi or whatever.

So here are just some examples of, of, you know, what you’ll find general sensors, temperature sensors, you can even get those for Raspberry Pi’s, and you can do interesting things with them. But you’re typically not going to find that same kind of caliber stuff out in the IIoT space where, you know, they’re looking for determining whether or not a pump is going to They’ll before it fits right, predictive analytics and things of that nature, then you got some of the players down here that they care about.

Now, I mentioned that you may need to go through a gateway in order to bring that data from the OT side into IT-type technology. We use PTC’s, Kepware, and Kepserver for that. That’s, that’s one of the platforms that does very well at bridging that gap to say, I have 50 different sensors that talk 30 different protocols. How do I make them? How do I get the data into something that’s common that I can then do a good analysis on? So Kepware is the middle.

And then many times as an IoT platform, we’ll use PTC ThingWorx, right, we can bring it in, we can create dashboards and mashups, and all kinds of interesting things that, that both show us data, and then allow us to analyze it and change the operational control.

Now, the way we typically go about starting one of these processes is we prefer to do what we call a Quickstart. We get everybody into a room, and we go through an exercise to make sure that everybody’s on the same page. And 95% of the time, they are not on the same page. Why? And it’s very eye-opening and very interesting to find the results of doing this experience where we can actually get people to then follow that one path down the road. That gets us to point B, right. And that’s how we start with a Quickstart.