How Do You Assess IoT Platform Software for Vertical Industry Applications?

TRANSCRIPT

David Greenfield, Automation World

Welcome to the Automation World Gets Your Questions Answered podcast where we connect with industry experts to get the answers you need about industrial automation technologies. I'm David Greenfield, editor in chief at automation World and the question will be answering in this episode is how do you assess IoT platform software for vertical industry applications. Now joining me to answer this question is Steve Ward with Emerson.

So thanks for joining me today, Steve.

Steven Ward, Emerson

Hi, thanks for inviting me. It's great to be on this meeting.

I've got a lot of experience in various different aspects of the automation industry.

I've been with Emerson for several years and previously with General Electric and previously with Rockwell. When I was with them, it was Allen Bradley, so lots of experience both in the PLC part but also in HMI/SCADA connecting to higher level systems including MES and ERP. So hopefully I can give some really good answers, that will help our audience.

David Greenfield, Automation World

So Steve, I guess you know just to start, you know the general description of an IoT platform is that it's, you know an application or a service that enables a company to connect all the various things in their Internet of Things ecosystem. You know which in manufacturing tends to be devices and software. Of course. Now the idea here is that data from all these devices and software can be aggregated, analyzed and you know shared for a better overall understanding of the manufacturing operation. So to start, you know, I guess, do you agree with this basic explanation of an IO T platform or do you view it differently?

Steven Ward, Emerson

I agree at a high level, platform can mean many things ranging from an item of hardware that can be used to run IO T applications through hardware and software combination all the way up to preconfigured applications or services. It's very difficult for some users to choose because vendors don't agree on the common term of what a platform is.

So for example, some people might say this is my platform and it's just an industrial PC that's ready to run. Whatever software you want, other people might say here is our complete solution and you pay us some money every month or every year and we'll come along and do everything for you.

So there there's a big range of different platform definitions between those two.

Yeah, it it's a sort of difference between buying a phone to make phone calls and then going to Apple to buy an iPhone and buy all of the apps and all of the services that go along with that. There's a huge range of different possibilities there, depending on what your requirements are, how much money you want to spend, what level of complexity you have in your plant. So it it's quite a tough question to answer simply.

For example, yeah, you can have an industrial PC which just provides software only and certainly Emerson can provide that we can increase that to having a software stack that's ready to run applications.

And I think that's very beneficial for users because if you just have an industrial PC, what's of whether you use what operating system do you run, particularly because these are often connecting remotely, how do you make it secure?

So having a predefined software stack can be a big step forward, but still give the user flexibility to create their final solution. But some people want the solution that's ready to run. They want the vendor to implement them for them, so again multiple options depending on who you are and what you want to do and what your level of technical ability is and and of course the fundamental issue is often how much money do you want to spend.

David Greenfield, Automation World

It always comes down to that one -- always. So you know from that basic generic explanation of an IO T platform which you generally agree with but noted you know the different vendors approach it different ways. You know, based on what they had to sell, you mentioned several things that Emerson can do and does with its technologies.

Can you explain how Emerson approaches it specifically with their technologies?

If someone comes to you looking for an IO T platform strategy.

Steven Ward, Emerson

Well, again, we can offer multiple solutions. You know, we could start with a piece of hardware, the industrial PC running, say Linux or Windows, and that's ideal for people who've already got their own software stack and know exactly what they want to do.

For people that want a bit more structure, we can provide our package solution which gives you preconfigured software and it's all containerized for cyber security reasons, along with some configuration in there to force you to use passwords and to protect those containers from being hacked.

So the benefit of that is it means you don't have to worry about what software stack you use, because Emerson has already chosen best of breed type solutions for that.

There's also the possibility of going up to a full hosted solution.

We have Emerson's Plantweb and that solution meets the automation world definition, so that includes cloud based hosted solutions that solve specific problems. For example, you might want to be doing steam track, steam trap monitoring, or monitoring of corrosion, and the platform can even extend beyond the single vendor into an ecosystem where you can have multiple people all contributing together.

So again, it's a very wide definition and we certainly have a wide range of solutions and it's not just us.

You know there are multiple vendors out there who can meet a lot of those requirements and even work with Emerson or work with the end user or configure that software or configure those applications. And also of course, yeah, let's not forget there's products like HMI, SCADA, HMI, SCADA can meet a lot of those requirements for IoT and HMI.

SCADA is not new. It's been out there for a long time. I think what makes IoT more interesting is that we're now being more open. We were adopting more specific open interfaces and protocols, and of course we keep coming back to cybersecurity when we have this wider connectivity that cybersecurity becomes very important.

David Greenfield, Automation World

Yeah, I like your reference there with HMI/SCADA because so much of this while it is, you know, IoT and digital transformation concepts are quite new to industry there, it's still being built on a lot of traditional technologies, just kind of extending our traditional uses of them.

So thank you for noting that. So you know, I guess before we move on a little bit before we get too far down the road into this discussion, you know what are IO T platforms most helpful with. And by that I mean, you know, why should manufacturers, you know, be considering the use of IO T platform software to begin with.

Steven Ward, Emerson

And that's a great question. And I'm sure that's one of the fundamental questions that gets asked because people want to know what's in it for me, you know, or more specifically, what's the return on investment. And again, if you go back to HMI, SCADA, if we look at traditional large plants, they've had large systems for monitoring that have been in place for years.

If you look at the process plant, it might be a DCS. If you look at more of the discrete or hybrid plant, it could be a combination of PLCs with HMI, SCADA and local operator interfaces.

But where people have always struggled is connecting remote assets. So if you've got a fleet of assets, it's widely distributed, maybe those assets move about, maybe those assets are a long way away from the central location. That's really where IoT comes into it because IoT it's that things or more specifically IoT industrial into the things, but it's also about edge connectivity and it's about getting the data from those Edge devices from wherever they are into those central systems and it goes beyond HMI/SCADA though page my starter is great for visualization, creating pretty pictures of the plant but more and more data scientists and data engineers, they're not interested in what the screen looks like, but that's still important, but they're more interested in what is the data coming from the plant and how can I use that to optimize the plant.

And the typical things that people want to do that the two main things that we always come across is they want to increase efficiency and they want to reduce unplanned downtime.

Now, if you have a machine or process, then you need to understand how that machine is running. You need to maybe set a baseline. You need to understand how well it can perform when it's operating correctly, and then by collecting data from that machine you can then analyze its current performance and compare that against the baseline performance, and then you can see is there a problem with the machine? Is it running worse than I'm expecting? Is something going wrong? Is there a problem somewhere and then with those data that you've collected, you can then do more detailed analysis? It might be a raw material problem. There might be issues with running the machine outside its normal parameters. It might just need some maintenance, but until you've collected that data and started collating that data, you don't know.

When I was with GE we did various approaches around Six Sigma and lean manufacturing and there are processes you can use.

One of the ones I used to use a lot as part of my Six Sigma training was the DMAIC which is define, measure, analyze, implement and control.

Of course, for define part is you have a problem and that may be reported from your user base or even with an end user who's phoning in about a quality problem. The next thing you need to do is M, which is measure, and of course that's where the IoT part comes in. It's about collecting data from the plant, getting it into a central location so that then you can perform some sort of analysis. And I think that's really where IoT platform as are most helpful with they bring in data from all these remote separate systems and then you can put the data in the central location and make sense of it. You can turn that data into information and use that to optimize your operations.

And it's not just about, you know, modern protocols. You know, we think about modern protocols like MQTT and OPC UA. That's great if you're creating a system now, but many of the assets in a plant or or in the customers processes, they may be old PLCs, they may have legacy protocols, some of them might not even have any protocol at all.

It might be something that's so old or not even very well automated, but you can't connect to it, so the whole point of IoT is being able to somehow gather that data and then bring it into your central location. And I think it goes beyond that as well.

Data is one thing, but then if you can turn that into information, if you can, then create some sort of analytical model. You could then close the loop and you could have a local analytical model. Some people refer to this as like a digital twin, but the idea is you can run their analytical model locally and it could take its own corrective action. Maybe if the machine starts drifting you could do some sort of closed loop control to bring it back under process control. Or maybe there's a a problem that's going to occur. So you could send an SMS or an email to a maintenance person to come out and fix that machine before it breaks down.

So that that's the promise of IoT—it's first of all, collecting the data, making it available, and then the real promise is that analytical capability to improve operations.

David Greenfield, Automation World

OK. Thanks for clarifying. That's a key point I was glad you were able to help explain about how these IoT platforms are not necessarily some advanced add-on or nice to have, but it's more of a core building block that enables everything you're doing with the data from digital twins to potentially AI analytics using whatever you have in place currently.

Steven Ward, Emerson

Indeed. It’s like the growth of smartphones. So it used to be you'd get up in the morning.

You'd get dressed, you'd go to work and it was all uncontrolled. Nowadays what probably happens is you get up, you check your smartphone. So the first thing is what's the weather like? What do I need to wear today? So already you've optimized your behavior for that day to wear something that's appropriate for the conditions you know. Is it gonna be hot or cold or so on the next thing you might do is check the route. Or maybe you're using sat NAV and that you will plan table you're sat NAV will guide you the best location to get to the office or get to where you're going, avoiding traffic jams and breakdowns and anything else that might occur.

So already you know, with this smartphone analogy, we've optimized our lives to reduce the amount of time it takes and the things that might go wrong in the day. And that allows us to plan much better what we're doing and that's the analogy we use. It's like having a smartphone, you know, the IoT approach allows you to connect your process to the Internet and be more intelligent, just like you and I are using our smartphones.

David Greenfield, Automation World

So one other question to kind of help establish the proper framework for our discussion here, Steve, is just over the past year or so the term IoT or Internet of Things has been overtaken a bit by the term digital transformation. Now from my view as a journalist covering the industry, I kind of tend to see IoT as a key ingredient of digital transformation, meaning that digital transformation doesn't really happen without IO T and that digital transformation I guess is more of a general call to arms, so to speak, for the manufacturing industries to update their legacy equipment or install new equipment if needed, that's digitally capable of being part of an IoT infrastructure. Do you generally see it this way or do you see IoT and digital transformation as being separate issues? Totally different things altogether.

Steven Ward, Emerson

No, I agree with you. Yeah, IoT I think is a is a part of digital transformation. Digital transformation can have a huge scope that could include anything that digitizes information in an organization. For example, you might implement an enterprise resource planning or ERP system. Maybe you're putting in a customer relationship management application.

Those who examples of digital transformation, they may or may not directly affect your manufacturing process and then you can even go beyond digitalization, go digital transformation could mean, but when you're implementing your ERP or CRM, you will not just digitizing what you do now, but you're completely reengineering your business processes so they better fit what is required by the modern digital world.

So of course that that can be huge in terms of the impact on the organization and organization, you might want to complete the overall all of your processes so they fit better with what the software is able to do and it also allows you, I guess to understand better what are you doing, how do you do it is, are there better ways of doing it?

So I think you're right in that IoT is a part of digital transformation. It can certainly help you get access to the data that we mentioned previously. It can help you understand your operations. It helps you understand where you can reduce downtime and how to reduce scrap and increase efficiency. So, I think it's really important to have IoT as part of digital transformation.

And there are other IoT projects as well that maybe they're not digital transformation projects as such, but a lot of organizations are looking at net zero and carbon reduction efforts and IoT can be a part of that as well. So you're right, IoT is a part of digital transformation, but it can be beyond that and digital transformation can be a much larger subject as well.

David Greenfield, Automation World

So with that basic framework established, now that we're on the same page with this and hopefully cleared this up for our listeners as well, what we mean by the two terms and what we're discussing here, you know, I guess to get to the core of the initial reader question that prompted the topic of this podcast about assessing IoT platforms by starting with the vertical aspect of the question from the reader and the reader asked for advice on how to assess IoT platform software for vertical applications specifically. Are there key differences about IoT platforms when it comes to implementation in different industry verticals that our audience should be aware of?

Steven Ward, Emerson

Yes, there are differences that are readers need to be aware of.

Since my role is technical, I tend to focus on the technical differences rather than the different verticals. If you look at the pure technical solution, it's often the same or very similar across different vertical industries because the underlying technology is pretty much the same.

You know, you have an industrial PC and the size of that PC can vary from something that's the size of a matchbox all the way up to quite a large box or even the rack mount server that the software that runs on it tends to be the same sort of software that we used to running on our desktops, although maybe in a slightly different format, maybe optimized for those industrial applications also then you might have other software that runs on top of that.

And again, that could be very similar to software that we used to in other applications, regardless of what the vertical is, but where it's different maybe and this maybe is where users maybe need to be a bit more advised is, does your integrator understand your industry?

And I found this before working across multiple different industries. Different industries have their own terminology. Some industries have their own particular requirements or particular issues that they consider to be more important than others, and that's probably where a user needs to take a bit more care.

It's not so much about the technology, it's more about does the person who's implementing this understand what my specific issues are?

Take for example, if you in automotive, a lot of the automotive optimization is about making as many vehicles per hour as possible, and then for those vehicles optimizing the price I I've worked with various automotive customers in the past and what they wanted to do was get as many cars manufactured as they could. And for the price of each car, make it as cheap as possible. Even just saving a few cents per vehicle, it doesn't sound like much, but when you multiply that over multiple lines, multiple plants, you multiply that by the hours in the day, the days in the week, the weeks in a year back and head up to huge savings, that's very different from maybe someone who's in the medical industry.

If you are life sciences manufacturer, it's less about the productivity. Although productivity is still an issue, but it's more about validation. You need to make sure that what you're making is absolutely conforming to your original plan or design or recipe, and you need to be able to document that. So that's very different from the automotive industry.

Having said that, increasingly the automotive industry is tracking things in a similar way to what the life sciences industry is. So there's a lot of overlaps between these different industries, even if they have some slight differences. And, of course, you might be working in Indian industry where safety is important.

So maybe there are particular processes and procedures you need to follow there.

Having said that, I think for most IoT applications it's more about monitoring rather than control. So safety is less of an issue there, but yes, you do need to be aware of what you're vertical industry is. You do need to be aware of your particular requirements and you do need to be aware of that. Whoever's providing your solution understands your industry and and that could be because maybe your assets are more distributed, maybe you're more averse to cybersecurity, maybe you've got particular risk aversion. Maybe there's some other the thing that you do that you need to be aware of that the the integrator or the implementer of your systems also needs to be aware of.

So they're very difficult to answer your question with a simple answer. But yes, that there are certain aspects that people do need to be aware of where we're implementing IoT.

David Greenfield, Automation World

So it seems like the more important aspect of that, regardless of the industry and is the domain expertise uh involved with those who are implementing the system for that particular business or in that particular vertical more so than it is about anything about the different types of automation technologies that might be used across the different verticals, is that correct?

Steven Ward, Emerson

Yes, I would agree with that, yes.

David Greenfield, Automation World

So when it comes to software platforms in general, they're usually more of an IT project than an operations technology product. But with industrial IoT and the IoT platforms, it's supported, of course, you know in the connections of plant floor systems into the platform OT clearly has a stake in this process. So how do these IoT platform implementations typically play out across it and OT in a manufacturing plant?

Steven Ward, Emerson

That's interesting because I've seen IT and OT have been converging for some time, but there are fundamental differences and again this is a difference between some verticals, but even within some of the same verticals, we see some organizations have war convergence of OT and it even others. So yeah, one organization that's doing exactly the same thing as a different organization, although they're in the same vertical, may have very different setups to how they do this.

So for example, the biggest difference for me between IT and OT is there OT is very focused on speed. We have systems that are running with the response time of a few seconds at most, but in some cases the response time is in the milliseconds, so OT often has very high-speed response requirements and also the amount of data.

If you think about the potential for an OT system, it can create huge amounts of data.

I've worked on power generation where you might have a gas turbine. So first of all, those gas turbines are rotating very fast, but also there are lots of data points to collect. So you can be collecting megabytes of data per second, and IT isn't necessarily familiar with that sort of data collection.

Again, there's also a difference in cybersecurity. Traditionally, OT people haven't really worried about cybersecurity because they've always said ohh, my system is air gapped, it's not connected to the Internet. Cybersecurity isn't an issue, of course. I don't believe that, and I've seen many OT systems which are supposedly air gapped, but they were connected in some way or other to the enterprise network. Maybe there were firewalls in the way, but yeah, you'd only have to compromise a few boxes and the OT system would then be exposed to a potential hacker. IT of course is is much more concerned about cyber security, but again it varies from vertical to vertical and it varies from organization to organization.

And I think this is where it and OT should be converging because OT wants to have that performance and they need to be able to explain to the IT team what that performance means and how much data there is. But equally, it has the expertise to keep these systems secure and also in some cases to simplify configuration.

For example, for a lot of OT systems, yeah, you have to create your own user database and you have to keep all your passwords up to date. It that's one of their main jobs. They already have the systems to implement that, so if you can take a note system and integrate it into your IT system for user management, that saves those OT people a huge amount of time.

So I definitely agree that yeah, we should have IT and OT convergence, but there are still differences between those two teams in terms of how are they see the data and how they want to handle it.

David Greenfield, Automation World

So, given those differences that you just referenced there, even in an IT OT converged department to some degree from an OT perspective, do you have any recommendations of that? OT personnel should be sure to ask about any IO T platforms that are being assessed by the IT department. I know you mentioned the real time or near real time capabilities for data processing or data handling. Are there any others?

Steven Ward, Emerson

Yes, that that's one of the things that OT people need to be careful of.

If IT are involved in the project, they may not understand how disruptive that can be to be OT operations.

So in the OT team they need to make sure that any changes that are made do not interfere with their operations. For many organizations, productivity is key, and even the few hours of downtime can be extremely expensive. So if the IT team is doing anything that disrupts OT operations that needs to be prepared and planned for upfront so that it minimizes any potential loss of production and that could work the other way of course, because we don't suddenly want the OT team flooding the IT network with data and crashing the IT network.

But I think for many organizations, you know, it's the OT aspect that's more important there.

It would be very easy for someone in it to come along, unplug a network port, connect their device and configure it, not realizing maybe that they've broken the connection between two processes, which is vital for the operation of those processes.

So that's very important for OT and IT need to communicate to make sure that they minimize disruption.

David Greenfield, Automation World

OK, so you know Steve, when we started our discussion here today, you mentioned your experience working across multiple different industrial technology providers over the years.

So, you know, based on your work working with different manufacturers and users across industry, how far along would you say industry is in terms of IoT platform adoption today?

Steven Ward, Emerson

Well, I think IT has dominated investment. Many manufacturers are seeing the benefits of that are large. IT implementation can offer so that covers applications like enterprise resource planning or ERP manufacturing execution systems, even HMI, SCADA though in in some organizations that's now considered as an IT investment rather than an OT investment where I think we can see more scope for improvement in these systems is future developments like artificial intelligence perhaps as a way to reduce the time and effort it takes to perform some tasks.

So that there's certainly, you know, a lot of IT investment there and definitely scope for the future. But I think manufacturers need to be aware that their primary purpose is to make something.  So whatever the output of your plant is, that's what you do and you need to make that product or whatever it is as efficiently as possible and also to the required level of quality and quality can mean different things to different people. Of course, you may have a very high-quality product and that could mean is expensive. You may have another product which is not as high quality, but it has a lower price. So yeah, depending who the end user is, they will be willing to pay different amounts for different levels of quality, and that's fine.

But the users need to be aware of that and that's where IO T can come in, because many for manufacturers don't have that good connection between IT and OT systems. You might get an output from an ERP system, for example, that says I want you to make this product on this machine because that's what the ERP system does. It does the planning, but it might be that that machine isn't best suited to make that particular product. Maybe it can't meet the quality requirements, maybe it can't beat the throughput requirements.

So I think that's where there's a lot of improvement there for manufacturers to start gathering data on their machines or processes and start matching them to the levels of quality and throughput they need. So that's a big benefit for IoT can bring you is having that data and being able to understand the capabilities of your machines and processes.

And then I think IoT also allows manufacturing data to be, let's start again on that one.

IoT also allows manufacturing data to be available to multiple departments in an organization, and then those departments can use that data to be efficient. And to me, whatever their requirements are, so again you may have quality teams that are looking quality data you may have sourcing teams that are looking at raw material inputs and how good that raw material is. You may be having other departments which all have their own requirements and their own particular functionality. They want to achieve and having access to that data can help them.

David Greenfield, Automation World

So one last question here for you, Steve, and this comes off, you know based off the original reader question about IoT platform applications in specific industry verticals and the differences there, would you say are certain industry verticals further ahead in their implementation of IoT platforms? And if so, are there lessons to be learned there for other industry verticals in terms of what to do or what to avoid?

Steven Ward, Emerson

Yes. Again, there are differences between verticals, but there are also differences between organizations that are operating in the same vertical. And then each organization needs to look at its own bouquets and determine what's important to it.  Are they positioned as a high-quality vendor, low cost, high throughput and also you know how do they want to achieve that? Are there particular manpower requirements? You know, do they want to have, like, some sort of quality stamp like handmade, or do they just want to maximize whatever their operations are?

So an organization needs to have some sort of mission statement to understand what it's doing and what differentiates that organization from other organizations. But then beyond that, we do see different adoption levels in different industries and even different users have different levels of adoption in the same industry.

So, it it's very difficult to come up with a simple answer to this. I think in my experience and other people may disagree, but in my experience, I've always found that automotive to be very advanced, but they have the benefit of a large centralized factory where it's easy to implement large solutions and they also have the scale factor where they can easily demonstrate return on the investment.

If you're making something like 50 or 60 vehicles an hour, if you can optimize even by just a few cents on each vehicle you're making, it's very easy to calculate how much money that's going to save you over the end of the year. So, I think automotive is probably one of the most advanced in my experience, but maybe that's not what we would call an IoT solution.

It's more of a sort of HMI, SCADA, MES, ERP implementation. So maybe that's not a true IoT implementation.

I do see the oil and gases also advanced and getting more so because there's oil and gas operators implement IoT across different assets. They're looking to improve operations and reduce carbon emissions, so you know, in the past it used to be that you would pump oil and that's all you're worried about. But with net zero and sustainability, oil and gas manufacturers are looking to be more efficient in their operations and they want to reduce leaks. But also improve their return on investment. Again, yeah, depending on what the price of oil is. But we do see what oil and gases is very focused on introducing IoT solutions.

I've also seen water and wastewater companies, you know, traditionally of course they are very distributed, you can have pumping stations and water plants, which are spread across wide areas. So again they they've implemented quite a lot of solutions there to bring all of those into centralized locations despite their distribution. So they're probably the three main industries, I would say, which are more advanced than others. I think if we look at other industries, the laggards tend to be those where the cost of IoT is difficult to prove.

Yet again, how do you calculate that return on investment? For some companies, that's very difficult and also for some companies, the technology has just been too expensive.

In the past, for example, food and beverage manufacturers, they often operate a very low margins and traditionally there are high levels of people in the plant and less automation.

So IoT is less adopted in in those sorts of industries. But again, that varies.

There are some plants where you've got high volume of making a particular product.

For example, I've worked at some companies making chocolate, and that's more of a process application where you do worry about throughput and so on.

But again, large, centralized plants HMI, SCADA ERP, MES, you could argue that's not true IoT, but equally there are those other plants where they're looking at understanding what the machines and processes are doing and also what are the people doing because you know, let's not forget in most applications there are still people. So you might have two or three shifts per day. You might have different plants which have different shifts and you might be looking at how are those shifts of people operating. So you can compare them and then share best practices, but I would say at food and beverage is is probably the ones that are lagging and also small amount of factors that they tend to struggle to justify the cost of IoT and they're changing now because you know one of the big benefits of IoT is the costs are dropping hugely.

Yeah, when I started getting involved in remote data collection many years ago, it was extremely expensive. You know, we used to use things that ISDN and you could be looking at leasing align and costing thousands of dollars a month, whereas now you can set up a cellular link and it can just be a few dollars a month for quite large amounts of data.

So the costs are dropping hugely and I think that's opening IoT up to a lot of people who previously couldn't afford it or couldn't justify it.

I also see a lot of industries that have resisted IoT due to perceived cyber security issues.

This always used to be the big issue when we started working with companies looking to adopt solutions they would say Nope, I'm not gonna implement it because I'm scared of being hacked. I think now that the risks of being hacked are better understood. I don't think they've, you know, abated at all. In fact, if anything, cyber security risks now are worse than they ever used to be. But I think now a lot of people have got a better understanding of what those risks are and how to mitigate them. So yeah, that's removing itself as an obstacle from people implementing IoT. So as cost come down and the benefits become clear, I see that more and more industries can now adopt IoT to improve their operations.

David Greenfield, Automation World

Well, thank you again for joining me for this podcast, Steve, and thanks to all of our listeners, of course. And please keep watching this space for more installments of automation world and get your questions answered. And remember, you can find us online at automationworld.com to stay on top of the latest industrial automation technology insights, trends and news.