In this episode, we connect with Sam Russem, director of the smart manufacturing practice at Grantek (a system integration company), to answer a reader's question about whether it's possible to replace supervisory control and data acquisition systems with manufacturing execution systems technology? Sam address the overlaps and differences between the two technologies and how their functions can be combined, with an example of a manufacturer that has already done it.
Wed, 4/21 11:49AM • 19:52
systems, skater, scada systems, combining, MES, add, layer, spc, replacing, production, benefit, supervisory, equipment, mes, business, technologies, real-time
Welcome to the Automation World get your questions answered podcast where we connect with industry experts to get the answers you need about industrial automation technologies. I'm David Greenfield, Director of content for Automation World. And the question we'll be answering in this episode is can MES replace SCADA? And joining me today to answer this question is Sam Russem, the director of smart manufacturing at Grantek, a system integration and engineering services company. So thanks for joining me here today, Sam.
Yeah, thanks a lot for having me, Dave. Happy to be here.
Yeah. All right. Great. So let's start with the differences between the two applications, you know, and this basically describes them, you know, what our manufacturing execution systems designed to do? And what our supervisory control and data acquisition systems designed to do?
Sure, well, I mean, first of all, you've already given me a leg up and that you you d acronym to each of those when you're exactly right. And yes, is more the execution systems and data that supervisory control side. Um, both of those systems are pieces of software tools are designed to perform a lot of different functions, right. So MES is going to do things like manage your production orders and the relevant data to them, analyze some of your raw production data and turn that into more useful management information like track and trace information or kind of summarizing raw data into performance KPIs, and it also needs to communicate in real time to your SCADA systems. But it also needs to be able to work transactionally with some of your business and ERP systems. So those are some of the main things that define me. Yes. And when you flip over to the SCADA side, skaters are really defined by a, like their ability just to connect to plant floor equipment, particularly PLCs, intelligent sensors, other shop floor devices, SCADA layers, also where you're usually doing things like your historian and kind of your raw data records are often kept in that scale layer. But most importantly, you know, this is where you have those supervisory controls. This is what lets your your human operators see what is happening with the plant floor equipment, and help to control it. So usually, you might have saying, Hmm, I that, that is associated to a specific piece of equipment of screen right there at the machine. But a scaler is going to give you more of that supervisory level of view of a line or a process area, or an entire factory, and kind of give you that overall summary of how you're performing.
since both systems, you know, we're obviously focused on, you know, equipment and device, you know, data acquisition, and you know, visualization, etc. What would you say are the key differences really, between the two types of systems?
Sure, so it's probably going to be hard for me to talk about that too much without getting into is a 95. A little bit. So let me do a just a quick explainer of ISA 95, or the Purdue model and how it kind of categorizes some of these solutions. So real quick, because I know a lot of listeners probably do know it. ISO 95 categorizes levels zero through four have some of your manufacturing systems and technologies right there. level zero is going to represent the physical production process. Level One is the control layer, the systems that are physically manipulating that process. Two is going to be your supervisory layer, which is more of that human interaction area where SCADA in historian lives. Level Three is your manufacturing operations management. And that's where MES and other management technologies like warehouse management or quality management systems could live. And then up at level four, you have your enterprise layer, which is going to include business systems like ERP, supply chain management and PLM systems. So I mentioned that to kind of help clarify some of the differences between those two, why do we say that MES is a level three manufacturing operations management system, and that SCADA is a level two supervisory system, right? So as you go through different layers of that hierarchy, the higher the number, the longer the time span of the data that you are working with, right? So when you're talking about level zero, the physical production process that happens in real world time, things are really happening, right? You're talking all the way up at level four business systems. Those are usually thinking in terms of weeks and quarters. and things like that, right, like financial levels. So a SCADA system at level two needs to be able to communicate a lot faster with PLCs. So it can communicate at these sub second rates. But then it presents a lot of that information, it kind of more of a human speed, it's kind of more of a second to minute mark. Because, again, it is a human operations interface. So it kind of needs to work at the speed of people. And MES is going to work on a slightly longer time scale, it is not usually going to be getting into sub second level control data, it's more talking about hours, or shifts, or sometimes days or weeks. And it also needs to be able to work transactionally with those business systems, right? So a SCADA to a PLC, that's working really fast in real time, they're in constant communications, the SCADA and MES layers are fairly regularly updating, but that MES when it goes to a business system, like an ERP, it's only pulling down data, you know, maybe once a shift or, you know, it's it's occasional transactions as opposed to a constant feedback loop between those two systems. So, MES is are going to do that a lot. skaters usually are going to work a lot more in in real time. Also maybe mentioned that does also kind of affect the protocols that each of those systems talk, right. So SCADA need to be interfacing with a lot of industrial protocols like OPC, or Ethernet IP or Modbus, things like that. Whereas MES actually has probably an even wider range of communication protocols, it needs to support because it needs to talk to SCADA systems, which is usually going to be through OPC, or database connections. But it also usually needs to go out to your business systems, which means going through the firewall using web services calls using kind of other protocols to inter interface those business systems. Okay,
thanks for that really good description that I think it all helps to, to almost visualize the the Perdue model stack there to see where things fit, and you know, why things reside in certain areas versus others? And, you know, it is one of the things you mentioned there, you know, so basically, you know, the the main differences we're talking about, I mean, there are a number of them, but some key differences are, you know, the real time nature of SCADA versus the more transactional level of operation of an MES system, and then the protocols as well, that you mentioned there. So given these very distinct differences, are there overlaps between these two systems that would make it even possible to even consider replacing SCADA with it? Yes. Yeah,
there are for sure. And, you know, yeah, this is a podcast double topic for a reason is that there are people that are out there looking into this and considering it right, and there's even a term for it, it's occasionally called a flattening the stack, right actually to try to elicit that image of crunching the Purdue model a little bit and combining that level two with three into one MES and SCADA layer. So yeah, I have people doing it. Um, and again, it would be debatable if you're kind of using bringing skater functions into an MES or an MES into a SCADA those lines start to get a little bit blurred. But as far as the overlaps between the two, I'm one of the things that really stands out to me is that they are both human interfacing systems that you do need to have a human user interface interface and the user experience that goes along with this stuff. So if you can present your control layer and your management layer in a similar platform in a similar way, where it's kind of seamless between those two functions, there's definitely an opportunity to streamline that human interface. They're also of course, both managing your production assets, right? They're just usually concerned about that at different scales, and maybe worried about different things where the SCADA is more worried about specific controls, turning equipment on or off. And MES is more supervisory and concerned with results. Like for example, if we're thinking about in terms of data, let's think about the temperature of a batch tank, right? a SCADA wants to know the temperature of a batch tag, all the time it wants to monitor monitor that every second because if it starts to drift in the bad direction, the SCADA is going to be where you're going to issue your correction to that and try to bring that temperature back into control. The MES is going to care about that temperature the tank too, but it probably only cares if you actually went out of spec and it needs to know an exception for future quality review or something like that. So again, SCADA is more focused on the the second the second real time control level, MES is going to be more concerned supervisory and managing exceptions, but they are still connecting to the same type of data.
Right. Thanks for that clarification. So for The end users operational perspectives, what would be the benefit of replacing SCADA with MES.
Well, everyone's favorite benefit for these things is always reduction of costs, right. So of course, I think that the main driver really is this idea of, I can reduce my license costs, my hardware overhead, things like that, if I can get, you know, multiple birds with one stone in combining my MES in a SCADA. So yeah, there's definitely opportunities, reduce the licensing, reduce the support costs, and reduce the hardware required to come require to run both of those systems. The other benefit that I see a lot is around just reducing the amount of screens and process complexity that you're asking your your operators to deal with on a daily basis, right, you walk up to these machines, sometimes when there's five different screens just to run a single piece of equipment, right, there's an HMI for one piece, and HMI for my SCADA, downtime entry system there ERP terminal, there could be all sorts of stuff out there. So anytime you have an opportunity to streamline operations, bring things to a single control point, make sure that people don't need to be monitoring multiple physical locations kind of in screens to get the information they need to do their job. huge benefits to things like that. Now, the other thing that does kind of come up in this, that, you know, because maybe a little bit of a word of caution, as you're thinking about combining some of these systems is if you're talking about saving in terms of licensing and hardware, this MES and SCADA together can ask a lot of compute power of your system as both a SCADA and an MES on their own can be pretty heavyweight systems. So putting them all together, you're kind of making a super system and you need to make sure that you have the physical computer power to monitor that and to maintain at the district level, especially since again, that SCADA working in real time, it becomes absolutely mission critical. You don't want slow network speeds your OS to affect your ability to actually control your process.
So you know, I guess, since we've looked at, you know, what the benefits of replacing the word you, you noted that note of caution there around the physical computing power aspects of combining these, are there any other downsides to doing are you missing anything by replacing you SCADA with MES? Yes, that you know, might be a value to keep them separate.
I think a lot of that comes down to your your design and making sure that the the functions that you need of each of those systems are checked, regardless of what actual platform you're putting that function in, for example, you know, the, I feel like the the requirements of the SCADA are a little bit more clear. So you know, you need to be able to visualize equipment and it's real time state, you need to be able to issue commands and get that down to equipment, you need to monitor for alarms and alert operators with things will add a specifications. So as long as you're kind of listing out your your features and the things that you actually need of the software, and making sure that whatever platform you're selecting still checks all of those boxes, then then you should be fine. So you know, it is just kind of a matter of making sure that what you need is there. So it's not necessarily Yes. And MES can always do what your SCADA needs or a SCADA can always do what your MES needs. It's a matter of making sure that you have the right features and designing for that. Okay,
Thanks for clarifying that. So, you mentioned earlier that you are aware of manufacturers who have actually done this they've replaced skater with MES, or that you've worked with them. Can you describe to how they did it, how they went about that process of replacing the system or, you know, narrowing it down to just one?
Yeah, sure. So the the one that comes to mind for me, is a customer that did add, I would say they added their MES functions into an existing SCADA system. So in this plant, they had a pretty comprehensive scale layer that was connecting to all their PLC isn't was the main terminal that their line operators would use to run material through their process every day. Then they had a an MES initiative, and they ended up building that MES using the same SCADA platform and building it all directly within the same interface on the same platform. So they had a SCADA platform that did allow for this. So it was a SCADA that did have a set of MES modules that you could be adding into the platform to add things like OEE capabilities or SPC and production planning and things like that. So so they picked the right software that allowed for them to do this. They also you know, it was done with off the shelf modules. You could of course, program a lot of these SCADA systems from scratch. To do a lot of these meds functions, I wouldn't recommend it, I would prefer to go something off the shelf that's going to be more universally supportable. So they had a platform that didn't allow for them to do that. And they did. So over time, they would add more meds like functions into the SCADA system. So first, it was more around establishing a connection to their ERP to pull down production orders, and schedule those orders on a line right there in your scale system, then it was kind of around adding OEE performance monitoring and SPC quality checks within the system as well. So now that I know what material I'm running, I have critical UI data, like expected run rates and things like that. So I can log all of that again, within there combined, SCADA has platform, similarly, with SPC checks, right, you kind of need to know what product you're running to know what quality checks you need to run at what time. So having that all in one system works very well for that. And so, yeah, I definitely do think it was a benefit for this customer. To have all of this within one system, it was also very beneficial to take it in this kind of agile approach where you're you're starting with your schema and you're iterating. On that over time, we're going to add production scheduling, now we're going to add OEE. Now I'm going to add SPC. So managing your risk in each of those stages. I would say that, again, the only downside that we really saw was that the system got a bit bulkier, and there is a little bit of a risk to the business. Because as we continue to add more features, more lines, things like that, there might be a place down the road where we're going to need to split that into multiple servers to run it, at which point it might actually make sense to kind of split out your meds in the scheme of functions. Again, I'm not quite sure if we're going to get to that point. But we are looking at a horizon where it could make sense to split those again, and we'll see how it how it all works out.
Interesting. So since you since you've had this experience, you know working with a customer has actually done this but uh, you know, I guess the the proof is still in the pudding, as you said it, you know, it might still need to be broken back out later, it's not a done deal. And this is set forever as the way to do it. Would you have any particular advice for manufacturers who are considering replacing skater with a with MES? You know, things they based on your experience with this application, things they would want to be sure to do and things they should avoid doing to, you know, make this transition, you know, as effective as possible?
Yeah, I mean, so pretty much anytime we have a question, you know, around making these types of design decisions, I always, my first reaction is going to be to bring that back to some of your business goals and what you're really trying to accomplish, right. So a lot of time, what you're going to be balancing in terms of combining your skate and MES is going to be the benefits of reduced license cost and complexity, with probably some increased risk and some more and more of a single point of failure, right, having that system go down, is really going to crash a lot of your operations, and it's gonna be pretty hard to recover from right. So are you okay, with that risk profile to the benefit of combining those two things? And there's a lot of ways that you can control for that risk as well. Do you install redundancy? Do you break it into smaller systems, there's a couple of options around that. Um, the main thing that I would recommend people do if they are looking at combining these are really for any MES project is to look at kind of more of an iterative design technique, you know, the to not bite off more than you can chew in one initiative, but to really think about this as an ongoing process, where you're going to be adding more functionality and getting more value out of the system over time. Just because if you do you know, an MES project can be very big on its own. A scatter project can be very big on its own. combining those two and trying to get it all done at once is a pretty monster project. So it'd be best to kind of think about what is that minimum viable product? Can we start at something that you know, was just going to get us up and running? And what are our options to add to that over time to eventually get to the system that looking for?
Well, thank you very much for joining me for this podcast, Sam. I do appreciate it. And thanks Of course to all of our listeners. And of course please keep watching this space for more installments of Automation World get your questions answered. And remember to visit our website at www dot Automation World calm to stay on top of the latest industrial automation technology insights, trends and news