Ignition: The New Enterprise Connection Platform48 min video / 36 minute read
President & CTO
Cirrus Link Solutions
The quest for greater productivity and reduced costs is driving market forces and investments into new projects trying to combat today’s challenges from the supply chain, labor, and inflation. Learn how Ignition has advanced from the “New SCADA Platform'' to become the standard tool for OT-to-IT Enterprise Digital Transformation. The session will discuss and demonstrate how Ignition with MQTT/Sparkplug is the “Swiss Army knife” Digital Transformation platform from the edge to the cloud to achieve these goals. Get your Enterprise ready to Xperience and Xplore the serendipitous nature of your OT data!
Mara Pillott: Hi everyone, welcome. My name is Mara Pillott. I am an Application Engineering Manager. I'm in our Sales Engineering Division at Inductive Automation. I'm your moderator for today. So welcome to today's session. This is “Ignition: The New Enterprise Connection Platform.” So to start things off I’m gonna introduce our speaker. This is Arlen Nipper, he is the President and CTO of Cirrus Link Solutions. Arlen has been designing embedded computer hardware software solutions for 45 years. In 1998, he co-invented MQTT, this is a widely adopted IIoT protocol you might have heard of that's been used across industries, it's designed to optimize and make use of data for both OT and IT. Arlen is the Co-Founder and President and CTO of Cirrus Link Solutions. In recent years, he has focused on how the Internet of Things affects the entire ecosystem of hardware, software, security, infrastructure, IT, and people. So please help me welcome Arlen Nipper.
Arlen Nipper: Thank you.
Arlen Nipper: Okay, thanks everybody. It's great to be back in person, I've seen you all for the last two years on Zoom meetings and Webex, and so I thought I'd bring the troll doll that always sits behind me on my desk in front of the demo equipment for doing the demo. So it's great, I'm glad we can have this opportunity to get together and really talk about the... Over the last two years, the realization that Ignition for Cirrus Link, we're seeing it with a lot of customers as really the new enterprise connection platform. Just a little bit about Cirrus Link. We were founded in 2012, and we really wanted to look at providing MQTT-centric software for the industrial market, as you all probably know, Andy [Stanford-Clark] and I invented MQTT on a project for Phillips 66. So it really was a OT-centric technology that we invented, but it kind of got adopted by IT, and OT kinda forgot about it, and we wanted the opportunity to start a company where we could go in and take the expertise that we had in applying MQTT to OT systems, and be able to productize that and bring it back into the market sector that it was intended for in the first place.
Arlen Nipper: Now, it's been an interesting evolution. So in 2015, I showed up at ICC, I think it was... Cirrus Link had one table out here, and I think across the hall there was Kepware booth and that was basically it. And I came into the auditorium across the road there, across the hallway, and I did the first ever MQTT into Ignition demo. We had MQTT Engine and we connected 300 or 400 PLCs, hundreds of thousands of tags, and that was pretty cool. That was great. We were able to get that in, and that was the first introduction and we started getting it out into the market and we got feedback from customers. So that, in 2016, we started to have these tenets of, well, if we were gonna do IIoT and Digital Transformation, what were the steps to Digital Transformation? And number one is decouple wherever possible, we've gotta look at connecting devices to infrastructure and not to applications, because that is where you kind of lose the ability to be very serendipitous with the data, is if you have a protocol hard connected to a PLC and you're polling it.
Arlen Nipper: Then you can't get that information out to other applications, so decouple, think of connecting devices to infrastructure, applications to infrastructure, and now you've got this message middleware or setup, if you will, of being able to publish data and you don't care who subscribes to it, and you can subscribe to data and you don't care who published it. And then step two, we said, is be able to demonstrate a superior OT solution, first and foremost. It's gotta be better, faster, more secure, more scalable, more available, because if we couldn't demonstrate that with MQTT infrastructure, nobody was ever gonna migrate to that infrastructure, and we would be talking about IIoT for the next decade.
Arlen Nipper: Then in 2017, we basically came out with some more modules, as Don [Pearson] showed you this morning, there was this notion now of Ignition Edge, since Ignition could run on just about any compute platform. We came out with the Ignition Edge, we started getting requests to get information into cloud applications, so we came up with the AWS Injector Module and the Microsoft Azure Injector Module. So we could take Ignition tags and get those into data lakes, basically with no programming just by connecting to their services and enabling these injector modules.
Arlen Nipper: Now notice number three, or number two, I'm sorry, we added another tenet. So, now we've got this notion of decouple, but number two, you should be able to provide a single source of truth for all your process variables. So, I've got a pressure in the field and I've got... We know that it's zero to 100 PSI, we know it's engineering high, we know it's engineering low, we can put all of that in the context in Ignition tag, so that should become our single source of truth all the way through the system. So that we're not taking a register from the edge up through another layer and then having a human being edit that Modbus register. Hope he gets it right. Given engineering units, given engineering ranges, given alarm values, we want the single source of truth to go as far out in the edge of the network as we can, so that became our second tenet, and then of course, number three, still be able to provide a superior OT solution.
Arlen Nipper: 2018. We came out with the first version of Sparkplug. Now, up until that point, and I think probably everyone's heard me say is that the great thing about MQTT is you can publish anything you want on any topic. The bad thing is you can publish anything you want on any topic, and to get some interoperability, we started taking the experience that Cirrus Link had with the 90-plus man years out there implementing systems and take best practices, lessons learned, what to do, what not to do, and start putting together a specification that says, if you wanna use MQTT in an industrial environment, Sparkplug is a good place to start if you wanna start looking at that. At the same time, we were doing a lot of work with the oil and gas companies, and to Colby [Clegg] and Carl [Gould]'s point this morning, the customers started to ask, “Can we get the EFM protocols actually running on the Ignition Gateway or Ignition Edge?” So we came out with the first EFM module that we had developed for the Emerson ROC Flow computers.
Arlen Nipper: At the same time, we started a working group with the Eclipse Foundation called Tahu, where this open-source Sparkplug spec could live and people could go out and access that. 2019, we still had the same three tenets, but now we're talking about... Let's kind of look at Ignition, and one of the advantages is that it's really tools on platforms, not coding on operating systems. So, with the Ignition tooling, we could very rapidly go develop an application and then be able to deploy that, so it's tools on platforms, not coding on operating systems. And the Ignition-based Digital Transformation uniquely addresses the fundamental issue that current OT infrastructures, legacy infrastructures, if you will, aren't conducive to the requirements of Digital Transformation. Notice I haven't said, let's go up in the cloud and go IT down to OT, what I'm saying is, let's look at the OT infrastructure that we've got, let's get that modernized and let's get that to the point where we can then scale from there, so it's not IT down to OT, it's OT up to IT.
Arlen Nipper: 2020. This is where we started getting the notion of putting a bunch of tags into a data lake, the data lake becomes a data swamp. Because, what we were finding out is that customers were spending a lot of their time writing code on the backend of the data lake to take the process variables out of the data lake and put them back together. Guess what? Like we had them in the UDT down on Ignition to begin with. So they were spending man months, man years, basically rewriting what we already knew at the edge. Now, at the same time, Amazon came out with a very unique service, up until that point, there was really no SCADA or OT notion of how to represent data in the cloud, but they came out with AWS SiteWise and SiteWise lets you create a model in the cloud and instantiate that model to create your asset, and then that asset has measurements associated with it. So, guess what? That looks just like an Ignition UDT.
Arlen Nipper: So why don't we go all the way to the edge and publish that UDT over MQTT using Sparkplug, so that we still maintain that single source of truth. And now notice tenet number two has changed a little bit, so now we're saying, provide a single source of truth for all models, assets, and process variables. So, we keep evolving our tenets on what's the best practices to get out there and get a system put together that's ready to plug in the cloud infrastructure. So in 2020, we got to demonstrate that we could take a UDT created in Ignition and get it automatically with no programming required all the way as a model instantiated in Amazon SiteWise service.
Arlen Nipper: 2021. Again, we kind of, the same notions we’re starting, now we've added another aspect to Ignition, so I think it was ConocoPhillips, Aera Energy, they're going, "Arlen, this is great. You can publish commands out to the field, but we want to actually secure those commands." So now with MQTT Engine and Transmission, we got the ability to actually look to see the person that sent the command out to Ignition Edge validate that before that command goes through and then also be able to audit log that command as well. And I kinda had the notion at that time, MQTT: invented for SCADA, adopted by IT, solving Digital Transformation today.
Arlen Nipper: So that takes us to 2022. ICC X. Oh I forgot to mention, this is also Cirrus Link's 10th anniversary. We've been in business for 10 years now, so that kind of lined up quite nicely. So the enterprise connection platform is... What is the definition of a platform? A platform is a set of software and a surrounding ecosystem of resources that help you grow your business. A platform enables growth through connection, its value not only from its own features, but from the ability to connect to external tools, teams, data, and processes.
Arlen Nipper: And believe me that when I wasn't synced up with Colby and Carl and Travis [Cox] and Kevin [McClusky] when they talked this morning, but that's exactly what we're talking about. We're realizing that Ignition, yes, it is a SCADA system, it is HMI, but it is a really cool tool that you can use for a lot of things. So, we're literally seeing customers go out in particular markets, so one of them is an oil tanker vessel control system. Well, there's companies that do black box applications that control vessel, but the customer understood that if you buy an off-the-shelf package like that, they typically tend to be black boxes, your improvements don't come, you can't ask... You can't change it, you got what you got, you probably gotta live with it. But, customers are starting to realize, especially the large enterprise customers, that if they go build that particular solution using the Ignition platform, now they know how it works, but they can extend it and they can make it better and they can make it better and they can make it better, and that's where Ignition, the notion of start. Yes, it does SCADA, it does connectivity to PLCs and RTUs, it does MQTT, it does SQL database, but in its entirety, it's more than that, it is a solution architecture that lets us connect all of that information and do our Digital Transformation.
Arlen Nipper: Now, the result of that exercise that I just went through those seven years of ICC. What was the result of that? If you looked at our pie chart here in 2017, it was probably... And we came from the oil industry, so we probably were 80%, 90% oil and gas. But you can see here that we've expanded greatly across all the other industry sectors as well, with manufacturing being the biggest growth sector that we're seeing here and recently.
Arlen Nipper: The other thing we've done is we've driven awareness, we've worked very hard in getting awareness of MQTT of donating the Sparkplug specification to the Eclipse Foundation, and we keep improving that, and I'm gonna have Wes [Johnson] out here in a little bit just to give everybody an update on where we're at with the Sparkplug specification and some of the advantages that we're gonna get out of that going forward. But now, what we're seeing every day is there another announcement of an OEM device, a PLC, a sensor manufacturer, the support Sparkplug, MQTT Sparkplug natively out of the box. And again, that is plug into a broker, plug and play autodiscovery. We've also seen this notion within the manufacturing sector of this whole thing called the Unified Namespace, in other words, what we're doing with MQTT is we're publishing information on topics, those topics can create a hierarchical namespace, so that now you can start doing naming within your factory so that you come up with this notion of a Unified Namespace, and going back to...
Arlen Nipper: I still remember the IIoT Manager with TE Connectivity, Carl Voss, and he said, "Arlen, what my vision is for my factories is I should be able to walk into the factory with my laptop with proper permission, plug into the network, and learn everything about that factory in a few seconds." And that's really what we're talking about with Digital Transformation.
Arlen Nipper: Now, it's interesting, Don and I, we have discussions with customers and with analysts and with things like this, and a lot of the pushback we got this year is that... “Well, Arlen, Don, SCADA is not IIoT. If you have SCADA in the product then that can't be IIoT,” because the proposition is that we've got all this cool cloud tooling and we can do all these cool things in the cloud, it looks simple. But when somebody like myself, Travis, probably a lot of you go, "Okay, that's really cool. I see all that capability you've got up there in the cloud." I'm gonna raise my hand, "How do you get that?" "Oh Arlen, it's easy, we'll just add some tooling and connect it. Done." That's fine. Let's look at that little blue line there. Okay, there's my blue line, and all I've gotta do is start pulling in some cloud tooling, I gotta get some connectivity, and there's an API for that, and I got write some configuration, and then that's gonna go to some Python scripting that's gonna go into my data lake and from the data lake I have to pull it back out of the data lake, put it back together like it was originally at the end, and I can get that in.
Arlen Nipper: Now, this is great, and I'm not knocking any of this, but what I've just done here is I have hard coded myself to one cloud provider, and if I want to move, I gotta start all over with all of this. So, what I wanna propose is you, without SCADA, without SCADA, without OT technology, you cannot do IIoT. So the reality is, it's difficult, the reality is we need to go into our legacy OT infrastructures, and we need to make those infrastructures ready to plug into the cloud. We need to give them modern messaging technology. We need to do our decoupling, provide a single source of truth, demonstrate a superior OT solution, first and foremost, and once that's done, once we've got the infrastructure, once we've got our Unified Namespace, once we've got that up and running, then we're ready to plug into the cloud.
Arlen Nipper: Now, to get ready for this, I was watching some of the TED Talk tutorials. It was kind of good. So this is the point where I pause. You put your hands together like this. You look out over the audience, you get a drink of water, and then you put a picture up. Black holes, what do black holes have to do with automation technology? What do black holes have to do with automation technology? Black hole, another black hole.
Arlen Nipper: Using the Ignition platform to access black holes. So one of the things Don mentioned, we did the state of Indiana program. And I'm not an integrator, but I play one on TV, and I got to go to 10, 15 different customers going around these factories in Indiana. And as I'm going through them, it's kind of getting an epiphany. I probably should have known this and probably a lot of you already know this, but as I'm going through the factory, I said, "We're trying to do energy monitoring in the factory, and do you have any meters already in your transformers?" "No, no meters." So walking through, I go, "Well, that's a Siemens meter, talks Modbus." Oh, we can't use that, it's hooked to our BACnet IP router." "Well, what's the BACnet IP router hooked to?" "Oh, nothing." And so my point is, is that we've got in... With all of the equipment is... So we're talking about Digital Transformation and Industry 4.0, but I really don't think we're talking about adding more sensors, we're talking about getting to the sensors and to the equipment that we've got today, that we're just leaving stranded on the floor.
Arlen Nipper: We have to put cowboy boots to walk through all the data that's laying on the floor. So one of the opportunities I just wanted to point this out is that whether it's... Now we've got BACnet drivers for Ignition, we've got the 61850 drivers, we've got really cool tutorial on our Cirrus Link website on using some really cool Moxa Ethernet or Modbus TCP to RS-45, RS-422, RS-232, to go get some of that stranded Modbus data that's out there. And it's the, I-don't-know-what-I-don't-know device. And what we found with the engagement in Indiana is, "Well, Mr. Customer, go ask your OEM what protocol it talks." And we had one customer, he had a 600-ton press, it cost I don't know how many million dollars. And they go to the manufacturer and go, "Can we get data off of this?" "No." It went up another layer, "Can we get data off of this?" "Well, we have a package that we can sell you." And go up another layer, "Can we get data off of this?" And then finally three months later, "Oh yeah, it talks Modbus, and here's the memory map."
Arlen Nipper: So, it's just asking those questions and getting engaged with the customers. I think this Indiana program, we're finding a lot of people really wanna get out there, use Ignition and modernize their factory going forward. So Ignition is the tool for enterprise connectivity. And you've probably seen this drawing somewhat within the marketing, but what we're seeing is the Ignition platform is that middleware that you need for the connectivity that we've got, and all the cool stuff we can do with UDTs, with all of the tags, and the ability to get that out and going back to my notion of a single source of truth for your models, your assets, and your measurements. So today, we are announcing a new product. Like I said, two years ago, we did the AWS IoT Bridge, today we are now announcing the IoT bridge for Azure Digital Twin. So basically this is going to give you the ability to create a UDT and literally point that at an MQTT broker with IoT Bridge for Azure, and that will automatically take and create the digital twin and then take that on into Explorer. So this is the topology of that. So in my Ignition UDT, I create my model, I instantiate that, create the twin, and then the real-time data that goes into Azure Digital Twin Explorer through Event Hub and to the Azure Data Explorer.
Arlen Nipper: So really excited about this. I think it's frustrating, you do the AWS demo and they go, "Oh, can you do this on Azure?" And you do the Azure demo. "Can you do this on AWS?" So I think we're gonna see this, but now what we can do is we can create models in either one, we can compare them. And that's what we'll be demoing here in a little bit. The other thing we wanted to announce that's new is that we were approached by Quorum Software, and they had bought Flow-Cal. So for all of our oil and gas customers, they kind of done a survey and they go, "Well, there's a lot of Ignition out there." So they called Kurt and they said, "Well, why aren't you guys publishing CFX files?" And then we said, "Well, you never told us how to do that." So now we've got the ability to publish CFX files from our EFM drivers, we can use the transmission file publishing to be able to do that. We can drop them directly onto the operating system and you can come and get them, but at least now we've got that. So probably into this month, in October, we'll have all the Emerson ROC, the ROC and ROC Plus, and then Q4, the ABB Totalflow.
Arlen Nipper: So really that last piece we needed to be able to wrap up everything that we needed to be able to do and to support EFM. So, we're all about OT. This is the Internet of Things, we do automation, we're integrators, so this is all about OT. But let's take and look at it at the next level of, "Okay, well, it could be a enterprise connection platform, and then let's take... " Now we're showing... Again, show the stakeholders that this is a better, faster, more secure, more scalable, more adaptable SCADA system, first and foremost. And then finally, let's look at the third leg of that is getting that up into IT and whatever you wanna call it, Digital Transformation Industry 4.0. So at this point, I'm going to jump into the demo.
Arlen Nipper: Alright, for this demo I've got the left side. So what I've got there is I have got, I believe, hundreds of Ignition Edge and Ignition Gateways, some real, some simulated. And we're gonna discover all of those by connecting MQTT Engine into that infrastructure and discovering everything about those. Now those, what you're gonna see, unlike the demo that I did, the first-ever demo, instead of just a bunch of tags screaming across, we're gonna see that, discover all of those models, all those UDTs that we built in all the simulated and real edge devices we have out in the field. Then we're gonna take that and then we're gonna look in the Ignition platform of what we can do with something like a smart factory. And then we're gonna take that entire smart factory and we're gonna reproduce it in the digital twins on Amazon and on AWS. So, the first thing I'm gonna do here is notice that I don't have any UDTs and in my edge nodes, I don't know about any devices out in the field. So I'm gonna click on MQTT Engine and notice up here, we're gonna keep track of the total number of edge devices, total number of tags. We have 46 metric tags that just happened to be sitting in the engine tag provider. And we have zero UDTs that we know about. So we're gonna enable Engine right now.
Arlen Nipper: And now we know about 377 edge devices, 25,000 tags, and 1,900 UDTs. Let's jump back over to our UDT definitions and notice that all of our Ignition infrastructure at the edge providing us a single source of truth, we're now publishing those models up. So now we can do really cool stuff with that. So if we go back over to our tags, and notice that we have a camera group, we'll go into that real quick. We've got smart building group, with a campus, with a facility with a UDT representing an area that has a UDT of a chiller, that's got the compressor and the cooler. So now I should be able to publish that model up and be able to take advantage of that in the cloud. You can see here from an oil and gas perspective, I'll go in here and go into my ROC devices, ROC Edge, FloBoss 107, and I've got a meter configuration. So let's go in, to the designer and of course, as everybody knows. Now, you'll notice I'm still using Vision because you guys are gonna see all kinds of cool stuff in Perspective. And I would be embarrassed if I tried to do this in Perspective. So I'm gonna use Vision.
Arlen Nipper: And I'm gonna go into run mode. And what I just did was I took a very complex EFM device in, and with all of the cryptic configuration and all the cryptic tag names for that ROC Flow computer. And I just did a form that said, "Oh look, this is my configuration." Now remember what I was talking about, one of my pet peeves is enumeration, is that I'll bet all of the oil and gas customers know that the orifice material for my Flow computer is in enumeration 0, 1, 2, 3, 4, 5, 6. And I don't know what the heck it is for any brand. But notice up here, with the UDT, I was able to take that representation and say, well, that's Monel. Well, now that I've built a dashboard, I can control that and I don't want it to... I want it to be carbon. Now I just published that command up, it went into Portland, Oregon where the ROC Flow computer and that I have now changed that configuration of carbon. So again, these are just... They're basic advantages of how we're building out our UDTs.
Arlen Nipper: The other thing is that, with Ignition, we now, in Engine and Transmission, we can now publish files. So you pick any directory you want, where Transmission is setting, set up a couple of clicks and it'll take that file and automatically publish it to Engine. Engine will drop it down into the operating system where the Ignition Gateway is running. What I did here was I literally took an image component from the designer, I dropped it on the screen, I went into Web Dev and opened up /OPT/images and it said I had got a camera roll. And for everybody that's publishing a new JPEG, we're just rolling through and going up and using WebDev to go down and pick up that new picture. So this is just kind of a cool demo of what you could do if you could publish files from the edge and be able to manage them and control them. But let's go back to our demo topology and let's finally go to our smart factory and let's take one of the smart factories that we discovered and just take a look at what we've got. So we've got these UDTs of bunkers and dryers and conveyors and extruders.
Arlen Nipper: So I'm gonna take those. And kind of like what we're doing for the state of Indiana, I want to take maybe an Opto KYZ, and I wanna measure the power coming into my factory, the total power. That's pretty cool. And then I want to go over and let's grab an extruder, 'cause that's what this factory does. And then that extruder is gonna feed parts into a bunker. And as the parts come out of the bunker, we're gonna feed those into a... Whoops, we're gonna feed those into a CO2 dryer.
Arlen Nipper: Pretty cool. And then maybe measure the power that our extruder is using. And now I've got all of this operational data. Single source of truth. It's down here and I wanna take... And I wanna get this up into the cloud. Now, of you that have looked at what you have in so far as the learning curve on being able to get your connection to the cloud, get it established, go create a model. Just all of the infrastructure. There is man year’s worth of learning to be able to take this and throw it all up in the cloud and actually get some digital twins. So what we're gonna do, we're gonna go back to our demo topology, and Wes, you wanna come out?
Arlen Nipper: So, can we go to split screen? Cool. Alright. So, on the right, as soon as I hit start transmission, what's gonna happen is we're gonna make one connection to an MQTT server where we've got IoT Bridge for Azure tide going north. And that's gonna take that smart factory and create all of those UDT. All those models in Azure Digital Twin. And using that same data stream, not another trend, but using the same data stream, we're gonna have IoT Bridge for SiteWise subscribing to that and building models and AWS SiteWise. So I'll be displaying how long it's gonna take on the right screen, and Wes is in the Azure Digital Twin, and he'll be showing how long it takes on the left screen.
Arlen Nipper: So let's see. Let me get my browser up here. Okay, and before... Yeah, let's do that. No magic under the covers. So for me, here's my SiteWise service and I go in here and the only models I've got are the metric models that we built. So our IoT bridges actually build informational models, so we can tell you, the customer, what's going on inside of our bridge. How many devices are connected, how many tags are we updating into the digital twin, how many digital twins did we create, things like that. And that's all I've got and, Wes?
Wes Johnson: Yeah, so it's exactly the same for Azure Digital Twins. So the models here on the left. So the models are basically in Ignition, what is a UDT definition, and the instances are in the middle here. And so again, these are just informational and there's not much information because there's no actual edge nodes connected in publishing data.
Arlen Nipper: Okay, so what I'm gonna do is I wanna pop over here and I'm going to enable... I'm gonna start transmission and then we'll just watch both screens. And I'll let Wes start with the Azure Digital Twins and then I'll show the Amazon Digital Twins. So, we'll start right now?
Wes Johnson: So, I have to keep refreshing because there is no... Well, I have to poll it basically. So there...
Arlen Nipper: It's not using MQTT.
Wes Johnson: Right.
Wes Johnson: So on the left side here, those are all the UDT definitions. So they've already arrived and those models were automatically created. So the actual IoT Bridge Module took all of the incoming UDTs that are in a standard Sparkplug format and created the DTDL which in turn created the actual models in ADT. And so, then I can run a query here on the instances. And so, those are all of the individual instances. And if I find one, I can show an example.
Wes Johnson: So, this UI is not the best. They show up in different orders all the time. And when you actually scale this out, the UI becomes pretty useless. But there's the die inlet pressure for the extruder and the current value. And so we also had to build in things like the notion of quality because ADT doesn't already inherently have a notion of quality. So, all of this is here now. Data is flowing. So one of the other things that we provide is the ability to take all of the data that flows into ADT and then take that into ADX, which is basically a historian in Azure. And so, with ADX we can go in and then query that data afterwards. So, as soon as Arlen started, we can have all that data flow in. And so, this is just a query tool that's built into Azure. And so this can sometimes be a little bit slow to update but we have data now. So, that was just a query of everything that's in that table. As soon as it shows me everything, I'm gonna have a little cheat sheet here with an individual query. So now I can go and fetch that same extruder die inlet pressure.
Wes Johnson: Right. And so there's all of those values as they've come in since the transmission was starting. So in addition, they have a little dashboard. So it's nothing like Perspective or even Vision. But we can also create a query attached to a graph. And that appears here too. So, all the data directly from Ignition, from UDTs with no coding. Because we do all the mappings between what a UDT is. We map that to a Sparkplug template which in turn then gets mapped to an Azure Digital Twin model and digital twin.
Arlen Nipper: So, I wanna stress, what you just saw, what you just saw, there's no platform on the planet that will do what we just showed you. In other words, create a UDT in Ignition, connect, make an MQTT connection, publish the models and they are built. So I'll show you over on the right, what we learned. We had all of these models, and if I go into my extruder.
Arlen Nipper: And the cool thing is, is we're also creating the asset models here as well. So we have an attribute definition, and guess what? The attribute definition, the location ID, the asset ID, the asset serial number, those were the properties in the UDT. So we literally take the properties in the UDT, you can start putting those in as well, and they get published up and recreated in the model as well. So if I go over to the actual instance of that extruder model and look at it, you'll see that my asset ID is Wile E. Coyote, the serial number's BR549 and the location's Casey office, which I need to change that Stillwater, Oklahoma now, but anyway... So we were literally able to build those models with zero coding and minimal cloud expertise to be able to get that information up there. And the same on the SiteWise side is I can come into the portal and just to show that everything is happening. We come in here and then look literally into the time-series database from when we connected just a few minutes ago.
Arlen Nipper: Find the extruder, which is right here. And now we've got our die inlet pressure, our die outlet pressure, melt pressure. And if we go back to our Ignition dashboard and look at the smart factory, we can see that follows the trend chart as well. So, that's pretty much the end of the demo. I'll open it up for questions here in a little bit, but I just wanna make sure that what you just saw is really disruptive, I believe, in the industry because now can you imagine... I can, hey, today I wanna go to Amazon, I wanna do this, tomorrow I wanna go to Azure, now I'm gonna go back up here. That gives us the ability not to hard code ourself into any one cloud platform and still provide that single source of truth for what we're doing and we know on the Ignition platform.
Arlen Nipper: So, one final thing I wanted to do, I wanted... While I had Wes up here, is give us an update on the Sparkplug Working Group. And that's alright, I'll just, yeah, I'll leave it there. So just give us an update on where we're at with Sparkplug.
Wes Johnson: Alright, so I think most of you were already aware of Sparkplug and the work that's been happening there. So we gave the Eclipse Foundation a Sparkplug specification and trademark back, I think in 2018. And so, with that, we've been doing a lot of work to make it much more easily adopted. So one of the things that we wanted to do was to actually clean up the specification, make it a bit more spec-like, actually put assertions in and clean up the language, fix some of the issues that were in it. So most of that work is done, but in addition, through the Eclipse Foundation, we're really trying to formalize it and also work towards making Sparkplug a global standard, which that process is now happening. So with that, there are some additional things that we have to do in order to release the next version of the Sparkplug spec. So on Friday of last week, we actually just posted the first release candidate of the specification itself. It's now basically in code form, it's written in AsciiDoc, but with that, we can tie the assertions in the specification to the TCK. And so the TCK is there to validate people's implementations of Sparkplug.
Wes Johnson: So the idea is, if you implement Sparkplug, you can go to the TCK, actually get it fully validated, and so then everybody in the industry knows that that particular implementation of Sparkplug is good and compatible. That is getting very close toward, I think, 92% or 93% coverage of the assertions in the specification. And then finally, we have compatible implementations that are in Java through the Tahu Eclipse project. So if all goes well in October, this will all officially be released. We also have kicked off the process of getting ISO/IEC global standardization. And that process should be done in May of 2023, which means that Sparkplug at that point will be an official global standard.
Arlen Nipper: Great. Thanks, Wes.
Wes Johnson: Yep.
Arlen Nipper: Appreciate it. So again, making great progress, I know there's been a lot of committers on the Sparkplug committee that Wes works with, but Wes has been... He's been a champion on this. As a lot of you know, he's putting in a lot of his own time and effort to make the whole Sparkplug spec the way it is today and we really appreciate that. So with that, I'm at the end of time, I think we got time for a few questions, but other than that, that's my presentation. Are there any...
Arlen Nipper: Thank you.
Mara Pillott: Thank you Arlen and Wes. We do have time for maybe a question or two, there should be some mic runners out here for us.
Audience Member 1: Hi. So we can get that up into the cloud, into AWS or Azure, can we... Is that a one-way street at the moment, or can you use that as the historian to write back down for trending in Ignition historically as well?
Arlen Nipper: Well, we're working on a SiteWise Engine Module that would let us bring that back and take those discoveries from AWS or from Azure, and they'll be able to write those back down into Ignition. We purposely didn't leave it closed loop right there at that boundary, 'cause we really didn't want the cloud application being able to write straight back. So there's a lot of different techniques that we can use to take those discoveries out and get them back into the Ignition Gateway.
Audience Member 1: Yeah.
Arlen Nipper: So right now, not through transmission.
Audience Member 1: Sure, but not in terms of writing back just as a historical provider, so double-click and get a trend. Do we still have to get it from the SQLite database in that instance that it's going to in Ignition or could we potentially pull that historical trend view from the cloud to display in Ignition?
Arlen Nipper: That's what we're working on with the product called SiteWise Engine, it's not ready yet, but it is out there. We are thinking along those lines, yes.
Audience Member 1: And link to that, then do you have any scalability numbers on what you can get up into the cloud out of there at the moment?
Arlen Nipper: Right now we've got customers with millions of tags, where it scales beyond that, I really can't tell you. I think both Amazon and Microsoft Azure will tell you, these are early kinda entries into these digital twins and how many tags per second or milliseconds you're gonna put into them, but we're finding those out. And again, it's because of the Ignition community that we've got customers that are giving us really good feedback on that.
Audience Member 1: Okay, thank you.
Arlen Nipper: Yeah.
Mara Pillott: Sure have time for one more?
Arlen Nipper: Great, I did such a good job. There's no more questions.
Mara Pillott: I think you did.
Arlen Nipper: No, really, I really appreciate it. Again, Cirrus Link, we've been very successful, and this event is just awesome that we finally get to do it in person again. But it's from customer feedback that we get a lot of these ideas. We really appreciate the whole community, you guys are great to work with, all of you. Nate and Wes and Chad, and all of our support guys that you talk to. Kurt, that you talk to in sales, we just love working in this community, so thank you all.