Search
Close this search box.
Build Your Digital Infrastructure. Register for the ELEV8 2024 Event Today.
By Clarise Rautenbach
23 July 2021

Where’s My Data? Dataops and the Unified Namespace with Ignition, Canary and Flow

Share

Watch the video to learn more about the value of the unified namespace.

Bridging the OT/IT gap. How to democratize your data, build your data model and facilitate cooperation between departments. We’ll build this with Ignition, Canary and Flow.

SPEAKERS

Leonard Smit
Customer Success Manager
Flow Software
Jaco Markwat
Managing Director
Element 8

Transcript

00:00
Jaco
Good morning, everyone. I see we have a couple of people still joining, but we ready two minutes over. So we’re going to start off this morning. I hope you can see the screen clearly. I hope you can see us sitting in our office clearly. Maybe raise a hand if there’s any technical challenges. You would think after over a year of focusing on this kind of thing, we would have it perfect, but always a chance for gremlins. But as I said, my name is Jaco. Thank you very much for joining us for the session. We’re really looking forward to chatting with you today and maybe share some of our learnings and things that we’ve learned over the last little while that will maybe give you a little bit of better context around some of the things that you’ve been asking and wondering about.

00:43
Jaco
So the topic for today’s session is where’s my data? Maybe a very simplified question, but I think very often quite a common one. And we’re specifically talking around concepts around data ops, unified namespace. And we’re definitely going to give you a view of how we build that with ignition, canary and flow. Lenny, you want to introduce yourself?

01:06
Lenny
Yeah. Welcome everybody. My name is Lenny, customer success manager here. And after Jaco is going to go through the presentation, I’ll go. As I say, this is live radio, so I’m going to do a live little demo just showing some of the concepts of building out this unified namespace, and I’m going to do it with all three of the products. So I’m going to utilize ignition, canary and flow, push a bit of data around and see how we can use that data in one central space by utilizing this concept of unified namespace. And then we’ve got Tebe.

01:41
Jaco
Yes, I am Tebe. Like Lenny said, I am a customer.

01:45
Jaco
Success engineer, and I’ll be handling the.

01:48
Jaco
Chats during the session today.

01:51
Lenny
So yeah, please give Tebe a lot of work, a lot of questions, and he will answer those questions in the chat for us this afternoon or this morning.

01:58
Jaco
We don’t have the facility to all speak together. We would love for it to be one big conversation, but I think it might be a little bit hairy with the number of people we have online. But absolutely, if you have any questions, comments, please just type it in the chat, ask away, and we will make sure that we monitor that and give back to you. It is going to feel like a little bit of a monologue to us, so any comments or questions in the chat would be appreciated. So maybe just the agenda of what we want to get through today in just under an hour that we have, we’re not going to get into a great detail around any one of these topics. I think that’ll be absolutely impossible to do in an hour.

02:34
Jaco
And I think, again, what we want to do is maybe just share some of our learnings and understanding of a couple of these things, and we hope that will be helpful to you and valuable to you, combined with a little bit of a demo. Not a little bit of a demo. Quite an awesome demo that Lenny put together on how we build these things. We’re going to start off with maybe just digital transformation. What is a PowerPoint or a webinar in 2021 without talking about digital transformation, we’re not going to do that. We’re just going to focus a little bit on the why we’re even speaking about these topics today and how it ties in with digital transformation. We’re going to chat a little bit about IIoT industrial Internet of Things, the unified namespace data ops MQTT.

03:15
Jaco
And again, we’re going to finish that off with the demo and features. If you’re not familiar with some of these terms and terminology, 100% correct, you’re in the right place. We’re going to demystify some of that for you. Before we do that, though, we want to do a little bit of a quiz. So this is the first time we’re doing it. Please bear with us. Hopefully it works. If it’s a dismal failure, at least we tried. But we’d love to get some views from our audience, people on the call to get their views and opinions on some of these concepts and notions and things and what they mean. So we want to do a bit of a live quiz. It’s going to require you to, on your side, open up a web page from your phone or your laptop.

03:54
Jaco
But we do invite you to participate and I think the result is going to be fun to see and interesting. So I’m going to stop the sharing of this and I am going to share a different slide and it is.

04:10
Jaco
Going to be this one’s cool.

04:22
Jaco
So you should see my Internet explorer at the moment. And I am going to open up this nice and big so you can get the full view. Hopefully you see the screen that talks about how to join this quiz. So from your phone, from your laptop, please go to joinmyquiz.com. It’s all one word. It does look like spaces on the screen, but it’s actually one word. Joinmyquiz.com. Enter the game code on the screen that comes up once the page is loaded, which is eight double 3693. And you will then be required to enter your name. I see Tebe is on there already. See who else joins. Promise you it’ll be quick, it’ll be valuable, it’ll be fun. Thanks to my team leading the charge, Lenny, Tebe, Laura, all on there. Vanesh. Thank you. Vanesh. Very formal, full name and surnamesh. Very pwc of you.

05:20
Jaco
Vanesh. Fantastic. Zachariah.

05:23
Lenny
Awesome.

05:24
Jaco
Zachariah, thank you for joining. Jules, nice to see you online.

05:27
Jaco
Cool.

05:27
Jaco
We’ll give it about another 30 seconds or so just to give everybody a chance to get on there. Shrez. Fantastic. Rudy is on. Good. All right, maybe another 15 seconds. It’s going to be fun. Promise you don’t want to miss out. There is a prize. We’ve just decided there is a prize. We don’t know what the prize is yet, but there is one. We’ll let you know by the end of the call what the prize is for the winner. All right, so how it works is we’ll basically go through a series of questions. You follow on your screen, on your side, on your phone or on your laptop. You follow the instructions, you make your selection around the answer, and then at the end of that answer, it’ll give you an idea of where you are on the leaderboard.

06:14
Jaco
And we’ll have a bit of a recap at the end.

06:16
Jaco
Cool.

06:16
Jaco
Everybody ready? Let’s get going. All right, so the first question is going to come up on your screen in 1 second. Why do most digital transformation initiatives in the industry fail? So we have one answer already. You just tick the appropriate answer that you feel is best and you submit three answers. 14 seconds to go. You only have 30 seconds to answer. Maybe I should have mentioned that. Five people so far, six out of ten that are participating in the game. Fantastic. Four, three. Quickly make your selection and submit. We didn’t have three submitted. All right, perfect.

07:02
Jaco
Cool.

07:02
Jaco
So the leaderboard, you will see that on your screen. Now we have Q. I wonder who that is. 950 points.

07:10
Jaco
Good.

07:10
Jaco
Fantastic. Next question coming up.

07:12
Jaco
Now.

07:16
Jaco
I’ll read it out loud as well, but you should see it on your screen. ERP systems natively understand proprietary protocols. Tags from a plant floor. That’s a long sentence. Maybe I have to read that quietly on your side again and you’ll see the options there as true or false. Easy one to get to. Nine out of eleven people answered so far. 8 seconds to go. Storm missing two replies.

07:43
Jaco
Kind of disappointed.

07:44
Jaco
Only eleven people are playing the game. How often do you have to play a game?

07:50
Jaco
Right? Time’s up.

07:53
Jaco
Cool leaderboard so far. Q still in the lead. Thousand, 870 points. Lenny. Second, Tebi. It feels kind of unfair that you guys are participating, but you didn’t have a view of the questions at least?

08:04
Lenny
Just a reminder, the faster you answer, the more points you get.

08:07
Jaco
That’s a good point. I didn’t mention that. Fantastic.

08:09
Jaco
Next question.

08:10
Jaco
The data ecosystem is the foundation of the industrial Internet of things. The data ecosystem is the foundation of the industrial Internet of Things. True or false? Question. Very easy. Already have seven responses.

08:25
Jaco
Eight.

08:26
Jaco
And we have about 10 seconds. I think we may get to full House of answers before we get to the end of the timer. 7 seconds to go. If you haven’t selected yet, please do so. Three, two, one. All right, time’s up. And the correct answer is. Of course, I should probably tell what the correct answers are. Doesn’t give you a view at the end. Like I said, this is the first time we’re doing it. We thought we’d have a bit of fun. You’ll see that the correct answer is in fact true. It is the foundation for the industrial Internet of things. Next question. You can use any Scada as a data hub for the industrial Internet of Things. And once again, this is a true or false answer. You can use any Scada as a data hub in the industrial Internet of Things.

09:15
Jaco
Maybe this one’s open for a little bit of debate, but maybe by the end of the call, once we’ve been through the demo, you’ll understand why we’ve gone with the option that we have. 5 seconds to go. Nine out of twelve responses. Time’s up. All right, let’s look at. The correct answer is in fact false. It was a guess. It is a guess. Hey, ladies in the front, Vanesh. It is in fact false. You cannot use any scada as a data hub in Iot.

09:48
Jaco
Right?

09:48
Jaco
Next question. We only have. This is the last question, in fact. Question is, what is the primary feature of an industrial Internet of Things platform? The primary feature, is it edge driven reporting by exception, open architecture, or all of the above? It’s a bit of a loaded question, isn’t it?

10:09
Lenny
I’m not going to answer. Give someone else a chance.

10:14
Jaco
Cool. 10 seconds to go. Eight, nine out of twelve responses, 3 seconds to go. And I think time is up.

10:27
Jaco
Righty, fantastic.

10:28
Jaco
Let’s look at the leaderboard. And of course the correct answer for that question was all of the above, all of those things. And we have Vanesh, congratulations. That was really good going. I’m quite surprised. You’re definitely a strong finisher. Congratulations, Dinesh. So I think the objective of this was basically just to give you an idea of, or at least for us to get an idea of where some of the thinking is. Maybe for you to understand a little bit more around some of the concepts we’re going to speak a little bit about today, MQTT, and also gives us an indication of your understanding of these things. But that was the first time we tried that. I think it was good fun. Thank you very much for your participation. The twelve people that were brave enough to play the game, we really appreciate it.

11:12
Jaco
We can also share some of the results and questions of this with you afterwards. But congrats, Vanesh. Vanesh, you win lunch with us. You’re welcome to join us tomorrow. It’s pizza Friday. Swing around the office and have some lunch with us. Right. Let’s get back to the presentation. I’m going to reshare this. I hope it works again.

11:33
Jaco
Should, right?

11:36
Jaco
Hopefully you can see my slides again. Fantastic.

11:39
Jaco
Cool.

11:39
Jaco
So before we get into some of these definitions and things, we really want to keep it light. Not very technical, not very theoretical or academic. Maybe just to kick off with a couple of three letter acronyms. You will all agree that our industry is really riddled with three letter acronyms. We find them there is literally for everything. There is almost a three letter acronym, a TLA, before we have a full understanding of what it is. If you’re not familiar with any of these three letter acronyms, we’re going to try and explain as many of them as we can. We’re not going to cover all of them. But I think you often look at the TLA as our industry and you think, WTF? But don’t be worried. We’ve got you covered. We’re going to and some of these things for you before we get going.

12:23
Jaco
But again, if you have any questions, if we miss any of these, please let us know in the chat and we will definitely talk through or explain what they are that we’ve spoken about. All right, so we’re going to kick off the slide with two definitions of digital transformation. Again, we can probably spend an hour talking about digital transformation. I’m not even going to read these two definitions on the screen. I think the one is from Gartner and the other one is from accenture. But the reason we have this on the screen, and again, this is not a digital transformation session, is maybe to identify key similarities or at least a golden thread. Through all of these definitions of digital transformation, we try to narrow it into a couple of key words.

13:04
Jaco
Maybe you can see some of those words already, but we’ve highlighting them on the screen and essentially we’re talking about something that I know Vanesh speaks about very passionately about as well. When it comes to smart manufacturing is technology, process people and enterprise wide. Those are some of the, not some, those are keywords that you’ll find in most if not all definitions of digital transformation. And the reason why this is important is because this gives us context in terms of why the unified namespace and data Ops is important. And we’re obviously going to talk through each one of those things now. But at least you have an understanding of per definition of digital transformation. What are some of the key things that are absolutely imperative to reaching, or at least enabling digital transformation, technology, process people and enterprise wide.

13:52
Jaco
So please keep those four things in mind as we progress. So digital transformation is driving edge to enterprise and the industrial Internet of things. We’re not going to go into those details right now, but I think some of the benefits why it is driving that is on the screen again, there are probably experts around this that can probably explain this far better than what we can. But these are some of the very real and proven benefits that you get from digital transformation. You can follow any of the lighthouse reports as an example of very successful efforts around this by various manufacturing companies and businesses that have done this. But these are some of the benefits. And the gentleman on the screen that you see there, to give you a little bit of context, his name is Keith Weed. Lovely name.

14:42
Jaco
So Keith Weed was at this time that the picture was taken. He was the chief marketing officer of Unilever. And Keith Weed is a very celebrated and award winning marketer. And during an interview, he said that his objective as chief marketing officer, or the objective of Unilever at that time was to have a direct relationship with every one of their consumers. Every single one of their consumers. Now, not customers. They have a lot of customers, but they have millions of consumers. What that direct relationship requires is an end to end understanding of how every one of those consumers interface with their business and their product from a raw material state where they’ve actually bought that product.

15:24
Jaco
If there’s any possibility of a recall, the frequency of use, that was quite a bold vision that they had for Unilever, and that, according to him, hinged on their digital transformation efforts. And I’m sure that Unilever is not the only one of those companies there are many, I would say majority of the companies that are in that exact position right now where they understand the realized benefits or they understand the potential benefits of digital transformation. But how to make it happen is very often a little bit more tricky. And again, we can go into some of these slides probably for another couple of hours, but very often the proposition is that you have multiple applications, multiple systems, multiple even hardware. And all of those things from an it perspective, whether it’s cloud or on premise, has to interact with your ot data.

16:18
Jaco
It sounds really simple, it looks really simple. But the reality is that it’s not interfacing some of your plant floor ot things with it. And it data requirements is very often the reality is not that simple to do. And that’s exactly the reason why we’re speaking about the concept of the unified namespace today, as well as data ops, and where that fits in with the unified namespace. So hopefully that’s a little bit of context in terms of why the unified namespace and data ops is important. So some of the challenges we’re going to cover a little bit today is problems with bridging the Otit gap. I think we’ve alluded to that on the previous slide, the perception of needing to rip and replace as an only solution, and very prominently, data problems.

17:06
Jaco
80% of the challenges that people have Today is centered around data and data problems and data challenges. So bridging the it ot gap, I hope we have it people here with us. I think for some organizations it’s a love hate relationship, but I think very often it comes down to the requirements. There’s some of the challenges that we see scalability, not cost effective, but I think what it comes down to at the end of the day is that OT data consists of protocols, formats, uncontextualized information, and really data that is designed to something that doesn’t really fit with the IT data needs around modeling, having the context of source to enterprise. And I think that’s where a lot of the, it feels like, that’s where a lot of the friction comes from.

17:57
Lenny
I’m going to use a very simple example, and it might sound too simple, but something as simple as an engineering unit, just adding an engineering unit, great example, and a range to OT data is something that if you start using it in data modeling and on objects on the IT side, it’s something that.

18:16
Jaco
You need to have.

18:17
Lenny
And a lot of the times we.

18:19
Jaco
See is just get me the data.

18:21
Lenny
Just push the data into the cloud, we will massage it there, we will do whatever we want. But actually we’re just moving the problem around, right? If we don’t model it on the OT side already with the correct ranges, engineering units and push that up, then we’re just moving the problem. And now the IT guy needs to scale and think about it. If he’s got multiple plants, multiple centers pushing into the cloud, the scalability is not just therefore, to now sift through data in his cloud environment or IT environment to apply ranges, scaling, et cetera.

18:51
Jaco
Definitely. And like with many other things in life, context is everything. And I think that context is where the problem comes in and the data not being in a usable form, as we have on this slide. And I think very often it’s just throw everything into a data lake, and we always refer to a data swamp, which it very quickly will become. It’ll become smelly and not very pleasant because nobody actually knows where a lot of the context of the data should be coming from and why it’s missing, making that data fairly unusable. I think Rowan from ABMBef spoke about that very nicely during our last community live episode. We spoke about their approach as Abmbef to data and the value of data and how they actually contextualize the data. So that’s really the problem that we have with the data.

19:34
Jaco
And I want to almost call it the data problem. It’s not a very positive way of putting it, but that is the data problem. So the unified namespace, if we look at the classic otit infrastructure and how these two systems interact with each other, so the best way to kick off what this unified namespace aims to achieve. And again, the unified namespace is probably far more eloquently described by Walker Reynolds. If you go to YouTube and you look for Walker Reynolds and his explanation of the unified namespace, he does it a lot better than us. He does a lot more detail than us. And like I said, a lot more eloquently, but that’s a great video. We can also share that link with you. Fundamentally, this is what we have. We have these different systems wanting to interact with each other.

20:15
Jaco
So if you’re familiar with the automation stack, you’ll find on the left hand side your typical hardware device, sensor, PlC to HMI to scada kind of environment. And the challenge is that the PLC connects and communicates with the ScADA. There’s a discrete connection, there’s a discrete connection between the PLC and the HMI. Now there’s a discrete connection between your ERP and your mes. There’s a discrete connection between your m s pushing down whatever that instruction or message or data may be to the SCada. And in turn, what you end up with is a myriad of hundreds, if not thousands of discrete connections. And that’s fundamentally the problem that we have in this environment.

21:00
Lenny
And that’s the problem with scaling, and that’s also the problem of addressing future needs. If we look at this example of the discrete, we call it spaghetti integration, but the discrete point to point integration. What if your analytics software needs data.

21:14
Jaco
Now from the MES solution, currently you.

21:18
Lenny
Don’T have that link. So again, it’s another point to point integration that you need to go and configure just to share data between those two applications, although they’re potentially on the same level on the IT side.

21:28
Jaco
And the scalability is the problem. The scalability is fundamentally because this exists, this environment exists. The spaghetti scalability is the challenge. And remember that for digital transformation initiatives to be successful and for it to be truly enterprise wide, the scalability is an absolute fundamental and non negotiable. And that’s where I think a lot of people feel stuck, and that scalability actually prevents them or a barrier from actually proceeding with any of these amazing potential opportunities. So if we look in the world of the unified namespace now, again, the unified namespace walker does it absolutely amazingly well. The unified namespace is a concept right now. Nobody is going to deliver a CD with a unified namespace 10 or a dvd or a download link. It’s a concept at the moment, it’s an amazing concept that can be achieved in a number of ways.

22:22
Jaco
But if you look fundamentally at that picture right now, the gray area in the middle, and it’s quite aptly gray, but the gray area in the middle representing the unified namespace. The unified namespace looks at all of these different nodes as absolutely as individual nodes and independent nodes, but it provides the ability for them to connect, communicate, push and pull to each other through the unified namespace. That’s fundamentally different to a number of discrete connections between various systems that is quite rigid. And this area in the middle, as I said, is the unified namespace. And where the data ops part comes in, is really providing a data platform that is massaged, has the context, and is the quality that will enable this unified namespace to be successful. I hope that makes sense.

23:16
Lenny
Very important concept here as well is that you’ll notice in our example, all of these components have the little gray word spark plug enabled. So we’re using the concept of MQTT.

23:30
Jaco
With the spark plug.

23:30
Lenny
The implementation of the protocol to enable this. That’s not to say that is the end all and be all of the technology that you use. We find that’s the one that we will promote. But the point is that the namespace utilizes open protocols. It’s not hard coded device protocols that you need to enable to get into the unified namespace. We’re utilizing a single open protocol with a broker centric architecture. So that gray dot there in the.

24:00
Jaco
Middle, that’s your broker.

24:02
Lenny
And if you need data from your mes, you can pull it from the broker. And if your MES computes a KPI or a thing, it will again back to the broker. What that allows us to do is that any other of the applications can literally just go to the broker and fetch the data if it wants it. If it doesn’t want it, so be it. But the data is available and ready for you already in the unified namespace.

24:23
Jaco
Correct.

24:24
Jaco
So again, the unified namespace is a concept, and Walker refers to it as the holy grail because it is the ultimate enabler of end to end business data. It is where your business data and context lives, right? Where data Ops, again, is not a product. It’s certainly a segment that came to probably say over the last six months to a year through companies such as Hibite. Data Ops has really been put on the radar in terms of what it offers as an environment to provide that data quality and massaging the data. Dylan asked a very good question. Where does the unified namespace actually live?

25:02
Jaco
Right?

25:02
Lenny
So yes, the unified namespace can live. Well, not in all scalar solutions like we said in our quiz, but we are going to utilize ignition as a framework for the unified namespace. As Jaco mentioned, there’s other companies like Hibite, which also has data ops or platform where they can build the unified namespace. So there is definitely a lot of products and vendors out there that is going to start playing within this data field. We’re going to use ignition in our example with MQTT, there are other scalar solutions that will say that they can do it. All that we’re going to advise is just do your homework properly, make sure only of pretty much three things. Make sure it works on open architectures, make sure it’s scalable, and make sure that it’s got a very good itot bridge for bridging these two components.

25:59
Lenny
But I think the biggest thing there is definitely has to use open protocols to build up the next.

26:04
Jaco
Absolutely. It is only those open protocols that will allow you to democratize your data I know Arlen, often from Siriuslink, often speaks about that also quite passionately. That is a fundamental, and these are some of the guidelines around the implementation of the namespace and data ops. And as Lenny said, that is an absolute fundamental. Without that, you will not be able to achieve the kind of openness and integration and interoperability that you’re looking for. I did skip a slide. I just want to go back to this one, which I thought was a bit of a last minute ad. And sorry guys, I saw this and I had to giggle on. It was actually on LinkedIn, I think it was. I think Samsung posted all the Internet of things.

26:43
Jaco
Somebody posted this where Samsung’s new fridge will ping your phone if you leave the door open. And then one of the comments of this post, the person replied, he said, why? I don’t know what that means. Why doesn’t it just close the door itself if it’s so mean? It makes me smile because I think about, yes, it’s absolutely amazing that it can let you know that your fridge door is open, and there’s probably reasons why it can’t close the door. Your cat might climb into the fridge or something like that. But it made me smile because it made me think about the actionability of our industry. And I think that is another something that MQTT and this unified namespace provides is the actionability of Internet of things. Industrial Internet of things.

27:24
Jaco
There is such a massive amount of value in that ability to also act on something as opposed to just being notified. But sorry, there was not prompted. I just threw that in there at the end. We’ve spoken about this. We’ve spoken about some general guidelines for implementation. We’re probably going to go through a couple of these slides really quickly. Maybe we’re not explaining it as best as we can. We’ll share this afterwards. But we spoke about MQTT and more specifically MQTT Spark plug B. That is the open technology that empowers transformation. It is our enabler of choice combined with something like ignition. And once we build this out and go through some of the details, you’ll understand why it is our enabler of choice. So Lenny’s going to attempt. Not attempt. Lenny never attempts. He just does it very successfully.

28:13
Jaco
Give us, and again, this is our very humble view of the space. It is our very simplistic way of building this out based on a little bit of interaction that we’ve had with a couple of people over the last few months. We would love to get your opinions and your feedback on this, but we are going to build this now with ignition, canary and flow.

28:34
Lenny
All right, so just to quickly highlight eclipse.

28:36
Jaco
Forgot about the Eclipse foundation.

28:38
Lenny
These three products are three individual products, right?

28:42
Jaco
Yes. So three different solutions.

28:44
Lenny
Three different solutions, three different companies in their underwrites. Yes. They do integrate with one another, but the reason why they integrate well with one another is because all of them use the MQTT open standard protocol for communication. And all three of these companies belong to the Eclipse foundation, more specifically, the actual Sparklab B working group of the Eclipse foundation. And they’re all three part of that. So although it’s going to be three separate products or three different products that I’m going to show and just show how they integrate, it’s going to be completely loosely coupled from one another. I’m just going to use open protocols to try and build out what this Unified namespace could potentially look like.

29:29
Jaco
Cool.

29:29
Jaco
If you’re not familiar with the active foundation, it’s worthwhile checking out the work that they’re doing. Phenomenal work for our industry as a whole. And they are really the home for the open source implementation of MQTT since 2011. I think other members include IBM, I think is on there.

29:43
Jaco
Yes.

29:44
Jaco
Aviva.

29:44
Lenny
And exactly same end users as chevron.

29:47
Jaco
Chevron. Yeah, Chevron is on there as well. Chevron. I’m trying to think hive. MQ I think is on there. So really, I think it’s a unified effort to put something in place that I think is really going to benefit the industry as a whole.

30:01
Jaco
Cool.

30:01
Jaco
So we do have some further learnings. We’ll share this with you. We did a podcast with Arlen Nipper, who speaks a little bit about his vision for MQTT and the role of better players in the IIot industrial Internet of things. And then obviously we spoke with the master and really the best teacher around the unified namespace, which is Walker. You may find those two podcasts interesting as well.

30:26
Jaco
Right.

30:26
Jaco
I’m going to stop sharing. That was me. All the boring stuff. I know. Everybody is here just to see Lenny in action and see.

30:33
Jaco
Cool.

30:33
Jaco
Lenny, you’re claiming that you’re going to build the unified namespace with these. Let’s see how you very boldly claim to do to build this.

30:45
Lenny
Sorry, I’m just going to switch over.

30:50
Jaco
It. Sure.

31:28
Jaco
That wasn’t a very smooth handover.

31:30
Lenny
All right, sorry about that, guys.

31:33
Jaco
Screen share the video as well. Very elegant. Hand over there.

31:48
Jaco
Rebecca.

31:51
Lenny
Can you just give me a confirmation?

31:53
Jaco
See everything as well when I’m ready to.

32:00
Jaco
Cool.

32:00
Lenny
All right. Sorry about that.

32:03
Jaco
Not a very smooth hand over there.

32:04
Jaco
Very elegant.

32:06
Lenny
All right, cool. So this is my little perspective app that I’ve built. It’s going to mimic a little bit of an enterprise wide solution. And I’m going to use the NQTT broker in this case to publish and subscribe to some of my plant devices. I’m going to publish and subscribe to some of my story ending applications, a little bit of industrial applications, as well as some mis applications as well, to see how I can utilize the unified namespace for all of these things. So the architecture that I’m trying to try and mimic is the typical architecture for IoT that’s on inductive’s website where we have a whole bunch of field devices. And these field devices will now bridge the typical OT protocols to actually get that data from your plcs and devices and machines.

32:57
Lenny
And it’s going to utilize the MQTT protocol to push that data out to an MQTT server. Some people record it as a server. We call it a broker. That’s literally that gray hub that Jaco showed in the presentation. And from that gray hub or the MQTT broker, that data can be made available to other types of line of business applications. I know this is a little bit more ERP and upper level applications, but we can make that data available to historians. We can make it available to analytics tools. We can make it available to mes and mis applications. And what I’m going to utilize for those are going to be the following. So I’m going to use ignition for the edge part. So I’m going to use a little bit of ignition edge to do that decoupling from our machine to MQTT.

33:48
Lenny
I’m going to use the Canary historian to historize that data, do some calculations on that data, push that back onto the broker, and then I’ll also show how I can do some KPIs in flow and get those flow KPIs back into that, but still utilizing my ignition as the building block for this unified mapspace. All right, so I’m going to jump a bit here.

34:09
Jaco
Cool. This architecture, even though it’s a very specific ignition slide that we’re looking at, hopefully this architecture is quite a common one. It feels like it is. Let us know if you can relate to this kind of architecture need.

34:22
Jaco
Cool.

34:23
Lenny
So I’m going to have three components here. I’m going to have my edge device. I’m going to have my main enterprise server, and then I’m going to have canary and flow. So I’m going to jump between three vms that’s got all of this product, but utilizing only the MQTT protocol to push data around and to build up my unified namespace.

34:44
Jaco
Cool.

34:45
Lenny
All right, so let’s look at the edge first. And I think this is probably one misconception about MQTT, that people think MQTT is only used for remote applications. Yes, it does massive bandwidth reduction for remote applications, but it’s not just used to get remote asset data to your enterprise. In this case, I’m going to use it for that as well. But when you talk about the canary and the flow piece, then we’re going to use it for actual line of business applications to integrate with one another.

35:16
Jaco
And again, the objective is to get.

35:17
Jaco
That contextualized metadata, almost want to call it within these line of business applications. You want it to exist there with that context already, not having to build or create or understand that context once it’s there.

35:31
Lenny
All right, so let’s have a look at the edge. All right, so I’ve got a site, the indoor site, and that’s running ignition edge. And this is the guy that I’m actually going to connect to my plcs and my devices. All right, so I’ve got a few drivers here that connects to a modbus device. I also have a top server running here that’s actually connecting to my PLC. And there I’ve got all the typical tags and plant information that I’ve got from my plcs and my devices. Now, the first thing that I do, and we understand that this is not always possible, we understand that at some point, the lower you go, the more difficult it can become to actually build out a model because of tag naming, infrastructure challenges, et cetera, legacy devices.

36:17
Lenny
But what I’ve done at the edge is I’ve already created a definition of what a machine looks like. In my case, I’m building out a filling machine. So I’ve created a UDT for the filling machine and the actual data that I’m going to get from our different devices. UDT is a unit defined template. So this is a template of what this device can be. And you’ll notice I’ve got power consumption. Power consumption comes from my Modbus device, right? Because that’s where I get all my power information from my modbus reader and all the other information I actually get from my. Okay, now what I’ve done on the edge is I’ve already modeled my edge device right here. So you’ll notice that I’ve created a very nice structure to my edge device.

37:08
Lenny
That’s important when I’m going to push it back to my enterprise solution to use the unified namespace. So I’ve got some buildings, I’ve got some control rooms with all their power information in it. I’ve got a packaging line, and you can see all the live data is now updating via that UDT that I’ve built for that. So I would like to push this data that’s already got context, right? It’s already got a model in it, because I try to do that modeling as low possibly as I can, which is not always possible. And I would like to push this out to the enterprise using, okay, so let’s go to the MqtT configuration for the edge. So here I go. This is the edge, this little home page where I can go and configure it. And this is very simple.

37:58
Lenny
All I really need to do is I will go to the configuration of this guy and I log in very secure. All I need to tell my edge device is what part of the edge model would I like to push. And I’m using the MQTT transmitter, in this case the ignition H Iot device, to actually do that. You notice I’ve got an MqTT transmitter module here. Go to the settings, and on the settings side here, all I need to tell it, if I look at my transmitter is I need to tell it which path in your model would you like to push? And I’m pushing right from the root, which is the edge part. So you’ll notice if I go back to my site, there we go, right there from edge. So edge going to push this entire model for me.

38:53
Lenny
It’s going to push onto my broker. So this is the edge. And remember Jaco said that rip and replace is not always the thing you need to do 100%. There could have been a legacy scalar, there could have been legacy devices on there. But I’m using just the ignition edge component to bridge that gap between legacy without having to rip and replace. All that this does is it creates that bridge between the OT layer and my MQTT broker in the front, but already with a model in place.

39:25
Jaco
Cool.

39:26
Lenny
All right, let’s go to the enterprise. So this is the enterprise, and in the enterprise I’ve got two things. I’ve got the engine, the MQTT engine. The MQTT engine is going to connect to the broker and it’s going to build up this namespace for me automatically.

39:46
Jaco
All right.

39:48
Lenny
On the enterprise, though, I also have a UDT, a udefined structure.

39:52
Jaco
All right.

39:53
Lenny
But this looks a little bit different now. And the reason why this looks different is because this structure now contains pretty much everything that I want from this filling machine. And this is the UDT for the unified namespace. There’s only going to be a very small portion of this that comes from the edge.

40:13
Jaco
Right.

40:14
Lenny
There is the state that the machine is in, the current product that the machine is doing, the power consumption only those things are real time data that’s coming from the edge. All of these other things, all the calculations around the total bad product, the availability, the quality, these comes from different systems. And I don’t really care.

40:36
Jaco
Right?

40:37
Lenny
I don’t care where they come from. The point is that the data is available in this namespace and whatever other system or user that needs to get it, they can just subscribe and pull that data out. And this might not even be complete at the end of now, let’s do the following on my enterprise solution. I’ve got, obviously the broker, right? So this is where my MQtT broker lives. We’re utilizing the series links modules for that perspective. So I’ve got my MQTT distributor here, which is in essence, the MQTT server running on this. Okay, I’m going to disable it.

41:20
Jaco
Cool.

41:21
Lenny
I’m going to save it. All right, so I’m switching off the MQTT server at this point, and what I’m going to do is I’m going to go to my enterprise and I am going to delete all my tags.

41:34
Jaco
Are you sure?

41:35
Lenny
This is the part that gets me a bit nervous, but let’s do it. I’m going to delete everything. All right, so remember we said this thing must be scalable at one point. It needs to be scalable, and you need to have as little bit of coding as possible when you try to create this concept of the next. So I’ve just deleted those two components right out in my solution. All right, so we know the real time data is going to come from the edge. We know we currently deleted the data, so there’s nothing in there. But we also saw in that UDT structure that I’ve got some data coming from other sources, historian, as well as from flow, which is going to act as my NES solution. For now. Let’s go to Canary. Canary story.

42:30
Lenny
All right, Canary historian is a time series database or historian that I can store my time series data in and do fancy calculations on that data as well.

42:41
Jaco
So in this case, Canary will be obviously a consumer, it can be a calculation provider, and it will be a.

42:48
Lenny
Publisher, it can be all three of those things. So the first thing is we need to let Canary connect to my MQTT collector.

42:57
Jaco
Right.

42:58
Lenny
So you’ll notice that Canary can go and consume data from an MQTT broker. It also uses the Sparklac B implementation of that, and in this case, anything that it sees that it’s going to throw onto the sparkler B topic, it will automatically start creating tags and start historizing those tags. Obviously, currently it is not connected because I switched off my broker. So it’s not connected. So currently it’s not storing any data. But if I do have the data in it, Canary can go and create calculations for me. This is one very important thing is that, yes, there’s a lot of data running around, but obviously we want to add as much information and context to that data as what we can.

43:46
Jaco
Right.

43:46
Lenny
So even on the historian level, I can go and create a calculation. In this case, what I do is I create what is the maximum power consumption for my filling machine every hour, and I’m pushing that out as a tag.

44:01
Jaco
Right now.

44:02
Jaco
You want to get this calculation with the context back into the broker.

44:06
Jaco
Correct?

44:06
Jaco
Into your unified nervespace.

44:08
Lenny
So let’s add another calc just very quickly here. So I’m going to add a calc on my device, on my filling machine, and I’m going to go and select my filling machine here as an asset. So I’ve got five filling machines in this.

44:25
Jaco
All right?

44:26
Lenny
So I’m going to create a calculation based on this filling machine in Canary little bit of school.

44:33
Jaco
X equals.

44:34
Lenny
And then you can do your little bit of equation here. But there’s a whole bunch of functions that I can utilize in it. So let’s look at the minimum. I had the maximum power. Let’s look at the minimum power just as an example. So I’m going to add that minimum power equation here because I’m running this off my assets. It’s been identified. You’ll notice that I’ve got some asset tags here, and there’s the power consumption that I’ve got available. And I’m going to go and add this power consumption into my equation there.

45:05
Jaco
Sorry. Here we go.

45:10
Lenny
So there’s my equation. I’m going to utilize that tag, and I’m going to create the minimum for this every hour.

45:18
Jaco
Right.

45:19
Lenny
So I’m going to specify just the time frame to actually calculate this every hour, and then I just need to tell it what is the output? So what is the tag where I want to write it to? I’ve got a calculation block where I store all my calculations with, and I’m going to write this same tag out again to that calculation block. And in this case, obviously this is going to be my minimum power consumption.

45:47
Jaco
Right?

45:47
Lenny
It will automatically change the asset number to the asset that I’ve got. Let’s evaluate this and see if it works. There we go.

45:55
Jaco
Right.

45:55
Lenny
So two is the minimum power consumption that it’s going to create for us, and it’s going to push it onto.

46:02
Jaco
This very easy, quick calculation.

46:05
Lenny
Very easy, very quick calculation. Going to apply it. And I’m going to go and close this here and just run my application as well. All right, so there I’m running my two application, my calcs, and it’s busy building up this tags. Now, there might be other applications that want this calc, right?

46:22
Jaco
This calculation exists within canary.

46:24
Lenny
This calculation only consists in canary. So in the old days, if I wanted to have this calc now in my scada or in other application, I had to do some integration with.

46:34
Jaco
Okay.

46:35
Lenny
Luckily for us, Canary has the concept of a publisher, and what I can do with the publisher is exactly the same. I can take all the data that sits in my calculation block. So all the calcs that I’m performing in Canary, I can push that out again onto my NQTT broker. All right, so this will create data for me in my NQTT broker.

47:01
Jaco
All right?

47:02
Lenny
And the last one that we’ve got, obviously, is flow. All right? So flow is our information platform. Flow will go and create a little bit more around OEE in this example. So there’s my five filling lines that I’ve created. I’m pulling in the bad product and the good product, and I’m doing some availability and performance calculations, basic OEe, as well as a little bit of KPIs around. What is your previous availability? What is your running time, what is the actual quality from your product? So yeah, the three components of OEE, overall equipment efficiency. And I would also like to take these calculations and push that out for other applications to utilize. All right, so notice here flow can obviously connect to multiple data sources to give context to that. And we also have the concept as a consumer to push data out.

47:55
Lenny
So any one of these KPIs and calculations that I’ve got in flow, I can go and push that data out back to the NQTT broker as well. And flow’s got a little bit of an integration engine to do that. And I’m just going to deploy this integration engine this.

48:16
Jaco
Now.

48:16
Lenny
So three products that’s going to start pushing data to my broker and start to, first of all, populate my tags, and then hopefully it is going to start to populate my unified namespace with all the data that I potentially need.

48:38
Jaco
With the plant area line, correct?

48:41
Jaco
Sure.

48:41
Lenny
So we utilize the ISO 95 as a structure to build out this notion, but it’s going to start populating these tags for a unified namespace of water filler can be, all right? And from this point forward, I don’t care what it is, all right? I don’t care where it comes from. All I want to be able to do is use these in visualization screens, dashboards, and build out visualization concepts on top of it. All right, so this is the part where I hold my breath. So let’s go to the engine. Remember, there’s no tags there. Let’s go in and enable that broker again. So on my main ignition system, I’m going to go to the MQTT distributor and I’m going to enable it.

49:28
Lenny
All right, so what’s going to happen when I enable the broker is all the components is going to start connecting to the broker and start publishing their dates on top of the broker.

49:36
Jaco
All right?

49:37
Lenny
So I’m going to save the change here, and that’s going to go and enable my broker. Now, what I’m going to see if I go to my enterprise here is already, this node can be expanded.

49:49
Jaco
All right?

49:50
Lenny
If I expand this node, you can see that there’s already components starting to push data onto my namespace here. This component is the component of the real time data coming from the edge device. This component here is flow, starting to push all the production information. You can see how it grows while all of these components are starting to push data into it. All right, one thing that I see is still needing to push is canary. Let’s just quickly have a look at my canary historian. Go to the NqT publisher here and just make sure that I’ve actually got it enabled. All right, there we go. It connected to the broker. So technically I should see my canary starting to push up 100%. So there we go.

50:42
Lenny
There’s my canary data starting to pushing into my cloud as well, again utilizing the canary structure to build out the namespace. All right, so I’ve got my filling machines for all five of the filling assets, and I’ve got my power consumption that’s starting to push there. The minimum calc will also start to push into this as well. Now all I’m doing in the namespace, the unified namespace is I’m taking all of this data that gets pushed into my engine, and I’m building out a default tag structure that combines all of that data. So if I now look at my filling machine for filler one, you’ll notice.

51:23
Jaco
Hey, look at that.

51:24
Lenny
A lot of my data is already starting to come in and starting to populate, and now I can utilize this data to start visualizing dashboard.

51:35
Jaco
Right.

51:35
Lenny
Not all of it will be there. There might be other systems that still needs to push into this to complete the picture. Remember, we said we don’t know what it is, we know what the big picture is going to be, but this is all the data that’s already being pushed live data coming from the field, aggregated data coming from flow and calculations, also coming from canary potential. So I built a little bit of a demo within perspective. So if I go back to my browser here’s my little namespace demo, and let’s have a look at one of my packaging machines. So there’s my packaging lines. Package line one. And there we go. There’s my current Oee information. Now, I’m utilizing ignition perspective to visualize as the front end. But this data actually comes from flow at this point.

52:25
Lenny
Do you actually even care where the data is coming from?

52:27
Jaco
You don’t.

52:27
Jaco
And I think that’s very often that some of the challenges is understanding where the data should be coming from or is coming from.

52:32
Lenny
You don’t care.

52:33
Jaco
And that’s why the facilitation of that scrubbing and hosting within the unified namespace is the big value.

52:40
Jaco
Correct.

52:40
Lenny
All right, so we can utilize the mobility and the ease of use of creating these dashboards within ignition perspective. But what I could have also done is because perspective is HMR five, and it allows for iframe components, I can actually utilize other visualization tools right here into the product as well. In this case, this is canary axiom, but I’m actually utilizing it inside of perspective, and I can actually go and add my historical trends right here from there. So there’s my filling machine. There’s all the data that I’m starting to push into the namespace again, having a nice namespace look and feel to the data. Add that on. And I have my trending capability right here in my solution as well. So real time data coming from the different solutions in my unified namespace, but then utilizing Canary’s capability of the historical.

53:31
Jaco
Part to do the historical training, no.

53:34
Jaco
Dependency on any discrete connections between specific components.

53:39
Lenny
I didn’t code anything in this, literally, guys, it was literally, as I just done it all by a configuration, I didn’t have to code anything for any integration at this point to happen. The report is not going to show all the data because we’re only running it live a little bit now. But this technically is a report in flow that I could have also embedded in it. We’re obviously waiting for the data from the different other filling machines, but there you can see your historical KPI information from flow also embedded as a report within the solution. So yeah, that’s our very simplistic view of it. Getting data from three products into one central namespace at my enterprise, building out this unit, and start populating the data from various systems into this model so that I can obviously use that.

54:30
Lenny
And obviously because perspective is mobile enabled, the last thing that I can just show is I can obviously get my phone in here. Yes, Jaco next to speak to Santa.

54:43
Jaco
I didn’t say anything, but I can.

54:45
Lenny
Also now browse my namespace demo, exact same namespace demo from my phone, and look at the exact same data available on my device.

54:56
Jaco
And in this case it’s through the.

54:57
Lenny
Native perspective, and this is through the native perspective. But the point is, it’s the same data, same namespace, and Jen is a.

55:04
Jaco
Consumer of the data with information.

55:06
Jaco
You don’t care where it originates from.

55:08
Jaco
Exactly. Fantastic. Cool.

55:11
Jaco
I think that was probably our very.

55:12
Jaco
Quick and very simplistic view of, and at least our version of the buildout of, again, the concept of a unified namespace and the approach, which is data ops, we’ve done it. In this case, we’ve done it, as Lenny has just done with ignition Canary and flow, and obviously MQTT as a very key enabler or protocol there.

55:35
Jaco
I’m sure there’s some other examples out there.

55:37
Jaco
We do have a little bit of time for questions or comments. Please let us know if you have any. We’d love to get your thoughts on.

55:45
Lenny
Any of this stuff.

55:46
Jaco
We’d love to get some other examples. This is our approach using these three solutions. Please let us know if you have any questions. We look forward to some comments and questions. Come in. Cool, Belinda, I hope we answered your question in terms of where the unified namespace lives. I see you’ve just asked another question. So the next question, Teddy, what is Dylan’s next question? Dylan asks, looking at the PLCs on the edge, do you put MQTT gateways on the PLC layer, or do you put the MQTT gateway on the SCADa layer?

56:25
Lenny
Good question. So the broker, let’s go just back to the architecture here.

56:31
Jaco
Let’s just go back to this architecture.

56:34
Lenny
All right, so there are some PLC devices and there are some hardware that now natively can speak spot plug b, but it’s only the transmission of that data. So your devices need to be able to push data out to a broker, but the broker does not necessarily live on the device or the SCADA. The broker can actually be hosted in the cloud as an example. And you just need to have obviously a connection to that broker in the cloud.

57:11
Jaco
Should we turn on the camera?

57:12
Jaco
Sorry, I just realized you’re listening to a radio show. I think people would have liked to see you actually really do this and you didn’t just play a video.

57:26
Jaco
Hopefully you can see us again.

57:28
Lenny
All right, cool. So the device just publishes it out, right? It publishes to a broker. The broker can live anywhere. The broker can live anywhere as long as you’ve got a connection to it, obviously via mobile, stat or wan or even a direct link. But the device needs to be able to push it. Mostly we use poplet b as a mechanism to push it out. But the MQTT engine within ignition does have the capability to understand different protocols as well. So you’ll notice if I go back to the ignition edge here. So this guy now runs my broker. And guys, there’s other brokers out there. I’m running the series link broker.

58:13
Jaco
Absolutely.

58:13
Lenny
And there’s brokers from IV and Q. Mosquito broker is literally just the post box. It’s where you publish the data to a specific topic and then clients can subtract that data from that topic. The engine is the guy that can go and take that specific namespaces, in this case sparkle B, and then create that tag structure that we’ve got via the sparkle b implementation. They do support some other implementations of the namespace via NQTT. You can even go and create your own custom namespace based on your topic and payload and it will actually create that structure for you.

58:54
Jaco
But I think the spot beep payload is really the secret source.

58:56
Lenny
The software beep payload is the secret. Alan said it so nicely. It’s the HTML for HTTP.

59:07
Jaco
Yeah, for the Internet of people.

59:09
Lenny
The Internet of people. You have HTML on top of HTTP gave birth to the Internet on website. And this is exactly the same. NQTT is just the delivery mechanism, but the spark plug beep protocol is literally that enabler to make this scalable, easy to implement without doing any coding on.

59:28
Jaco
Top of that data.

59:29
Lenny
So Dylan, I don’t know if I answered 100%, but some of the devices can publish out with Spark plug B. Some of them can do MqTT as well. In our scenario, we just utilize the ignition edge component and all that the ignition edge component does is it pushes the data out. It takes the data that we get from the PLC devices, wraps it into an MQTT payload, and pushes that out via a spotload b protocol out to the broker. So we’re using this as our translation layer between legacy plc devices and legacy SCADA solutions to be able to push it into the namespace.

01:00:08
Jaco
Fantastic. Cool.

01:00:10
Jaco
Veddy, thank you so much.

01:00:11
Jaco
And again, thank you to everybody for joining us. I’m still reading Dylan’s question. Sorry, Dylan.

01:00:21
Jaco
Cool.

01:00:21
Lenny
Dylan.

01:00:22
Jaco
Yeah.

01:00:22
Jaco
Let us know if that answered all your questions or not. We’d be happy to chat with you afterwards.

01:00:27
Jaco
Thank you so much for joining us.

01:00:28
Jaco
Again, this was our very possibly simplistic view on the topic of the unified namespace. I think for our team, like for many other people, we’re trying to make sense of not only the unified namespace as a concept, but also this new category of data ops that has been introduced and trying to figure out where.

01:00:49
Jaco
It all fits in.

01:00:51
Jaco
This was our very simple explanation of that and maybe give you a bit of a practical view of how we’ve built that with ignition canary and flow. Let us know if you found it valuable. Let us know if you have any questions. We will share the recording with you afterwards. And thank you again very much for joining us. And thank you for sticking with us. Five minutes over.

01:01:11
Jaco
I just want to make sure we.

01:01:11
Jaco
Don’T have any other.

01:01:15
Jaco
Questions. That’s it.

01:01:18
Jaco
And we’ve got a couple of comments directly to us. Durwin, thank you.

01:01:23
Jaco
Thanks for joining.

01:01:24
Jaco
Even though it was late, we don’t expect anything less from Cape Town, but thanks for joining.

01:01:30
Jaco
Dylan.

01:01:31
Jaco
Happy we could answer your question. And awesome. Thank you very much.

01:01:34
Jaco
Lenny.

01:01:35
Jaco
Great. Love that example. Thank you very much.

01:01:37
Jaco
Awesome.

01:01:38
Jaco
Thanks, guys. Cool.

01:01:41
Jaco
Is that most of the questions Durban chairs love?

01:01:46
Jaco
Right.

01:01:46
Lenny
Thank you very much.

01:01:47
Jaco
We will send out a recording of the session afterwards.

01:01:50
Jaco
And absolutely, as a community, we definitely want to continue having these conversations in our industry. We don’t think we do them enough. I know somebody like Kutzai, industry four o tv. He does it really well. He’s on to check out as well. And yeah, we’re just looking forward to some further conversation. But thank you very much for joining us today. We’re going to end it here. Let us know if you have any other questions.

01:02:13
Lenny
Cool, thanks, everybody.

01:02:14
Jaco
Cheers. Bye.

You might also like