- Speaker #0
Hi to all of you and welcome. Good afternoon. Good evening. Good morning to those of you who are joining us from all the sides of the world. I'm happy to have you today with us for this new session of the WhatsApp UO. It has been a long time since we have discussed last. But today, it's quite special because we get one of our newest members on board, which is the company Barbara and they will drive us a bit more on their use of the on the technology the 649 and the way they are implementing it in their own use case with the company actiona so before we jump to them just would like to to tell you a couple of words to the next events which are coming because as you may know next week will be the the sps fair in nuremberg in germany and we'll have a live directly from this from this fair so I'm happy to to welcome you there and for those of you cannot make it and cannot join us on a physical basis you can still follow what we are doing using this this live session so now I'm happy to welcome the stage David Pouhon from from Barbara David hi Greg hi Greg hi everybody how are you yeah doing good yourself
- Speaker #1
Very, very good. Really happy to be here with you and all the audience.
- Speaker #0
Good. Tell us a bit more about your company, David.
- Speaker #1
Yeah, so Barbara is a company that has been around for the last seven years. And what we do is we provide edge management and orchestration, which is a topic very, very related to runtimes and IC64099. So when you have to deploy... a large fleet of software PLCs or software in general, you really need orchestration. So our idea is to tell you more about this using a use case and also a practical demo.
- Speaker #0
Great, sounds very interesting. So today you told me you want to talk to us a bit more about your use case with the company Acciona.
- Speaker #1
Yes,
- Speaker #0
I love it. about who is Acciona and more about your use case.
- Speaker #1
Yes, yes. I think that the best way normally to understand the technology is through a real use case. And this is what I wanted. So if it's OK, I can start sharing my screen and I can give you more information about the use case and how this is related to you and run times and everything. So let's start very quick with who is Barbara. graphically. So Barbara, as I was telling you, is a cyber secure industrial edge orchestration platform. And I will explain much better what it is after. But we work with different types of users of different verticals. We work in the smart grid, water treatment, food and beverage, pharmaceutical, oil and gas, chemical and many other industrial sectors, deploying industrial technology to the edge. In this case, today I'm going to use an example coming from one of our water treatment companies, which is called Acciona. Acciona, for those people who don't know, is the largest infrastructure management company headquartered in Spain, but they are operating globally in 25 countries. They have different divisions. In this case, the water division, Acciona Water, who is in charge of... maintaining water treatment plants, desalination plants, wastewater, drinking water plants and so on, wanted to execute a project. They wanted to have central management for their 150 plus water plants. So they have 150 water plants distributed all over the world and they wanted to have centralized management. But the problem, I mean, this is something that they've been doing for their whole life, obviously. But the thing is that they really needed real-time data. So the data that they were having in their SCADAS was coming in many cases from manual measurement, from low frequency devices and so on. And in some cases they had these measures every 24 hours. And this, for a water company, it's a problem because they may have even regulatory fines and regulatory compliance issues. If they are not able to report things on time or they report things incorrectly, it's a problem. So they wanted to have monitoring and management, but with real time. And this is very important. Another important thing is that they really needed cybersecurity, data privacy. So everything needs to be really secure because this is a critical environment. And they wanted to do... advanced data applications. So they didn't want to do the typical data applications to raise an alarm or something. So they wanted to do machine learning and really innovative things. So there were some challenges when you are speaking about the large company like Acciona, there is a heterogeneity of the plans. So there are plans who are using some industrial protocol like Profinet, some others are using OPC, some other Modbus, some other are using one vendor, some other are using another vendor. So that's a challenge. The other challenge is that the connectivity from the OT network to the upper part, to the IT network or to the intranet or internet is very, very limited in this case. I already spoke about the regulatory compliance. They need to be compliant with norms like NIST or cybersecurity policies like ICC 2443. Then another challenge is that if you are working in a, let's say, in a single location project, things are normally manageable, but when you have to scale from one location or from the laboratory from 150 locations is when the problems come. So this is scalability and orchestration and what we do in Barbara and the demo that you're going to see is very related to scalability. And last, but probably the most important, they wanted to avoid hardware dependency. And this is where UAO and the IC6-1499 standard comes into play. Acciona is a company that manages plants for third parties, for water utilities. And they cannot ensure that they are going to be managing the same plant forever. So the contracts change every four years. So they really wanted to do. everything with software. So if they want to change a contract to another maintainer, it's always simpler to take the software out than to uninstall 150 or thousands of equipment. So doing things in software, in portable software, is probably the key requirement for this project and the reason why they are using UAO and IC 61499. So, um... the architecture of the project is is this that you're seeing in the screen um in the in the left part you have what we call the edge computing nose which are um industrial pcs in this case industrial pcs from uh from schneider in particular the schneider harmony p6 model but uh this is just a commodity so the good thing about using uh uao is that um is that hardware is somehow a commodity and it could be Schneider, but it could be, of course, Advantech or any other manufacturer. The most important thing is the software. But how you push software to 150 plants. This is why we are using an edge management and orchestration platform. And this is manufactured by Barbara and is the one that we will be showing. in a few minutes, but what this platform allows is first of all manage the devices themselves, so manage the firmware, install security patches, change the network of the device, but more importantly orchestrate applications, orchestrate software PLC logics, orchestrate visualization tools, orchestrate data gathering tools, orchestrate, which is a word that you will probably be hearing a lot in the coming years. So orchestration and other forums like for example the Open Process Automation Forum are very aware about the importance of orchestration. In our case the orchestration happens using a secure channel with MQTT publication and subscription and this channel in the case of Barbara is completely secure. always outgoing, always, so there is no need to open any incoming port into the OT network. So all the connectivity is outgoing, but through this outgoing channel, we are able to deploy these software applications and AI models into the devices. So if we zoom a little bit into the architecture, I don't want to spend too much time here because you will see it on the demo. But from the Barbara Edge platform, we are able to deploy different types of applications, like protocol connectors. AI models like an AI model that Acciona is developing to predict the amount of chemicals that they need to use real time. And we deploy, of course, the runtime. In this case, you're going to see a demo deploying Snaders of Deepak, which is an implementation of the UAO runtime. But of course, we can use other runtimes the same way that you will be seen the the snader runtime here which is again it's a uao particularization done by by snader but the point is that from this um from this centralized management system we are able to deploy applications into one two three or 150 industrial pcs regardless the um the manufacturer of the industrial PC. So it's fully decoupling. hardware from software, making the software a commodity, putting the focus on the software, and making this in a very complex and scalability demanding environment. So one important thing is when you deploy runtimes, normally they are coupled with an IDE, with the development environment. In the case of the Snader environment, the runtime is called SoftDPAC, based on the UEO, it's called SoftDPAC. And the IDE, the development environment, is called EAE. It's where the developers are programming all these complex logics. So the thing is that normally the development environment needs to be connected to the runtime. using a local network. This is good because of security reasons, but it's not very practical because if you need to develop for 150 plans and each plan having hundreds of runtimes, it's basically unmanageable. It's impossible. With Barbara, with the orchestration, you can use a single runtime, a single IDE, a single, in this case, the EAE, the them. a construction automation expert from a sneider but it could be another ide as long as it's compatible with ic6 to 61499 and you can using the orchestration platform of barbara you can program from a single centralized ide any runtime any runtime even they are in different plans in in different networks, in different locations, in different countries. So for that, our orchestration platform has a built-in VPN service that you will see in action in a few minutes. And yeah, the idea with this project was to deploy runtimes as the central part of the logics to deploy, to execute control based on advanced algorithms and AI models. But with Barbara, one other important... thing is that you can deploy other applications that interact with the runtime so if you want to have for example a database from barbara we can deploy different types of databases so that the runtime can store the data in the in the device and have local persistency you can have data ingestors you can have MQTT brokers you can have ingestors to different cloud providers we we do have a good integration for example with aviva with azure with with aws you can have data visualization if you want to have data visualization on the on the node on the on the industrial pc you can deploy applications like grafana like ignitium from inductive automation also from aviva edge and of course you can deploy ai models which is let's say the the more futuristic use case, but which is a reality in Acciona. They are deploying AI models to the industrial PC and taking the data gathered by the runtimes and so on. They are predicting, for example, the amount of chemicals that they need to put and then they are executing the controls for doing that. So it sounds like a very futuristic scenario. but it's a reality. It's something that is already in production, it's already working. It's a project that we've been working in for a little bit over two years right now, and it's a reality. It's, I think, a very, very good example of how software can be decoupled from hardware and can bring real innovations to the field. And with that, I would like to pass the floor to our CTO, Isidro Nistal, who will be doing the practical demonstration.
- Speaker #0
Good. So then I call Isidro to the stage. Hi, Isidro.
- Speaker #2
Hello. Hello. Good afternoon. Good morning. Good evening, everybody. So very, very, very pleased to be here with you guys. I'm going to show you a little bit about what David was saying. So you can actually... put face and let's say understand a little bit about what we are talking about okay um can you see my screen now i think you yeah you can go ahead okay so uh this is barbara so in barbara what you see is basically a list of nodes i want to be very clear about what david was saying that we are hardware agnostic that means that for us a node is basically a computer that is deployed somewhere okay And with a computer, what I mean is that it can be anything that is running our operating system. So basically, we support lots of devices. We also support virtual machines. So all of these devices, what it does is that basically we remove the complexity of managing the operating system of a hardware and then having the drivers for that hardware and having all the things that comes with it. managing an operating system like updates firmware updates security updates all of those things you can just go here download the image for your for your device you plug it into your device and you get the device in this list so once you have the device on this list you can basically manage it as if it was in your in your local network okay like it's like you can manage it from a central point you can also do batches so anything that we're going to be doing today with one device will you You can also be doing... You can also do it with a batch or with several devices at the same time. So what happens here is that, okay, you have purchased this amazing computer and you say, okay, I want to run the UAO runtime. I want to deploy here my PLC. How do I do that? The reality is that the steps to actually make it work, it takes some time, but it not only takes time, let's say, from a provisioning perspective, from a starting point perspective. it also takes time during the life cycle of the device and what we know in the OT world, the life cycles are not three months, okay? The life cycles are more like three, five, seven, eight years, okay? So, Barbara will take you from the very beginning, from the provisioning stage, all the way through the life cycle of that device, so you can keep up and running not only the UI or RAM time, but any applications that you want to install on the device. For the demo today, we're going to be working with this Harmony P6. Okay, this is what Barbara looks like internally. I'm going to put a lot of focus here. I'm not going to go through the whole platform, but basically the platform allows you to deploy applications. So, for example, David was talking about deploying Grafana from our marketplace, for example. So we can go to our marketplace, for example. We could look for a Grafana and say, okay, I want to deploy Grafana. You will purchase it. So you can go here, add it to your device and select it. Basically, you will deploy it Agrafana, okay? But the whole thing here is that you are managing a device from, let's say, remotely. So one thing that you can deploy and that we have a lot of focus on is that you can deploy the UIO runtime, okay? So in this case, what I'm going to be deploying today for the demo, I'm going to be deploying it. By the way, this device... is not in my network okay so i cannot really get to this grafana through the through the ip of the device i have to go through the vpn so that means that the device this device could be anywhere in the world basically okay so i'm gonna keep the grafana running there and i'm gonna start deploying the softd pack so vpac as david was saying is the implementation of of a sneider of the uao runtime So it's a product from Snader, but internally what it has is basically the URO runtime. So I'm going to deploy the first step to actually deploy the runtime. That is the Genie, basically the initial application that basically configures everything on the system for the URO runtime to be able to run. So it will take some time here and says, OK, configuration completed successfully. so I cannot go and deploy the runtime. So this runtime is available on our marketplace only for private for people that has a license with the EAE, but you can deploy it directly with a couple of clicks. If you have gone through the runtime deployment, Ubuntu installation, you will know that it's probably quite a complex scenario. Here, we're going to do through a wizard. So we're going to deploy the runtime. Okay. I'm going to click on next. It's asking me about network configuration. So I'm going to put my network configuration that I have in my network card that I showed you before. So I'm going to put something here like an IP that is free in my network. This will be my local network. And I'm going to do it through the VPN. So I'm going to put this as true. So now I have the softdpack configured. So I can send the runtime. And it's going to start executing the runtime. So now, just with these couple of clicks, I have the runtime up and running. and ready to be managed through the EAE. Okay, so I'm going to go through to the EAE. By the way, the EAE is in my local network, so I can access this EAE. So EAE is one of the EDES that are available for the UAO runtime. So I'm going to add a server, for example. Okay, I'm going to be adding the ethernet card so here as you can see this basically this machine is not in the same network that that the runtime is so this this machine is connected to the vpn through the through the vpn client that barbara provides so i'm connected to the vpn and now i'm gonna be basically configuring a runtime that it's somewhere in the world okay so i'm gonna put the variables that I have from my network card. that I have it here, you know why. Okay, I'm gonna put the IP of the device that it's the 41, so I'm gonna add it. Okay, so now I'm gonna start the service. It will ask me for security basically credentials. So now we have the security provision, so we are gonna add the first soft tip. So I add it. I say that it's in the local network of the device, that is the 160, so that's okay. I also make VLAN, that is something I have to configure, and now I have to configure the runtime. So I'm going to put it for the runtime, I'm going to keep 30 megabytes of RAM. I'm going to add two cores to the runtime. or one core for example okay and then i'm going to tell in which ip it's hosted that is in the 78 so uh now i just have to tell the eae which version of the runtime we are running that at this moment this 23.1 is the one that is available on on our marketplace so i'm going going to apply it will fail because it fails always the first time Okay, we will apply it again. Okay, so now we have the SoftDepak configured and running. So I'm gonna link it to one of our logical devices. So now I can go to implementation and I can actually just like that. It tells me if I want to deploy the program, if I want to clean before deploying, if I want to execute it at the same time. I'm going to do everything. So I have read the messages, and I want to execute the program that I have loaded. Okay, so now it's up, and it's compiling. It should be... It tells me if I want to execute the logic, and it's done. So what we have seen here, it's very simple. I haven't really deployed anything. So I have just only basically connected to the runtime, deployed a... let's say the the empty program okay but the reality here is that now what i have here is basically one soft deep that i can be managing from anywhere in the world and it's running in an edge device in a in a in a plan that could be outside of my network it could be anywhere okay so if i want to add other soft deep acts i can keep doing that and i can keep adding stuff so deep by the dark. not only in my network, but also in VPNs. And I can manage all of them and deploy logics to all of them from a single point, okay? But not only that, if I go back to Barbara and I have my soft debug running, I could also deploy here an AI model that connects to the MQTT broker of the soft debug or the runtime from UAO and gets the data and do something on Python. That is another language. And then basically you can start extracting the value to a PC that is quite expensive because it's an industrial PC. that you have deployed so you can uh liberate all that all that compute power uh from uh from a platform okay from from barbara so you can deploy not only the applications that we have at marketplace that are many of them like databases not red mqtt anything that you want to deploy but also the runtime and and then do that in not only one node but then do it in several nodes in nodes all over the world. So that's... kind of the key element of barbara to say how to manage all those soft defects okay how do you get from buying the the industrial pc all the way to the uh deploying the logic that you want to deploy but not only deploying it because the playing is usually deployments are more or less easy the problem is how do you maintain it during the time because if you're going to have developers working on eae or working on the on the runtime logics they will probably want to update things, they will probably want to have different logics for different things, they will want to maybe... the life cycle of that application is not static anymore. So I think that what the UIO is providing is a logic that makes the application the software-based and then software is much more alive. So you really want to change that and you want to keep updating it. So Barbara will help you to do that in a centralized way. and also in a very secure way because the operating system that is below is the same same thing that we have done with the p6 we could be doing it with an advantage or with a virtual machine so we could be launching uh uh this this machine for example this virtual machine is from one of our partners stratus i could be deploying a um uh status that is also a member of uao uh we could be deploying the the ua runtime here in the high availability vm and i could have different capabilities depending on the hardware that is below. I don't know David if you want to add something?
- Speaker #1
No I think that that was pretty good Isidro. So this is a demonstration of how a project as complex as deploying things in 150 plans can be much simpler with a platform like Barbara and with fully decoupling software from hardware. So I'm going to share my screen and do a kind of small wrap-up. And by the way, if there are questions, I think that you guys can put your questions as comments into the video and we can address them either now or offline. But as a wrap-up, the first thing that I wanted to...
- Speaker #0
Do you need to share your screen again?
- Speaker #1
Yeah, I am on it right now. Yes, so one of the things that...
- Speaker #0
that really make a proof that this is really saving resources is that to put the first intelligent application into the field, it took four months. So a project like this, a couple of years ago, without UIO, without orchestration and everything, maybe it would have taken one or two years, but with the decoupling from hardware, orchestration and everything, it took four months. The KPIs are amazing and this is public information. I'm not revealing anything confidential, but Acciona is now saving over $250,000 per plant. So you can do the math and assume how much impact this is having in their P&L. And using orchestration and software virtualization and software runtimes and everything, the reduction of the total cost of ownership of the platform, we have calculated 74%. So something that without orchestration and without UAO interoperability will have costed 74% more. So this is the power of Barbara plus IC61499. We can define software-defined infrastructure with no hardware dependency, fully flexible to deploy in the same piece of hardware, whatever you need in terms of runtime control logics, visualization, databases, everything from a centralized point of view, and all with the IEC 62443 cybersecurity compliance, which is the norm that we normally adhere to. So also why Acciona chose Barbara? If you are familiarized more or less with the orchestration market, there are obviously more orchestration products and IT has been doing orchestration for a long time. So when you deploy websites, when you deploy web applications, this is already done remotely all over the world using orchestration. platforms like Kubernetes, platforms like the hyperscalers, AWS, Azure. they all claim that they have orchestration. But as the Open Process Automation Forum says in their documents, the OT orchestration is very, very different. So in OT orchestration, the focus shouldn't be the scalability and everything. The focus should be having the system running 24-7 every day and having security, high availability, and a user experience that do not require IT expertise. If you want to orchestrate runtimes using, let's say, IT orchestration products, you're going to need probably very, very high IT experience, command line interfaces, Linux programming, and everything. As you've seen with Barbara, user experience is like any other industrial platform. It's visual, it's simple, and you can do things without coding expertise. And this, and this is, with that I will finalize, this is basically possible because the technology that we've been developing in Barbara the last years. So it all starts with an operating system, which is compliant with 62443. real time and with high availability. So this is a development that we are completing as we speak and that it doesn't require any IT expertise for installation. On top of that, we have the runtimes, the runtimes, the any UAO implementation runtime. We've seen the demo with Schneider, but any other runtime that follows the standard can be orchestrated the same way. And on top of that, our management panel, that hopefully you've seen useful and our marketplace of applications, including not only runtimes, but other applications that can include, that can interact with those runtimes. So with that, I think we're done, Greg.
- Speaker #1
Good, David, thank you. So as said already, please feel free to ask your questions. We'll still have a few minutes to go. For now, I don't see... any question and maybe check directly on the website if there are some. For now, I think you have been quite clear on your presentation. So maybe one question from my side, which is, let's say, more linked to the technology. So we have seen here that, in fact, this implementation, what you have done for Axiona is like deploying the 64999 for, let's say, on multiple sites for Axiona or making this possible. let's say, but thinking about the 1499 and for ACCIONA as well, maybe just a question, because I think you mentioned it, but just to make it clear as well, why choosing 1499, in fact, for this kind of application? What do you see as real, really benefit for the,
- Speaker #2
for ACCIONA in this in this choice but the main driver here for the for the standard was the whole software approach. Okay. So basically imagine that the Acciona, every time they get a contract with a plant and it's going to be a contract for five years, they need to go there and they want to put their technology. Okay. So in order to do that, they have to have this software approach. So they need to actually, let's say they have to adapt the software that they already have to that specific plan. And that has to be done. in very short time because they get the contract that they have to start operating in a few months if they want to do that with all the technologies or with other plc's they will have to actually first of all the the the implementation for the implementation process would be much longer but also the time to market but also when they leave and the the contractor the contract the contract period ends the removal of that hardware is very difficult So being with the software PLC and with the technology from UEO, what they can do is basically arrive, tend to market very fast. They can stay for the four years, do the operations the best way they know, that is with AI models and AI models that are their own property. And then after the four or five years, they can remove that software and the plant stays as if it was before. So they are always like, I come, I operate this plant for four years and then I leave. in order to do that that complexity of of arriving and living it's uh what makes uh the the platform very very uh suitable for so this flexibility this possibility of of adding let's say functionality on top of the existing and
- Speaker #1
and well having the possibility to remove it afterwards if uh if uh let's say the concession or the contract is uh is done it's one yeah
- Speaker #0
the flexibility of keeping adding new functionality without needing to add more resources. So they started with one use case and now they are even moving into other areas like energy management and things with the same infrastructure. So this is super, super powerful. And as I said, the calculation that we've done is that we are saving like 74% of the infrastructure cost.
- Speaker #1
Yeah. yeah so this this this scalability basically uh from one uh let's say to to multiple and as well not only into multiple in the same uh domain structure but multiple into different domains so not only water but maybe energy management and some other domain in the in the in the future so yeah capitalizing on that yeah since uh this is interesting and this is i mean as as universal to mention the talk this is something we've been pushing we've been explaining uh people but it's it's it's always good to see as well uh from the market that others sees that as well the the the same way uh one other point i saw as well uh and and we see that with other of our uh members especially those um um let's say operating for for utilities for example especially in the water segment uh the feedback was Yes, the issue is that we need to train people in multiple technologies or multiple hardware. So every brand has its own way of functioning. And then it means that you need to have people with knowledge in all of these brands because you don't necessarily, let's say, own or master the whole thing. So, yeah, kind of reducing as well this need and being able to bring more. more flexibilities as well. Good, a good point.
- Speaker #0
Yeah. So this is, this is something that has happened in the IT world for, for many times. And, and we, we never know why this didn't happen yet in the OT, but hopefully initiatives like, like these are changing. And with standards like UAO and also with platforms like Barbara, which is using like standard things, standard protocols, standard APIs to, to do complex things in the OT. hopefully this can be democratized which i guess is the objective of everybody
- Speaker #1
Yeah, that's necessarily, that's the goal. That's the goal. So I'm checking the chat to see if there is any question between. So I don't see for now any questions. So I would say that we can close slowly this live. As you see on the bottom, you have the contact details of both David and Isidro. If you wish to contact them and contact the company Barbara, which is anyway linked. in the comment of LinkedIn. So don't hesitate to reach out to them if you have any question on these use cases or if yourself you have some other use cases on which you want to partner with Barbara. I think, David and Isidro, you'll be happy to do that. Any last word before we close the session?
- Speaker #0
No, no. Thank you very much. It's been a pleasure to be here. And as you said, Greg, if anybody has projects with distributed runtime distributed logic in industrial environments please call us because we we're very happy to collaborate with the community to bring this to life.
- Speaker #1
Yeah then thank you. Just another reminder I said it in the beginning of this live but I will repeat again because maybe some of the viewers were not there at the beginning but next week we have the the sps fair and we have a live for this fair for those who cannot make it to to the fair so please uh just go on our linkedin page and you will see this event and don't hesitate to sign up to it so i thank you all thank you isidro thank you david and uh talk to you soon and uh see you to all the ua viewers bye bye give it a match right here bye