Revolutionizing software in the robotics industry: Insights from Dr. Alexandra Mikityuk of Staex

2023-02-22

Get ready to dive into the future of industry with the «Robotics in the Industry» podcast.

Get ready to dive into the future of industry with the Robotics in the Industry podcast. Helmut Schmid and Robert Weber sit down with Dr. Alexandra Mikityuk, the CEO of Staex, to learn about the cutting-edge technology behind their distributed networking and orchestration software. Discover how Staex is revolutionizing the way machines communicate with each other through reliable and encrypted peer-to-peer connections. From the latest advancements in robotics to the impact on the industry, this podcast is a must-listen for tech enthusiasts and professionals alike. Don't miss out on this informative and engaging conversation on Robotics, Industry 4.0, and Staex.

Transcript

[Music] Robotics in Industry. The podcast with Helmut Schmid and Robert Weber [Music]

Robert:
Hello dear listeners and welcome to a new episode of the podcast Robotics in Industry. My name is Robert Weber and joining me again in Munich in the new year is...

Helmut:
Helmut Schmid. It's my pleasure.

Robert:
Today we have invited a great guest to kick off the year. Alexandra. Hello, Alexandra!

Alexandra:
Hello, greetings. Thank you for the invitation.

Robert:
You're welcome. We're glad you're here. Would you like to introduce yourself very briefly to the audience in two sentences? Who are you, what do you do, what do you actually have to do with robotics?

Alexandra:
Yes, sure. My name is Alexandra, I am the founder of Staex GmbH and also the CEO. I actually come from the field of Computer Science and from 2013 to 2017 I wrote my doctoral thesis at the Technical University in Berlin on the topic of Security in Computer Science and then took over the management of the Blockchain team/department at Deutsche Telekom, where I researched the topic of distributed systems for a very long time and discovered my passion for how you can actually make the world of machines a little more secure, and so I ended up in the field of robotics last year.

Robert:
And just to get this right, so that everyone can find you: your full name is Dr. Alexandra Mikityuk. Is that pronounced correctly?

Alexandra:
Yes, that's exactly how it's pronounced.

Robert:
Very nice, thank you. You were at the Robotics Festival in Leipzig. That's where Helmut met you or became aware of you. What brings you to the Robotics Festival as Staex?

Alexandra:
I might go back to the story of how the Staex software came about in the first place. In 2017, I actually got to know all the security topics (the generic term would be Web3.0), what's going on and what's even possible. Besides that topic, I worked a lot with companies. At that time, I was still part of Deutsche Telekom, which dealt with IoT and machines and everything that has to do with it, and I actually saw how much efficiency and security you can bring into it with these new technologies. The issue is really What do I do when I'm no longer in the Cloud environment? That means I don't have my application in a secured Cloud environment, but what do I do when it becomes dynamic? And that's actually one of our main use cases of Staex. These are dynamic networks, dynamic machine networks. For example, drones that fly or mobile robots that move. The problem is on the IT side. It's really not that easy to deal with the whole IT issue the way we do it today in the Cloud, and that's how we found our use case, because robotics actually has exactly this problem. In the last year, we have spoken with more than 60 robotics companies and also confirmed our assumptions that it is indeed one of the difficulties. So we feel very arrived in the topic.

Helmut:
You said, especially in mobile applications, meaning when the robots are moving, that of course AMRs or AGVs are ideal, which move from position A to position B. Many of the companies don't want a Cloud solution, but rather on-premise solutions. Maybe you can go into a bit more detail about your solution to explain to us what makes you different or where the potential problems lie in the Cloud. Because there are of course applications for AMRs or AGVs for navigation. Is your solution better or why should one think about changing?

Alexandra:
Sure, thank you very much for the question. You actually brought up the word Cloud again and that's where I might start. Staex comes from the field of distributed systems. It's actually the protocols that have become very popular over the last decade. All these crypto topics. It all started with Bitcoin. So, with our software we don't have anything to do with the whole crypto topic. But we actually have the algorithms that have emerged there over the last few years. We use that in our software stack and that's actually a counter approach if you compare it to the Cloud topic. We make the software that actually runs directly on the robot. We don't have any components in the Cloud and, as you said yourself, since a lot of companies right now, what we see in robotics, don't choose this Cloud route, Staex is actually a good alternative because we do everything on-premise and this significantly improves the whole security issue. We have created a system that you can't switch off. That means every component that runs on every robot acts in a system as an independent component. The whole Staex-based management system, which is responsible for networking or managing robots, will actually run until the last robot is running. So we are not dependent on the Cloud at all, which is of course very important, e.g. in topics like Smart City. When we think about critical drone infrastructures, and it is the case that the Cloud sometimes goes out, we actually offer this 100% resilience of the system, that you can't switch that off at all. And that's, I think, the first big difference when you look at the two worlds of Cloud vs. something that's on-premise and distributed on the robots (or you say Swarm Robotics), when that's actually running on-premise.

Robert:
How would you define yourself? What are you now: a connectivity layer or a platform? What is this Staex? Or is it where the threads come together? How would you put it up in the hierarchy in the factory?

Alexandra:
You described it very well: it is a connectivity layer that goes together with management. That means you can't just connect the robots in this layer. What we do as Staex is we do an overlay network. That means we use the physical protocols where the robots exchange messages with each other, and we run on this network, for example on an intranet or internet, creating a network and in this network you actually have functions, like you can manage your software, you can do updates or perform telemetry tasks to retrieve the data from the robots. And that's actually a layer. I don't think we have a good name for it yet, but basically, that's what's called DevOps nowadays with the existing large applications, that ensure connectivity to my services no matter where they are. They're globally distributed and I get access to all machines. I can always run software updates. At Staex, we are creating the whole new topic for the field of robotics DevOps for robots, but it's just not as easy as a Cloud world because it's not all side by side in one data center but it's distributed in a factory. It's not so easy to actually address the robots in a factory because there are a lot of firewalls etc. in between. And there are the problems that we solve for the manufacturers.

Helmut:
When you have AMRs and AGVs, they also have their own fleet management system somewhere or bring it with them. Do you now sit over it or under it or do you integrate it? How can I imagine that?

Alexandra:
Exactly. We are positioned next to it. There is actually a big issue of ROS which is also a very important component. We run in the operating system. Often it's something like Linux and it's like these fleet management systems, they have an agent on the robot. Always a counterpart on the robot that you very often have to somehow directly provision on the robot or make updates of it or have locator services. The maps that are on the robot also need constant updates. For example, what the ground floor or certain setups look like. We are also just like an agent as in a fleet management system. We run directly on the robot and control all the other agents. We make sure which version is on which robot, how to renew that and how do I actually make updates overnight. It's running alongside all the major agents that are actually supposed to be running on the robot and enabling the updates from those components.

Helmut:
May I ask again? Very close to your question, Robert. It's not just fleet management, it's also integration with the VMS system. You mentioned the Swarm intellect before. Agilox also works without a fleet manager just with Swarm intelligence. My question is, they need their mapping, they need their software update, but of course they also need from a PMS ERP system, if I'm talking computer robotics now, the allocations, what is prioritized, what has to go where. It is not yet clear to me where the benefit lies if you are only alongside. Then I have access to the VMS or the ERP system. You're running alongside and I might even have a fleet manager. Where is the added value really? Is it more towards security and not being hacked? Is it stability, because let's say if I have on-premise WiFi solutions, that's always the biggest challenge. Do I get a stable network? Are you contributing to the network stability? I'm still not quite sure, why do I need you guys and don't do it directly.

Alexandra:
You have addressed many topics. I would actually start with the topic of automation. What we have seen in the industry, especially with many drone manufacturers and real estate robot manufacturers that a lot of things are done manually nowadays. To begin with: you actually don't have good solutions. You mentioned that the robot is actually in the WiFi network and the network is not very good now, does not have good quality very often in the factory. How do I do the updates there? What we have seen is that a lot of this is still done manually because the network simply cannot cover it. For example, the issue of WiFi roaming: that you can get from one WiFi to another in the factory. Via the Staex network, you have a very secure and resilient network. And this network can actually cover all these problems that happen in a physical network. That means we solve the problem of WiFi roaming, so that the robot can move from one network to another without any problems, and an update, for example, is not interrupted. These are all issues that we as Staex are dealing with. Or, for example, it is often the case that the manufacturer has robots in the factory and then they talk about such a VPN solution with their own robots that are used in this factory and there it starts in the network layer that these robots have no fixed addresses because the IP addresses in the factory are distributed dynamically and that is also one of the topics where Staex actually assigns these global addresses so that the manufacturer always knows where the robot is and what other address this robot has. And for that we actually have a patient at Staex. This is, if you like, an add-on for a VPN solution, which enables the manufacturer to always be in contact with the robots and of course in a very very secure way. And that's where we start. Because the topic of dynamic networks is actually not that simple and in one doing. If you have the robots in one operation, for example as a robot manufacturer, in several factories, it becomes very complex how you can monitor this robot fleet on the network level. Then there is the whole issue of updates. You actually have to do them very often so that you have access to these ERP systems, etc., as you said, and that's just like getting in there and automating. That is exactly how you know it from the Cloud world. With one push of a button I can actually run it all overnight and our system automates, takes responsibility for the whole process and makes sure that every robot has got hold of the update and that in a secure way.

Robert:
Where are your customers? What is your target customer group? Is it only something for the big ones or do you say it's only worthwhile from X1000 robots, AMRs that are driving around. Where do you see your customers?

Alexandra:
In the last year, we've crystallized a good segment for us: there are always companies that have over 50 robots. And when they start to provision robots not only in one factory, but in two or three factories, the network setup becomes quite complex and you can't scale that easily. And that's actually a starting point where they start to hire more people in an IT team, where we as Staex say it's actually much more than you need when you use Staex, for example. And that's actually where it really hurts and they then start looking into Staex and trying it out.

Helmut:
You certainly offer this as a robotics-as-a-service, from what I've read. What services do you offer to support this when things go wrong? You just said that sometimes the IT departments have a hard time when things get more complex, or does the software solve the problem in and of itself or do you also monitor and support manually?

Alexandra:
In this case the software solves the biggest problems. That means that we actually have the topic of monitoring included. It is also automatically solved what happens on the robot, so that the manufacturer can get to all the data in the first place. Of course, we as a company also provide manual support if it is necessary, but essentially our USP lies in the topic of network updates, which is actually automated by the manufacturer and we solve this on the software level.

Robert:
And what is the next step now, where are you going?

Alexandra:
The next step are such issues as what we are doing today, that we become more and more known in the robotics world. We just started our journey last year if you want to describe it that way. We as a company are also learning because it's a new topic for quite a lot of robotics companies, a DevOps topic: how you can automate all that, bring that to the market and scale, to make more companies understand what services we offer at all, what works with Staex, what kind of problems we can solve and that we also work that out together with our customer. The platform we have is very powerful and just in the last year, for example, a topic like WiFi roaming is new to us and we notice every week that there are new problems coming from the industry, which Staex actually solves as a layer and that is very exciting for us. We want to develop even more in this direction.

Robert:
Do you think that robotics users don't have this DevOps topic on their radar or that they are using approaches and strategies that are no longer up to date?

Alexandra:
I would actually go out on a limb here and say that it hasn't really taken off yet, because my understanding is that the scaling wasn't there yet, because the topic of robotics started really taking off in the last few years, especially due to the Covid topic, all the intralogistics applications with robots, etc., and companies are just coming onto the market, and companies are coming onto the market that have larger robotics fleets in operation and they are now starting to say "Hey, this is all getting really complex somehow, and I'm losing my robots and I can't find them, and oh my God, the networks aren't that good in reality". We had the feeling that because of this, last year, it also took this long for us too. We actually come from the Smart City sector and have networked IoT gateways. That's why I think it has somehow started to take off, because the topic of scaling is coming, the topic of robotics has many more applications and will grow a lot in the coming years. I think the growth rate was something like 25% per year and that is of course massive.

Robert:
That's interesting, Helmut, what Alexandra says, that people don't realize "Oh, the networks aren't that good". We've actually had that on our robotics team's radar so far, this topic of connectivity, right?

Helmut:
It always plays a subordinate role and that would have been my next question. We have problems with connectivity everywhere. Many people are talking about 5G networks. In Asia, they are now talking about 6G networks and that would actually be my question to you, Alexandra. If I inevitably have to improve something on the network, I'm in the middle of Munich and I sometimes have 2G or 3G... there's nothing better. But apart from that, if you grow towards 5G now, does that support you, does that endanger you or would that be an additional enabler for you? How do you see that on the network stability, connectivity? What needs to happen? It's clear that it used to be way behind. How does that play into your cards?

Alexandra:
First of all, I'm very very excited that this is coming at all, because I believe that we actually need this in the industry. It plays into our cards because Staex as a global overlay network connects each of the networks. It is enabling the manufacturer to talk from the Internet to the Intranet in a much easier way instead of having to set a lot of VPN solutions. And now you add a 5G network, which is a private network, and it becomes even more complex if you want to do remote maintenance in this network, for example. It's not all that simple. Let's say I am the manufacturer, have 10 robots in this factory and 20 robots in another. And I actually have to talk into these networks somehow. Of course, the telecom providers will also offer such solutions over time, but we as Staex already have the connectivity solution, how you can talk between several 5G networks and actually do that with these new technologies, which are distributed, and of course also very, let's say, inexpensive, if you compare it with telco solutions. That's why it plays into our cards when you also have 6G networks, etc. The drones have to move somehow between all these networks, or robots. And that's also the handover story that we offer as Staex.

Robert:
Very exciting Helmut, isn't it?

Helmut:
Yes, very very interesting and I think I can actually come back to your question about the fact that we need to go into more depth about connectivity with one or the other. Alexandra also mentioned that ROS is becoming a topic that also plays a role now. I think we need to ask our listeners who have interesting topics, companies, solutions there, to please let us know. Then we can go into this in more depth from other sides. Otherwise, Alexandra, first of all congratulations on getting so far in such a short time, and during Corona, and I really do think that robotics is a good spot to start, and of course to participate in the scaling part. I'm keeping my fingers crossed for you that it works out the way you envision it. For you, I think it will be a very exciting year.

Robert:
Thank you Alexandra and best regards to Berlin.

Alexandra:
Thank you very much! I'm very pleased.

Robert:
And best regards to the cat.

[all laugh]

[Music] Robotics in Industry. The podcast with Helmut Schmid and Robert Weber [Music]

Staex is a secure public network for IoT devices that can not run a VPN such as smart meters, IP cameras, and EV chargers. Staex encrypts legacy protocols, reduces mobile data usage, and simplifies building networks with complex topologies through its unique multi-hop architecture. Staex is fully zero-trust meaning that no traffic is allowed unless specified by the device owner which makes it more secure than even some private networks. With this, Staex creates an additional separation layer to provide more security for IoT devices on the Internet, also protecting other Internet services from DDoS attacks that are usually executed on millions of IoT machines.

To stay up to date subscribe to our newsletter, follow us on LinkedIn and Twitter for updates and subscribe to our YouTube channel.