Transcript
Pablo Tomasi: Hello everyone and welcome back to the Telco Drift. I'm Pablo Tomasi and I'm an industry analyst and very excited for today's episode as we're bringing a new guest to the show.
Pablo Tomasi: So we have with us Mika Skarp from Cumucore Mika, welcome to the Telco Drift.
Mika Skarp: Thank you, Pablo. It's ah it's a great to be here.
Pablo Tomasi: Fantastic. And you know for those that you know that you haven't met yet, ah can you just provide a little bit of ah of a background about yourself and a little bit about your company? sort of you know Which part of the market are you guys working and sort of what makes you guys unique in what is a quite competitive private 5G market?
Mika Skarp: Yeah, so if i if i start start from myself, so I'm ah kind of a really typical Finnish guy working in ah in a telco sector. So my background is is in Nokia.
Mika Skarp: ah in I was there about 15 years working in the radio radio product management and delivering and developing new kind of ideas to the market. So my slogan is that you know from idea to production and then I left Nokia 2013.
Mika Skarp: We started a company called Cloud Street and developing the developing a network slicing solution, basically solving the same problem that what the private networks are doing.
Mika Skarp: But at that time, the frequencies, they were not available. So so we thought that network slicing could be kind of used for for these use cases.
Mika Skarp: um
Pablo Tomasi: Mm-hmm.
Mika Skarp: Then with KumuCore I've been now about five years. My role is is 10 years old company.
Mika Skarp: It's a university spin-off. Our home university, if you like, is Aalto. It's the largest technological university in Finland. and it's one of the largest technological universities in North north Europe.
Mika Skarp: um So our kind of background is is quite different to to compared to the many other companies that and we don't have any VC money kind of in the company.
Mika Skarp: So we have been cash flow funded from day one
Mika Skarp: And then we are doing, or let's say, supporting doing a lot of research. So originally the company was established to virtualize the core.
Mika Skarp: So that's where the name is coming from. Cumu it's coming from the Cumulus cloud.
Mika Skarp: Cumulus Cloud Core and
Mika Skarp: So then when you virtualize that, then you can, you know, do all kinds of stuff in the university. And then about five years ago, when I joined the company, then we started to look at the opportunities, you know, outside of the research area.
Mika Skarp: And that's where we are. that's where we where we are What makes the the other thing, what makes the different, you know, from the many other other core vendors is that ah we have done everything ourselves. so So we don't use any open source telco projects.
Mika Skarp: So everything is done wow from the scratch in in-house, taking the 3GPP specification and start coding. um and And then we have also the product structure where the one function is in one container.
Mika Skarp: So basically that and then the interfaces between the functions are also 3GPP compliant. So that basically means that if somebody wants to make some changes, so we can do those very rapidly that we know the code base.
Pablo Tomasi: Mm-hmm. Mm-hmm.
Mika Skarp: and we have the means as long as we don't you know change the interfaces so all kind of changes are fast so i would say that that's the kind of the heart of the company is to to provide open interfaces and and very very
Pablo Tomasi: Okay. I mean, you, you, you touched on many points that that I think would be interesting to, to further dig deep. One very quick one is that, you know, you said that as, as part of your history, uh, you know, you have looked into the network slicing thing when you were guys thinking, okay, we could do that.
Pablo Tomasi: And then you, you said, but, you know, spectrum was not available. And, and I, I found it extremely as an very interesting, a great point because, ah um, spectrum liberalization is the single trend underpinning private networks.
Pablo Tomasi: ah It is as simple as that. with that spectrum liberalization, I really struggle to see any market.
Mika Skarp: um
Pablo Tomasi: And and unfortunately, as as basic as the statement is, I still get a lot of pushback when I'm discussing sort of spectrum priority and liberalizations.
Pablo Tomasi: But the the other point that that I wanted to to quickly touch on, ah so it effectively, you're kind of fit for purpose if you want. ah So you are creating everything in-house and you have you know your target markets and you are creating a product that is specifically tailor-fit compared to what other people may be doing in the market. Is that a correct way of seeing it?
Mika Skarp: Yeah, it's a kind of a, it's in a company DNA. So, so that's ah because we are doing so many projects yeah with the universities and research institutes. So, so we are basically working on the future, future things so that we are bringing features on the market that that are the kind of for them.
Mika Skarp: the latest and greatest the technology. so
Pablo Tomasi: Mm-hmm.
Mika Skarp: so So working with the bleeding edge features, that's so that's what we do all the time. So that gives us a certain flavor also in our business.
Mika Skarp: So that they're doing doing things first is that is what we are really kind of what we want to do and what we what we are doing all the time so And then also, ah' there is a huge difference when you are doing some research.
Mika Skarp: Then if it works once or twice, it's enough. But when you take that into the production, there is a huge step. So so so that's also something what we do the time.
Pablo Tomasi: Yeah, and and and it's a little bit sort of the whole conversation about academia and getting sort of a development or a researcher indeed that sort of very good, but that you cannot really commercialize and actually getting something that is very good and can actually be commercialized because it serves a purpose.
Pablo Tomasi: within the enterprise clients. And that that is something, again, that is quite quite an interesting dilemma, if you want, that I've seen happening with um quite a if few companies. um But if if we look a little bit more sort of some of the key verticals that that I've seen for for private networks, for private 5G,
Pablo Tomasi: whatever we want to call it. And industrials, I would say, is difficult to argue. ah you know they are the the biggest drivers, the biggest verticals for private 5G, but by all means, correct me if I'm wrong.
Pablo Tomasi: And within industrial, and the big opportunity that lot of people are talking about is manufacturing. And partially is simply because when you start counting sites, there is like a X amount of ports.
Pablo Tomasi: And you know if you make it like that times millions, you get the number of manufacturing sites. But I'm curious to to hear your thoughts. Do you see something similar? Do you have activities within these industries or or you have a different picture?
Mika Skarp: ah Yeah, so so first of all, about the numbers. so So what we have studied is that there are about 20 million factories in the world, 2.2 million in in Europe, about 30,000 here in Finland.
Pablo Tomasi: Mm-hmm. Mm-hmm.
Mika Skarp: So the numbers are huge. um But when we talk about the manufacturing, so so I would like to kind of ah separate two different things that, you know, in the factory you have a traditional IT network.
Mika Skarp: And that's a something that that we have been kind of ah um where we have used the 4G and 5G private networks on the IT t network side. so So basically what does that IT t network do in the factory is that um you have a forklift driver and you need to give that ah person an information that go there, pick up that pallet, you know, take it to that other direction or you know that you have some um
Pablo Tomasi: Mm-hmm.
Mika Skarp: so some mean mean information that you you want to to distribute, for example, some kind of ah emergency alarm or something like that. So that's a kind of for the typical call IT t it t industry or i IT t network stuff in the in the factories.
Mika Skarp: But where I see a kind of a huge, ah huge and kind of a blue ocean opportunity is the on the OT side. So then the operational technology, so the factory automation.
Mika Skarp: And the factory automation today, it's based on the Ethernet cables. So my own background is actually actually in the from the um automation.
Mika Skarp: So I'm an automation engineer ah from the 80s.
Pablo Tomasi: OK.
Mika Skarp: And then, you know, there are still those same boxes and exactly the same kind of cables and so on that have been there for 40 years. And no, exactly. and um
Pablo Tomasi: it It doesn't change very, very rapidly, does it?
Mika Skarp: and no no exactly
Pablo Tomasi: So if it's not broken, don't don't touch it.
Mika Skarp: Exactly, exactly that's the that's the that's the point there. so so um So now when we want to change or you know move forward in the automation, factory automation especially, we need to get rid of the the signaling cabling, meaning the Ethernet cables.
Mika Skarp: and And there we see that the 5G can actually actually play a significant role of ah ah reducing the cabling. And well, that's the one thing, that's a one cost element. But then what what it also does, that is when you don't have those cables, your um factory ah layout will come significantly more flexible, that you can more easily move around machines, because then you know the only thing that you need to worry about is electricity.
Mika Skarp: And then you can add a new kind of things like drones or AMRs and those kind of things into the production process that are still controlled by the same PLC.
Mika Skarp: So what's happening on the PLC side is that it's also virtualizing.
Pablo Tomasi: Mm-hmm.
Mika Skarp: So we will have and virtual virtual PLC, then we will have a
Mika Skarp: connecting the sensors and the motor controls and valves and so on. So um that will come as a kind of a one package. so the if you like from telco words so the the the PLC will be an edge application in the network of course we can think about it another way around as well that you know that the 5G will be part of the PLC but doesn't matter so then they will kind of emerge together and provide them them completely new kind of and
Mika Skarp: ah flexible environment for for the factory automation. It's not going to happen overnight, but but during the next 20-30 years it will be a similar kind of a revolution that Wi-Fi and cloud has created for the consumers. So the similar kind of thing will happen on the on the factory factory side.
Pablo Tomasi: um I mean, I i like that you're, first of all, you're framing it in a quite quite a substantial you know timeline, which I 100% agree. ah But I also noticed that, you know, talking with a lot of telcos, that that overly aggressive expectation, overly aggressive timeline as something that I think has hindered a little bit of the of the appetite from them because when you're expecting to revolutionize an industry in two years um and when that doesn't happen, some people start asking questions. Well, you know, it it is going to be a long, like we we know these are very, you know, conservative industries, is industry that takes a lot of time to make changes. And at the same time, this is quite a big change. You know, it's not just a mild upgrade, you
Pablo Tomasi: or change of type of cable is is really kind of ah re rethinking the whole idea of manufacturing in a sense.
Mika Skarp: It does, it does. and And then it's also that, you know, them if you think about the lifetime of the factory, so it's maybe 30, 40 years, and then, you know, in in in the middle of there, there is this kind of a major...
Mika Skarp: Major um um change in the, or you know, you can make the changes is in the middle of that lifespan or so. So so they them the timing when we are talking about an industry is different than when we are talking about the consumers.
Mika Skarp: But even from the consumer side, so so if you think about that, you know, we got the first digital phones somewhere in the 80s and it took for quite a long time because it went to the mainstream.
Pablo Tomasi: Mm-hmm.
Mika Skarp: And then you know the mainstream was still the kind of a voice-centric world. So um even on the consumer side, it wasn't really you know that quick. Of course, you know now when we are thinking about it backwards, it feels like a very rapid change. But in in well you know when I'm remembering the time, it it wasn't that kind of... It went small steps forward. And I think that will that we will see also in that in the in the factory automation, 5G coming into the OT.
Mika Skarp: OTA networks.
Pablo Tomasi: and and And I think from from my point of view, sort of as an analyst perspective, as long as it's happening, that's good for the industry. It doesn't need to happen tomorrow. It doesn't need to happen in a couple of years.
Pablo Tomasi: ah it's it's ah It's about keep driving that momentum and it will take whatever time it needs to take. Effectively, I mean, from an analyst point of view, it is kind of easy to say that because I don't have revenues attached to it.
Pablo Tomasi: So I can be a little bit more detached. I know that, you know, if you're in the sales business, probably have a little bit more pressure to to hit certain targets.
Mika Skarp: yeah Yeah, there is that. But, you know, from the if we look at if we take a little bit you know higher views, so in the societies, in the Western societies, and even in China, you know, people are getting older.
Pablo Tomasi: Mm-hmm.
Mika Skarp: So so we are we are kind of seeing the situation and that... we don't have enough labor to do everything what we want to do. And then then at the same time in the Western countries, we are talking about that we will start manufacturing again, you know, all kind of stuff here.
Mika Skarp: So that's, you know, if we keep the same way of manufacturing, so then then we need significantly more labor and we don't have that even to keep the existing manufacturing.
Pablo Tomasi: Mm-hmm.
Mika Skarp: So it means that ah that those kind of stronger forces are actually driving the technology towards kind of a more automation.
Mika Skarp: and and And I think that will be also something that what what helps this to happen or it's kind of a forcing it to happen. And in China, what we What we, you know, the last Mobile World Congress when I was talking about this on my small booth and then there was one Chinese guy, you know, I was talking about this is going to happen in the future. And then there was this one Chinese guy coming to me and saying that, you know, there is Mika something I want to show you. And then, you know, but we went to his booth and all this thing that virtualized PLC, 5G connecting sensor, and it was already there.
Mika Skarp: in in production. So yes, it's happening and it will happen. Of course, the timing is always that, you know, I would say it's relatively easy to predict what's going to happen in the future, but it's very difficult to say when.
Pablo Tomasi: Yeah, I mean, definitely I I agree that it's difficult to pick up the right moment in time. on On China, I agree. I mean, i've talking I'm talking with a lot of Chinese ah players and they are ahead, I would say, for compared to a lot of companies. And I've read and I've seen some of the deployments that they have and they're They are quite impressive, to to be honest. is like It's good to see sort of what's feasible from you know a pure tech point of view, if you want.
Pablo Tomasi: And we all know that sort of the business model is different in China compared to to other countries.
Mika Skarp: Mm-hmm.
Pablo Tomasi: but But you can see, okay, once I I don't have the sort of issue of business model and maybe have some government incentive, this is how fast I can go. This is you know where I can bring my factory in terms of driving that automation.
Pablo Tomasi: ah But if if you're thinking you know factories, ah you know automation, private 5G, it's something that ah has been talked for for a while. It's TSN.
Pablo Tomasi: ah and And I'm curious to hear from you, how how do you see the TSN relation with 5G? Are we there yet to get the two things working together?
Pablo Tomasi: And maybe yeah if you can give a quick introduction on TSN for our audience.
Mika Skarp: Okay.
Mika Skarp: Yeah, so so TSN TSN is ah is a time sensitive network. so So basically what what the TSN does is that it synchronizes the UE's and it's working on the layer two, so you on the on the Ethernet ethernet layer.
Mika Skarp: So when we synchronize the UE's, so then we can also schedule when those UE's are transmitting. so ah So that way we are able to ah kind of ah prohibit gap that basically all the all the packet collisions.
Mika Skarp: Meaning that the bit error rate is significantly better than what we are seeing or what we have kind of seen in the mobile network. so In the mobile network, the bit error rate is maybe something like 10 minus 3 10 minus 4 With the TSN, we can we can go into the 10 minus 11 and so on. So so it's equally ah reliable compared to the cable.
Mika Skarp: um And that's the kind of prerequisition for that, that you are able to use.
Pablo Tomasi: Mm-hmm.
Mika Skarp: That if you imagine that ah there is a factory line and then there is an emergency stop button. So when you press that emergency system stop button, it has to stop. There is no you know option that there is some failure. So it has to be and significantly more reliable. And that's what the ts and TSN and does. So that's when we synchronize the UVs and then then we schedule them, ah or give them the scheduling when they are allowed to speak.
Mika Skarp: communication So But the TSN as such, it doesn't do anything. It's an enabler. It's a feature on top of the 5G network. and then the applications like emergency buttons, like the drones, like the sensors and so on, they are then able to to to use it.
Pablo Tomasi: Mm-hmm.
Mika Skarp: um The TSN, it has been an um kind of around for for a long time, so it has been on the on the fixed side and you can do that over the Wi-Fi and and different different other other medias.
Mika Skarp: so ah So as such, it's very robust robust technology. ah But when you kind of add it on 5G, then it's kind of a easy to to understand when you but but where or how how complex that system is. thats ah You want to have a standardized ah delay that should be very there is as small as possible.
Pablo Tomasi: Mm-hmm.
Mika Skarp: And then you need to kind of... um buffer to delay the way that the accuracy in the in the system is very high. So we are talking about a microseconds, we are dreaming about nanoseconds.
Mika Skarp: So that's ah so when we ah the nanoseconds refers to the accuracy, that how much there then how much what is the difference between the clocks inside that inside the network.
Mika Skarp: so they are not allowed to drift too much from each other. So and then you know we have an air interface where the delay is depending how far the the the user is and typically also the delay is depending on how much load we have a network and so on.
Mika Skarp: So it's a feature TSN is a feature that you need to have a UE supporting it, you need to have support from the from the base station, and and then you need to have a support from the core.
Mika Skarp: So it's truly kind of from the mobile network perspective, it's an end-to-end feature, complex thing to do. To take a look at it where we are at the moment, so So we were able to demonstrate live first time in Helsinki 5G ACIA conference end of June ah together with Qualcomm and Node 8 running e Askey hardware.
Pablo Tomasi: Mm-hmm.
Mika Skarp: So basically we have a commercial hardware where we were showing it live that you know It was stable over the over the the demo, one day demo there. so So it was a great success. It was really great.
Mika Skarp: um And now the the thing is that so we are on that productization path that ah that we are bringing in the features, making it you know easier to to to work with that that's ah um that you don't need to be a PhD to get it up.
Mika Skarp: up and running. So so it's still a work in progress there, but ah we are we are progressing. We are making kind of improvements on on daily basis. And and ah during the year during this year, we we will we will be come out with that with some new new demonstrations and and then hopefully hopefully soon soon enough we can we can go into them, a proof of concepts. things
Pablo Tomasi: Oh, exciting
Mika Skarp: but yes yeah But as I said, you know this is something that I want want the audience to kind of remember. TSN doesn't actually do anything, it's an enabler, it's something that ah you you have to start in building applications on top of it.
Pablo Tomasi: Yeah, I mean, when when I'm, and and it does it does feel complex, TSN I'm not going to lie. it's It's not one of the easiest things to, to to understand or to analyze. ah but But from my from understanding, effectively sort it creates that an ideal.
Pablo Tomasi: I don't want to use platform, but everyone uses platform for everything. ah So I'm going to just use platform. It creates the sort of platform, if you want, that allows you then to build better applications and to to have better results.
Mika Skarp: and
Mika Skarp: Yeah, well it's ah basically an enabler for 5G to be used in the in the factory automation. So that it comes from that... Well, it's actually two things that these if we think about like for example a profinet protocol so it's it's ah running on on on layer two so it simplifies the world significantly if we can have a layer 2 5g network even without without tsn so so there is a feature called ethernet pdu so uh so that's the first thing but then the TSN uh enables the
Pablo Tomasi: Mm-hmm.
Mika Skarp: the scheduling of the of the of the UEs basically then making them reliable of course you can also achieve a similar kind of effect by over dimensioning the network so that you know what has been what is the i would say it's the the most typical thing to solve the problem is over dimension that you know if you don't want to calculate something
Pablo Tomasi: Throw more stuff at it.
Mika Skarp: It's good enough. And who knows how it's going to be in the future? That's an also an alternative. that you know don't you know You don't need to think about it, just put enough capacity and then it's a route.
Mika Skarp: Obviously, it's more expensive.
Pablo Tomasi: And...
Mika Skarp: but so
Pablo Tomasi: Mm-hmm. But that's sort of like a quite a question that that came to mind. So when you were mentioning you know how how you can sort of deploy this ah the TSN on 5G, if you need to expand your network at a later stage, would that require you to do some significant re-tweaking of what you already have?
Pablo Tomasi: that That is just...
Mika Skarp: um
Pablo Tomasi: Because I would imagine you're fine tuning the network so that, soto as you were mentioning, the delay and everything is working a certain way. Now, if you want to expand the network, if want to add something else, would that require you to do some significant work?
Mika Skarp: yeah yeah Today today it's actually it's still the case that you know setting up the system it takes takes quite a lot of effort.
Pablo Tomasi: Mm-hmm.
Mika Skarp: But we are also, as I mentioned, that we are building it or we are making it easier to use. So there will be an application function that you know makes that management layer and and significantly easier.
Mika Skarp: And there, as I said, you know TSN as such has been there for a long time.
Pablo Tomasi: and Okay.
Mika Skarp: So there are those kind of... um those kind of ah management tools, ah kind of a management ah standards available, maybe tools available, but they are you know designed for the fixed fixed line.
Pablo Tomasi: who
Mika Skarp: And the the difference there is that in the fixed line, the delay is basically constant.
Pablo Tomasi: Yeah.
Mika Skarp: So there is no delay variation as such.
Pablo Tomasi: Mm-hmm. Mm-hmm.
Mika Skarp: but Let's say the cheater is significantly less problematic than you know what is in the mobile network. And that's the that's where where the the difference comes and and and and how do you need to kind of develop further or develop newer newer software software tools to to provide the management that too.
Mika Skarp: Because you know if it's too complex then I think what in general what we have seen in the in the in the private mobile networks is that you know people are considering them to be too complex.
Mika Skarp: And I have to admit that you know if you look at the Wi-Fi, basically, if you are an engineer, you can set up a Wi-Fi network at your home.
Pablo Tomasi: Yeah.
Mika Skarp: But it's still, you know in the 5G side, it takes and a little bit more than that. So then, you know, if you imagine that 5G is already quite complex, so then if you add, you know, some more complex stuff on top of it, then the complexity is too much for sure.
Pablo Tomasi: yeah
Mika Skarp: So we need to kind of reduce the complexity as much as we can.
Pablo Tomasi: Yeah, and and I agree on that point because I've i've heard a lot of enterprises that you know are are sort of complaining about the complexity and they don't know the technology. they may not not They may not have specialized experts.
Pablo Tomasi: And therefore, that that does become a little bit of a roadblock. but But it seems that you know when we're thinking about 5G and TSN, we are going in the right direction. As you mentioned, you know you're getting results, then the next step is sort of to keep building on them and then and making sure that you're taking the the more business-related challenges, if you want, of of hiding that complexity ah towards the the market.
Mika Skarp: Mm-hmm.
Pablo Tomasi: um So we are running pretty much on top of the time that that we allocated for this. But Mika, is there any any last ah last thought or recommendation that you would want to give to the industries regarding sort of the the relevance of automation, TSN, 5G, or anything else?
Mika Skarp: Well, um I think I'm kind of for my message is that, you know, the world is changing very rapidly. The AI is coming and them the productivity needs to needs to kind of increase. So um looking um kind of be um curious for the future, looking at the opportunities and and think about that, that's what can we do better um with the with ah with the less less resources. So I think that's the that's the thing, be curious and then, you know, go after the productivity.
Pablo Tomasi: a bit of a A bit of urgency. It's good good to put a bit of urgency at the end of the podcast. Well, Mika, ah thank you very much. I really appreciate it. And thank you very much for dialing in.
Mika Skarp: Thank you, Pablo. It was my pleasure.
Pablo Tomasi: And have a great day, everyone.
Mika Skarp: Yeah. Take care. Bye.
