By James Blackman

Another late entry in the MWC archives; this time with Finnish private networking specialist Edzcom, latterly of Spain-based Cellnex Telecom, which has blazed a trail for LTE and 5G in the Industry 4.0 market. It is the interview Enterprise IoT Insights has been seeking for two years, or more; its absence from these pages probably reflects its own work schedule, and ours (plus other failings on our part, perhaps). But clearly, it is a busy time for all.

As we meet, in a back room of the Cellnex stand in the Barcelona Fira at the end of February, the company’s chief executive, Mikko Uusitalo has just emerged from a photo opp, standing with some other execs, to go with an announcement about its first deployment in Germany, the home of Industrie 4.0, and also its first with the automotive industry. Edzcom, says Uusitalo, is to install a private 5G network, from Nokia and Athonet parts, for the automotive branch of engineering group SEGULA Technologies.

The others in the frame, it turns out, are Oscar Pallarols and Didier Pagnoux from Cellnex and SEGULA (pictured, left and centre with Uusitalo). This is different, says Uusitalo, to Edzcom’s deal for sports broadcasts and vehicle testing at the KymiRing race track in Finland; the new install is at SEGULA’s test centre and proving ground in Rodgau-Dudenhofen, near Frankfurt, and is geared for next-generation automotive development, covering advanced connectivity systems, right through to modish 5G-based vehicle-to-everything (V2X) experiments.

“We are looking at the transport sector very closely. It’s still early days [for driverless cars and intelligent transport systems]. But ultimately all these cars will be more connected than today, and 5G connectivity and edge computing will be a part of the package,” comments Uusitalo. It is interesting, for sure, but not as interesting, yet, as the rest of Edzcom’s 40-odd private network deployments, which have focused mostly in the Nordic region with port groups, mining companies, and industrial machine makers – as these other references go to the heart of Industry 4.0 today.

It is not as interesting, either, perhaps, as Edzcom’s assessment, apparently, that industrial 5G is best served via an Athonet-Nokia joint-ticket. Certainly, they are the common thread in its infrastructure choices, so far, although Uusitalo also makes clear they are not exclusive arrangements. He says: “The philosophy is we are neutral, but obviously we’ve selected partners that really perform. For the radio part, Nokia and Ericsson are more advanced; Nokia perhaps a little more so than Ericsson. On the core, there are many vendors, but, yes, we like Athonet a lot.”

So just explain; why not go with a Nokia core, as well, if Edzcom is already working with its private RAN components? Is it good business to disaggregate these things? Or is it just that Athonet just has a better core network than Nokia? Uusitalo explains-away the logic of RAN and core selection. “It’s not that black and white, clearly,” he responds, detailing issues of spectrum availability and radio capability, budget constraints and application requirements, as well as mutual business benefits.

He explains: “Everything starts with spectrum; that is still the key component. The market is quite fragmented in Europe, still, compared to the US, which has enabled it with CBRS for everybody. The EU should have done the same for enterprises. Instead you have to go nationally, because spectrum is regulated nationally, and pick the radio vendor accordingly. That is the first decision. Because not all the vendors – not even Nokia – support all the bands. The core network, meanwhile, is more of a commodity, but stability features and time-to-deploy are important. Because you still have to configure the boxes at the end of the day

“Even if it is becoming more automated, it still takes a little time. And all of it is driven by the use case and the cost – the demand for geo-redundancy, say, and the bidget availability. I don’t think there’s necessarily a lot of difference really between Nokia’s DAC core and Athonet, actually; we still work on a case-by-case basis. And obviously we like to work with these vendors on the demand side, too – so we get leads from Nokia sometimes, and Athonet as well. So it’s the sum of many things.”

We need to move on from Nokia; we have written enough about it, recently, and the Finnish firm is pitching to a niche part of the private 5G market, arguably, outside of the kind of ‘carpeted’ verticals that are lighting up CBRS spectrum in the US, momentarily. But Edzcom’s view on this counts. So is its selection influenced, also, by how ‘industrial’ the deployment is? Because the story Nokia tells, increasingly, is that it is going after ‘critical industry’, at the critical edge, where school districts, hospitality venues, and some logistics centres might be left to lighter-weight providers.

Is that right? Are these industrial 5G vendors occupying, and destined to occupy, different industrial 5G planes? Should we consider this when assessing their present and future successes? Uusitalo, to his credit, is relaxed enough to have a stab at an answer; it comes down to the size of the opportunity, which increases with both the complexity and scale of the operating environment, and which is only captured effectively on the vendor side with the size of the capability and experience. Edzcom, like Nokia (and Athonet), is focused here, he explains.

He says: “It is a difficult question because there are multiple dimensions, as well. I think the money today is in complex environments; the volume is in manufacturing and industrial enterprises. And I think Nokia is a leader just because it has been in this space longer than any other large vendor. It is focused today, the same as us, on large value opportunities. It is not doing proofs; it is doing real implementations. A lot of these other players, which are newer to the market, are still in the early days in POC environments.”

The “other players” are implied, rather than spoken. But Enterprise IoT Insights is happy to provide (some) names. Cisco and HPE have just entered the enterprise 5G market with Release 16-level core products; their positions, at the outset, appear to be defensive, to cross-sell 5G into Wi-Fi channels, in case of cannibalisation of their industrial networking estates. Microsoft and AWS are teeing up network components for (mostly carpeted) enterprises, as a baseline for a developing IoT horn section – to annotate a crashing Industry 4.0 chorus in software (for everyone).

It is confusing to know who plays where – or rather, who will play where. “It is confusing for somebody who doesn’t know the market,” responds Uusitalo. Ouch. But that’s okay; everyone is playing the same game, also, at the same time, he implies. There is work to do in the technical area, he says, for “highly competitive” vendor teams to simplify 5G so it scales more easily for Industry 4.0. “You need a super simple installation, and that is still being worked out for manufacturing. You talk to AWS, you talk to Microsoft, you talk to Nokia – they all speak about simplicity.”

He restates the order-of-service. “The drivers are still spectrum, radio, core – in that order. And then, how you design and how you operate [the infrastructure].” We can have the same conversation about private 5G design and management, and late-emerging contenders, as we have had for the private 5G parts, but we should pause, and ask about the Edzcom story, which traces back to 2014, when Uusitalo and chief tech officer Kari Lehtinen found themselves “just kind of by accident” in possession of a “few bands of spectrum” in Finland.