Integration, simplification, multiplication – three signs private 5G has come of age
This is a rehash (and extension, in the end) of a previous post, to an extent; but it distils the message from before, and says there are sure signs, at last, that the private cellular market, covering localised installations of enterprise-geared 5G, nominally, has moved out of the industrial sandbox. It says that, even if private-industrial 5G remains a complex sale, the supply-side has started to understand the game and the demand-side has started to trust the relationship.
But we should not get carried away; triumphant (or triumphantly-received) notices from analysts about the private-5G run-rate reaching 1,200 installations already, projected to stretch to 12,000 in the mid-term, seem laughable (and probably disputable) – considering the total addressable market is supposed to stretch to 15 million-odd venues. At the same time, the growth curve is certainly getting steeper, which might just be down to the following items.
1 | INTEGRATION
Private 5G is not supposed to exist in splendid isolation; if it does, it will fail. It must grow as part of a dynamic and evolving enterprise networking environment, crossing into the IT and OT domains, and priming both for a new generation of digital-change gadgetry. Recent announcements, then, from Nokia and Siemens must be considered as positive examples, if not unique innovations, of its progress as an industrial-grade ecosystem technology.
Nokia is adding Wi-Fi 6 and 6E to its seminal Digital Automation Cloud (DAC) private networking product, available from late 2022. DAC support for Wi-Fi is included for non-critical industrial applications, it says; the Finnish outfit maintains cellular connectivity – from easily-available LTE, through early 3GPP Release 15-era 5G broadband, and on to incoming 16-through-18 level industrial 5G connectivity – should be preferred for critical operations.
In tandem, it has introduced a performance booster (MX Boost) to allow enterprises to automatically switch between optimal local-area LTE/5G and Wi-Fi radio tech and spectrum. These developments – which may or may not be the first, but will be the norm – make clear 5G will co-exist with Wi-Fi in the same venues, systems, and solutions. Talk about private 5G replacing Wi-Fi (or Wi-Fi 7 replacing 5G, or 5G-R18 replacing Wi-Fi 7) is cheap, as we know.
The same can be said of Siemens’ parallel announcement (also timed for Hannover Messe) about integrating Profinet into its industrial-grade Release-15 5G devices – as well as into its prototype private 5G system, available next year. Profinet (process-field-network) is the popular layer-two OT protocol for time-constrained two-way comms between SCADA systems and industrial machines; Siemens’s Scalance MUM853-1 5G routers now support it,
It might sound like minor technical philandering, but the move represents an important consummation of cellular’s entry into the industrial domain. Siemens reckons no other industrial 5G device, as yet, provides off-the-shelf support for layer-two connectivity – which is the only way industrial machines will communicate deterministically over 5G networks, and therefore the only way ultra-reliable (URLLC) 5G will bring order and control to smart factories.
Nokia, and some others, might say different on layer-two OT protocol support, but it ain’t the point, just as it is not the point that Nokia is trumpeting its timing with Wi-Fi integration; the point is, only, in both cases that their work shows the direction of travel, that private 5G is being joined with entrenched and vital IT and OT systems – as a complement, more than an alternative. This is important if private 5G is to scale.
2 | SIMPLIFICATION
Another clear sign that the supply-side is readying to popularise 5G with enterprises is that complexity is being abstracted, both from the radio technology itself, and from the process to put it to work. More than this, complexity is being abstracted – or rather, process is being documented and digitised – within enterprises. This way, and only this way, the challenge to knock heads in co-creation is being more quickly resolved, recorded, and replicated.
Again, this is the way to scale; and there are some examples to illustrate this three-way simplification. But we have to ask Nokia again – and Nokia should, rightly, have its say, just because it has (either) worked this market hardest (or communicated it best). The classic private 5G case is from aviation services outfit Lufthansa Technik; again, with no apologies for editorialising, it is the best documented deployment of an (two-part) industrialised private 5G system.
Lufthansa Technik’s big challenge, to start with, was how to work with the macro-sized gear it took receipt of from Vodafone back in 2020, for one hangar in its two-hangar deployment; Nokia’s delivery for the other side was smaller, arriving in a hatchback. This anecdote tells how 5G has shrunk-down for industry; but the task to make it compatible for Industry 4.0 has required it to be de-telco’d as well – so the art of radio is for pop artists, as well as old masters.
This process is not exclusive just to cellular radio engineering; parts of the non-cellular low-power wide-area (LPWA) market have been dumbing-down IoT networks forever, and the best brands for IoT sensing and AI sense-making are doing the same with embedded computing and software engineering. Because their endeavours are interrelated; the end-game is the same – to popularise digital tech with the industrial set, and make Industry 4.0 a ‘no-brainer’.
In each case, the goal is to render once-rarefied technical knowhow into easy modular, visual, drag-and-drop control functions – effectively to do with radio planning and algorithmic software what Orange did with consumer 3G comms back in the day, just by getting picture-book illustrators to rework welcome books and instruction manuals. Again, Nokia’s sanitary work to tally enterprise itches with technical scratches in apothecary-like blueprints should be applauded as another balm for the headache of Industry 4.0 problem solving.
But in the end, the north star for any such discipline is the cloud model for compute services. Hyperscalers will bring private 5G to the mass-market, logic says. Their almost-total dominance of both digital mindshare and global economy, now conjoined in industrial revolution, makes them elephants in every boardroom in every sector. Their late entries into the private 5G market, through partnership or acquisition, means their background Industry 4.0 herding is amped-up and distracting, and liable to trample into the foreground at any moment.
The whole scene will be remade, and made simple, the story goes (and the global economy will be even more lop-sided, probably). But a third simplification, to document and digitise demand-side processes (in higher coal-face resolution than the Nokia blueprints), must also happen for any of this industrial magic to pass. NTT understands this, like others no doubt; but its quiet partnership with US software firm ServiceNow has rightfully turned heads.
The problem, says NTT, is that most factory engineers can show an industrial process on a scrap of paper (“on a two-by-four yellow card on the wall”), but not always in a digital file – and certainly not in a digital twin. Which is alarming, if it is true, or even just half-true; because, how do you sell – and why would you buy – a private 5G system if you don’t know where to attach the cameras? Or, how do you map the solution if you can’t map the problem?
The ServiceNow software resolves this with “AI-enabled end-to-end workflow automation”. In other words – like Nokia’s dogged work to cross-reference industrial process requirements with functional 5G capabilities in a takeaway menu for developers and integrators to spur the ecosystem – NTT can access and expand a down-and-dirty library of digital workflows to animate into its 5G systems. Again, all of these simplifications say cellular is getting easier for industry.
3 | MULTIPLICATION
Which brings us full circle to the first signs of this supposed mega-scale – which the cellular market is so convinced by, the broader IoT sector has always fancied, and these other progress-items make possible. Because when the system is lined up, the dominos will fall down; these are not national infrastructure projects by a few vendors and a few operators, but private pursuits, directed by money-minded enterprises, with money-eyed partners in tow.
And the first signs of proper scale are there, with serious-looking installations by the likes of Lufthansa Technik, most notably, as mentioned, but many others besides, and single multi-site orders now going in with Airbus, Dow Chemical, and Schneider Electric – to list only a handful of headline references, across a whole bunch of industrial disciplines, and also to stick with Nokia and NTT, which have helped carry this narrative along.
Nokia reasons that it has done the donkey work of industrial-learning and direct-selling already, and is in position now to take a well-worked private 5G system to market via third party channels. One of those – the biggest of them, a channel all by itself, practically – is New York-based managed infrastructure services provider Kyndryl, separated from IBM late last year, which the Finnish firm signed with in February to jointly sell Industry 4.0 solutions.
It is a move that says 5G can be levered into industry with confidence, without complicating the rest of the digital change architecture, and made to scale quickly. “Customers are signing up for 40-50 sites in one go, and buying the whole industrial edge; saying, ‘yeah, I’m all-in’,” says Nokia. “That is one of the big changes.” In a heartbeat, Nokia’s deal with Kyndryl will see a one-site installation explode into a 200-site deployment with Dow Chemical.
The project with Dow Chemical promises to multiply to other mega-sized industrial clients; Kyndryl has 4,000 of them, it says. Meanwhile, NTT has a new arrangement with Schneider Electric, which, if it can get an initial conveyor belt case to work, will scale private 5G very quickly into its similarly-specified plants – from a factory footprint of 200 sites across the globe. The deal also includes a unique resale component, which will see Schneider Electric integrate NTT’s 5G system into its prefab data centre proposition for its own customers.
Scale beckons, like we said; these look like the kinds of deals that will finally line-up and topple-down this domino-line of industrial 5G installations. These are the deals that make the hype look like it might be justified; like industrial 5G might just be the earth-quaking reset for industry. Suddenly, the cookie-cutter approach seems a likely story; NTT’s own system integration, about bashing heads in co-creation, looks likelier.
But what is left? The small/mid sized (SME) market, of course. Because industrial 5G will not amount to a hill of beans if the great engine room of Industry 4.0 is not well served, also. In the end, 5G will not matter much if Industry 4.0 does not go beyond well-heeled mega-sized early-adopter companies like Dow Chemical and Schneider Electric, except as a technology that spreads digital-divide into the industrial market.
So how does Mr. Man up the road with 50 staff putting rivets in steel joists get a role in this tale? How does the Rust Belt and the Northern Wall, and all these clichéd industrial heartlands which carry the Industry 4.0 supply chain, make a go of 5G-geared digital change – when they are just trying to keep the lights on? Come on, my brain is tired; tell me about the kind of scale that matters.
Ah well, we are back at square one (and two); to the simplification phase, with its hyperscaler model for easy enterprise services, and to the integration phase, which will dovetail with the enterprise reseller model for SME engagements. But for now, the radio play is probably too complicated, and the sales channels are not ready. NTT admits as much; Nokia is focused, first, on the big beasts of industry with Kyndryl, while the solution gets easier for everyone else – and properly comes of age.