By James Blackman July 8, 2024
Note, this article forms a (middle) part of a recent editorial report from RCR Wireless about ‘private 5G in Industry 4.0 – hype versus reality’. The full report, which extends the discussion further, is available to download here – for free.
To an extent, all of this – simpler systems, newer verticals – is being combined in the post-hype private 5G era with clearer thinking. The business of private 5G is quicker, all of a sudden, in terms of both the sale and return, says Stephen Douglas, head of market strategy at test and measurement company Spirent. “We have seen really rapid returns-on-investment. A number of end users – our customers’ customers – are posting an ROI in six months on a small set of targeted use cases. So you don’t need hundreds of use cases anymore to justify why you are doing it. With a low-cost implementation and a couple of use cases, the return can be very quick.”
But the biggest breakthrough in the Industry 4.0 game is just that the supplier market is match-fit.
He explains: “Enterprises have had this day-one impression that cellular is really expensive. Maybe it’s a historical thing; where they’ve engaged previously, and come away with the idea it is five or six times the cost of Wi-Fi. Which just isn’t the case. I don’t think the vendor community communicated that very well. Maybe because the operator community, in particular, wasn’t at a stage, and in a good place, where it could demonstrate how to achieve the ROI. It is much better at that now – because the ecosystem is working better, and these lighter-weight solutions are coming into the market.”
But is it match-fit, really? Is that the difference? Certainly, by its own account, Verizon Business has changed-up and limbered-up; other operators will have tuned-up, too. But the rest of the supplier market, comprising a right-on range of long-sighted equipment makers and laser-focused specialist integrators, know their Industry 4.0 onions very well and have worked these channels for years. Something else has changed in the background, to reorder the foreground – while the industry frets generally about procrastination in the manufacturing sector. And the thing that has changed, besides simpler 5G systems, is that the industry has hit on a killer app.
And – get this – it is called connectivity. Shock, horror; exclamation point.
KILLER APP
Rajeev Shah, chief executive at Celona, comments: “Easily 80 percent of deployments are just about providing connectivity to people. Literally, these refineries have these small pockets of Wi-Fi, and then absolutely nothing. It’s kind of funny to go there; they have these golf carts to take them around, and these weird speed limits of 17 miles per hour, and they take a notebook and pen and tour about, and then return to these pockets of Wi-Fi to do data entry. And they are telling us that, one in five times, they make a mistake – and so they have to go all the way back in this golf cart, record the data, and then return again to get a Wi-Fi signal. And it’s completely inefficient.”
It’s weird, as well, because Nokia tells almost exactly the same tale, in a proper discussion, when relating the origin story for its new generative AI prototype, MX Workmate, which attaches to its DAC and MXIE systems as a kind of industrial chatbot for factory workers. “In terms of product deployment, the easiest use cases to implement are the ones that focus on human workers,” says Stephane Daeuble, head of solutions marketing in the Finnish vendor’s enterprise unit. “Because they have the most impact.” He explains how the whole project started after observing a worker in a client factory go to retrieve a part, only to return again for a pallet loader to do the heavy lifting. “We were like: ‘Isn’t that where you need automatic dispatch of an AGV?’”
He adds: “And then we got with Bell Labs, and put two-and-two together – to have an AI engine figure things out, and an LLM to converse’. And then you’re like, ‘Okay, the pieces fit’.” He reasons, as well, that “the most successful apps meet a fundamental need”, and that private 5G generally finds its mark, initially, just with team-comms apps. It is the same logic, whether for the base-level network or the top-level application: manual processes are made impractical by smart connectivity and clever software. The ROI sum flows from that. “Because humans in charge of complex machines can call a remote expert – versus having a bloke walk 20 miles every day with a sack of books.”
The conversation with Nokia is played back to Celona. “Exactly. One hundred percent. Nokia is right,” responds Shah. “Once you connect the worker and deploy the network, then it’s easy to layer-in IoT use cases. Because you have the network, and it’s paid for. The great challenge with all these IoT / OT applications was to find that trigger case to justify the investment. Connected-worker is the catalyst. It has taken away a lot of the speculation – which goes back to the point about the hype. Otherwise you’re searching for these very fancy use cases – with augmented and virtual reality, and whatever. But all of that has gone away – because it’s as easy as workers-need-connectivity.”
Arvin Singh, global head of 5G solutions at Verizon Business, says the same, as well, talking about how 5G has found its métier outside of a strict Industry 4.0 footprint. “In most cases, to start with, it’s a coverage need, getting into the basics and starting with a small footprint,” he says. “And then expanding [use cases] in the same physical environment or expanding to multiple sites.” And so, even as his company scales “horizon-two/three” projects in oil-and-gas and pharmaceutical production (and discrete manufacturing is pushed to a horizon three/four schedule), plus in a bunch of ports, many of his company’s ‘land-and-expand’ gigs are in so-called ‘carpeted verticals’– in retail, offices, venues.
He responds: “Yes. It’s less of a rugged conversation, more of a carpeted one. There are lots of sort-of corporate next-gen campus networks [to start with, in industry, too]. But it’s modular – so you add capabilities all the time.” The point, in the end, is that the original Industry 4.0 hype around ‘smart’ factories – captured in the imagination as ‘lights-out’ automotive manufacturing, as the sharpest application in the sexiest discipline – still holds; it is just on pause. Industrial 5G capabilities and devices are in the works, and will come available (starting post 2025, in flow by 2030; see graphic, page 16 of report) and gain sufficient trust that hard-nosed Industry 4.0 (discrete manufacturing) scales serious-sized systems.
BLAME GAME
And in the meantime, there is good business to write with big-sized firms elsewhere and mid-sized firms everywhere – the story goes. So on one hand, the hype is probably justified in the general and long term insofar as private 5G has a grand (yet unknowable) role in a ‘fourth’ industrial revolution. And on the other, the hype is completely screwy in the specific and short term on the grounds that total market sales are still slow, however big the little victories for individual suppliers, and most private 5G deployments are not about AI whizz-bangery, at all – and hardly about 5G, actually. Most deployments and sales are LTE-based (just take a look at the Nokia figures on page 14 of the report). “Everyone wants 5G,” says Peter Cappiello, chief executive at US system integrator Future Technologies.
“But industrial companies have Apple devices with support for Band 48 – so they need LTE in there.” He describes a back-and-forth conversation where customers have to square the marketing promise about future 5G releases with the coal-face reality about existing ones before they commit to a deal. “We spend so much time on this. It’s almost like too much common sense,” he reflects. Indeed, the hype machine has coughed-up a hot mess of half-truths for resellers to unravel. But there is a clear sense that some kind of order has been restored, or realised, as well. Is the disillusionment because a developing technology has been blamed for the stop-start failure in manufacturing?
And is the difference, now, that the industry has just moved on – in the knowledge there is enough in it to address other opportunities, while it matures? Shah responds: “It goes a level deeper. The people making the robots and headsets for these fancier use cases all say the same thing: but is there a critical mass of 5G networks? They know it is the right technology, but they need the networks before they put it into their products. It’s chicken-and-egg; but it is a virtuous cycle, as well, right? Enterprises are justifying the investment because of this connected-worker case, and then going to their robotics companies to say, ‘We have the 5G network, and we want the 5G equipment’.”
It goes the other way, too. Celona is now selling private 5G systems to enterprises via AGV manufacturers; it has sold three in the last two months, it says. It will not be the only one. The wheel is turning; the chicken and the egg will appear together, at some point. Some version of the hype will be real.
MAPPING METRICS
The other aspect, referenced already, is that the cultural approach to selling digital-change has shifted, notably among mobile operators. Asked if closer alignment of its 5G, IoT, and edge componentry in its new structure has helped Verizon Business to better-design enterprise solutions, Jennifer Artley, senior vice president of 5G Acceleration, at Verizon Business responds: “It’s less about that. I mean, it’s a lot about that. But it is really about business outcomes.”
She explains: “We’ve gone from selling technology to having conversations that bring OT and IT together. And that’s a skillset; that’s a unique skill-set that a lot of general core [network] sellers don’t have. The ability to engage with customers in a different kind of conversation – that’s what has accelerated this.” Which is about zeroing-in on ROI, ultimately. Indeed, ROI is the proof-point for all the other market trends and sales victories discussed in this piece already, including about scaling 5G upwards and outwards with new applications and new sites, and migrating 5G across into new sectors and down into new segments.
But it is also about effectively mapping use cases to KPIs to business outcomes to SLAs for the first time. “It bodes well for the future, and how the market scales,” comments Douglas at Spirent, who has seen progress with KPIs/SLAs across the company workbenches as 5G systems are integrated with “applications around performance”. As per Artley’s comments, the supply-side is engaged with the buying-side in earnest at last about ‘business outcomes’ and ‘business value’. It is no longer just selling technology, in other words, but rather looking to ‘solve’ problems. But for this, and to get paid, it needs to prove it can hold up its end of the bargain.
As such, the private 5G community is working with the likes of Spirent to test how to match network KPIs – as delineated by business applications, as dictated by business outcomes, as determined by business problems – to vendor SLAs. “Enterprises don’t care about radio networks; they care about business outcomes, related to operational efficiencies and productivity gains, and things like that,” says Douglas. “That value discussion has completely changed. It is about how you map those outcomes back into the network. That’s where the bridge is, and what we are seeing more and more. Which is why these SLA-oriented use cases are emerging.”
Interestingly, he says the disjunct previously was not just about a failure of language on the part of vendors; the buyers, on the enterprise side, struggled to articulate accurately and reasonably what they wanted from 5G, as well. Spirent ran a survey with telecoms consultancy STL Partners about 18 months ago to figure out what enterprises would actually pay for SLAs. “The thing we found was that a lot of them have really unrealistic views about the network performance they actually need. A lot of industrial enterprises, particularly, didn’t have any real evidence that they needed this kind of heightened network performance. It was just a gut feeling.”
The research discovered industrial firms, paying heed to the hype, wanted ‘five-nines’ (99.999 percent) reliability, he says – “when the baseline was a best-effort network that could barely do two-nines”. Douglas goes on: “And it went the other way, too; a lot of telcos, say, were pushing these incredibly low latencies; when most enterprises don’t need latencies anywhere near as low. What they really want is a window of latency, or reliable latency, so they can guarantee a service does not go outside of a median latency range – which networks can deliver already today; just not very consistently. That exercise identified a gap between what is expected and what is needed.”
Leave a Reply