There’s a moment, late in conversation with Siemens at Hannover Messe last month, where RCR Wireless replays a friendly admonishment from a rival telecoms vendor for making too much of the German firm’s heavyweight influence in its coverage of the Industry 4.0 market. The charge was that it is overplayed; that Siemens’ pedigree in Europe counts for rather less in North America and Asia Pacific, say, where local equivalents are automating local industries just as fast, or even faster.
The argument is not entirely convincing, even if it is plain that Siemens is not the only one; but it is played-back anyway – and Siemens, scratching its head at the question, is robust in its defence, and the gambit also shifts the focus to what makes Siemens different in the 5G game. So, somebody told me off for making out like Siemens is the kingpin on the Industry 4.0 scene, and that other operational technology (OT) vendors hold as much sway in non-European markets. What do you make of that, Daniel?
Daniel, here, is Daniel Mai, director of industrial wireless communication at Siemens. RCR Wireless caught up with him a year ago at the same event, where he’d just taken over from Sander Rotmensen, who had shepherded the firm’s nascent private 5G system into existence. Mai’s remit, ostensibly, is to shepherd the product into market as an OT-made platform for OT-grade applications. He responds, a little nonplussed: “But we are the leader; we have the biggest market share.” He offers up a share percentage, and then retracts it.
He is reluctant to use statistics to make an argument, it seems – which is commendable given how many are out there, how they are used to construct all manner of ‘truths’, and how many ways the industrial automation market can be carved up. “We are a global company,” he says, simply, before changing tack. Because it is not a useful argument, he reasons – because Siemens is the only one with a 5G system of its own, anyway. ABB doesn’t have one, Rockwell doesn’t have one; nor does Bosch, Mitsubishi, Schneider.
That much is implicit. Mai comments: “We are the only OT supplier that produces 5G equipment, and we also produce Wi-Fi and wired devices – routers, switches, firewalls, everything. Because these things are needed for digital transformation… And we are on the shop floor already, so when customers want a 5G system for an AGV (automated guided vehicle), or a crane or a stacker, they come to us – because they have known and trusted us for two decades or more. Whereas they probably won’t go to a telco vendor.”
Time is almost up when Mai offers a more profound response about this battle for Industry 4.0 hearts and minds. The thing is that when that AGV stops in its tracks, he says, whether because the motor has conked-out or the network has blacked-out, Siemens gets the call. “They don’t phone the network vendor about the industrial equipment,” he says. “Because the telco will just say, ‘Well, the signal looks alright’. They don’t have a clue about the AGV. Whereas our stickers are everywhere inside the cabinet.”
He adds, just to be clear: “Even if it stands still because of a different network from a different vendor, we get the call; they say, ‘Hey, Siemens, your AGV is not running’.” So is Siemens taking lots of calls about malfunctioning AGVs on malfunctioning 5G from rival firms? He does not really say, making reference instead to some sort of Copilot-style self-help hardware fixes (“because equipment fails in the middle of the night, mostly, when trained staff aren’t around”), which sound like they apply as well to the 5G as to the AGV.
“The equipment is designed so any issues can be easily sorted. You don’t need to be a 5G expert to deploy our system. It’s the same with Wi-Fi, where we have a small caretaker feature – and as long as you can replace a piece of hardware and put the cables back where they were, then it starts up,” he says. Even so, the message is that, yes, Siemens is taking calls about failing 5G-AGVs, but also that it has a captive audience, anyway, which has waited patiently in the Industry 4.0 wings for a Siemens-made 5G system.
“A lot of customers waited for us. They told us, directly, they wanted an OT player that understands the industry to supply [5G to] them. Some now have their first systems. We are still holding hands – so the applications are implemented correctly on top. Which is what matters. It’s not about 5G; it’s about mission-critical comms [for solutions]. That differentiates us. And we have changed our message – so we don’t talk about 5G releases anymore, but about enabling applications and [specifying] the total solution.”
Just to recap: Siemens, in no particular rush, launched its own private 5G system, including core and radio (RAN) functions (built from scratch) in late 2023 in Germany, for use in the local 3.7-3.8 GHz ‘vertical’ band. It has since been testing with major customers, and recently been selling in more serious fashion. “It is going quite well,” says Mai, with the kind of reserve that is characteristic of his company’s whole approach to the technology (and unlike its grand pronouncements about industrial AI in Hanover).
So do these customers in Germany, which have waited so long for Siemens, categorically not-want private 5G from ICT companies like Nokia and Ericsson or IT companies like Cisco and HPE? Have they tested and rejected their products, or just bided their time for Siemens? “A lot was over-promised,” he responds. “And yes, big customers, especially, tested equipment from other vendors and weren’t happy. Because neither the technology nor the business model did the job as expected. That is what they have told us.”
He adds: “So they have been waiting for us, yes. Because we understand what OT needs and what the critical KPIs are.” There are a couple of things to unpick, here; about these OT demands on the technology, and also how it is served up. The exercise is never just about cellular, says Mai; it is less about the individual parts, and more about their total sum, as 5G is levered-into complex Industry 4.0 setups. For Siemens, and Siemens ‘shops’, it is about integration with layer-two protocols like PROFINET (Process Field Network), for starters, plus OPC-UA.
“We set parameters – for devices and cycle times, say. And we test extensively within these ranges and spec the system around them.” Indeed, the firm’s early work with big German industrialists to switch workloads from Wi-Fi onto 5G seems cautious, even painstaking – which goes against some of the messaging about rampant use-case stacking from rival firms. Mai says: “Most customers want to start with a proof of an application – like an AGV or a crane – to understand how to implement it with 5G.”
But the subsequent ramp-up will be faster, he suggests. “With Wi-Fi, it is done application-by-application, whereas 5G is powerful enough to support more than one application. So we can load the network with more use cases – which is the agenda for our customers too.” He tells of the company’s work with an unnamed steel mill, as presented on stage at Hannover Messe earlier in the week. “This steel mill started slowly, with an AGV – albeit a really big one, carrying 100 tons of steel,” he explains.
“But the plan is to integrate 5G into its whole infrastructure. The next step is to cover this huge outside area where it stacks and stores these slabs – because it cannot be covered with Wi-Fi, just from a cost perspective. And the step after that is to cover its crane automation. So for this steel mill, like for most others, the strategy is to start with a simple use case, and then expand with the company’s digital agenda.” But the pace is slow, arguably – albeit steady, and also thorough, as is the way of Industrie 4.0, it seems.
Asked if customers have the confidence now to multiply use cases more quickly, he responds: “Yes, that is happening as we speak.” But he also makes it sound more like they are confident, presently, to plot how to stack use cases, rather than to actually load their systems with them right away. He comments: “The first use cases have been implemented, and are being looked at in terms of their implementation and management. But most customers haven’t started on a second use case yet.”
The point about ‘management’ relates to the earlier one about how private 5G is being served up (the “business model”; Siemens versus the rest). Asked if he means money, Mai responds: “No, it’s not about being cheaper. But their solutions are not combined with mandatory services, for instance.” Which means, what? KPIs and SLAs? “Correct, yes. Again, we are not focused on just connectivity, but on the whole solution – on mission-critical applications for Industry 4.0. Which is how the market is going, actually.”
The point rather justifies its timing, than explains its difference. Mai zooms-out, in a couple of directions. Firstly, he talks about how even the telecoms market, carried away with its potential role in Industry 4.0, has come around to the same viewpoint, which is more measured and also more massive, and also more reflective of the tone in the German manufacturing industry. “Two years ago, at MWC, there were lots of drawings; last year, there were lots of boxes; this year, no one talked about private 5G very much.”
Instead, they talked about applications, he says. In particular, like in Hanover, the buzz at MWC in late February was about generative AI. “That is where Industry 4.0 gets real. Because AI helps to analyse lots of data, but you still need to understand what the analysis means. With generative AI, you get a clear direction about that – which makes the whole solution valuable to customers.” So there’s a point to these fat 5G pipes, at last – because all the data they pull out of a factory can be made sense of, suddenly? “Correct, yes.”
And are industrial customers therefore looking at 5G with renewed interest? Is all this caution, which characterises the Siemens experience, and makes sense for mission-critical operations, being thrown to the wind? “I wouldn’t call it renewed interest,” says Mai. “Everyone perceives [5G] as a foundation for digital transformation – or for digital evolution, actually. But they now see how quickly they can take advantage of it. Previously, only big firms had an idea of what to do with it, and even they didn’t necessarily have a productive outcome. Now it is happening quite quickly.”
He notes the significance of the Industry 4.0 double-whammy of licensed spectrum and industrial cellular. “Licensed spectrum [for enterprises] is a gift, especially in the OT space, especially where you have moving bits and pieces – because it eliminates, or drastically reduces, interference. I have been in this sector for 20 years, and it was always a big topic. And now we have it. And the cellular options [for industry] are part of the standard. We have sold industrial Wi-Fi very successfully for 20 years, but only with proprietary extensions, and only with unlicensed spectrum.”
So don’t knock it, says Mai; industrial revolution is not just another AI story. The combination of spectrum and features in private industrial 5G is the real deal, then – is the line from Siemens. Industry 4.0 will not work without it – on the factory floor, for mission-critical operations. Public 5G will not suffice, the message goes; as illustration, one operator in Germany is offering 10:1 downstream/upstream ratios on its national network, says Mai – when the primary Industry 4.0 discipline is about getting data out of the machines (on the uplink), into an edge AI system.
He says: “The system we have in place now is three-to-one. And with one of the next firmware updates, we will have a one-to-one, or even a one-to-three time slot ratio – which means three on the up. Because we are generating data, mostly, and not consuming it. And with some generative AI on top of it, you start to see the real advantage.” But while the buzz about generative AI, as witnessed in Hanover and Barcelona, is a rush, and while Siemens has undeniable influence on the global Industry 4.0 stage, its private 5G activity, as it stands, remains rather parochial.
It is only selling in its home market. Why not in the UK, where 400MHz of spectrum has been cleared (at 3.8-4.2 GHz) for shared and private usage? Why not in Scandinavia and Poland, where spectrum is recently-available? Why not the US? “We are still making our minds up about which markets to tackle,” he responds. Why don’t you just tackle all of them? “That’s the plan,” he laughs. “But it is not that easy. It is still a new technology for industry. It will get faster because of generative AI; demand will rise at the end of 2025/26. But it will take years – like it did for Ethernet and Wi-Fi.”
There’s that familiar reserve, again; but that reserve has started to look right.
This article is continued here.
