The ‘failure’ of private 5G – another telco bungle, or just industrial inertia? (Is the window really closing?)
Did you see the writeup last week (or the week before?), that the window of opportunity for telcos to reinvent themselves with private 5G, as more than just dinosaur utility pipes, is closing, almost before it has even opened? Sound somewhat alarmist? We thought so, probably.
The story was from ABI Research, which reckons the carrier community has to pull-up its proverbials, quick-smart, or be caught at the start line, as the industry’s promise of industrial 5G outruns its means, enterprises grow impatient, and system integrators overhaul them with all manner of other connectivity tech.
Before Enterprise IoT Insights broke for holiday, we caught up with Leo Gergs, senior analyst at ABI Research and author of the original report, to sense-check his findings. Here is a post-holiday replay of the conversation, conducted as an email Q&A a couple of weeks back, to chew over the findings. Either way, there’s a lot to digest. All the responses below are from Gergs; all the quotes in the questions are from the original story.
Your statement, “Germany’s interest in private networks is fading”, says there were 80 applications for private 5G licences to BNetzA in H2 2020, and only 20 in Q2 2021. Is this decline (of 50 percent on average per quarter) not just because the big industrial outfits – and whoever else – have all been naturally quick to hoover up these licences in order to start early to test industrial 5G? Is it not the case that there will be a natural slowdown, anyway, as tier-two / SME / laggard enterprises – comprising the large majority of the industrial space, just struggling post-Covid to stay in business – bide their time? After all, ‘industrial 5G’ is not the finished article, as yet, and most ‘vertical’ licences, so far, remain attached to PoCs. So the early interest is from ‘industrial leaders’, often with vested interests in selling solutions over-the-top, to kick the tyres on private 5G. And the number of fully-fledged deployments are limited, as it stands, however you look at it. What do you make of this?
“You are certainly right in saying that ‘5G is not the finished article’. Also, if you look at the names of the licensees, more than half of them are strictly speaking non-commercial. Either, they are research and / or proofs, as you rightly mention, or [else] they are system-integrator (SI) deployments – [all of which] want to test and showcase 5G solutions they are looking to provide to clients.
“And you are certainly right [that the] dynamic will change with more mittelstand and SMEs coming on board, for a number of reasons, as below:
1. With a local license you need to source expertise on network deployment and network management yourself. So maybe going with an operator (and operator licensed spectrum) might not be such a bad alternative after all.
2. This 3.7 GHz frequency can give excellent coverage within a respective campus. As soon as your use cases require connectivity outside the campus (for example remote maintenance) you need some interaction with public network resources, anyways.
“So in short, I am not sure whether a completely isolated private network (as a regulatory framework like the one Germany ultimately proposes) really will be the dominant deployment model. Quite the contrary; enterprises will realise that some public network integration is beneficial for them, either to drive down TO or guarantee connectivity beyond the premisses, or both. So any form of PNI-NPN will, over time, be dominant. The question then becomes what level of PNI-NPN will emerge as the most successful.”
In reference to your statement, that “in China, the industry assumes there are several hundred private network deployments, but only 40 are fully publicly disclosed” – I understand the number of deployments in China is higher; that there are 1,000-plus private networks deployments globally, and that most of these are in China. I also understand that most deployments in China use this public-private (non-public) network integration (PNI-NPN) model, as you reference above – so they are carrier-led ‘campus-netz’, effectively, where the UPF runs off the public network, allied with local-area private radios. Presumably, then, comparing success in Germany / Europe with China is difficult – because China is so heavily reliant on carriers pushing hard. What do you make of this?
“You are right, but I think this clarity in [the] supply chain might also be one of the success factors. After all, enterprises know that to get cellular connectivity, they have to go through CSPs. So they do not have to make up their mind [about] whether to work together with an SI, go through a hyperscaler, or work together with a tier-one infrastructure vendor. All of these decisions can be very technical, so they serve as additional barriers to entry for enterprises that ultimately do not have the level of expertise to assess the different alternatives.
“In other words, enterprises need either a clear supply chain or hand-holding for the different deployment opportunities. Here is an excerpt of a commentary I wrote on the perceived role of CSPs in China:
“Within China network operators are seen as the primary enabler for enterprise digitization. While part of this might be ascribed to the fact that mobile network spectrum is still exclusively licensed to network operators (hence, an external factor), there is an internal factor at play as well. After all, enterprises could decide not to deploy cellular connectivity at all if they didn’t trust network operators in this endeavor: infrastructure vendors as well as operators want to design cross-industry standards that are not only accepted [but] co-developed by implementing enterprises.
“To do this, a range of different initiatives are [in] play that bring enterprises and infrastructure vendors as well as CSPs together (such as Huawei’s 5G ‘deterministic networking alliance’). In addition, the telecom industry in China realizes that the true value proposition for enterprise cellular connectivity does not lie in the connectivity technology as such, but rather in the application it enables. Therefore, market education and commercial offerings are built along different applications. Infrastructure vendors Huawei and ZTE, for example, place a lot of emphasis on 5G enabled solutions around machine vision or cloud-based quality control of manufactured goods.”
You say that “the window of opportunity for enterprise 5G is closing.” Why is it closing? Will Wi-Fi 6 upgrades gain traction in place of industrial cellular? And why is the window closing for operators, specifically? Presumably, operators in Europe are playing the long game. They are less interested, ultimately, in selling networks, than in operating networks – which is a longer game. Is this right? Is the situation really that European industry, as with European-everything, is biding its time, and Chinese industry, like with Chinese-everything, is embracing change faster? Is the problem not one of industry, rather than telecoms?
“Enterprises are losing patience with the telco industry. They are still waiting for the exact [industrial] features – like URLLC and TSN – that they have been promised, for three or four years, will come with 5G. The telco industry is now sending out this message that ‘5G is here’. So enterprises are impatient; they realise that today’s 5G (R15) is not the real deal. This unreliable communication from the telco industry has not added any level of trust.
“Enterprises are now looking to deploy technology to enhance workflow. Bear in mind, they are used to investment cycles of several decades, so if they deploy a good-enough technology – like industrial wireless LAN from Siemens for example – they will not invest in new technology again for decades. So they will be lost to the telco industry. Certainly, you can say ‘this is a problem of industry’, and remove any blame from telcos. But in the end, telcos have to live with the fact they can not change enterprises’ investment cycles.”
Is industry not waiting, ultimately, for R16, R17, R18? Certainly for critical industry, is there not important work to do before industry can throw its weight entirely behind 5G? Even if China has made up its mind?
“Yes, in theory. I am not sure whether enterprises are patient enough, or whether they would deploy something that is available today to address their pain points. After all, enterprises do not care about what exact technology they deploy as long as it addresses their pain points.”
You state: “The telco industry needs to radically rethink its approach… [and] needs to embrace spectrum liberalization initiatives and consider flexible business models.” Is this not what Deutsche Telekom and Orange are doing? Is this not what Vodafone is doing? Or are these the exceptions that prove the rule?
“I think the key to this is the statement that ‘telcos need to realise that the true value of enterprise 5G lies in the enabled applications rather than the technology itself’ – meaning, for example, the real value is in machine vision, large scale adoption of AI, and so on. Telcos need to realise it is not enough to deploy a private network only; [they must deliver] a solution that provides applications enabled by 5G. You are right [that] this is beginning to take shape with Deutsche Telekom and Orange, which can provide this [value] through their SI arms. Vodafone on the other hand is still very connectivity driven.”
I feel like the message in this report speaks more of a lull in sales, and a lag in the cultural approach in Europe, than a crisis (“window closing”) for operators – that they will miss the boat. Your thoughts?
“Well, you are right, I think. Plus, China overall puts a lot more emphasis on supporting enterprise digitisation. Read the Set Sail Programme about China’s ambitions regarding enterprise connectivity and private networks – totally, the Chinese government is creating a programme that fosters enterprise application development and industry-wide standards. Also, it offers funding opportunities for enterprises looking to digitise their operations.
“So, overall, I think we have four factors at play here, as below:
1. More direct influence from regulatory bodies on enterprise workflows
2. Favourable technology conditions for enterprise digitisation
3. Funding opportunities to take away financial risk
4. A much clearer supply chain, where CSPs are perceived as trustworthy digitisation partners.”