Last month, the CBRS community in the US made a swift counter-strike, in the form of a joint letter to the FCC and NTIA, against a report by telecoms lobby group CTIA that said private and shared cellular is bad for the US economy – including for both the 5G-fixated telecoms industry and the 5G-interested industrial sector. Rubbish, they wrote; the CBRS model, aped by the rest of the world, is proof already of “America’s inventive spirit at its finest”.
Their forceful response can be read here, with a link also to the CTIA broadside. Meanwhile, Enterprise IoT Insights caught up with Rajeev Shah, co-founder and chief executive at California-based private 5G startup Celona, one of the signatories to the CBRS letter, and a major voice in the CBRS camp – to discuss the political jockeying by the old telco crowd, and to argue that attempts to stall or limit private spectrum for Industry 4.0 are wrongheaded, and worse.
It is a compelling dialogue, and a good read – which also covers Celona’s own ambitions in the international Industry 4.0 market, its constant hope that the neutral host model will “double and triple” the market, its thoughts on NTT as the template for service providers, and why warehousing is the sector to “mainstream private cellular” (and why manufacturing overthinks it). And much more. The transcript of the conversation is printed in full, below; all the answers are from Shah.
…
There is palpable tension around CBRS at the moment, and how successful it has been. Some months ago device vendors were complaining to Enterprise IoT Insights that the volumes weren’t there, and there has been concerted lobbying recently by the legacy telco crowd to declare it an outright failure. The letter in response, from the CBRS community to the FCC and NTIA, to argue the other way, contains strong numbers, and makes a good opposite-case in defence of CBRS as the ‘innovation band’. But there is a counter-narrative to this, again – as there is with the Helium network on LoRaWAN, for example – that, while half the market says how many radio cells have been deployed, the other half says how little traffic there is. Clearly, innovation takes time; but just explain your take on all of this.
“Yes, so I have multiple perspectives. I have been involved in the CBRS ecosystem since 2016. I led product and marketing for Federated Wireless in the super-early days, when the CBRS Alliance was just five members. And I have gone on to start Celona – which could only have succeeded as a startup because of CBRS. It wouldn’t have happened any other way. We have raised $100 million, and created all these products – and it is a great example of what CBRS can do.
“Because, if you think of the time before CBRS, the Silicon Valley VC landscape – not necessarily location-wise, but as a representative engine of innovation in the US – had essentially given up on telecoms. CBRS has made it possible for tier-one VCs to perceive innovation in telecoms, once again, and to go and fund a team like Celona’s to do this. So even just the existence of Celona is proof of the innovation that CBRS was meant to foster in telecoms, in the way the US should want innovation to happen.
“The numbers tell the same story. It is a pretty extraordinary number of deployments – close to 300,000 small cells. I mean, when we were first anticipating the rate of CBRS adoption, I remember (as a frame of reference) that the total number of small cell deployments in public networks [in the US] was in the 100,000-120,000 range. Which had taken 15 years, across multiple operators. And here we are, in just three years, almost-tripling that number. And quite honestly, we are only just scratching the surface.
“We are picking up momentum in everything other than infrastructure, too – in devices, applications. Enterprises are only just starting to get educated on what they can do. So, 300,000-ish deployments, even before it takes off, is extraordinary. There is zero doubt that the other [anti-CBRS] side is only doing this to lobby for the next 3.1-3.45 GHz to go to auction. There isn’t any logic to say CBRS has not done well. They just want to create doubt – so the next spectrum band opens up in a more traditional manner. That is the only reason.”
So, on balance currently, is the expectation that the CBRS model will be extended and multiplied out in part of the 3.1-3.45 GHz band? Or is there a genuine fear that this will be it – so far as shared and private spectrum goes for Industry 4.0 in the US?
“Both scenarios exist, unfortunately. We are not far enough into discussions with the FCC and NTIA to know which way it will be allocated. This [CBRS] community, which signed this letter, feels unanimously that adding 3.1-3.45 GHz to the CBRS band would do phenomenal wonders for what 5G can do for the US economy. I have just returned from the UK, and I am flabbergasted at the amount of spectrum it freed up in one shot, when it legislated on 3.8-4.2 GHz. All the activity in the UK, currently, is indicative of what can happen when you free-up enough spectrum.
“There is a real desire in the current environment, when everyone is rethinking their supply chains, to bring manufacturing back home, and to return more of the supply chain to home territories. There is no way you can do that without a whole new level of automation. And 5G is going to be the lifeblood that enables that. So the more spectrum we afford it, the faster we move that bandwagon down the track. So 3.1-3.45 GHz should be available in a CBRS-like model – for some version of private spectrum, to really spur US enterprises.
“I hear this on a weekly basis from enterprise CEOs and CIOs. Their parallel is Wi-Fi, actually – which was only in the unlicensed 2.4 GHz band in the early 2000s, and a success in consumer-land; but needed the 5GHz infusion to truly go Brix-scale with enterprises. Which is when a whole wave of Silicon Valley companies came in, and when the market took off. New spectrum will do the same for 5G, opening up this automation revolution. That is what should happen. But it is too early in the process, and the other letter is an attempt to slow it down.”
Is there any sense that the US has lost the initiative? Germany was quick with enterprise spectrum, but appears perhaps to be stuck on later 5G releases. But the UK released this big chunk, like you say, and France is coming around, and India is making moves. I mean, there is clearly a broad consensus that this is the way to go, even if the rate of regulation varies. But does the US, with 150 MHz of CBRS spectrum and a long-gestated access system, maintain its original lead, would you say?
“Yes, the US maintains its lead, mostly. But the 3.1-3.45 GHz band is required to maintain that lead, over a three-year horizon, say. But more than the lead over other countries is whether we continue to satisfy demand from enterprises – which are asking for more. That is the question. If you talk to wireless specialists inside enterprises, they will tell you, unanimously, that another 150-200 MHz would enable so many more initiatives than they are currently able to do. So there is strong demand, and it is about keeping up with that.”
Just on that letter; Deere & Company and JBG are signatories to it, but everyone else (I think) is in the CBRS supply game. Because if you think of how this went in Germany – where Siemens and Bosch and BMW, and others, lobbied the government, against incumbent telcos, to release the 3.7-3.8 GHz band – it looks very clear that local regulators picked sides, and went with national economic interests, as voiced by the house brands for the country’s Industrie 4.0 agenda. So why aren’t there more enterprises on this CBRS letter to the FCC and NTIA?
“There is a strong rationale for that, and it is bittersweet. Because CBRS is super-innovative, and it was by-design. Credit should go to the founding members of the CBRS Alliance, because they said, ‘We don’t need a paper-based licensing process; we are in the modern world, and software can do it’. And so hundreds of thousands of enterprises can use CBRS without ever filling out a paper application form. It is done in software. Which is fantastic. But the flipside is enterprises have no clue there is something happening behind the scenes.
“So they are kind of oblivious. They are really happy with their (cellular) products, and they just deploy them and use them. Unlike in Germany, where enterprises actually have to apply – and someone has to think about it. So unfortunately, one of the greatest benefits of CBRS could be perceived, also, as one of its greatest weaknesses – that enterprises do not care. Which is not the point; it is just we have abstracted-out [that process] beautifully. So hundreds of thousands of enterprises are deploying, without ever thinking about it.”
So, as this spectrum discussion goes on, would you hope and expect enterprises to stand up for CBRS?
“I would expect that. We are talking to our customers for that reason – that, ‘Last year, we told you, ‘don’t worry, that software will take care of everything; but we kind of need your help now’’. Because it would be helpful if they come in. So we are absolutely having those conversations. To an extent, this is a learning moment for the CBRS industry, to say, ‘We are not out of the woods, yet; that even though we are living through this success, we have to tell the rest of the world, as well, especially the political side’. So we don’t lose momentum.”
But won’t the big telco operators in the US – and certainly those who appear less comfortable with private and shared CBRS spectrum, and the prospect of more of it – do the same? Won’t they approach big and noisy industrial engines in the US, which are already taking fixed and mobile telecoms services from them, to ask the same – to represent them to the FCC and NTIA? Will it just come down to how progressive big enterprise is in the US, whether it backs private or public 5G?
“No. We are not seeing that at all. When enterprises think of their comms technologies, the implicit assumption is they will have wide-area network (WAN; basically public 5G) connectivity from their [traditional] mobility partners, but that their local-area network (LAN, including private 5G) will, quite honestly, be from their incumbents, today – from Cisco, Aruba, Juniper. They are the only really serious players for LAN, but now you have these new 5G players added to that. They don’t think of it in terms of backing telecoms partners.”
So, changing the subject; what have you been doing in the UK? Are you any closer to this international launch, to offer the Celona core network system for private LTE and 5G deployments in Europe and Asia, say – as per the mission statement that followed the $60 million Series C round in March?
“The cash injection was to fuel three things: to expand our product portfolio, which meant going beyond CBRS in the US, which is a huge focus for us; to build our go-to-market engine across multiple dimensions, and we have tripled our direct enterprise team, and gone from 20 to almost 100 channel partners since the start of the year, including with NTT and Verizon; and to start to think how private 5G fits into the wider enterprise architecture, and the role it plays with edge compute, and new sorts of applications and security models.
“On the expansion, we are not ready to announce partnerships just yet, but we are closer, and we are on the verge of releasing a whole new set of products that will cover all the spectrum bands in Europe. Next year is the year of going out of the US, to go global – with an initial focus on the UK, Germany, France, the Nordics, and so on. We are recruiting sales teams and engineers, and we have had a couple of trials, while we’ve waited for the radios to be ready. The platform was conceived – and is perceived, quite honestly – as unique in the private 5G space, so there has been demand to get acquainted with it.
“That third part – about how 5G fits into the enterprise – combines both external partnerships and internal innovations, and is an area we continue to invest in. One of our key partnerships is with Google, and we are bullish on that – because Google’s vision of the enterprise architecture, especially as it relates to edge compute and 5G, is closely aligned with what we are doing. But there are more partnerships like that.”
More so than AWS and Azure?
“Strategically we have aligned a little more closely with Google. Some of that is because AWS has its own private 5G as-a-service. So Google is a little bit of a closer partner from a business perspective – but our product performs on all three cloud platforms.”
Anecdotally, is the traditional telco view of this the same everywhere? Or have you noticed a cultural change in the UK, say, where operators have been forced to sublet spectrum and play nice?
“No, it varies operator by operator, more so than country by country. Verizon is a strong partner in the US, and we are very excited about how well it has embraced innovation in this space. It has used that combination of public and private spectrum very smartly. On the other hand, there are others that have stayed away from it. It really is about individual operators. It almost feels each has made a strategic decision about how best to use different assets – and whether private spectrum fits their go-to-market mode. The UK will look the same, and split down the middle, with some embracing private 5G more than others. That is my early read.”
Talk about the closeness with NTT, which uses the Celona system in its managed private 5G offer, and which has a significant investment in Celona (via NTT Venture Capital in the US). Because NTT looks, to a market watcher, to be quite good at this. Is the NTT approach a template for other operators?
“NTT is less of a pure-play operator; it is separated from the service provider in Japan because of the holding company construct. The global NTT business, which uses our platform, has a channel play on the back of a number of acquisitions via NTT Ltd and NTT DATA. So those parts of NTT look closer to a combination of a [global] SI [consultancy] and a [global SI] reseller. And it might be that could be the template for a lot of service providers – to untether from local spectrum assets, and offer private 5G in any part of the world. It is discussed within that community, but NTT, I would suggest, is further ahead because of these acquisitions.”
Can you share deployment numbers or market shares for Celona in the US? And where is all the action, at the moment?
“The thing to look at is the number of customers. Because that is what matters, right now. About 75 enterprise customers have deployed our product in the US – including a couple of the largest retailers in the world, a couple of the largest automotive manufacturers, and a couple of cities. We are starting to see a concentration of activity in manufacturing and logistics, without question; warehousing is a major sector, with a broad range of use cases – just because everyone has a warehouse, and they all want to modernise and automate their supply chains. There are early signs, a couple of early engagements, in the precision agriculture space, too.”
Just quickly on the neutral host model; Celona has written in these pages many times that neutral host networks will explode (“this is the year of…”). Why are we still waiting? When will it happen?
“As an industry, neutral-host is the right way to scale in-building coverage. It is the smartest way to do that, and it helps everyone – enterprises and operators. The technical problems are relatively well solved, but it will require the whole ecosystem to come to the table. It is moving in the right direction, but we are cautious not to put a pin in it. Because we don’t know, exactly. That is the reality. That said, 2022 has been a good year, from a very-selfish perspective. We have had commercial trials in the US with operators – which gives us a heart.
“When does that become a truly scalable business? It is still a little up-for-grabs. But I anticipate we will be able to take this to the market in the next 12 months in the US, pretty generally – even if there is work still to be done. One of the things that will drive it even faster is how 5G gets rolled-out in the macro networks. When that reaches a certain critical mass, then everyone’s attention shifts a little to the in-building problem. So that needs to happen, which aligns with a late-2023 timeframe.
“But there are two different business models [for CBRS-style spectrum]. There is private 5G for industrial IoT, as well – which continues to be our primary focus. If the neutral-host thing happens, we just double or triple the market opportunity instantaneously, and we would love that. But from a business standpoint, we are very happy with that first market – it is more than enough for a company like ours.”
And on private cellular for industrial IoT, how do you perceive different verticals – manufacturing versus warehousing, say, as the primary sectors for it? Because there is a perception that manufacturing is a kind of blue riband vertical for 5G, but that it is also paranoid and competitive, and probably after something higher-grade. Whereas warehousing, perhaps, is happier just with its own mobile broadband. Does that characterisation tally for you?
“Warehousing is the right mix. Warehouses have had wireless technologies for a while – Wi-Fi, and a whole bunch of stuff – and they continue to modernise, dramatically. The amount of robotics and IoT going into warehousing is very high. So that is the vertical I would bet on to mainstream private cellular. On the manufacturing thing, I would broaden the point. Some of those concerns are true, but they are also vendor-driven, to an extent, unfortunately.
“The barrier to entry to deploy and run a [manufacturing] network, either on a cost basis or an education basis, has been too high, to the point the network has to modernise the whole factory, effectively – or else it isn’t worth it. So the enterprise looks at it with crazy expectations. Which is the side-effect of trying to force-fit the traditional telco model into an enterprise, and was the way for a very long time – and was one of the reasons Celona started. We have gone in, from day one, and said, ‘Look, all of these other problems will get solved, eventually, and you will remove all the wires from your floor; but deal with your bread-and-butter problems first’.
“Which you can solve today with private cellular – with lower cost and complexity. The classic case is customers have these handhelds and tablets for data entry, and sometimes slightly-sexy AR apps for remote assistance. But that is it. They don’t need five-nines URLLC to solve that problem. But they won’t solve it across an entire manufacturing floor with Wi-Fi, either – because they’d need a gateway every 1,500 feet. So just the fact private cellular gives better range and service, and works with Celona like Wi-Fi-lite, is reason to do this.
“So even in manufacturing, you just need to change the discussion – to say you don’t need to solve everything from day one, but you can start to solve lots of problems right away. So we are doing that – solving bread-and-butter challenges for manufacturing customers with LTE and 5G, today, and setting up to solve the really big problems tomorrow.”
So with warehousing, is that conversation just simpler – just that they are less inclined to overthink it? That they are not thinking about five-nines reliability, and the picture hasn’t been muddied by some future vision of something?
“That is exactly it. In manufacturing, we have the same conversation, and it changes [the approach of the customer]. But at CXO level, the hype around 5G is powerful, and sometimes a little harder to untangle. But in the warehouse space, it is very straightforward: ‘I have Wi-Fi, and it doesn’t solve these problems, and so let’s switch it up with private cellular’.”
