Did you see the interview with Airbus this week? Well here’s another dynamite entry for your over-stuffed private 5G files, pulled and transcribed from an interview at Industrial 5G Forum at the start of the month, which follows the story of US process manufacturing giant Cargill, a business with around 160,00 employees and $165 billion in turnover, and a presence in 70 countries. It also has exactly two (or maybe three) private 5G networks, and a couple of pilot projects besides. But in ways, its adventures with private 5G are more representative of the wider Industry 4.0 market, fortified by a developing view of the opportunity, and complicated by the same old tangles with internal structures, legacy equipment, and more urgent priorities. But its vision (and candour) are very clear.
Unlike pioneers like Airbus, which told RCR Wireless this week it wants to place all of its industrial workloads across all of its industrial plants onto private 5G networks, and John Deere, which has a target of about 80 percent, Cargill is more restrained, and perhaps more realistic. “Eighty or 100 percent is way too high for Cargill,” says Robert Greiner, global manufacturing network manager at the firm. “Most of our companies already have traditional networks; they have fibre already. We’re not ripping all of that out… If you’re talking about a greenfield site, three years from now, then, yes, we might line up with an 80/10/10 strategy – for 5G, Wi-Fi, ethernet. But man, if you’re talking about going into an existing meat packing facility or a corn milling facility… we’re not walking away from that.”
Private 5G has to “figure out its footprint”, Greiner says – or else enterprises will give it a hard time, and “dig in against” the whole cellular charm offensive, as delivered in grinning fashion by an increasingly mad-eyed mob of mobile operators, equipment vendors, system integrators, and various other sorts. But it has a footprint, he says also; it is cheaper than Wi-Fi, based on lower (repeat) installation costs for fewer radios, and also better than Wi-Fi, giving total site coverage and scope for new applications and innovations. Cargill has proved this already, he says, with two large-scale deployments in the US and the Netherlands, plus a couple of more limited warehouse projects. But, like any business, it also has other concerns, and sometimes, as with the big US setup, these get in the way.
And more than this, as with any heritage industrial enterprise (and Cargill has been about since 1865), private 5G has to fit in; business, including entrenched legacy networks, will not be turned on its head just for the sake of new technology. The full interview with Greiner is copied below; there are are some other interesting anecdotes and asides in there, too, such as Cargill’s intended pursuit of globalised vendor/integrator partners (which precludes mobile operators from pitching), and also its expectation to take its future 5G networks from a third party as a managed service, and then to take their orchestration in-house after a time as familiarity and confidence builds.
As a footnote, you may recall RCR Wireless caught up with Cargill at Industrial 5G Forum last year (in 2023), as well, where a site director at a Cargill plant in Europe provided a reality check of sorts, having kicked the tyres on private 5G in his own plant, and suggested its value was unclear for process manufacturing – based on the idea that mobility, as a network capability for flexible production lines and automated vehicles, is not an urgent requirement for static industrial-style recipe mixing. That view has since been discredited, to an extent, by vendors talking up a storm in 2024 about new connected-worker sales in the process manufacturing sector; and Greiner, now, also makes a nonsense of it, by talking about warehouse robots, barge connectivity, and other monitoring cases.
But the points about mobility cases in process manufacturing are valid; and Greiner’s viewpoint is a total global one. Plus, 12 months is a long time in the cellular market, and it is probably getting better, little by little, at selling to (read: understanding) enterprises. But, then, Greiner, with Cargill for almost three decades, also suggests in the interview below that the whole co-creation discipline, to bash heads and make solutions, falls on the enterprise, too. “It’s like when you go to see the doctor,” he says. “The doctor’s going to give you an opinion, but you need to be your own advocate out there.” All the answers below are from Greiner; note, content and language are edited here and there for economy. Watch the full interview here.
Talk about the process to expose the underbelly of that OT iceberg, and how that’s gone. Talk about some of the projects you’ve been involved in so far – in the US, in the first instance, and further afield.
“When a [network] project comes into Cargill, it [is assessed in terms of] our standard offerings, and, in large manufacturing areas, those offerings [haven’t been] fit-for-purpose. There was [always] pushback on either the cost, or just because [they didn’t] fit the manufacturing floor… And vendors had come in – AT&T, Verizon, NTT – and had made really these cellular/5G presentations to the IT side of the house. And I was in those meetings, probably about three years ago, and a light bulb went off – that [private 5G] was something we needed to be looked at. The light bulb really didn’t go off for the IT folks just because it didn’t mean that much to them.
“But it resonated with me – that we could use [private 5G] on the plant floor; it seemed logical. And so we’ve been looking at how it would fit with Cargill, ever since. It’s just been a journey – reaching out to other people in the industry, other manufacturing firms that have started before us, to see what they’ve done with 5G, and then to find the right use cases in Cargill to pilot it. And we had a facility in North America lined up to do a pilot and a facility in Europe lined up to do a pilot. And we were pretty close to kicking them both off. And then the one in North America got shelved for a year because of budgetary concerns. But the one in Europe is steadily moving forward.”
So the US project has been paused essentially because of budgetary constraints, but it is scheduled still – and the use case and the value is well understood. Is that correct?
“Yes. The value was there, and the use cases were valid. It is just there’s only so much capital for each plant, and each business unit has its A, B, C priorities; and this was in the list of A priorities, but even that list got divvied-up pretty sparsely, and it just didn’t… move forward. So it was a casualty [to other priorities], and was not among the projects this facility moved forward with. But it will probably go into the next budget cycle. We left it in so it could be picked up and started again fairly quickly.”
And is there anything else going on in North America? I understood there may be a couple of warehouse projects which are in process.
“Yes, we have some smaller warehouse locations that were looking at a Wi-Fi solution, which hadn’t bubbled up to my level yet. They were just at a BU level, and one of the BUs asked if it was an opportunity for 5G; he’d heard some of our ‘town-hall’ presentations on 5G. And I said, ’oh, absolutely’; we did a quick bid on one location to show the cost, and it was very attractive compared to the Wi-Fi. So yeah, they will pivot to 5G… [But] it’s a smaller size implementation… The first two we had lined out were ‘fence-to-fence’ [of the facilities]. We were [looking] to do the outside areas, as well – which was a strategy near-and-dear to my heart
“[Because fence-to-fence offers a way to] snowball into other things – to all these use cases that haven’t been talked about, whether wireless cameras or outside tank monitoring, or just, you know, getting connectivity to a tablet outdoors… [Which would mean the OT team has all these extra] nugget[s] it didn’t think was getting… We wanted that for the first couple of projects, so they snowballed… I was pretty hard-nosed about lighting up fence-to-fence. But the BU in these [smaller] warehouse [deployments] can’t light up fence-to-fence. They won’t entertain that; because they only need this [much] done; it’s not in their scope. They can’t do scope-creep.”
And talk about the European project in the Netherlands, which is a little bit further down the line. What’s the latest with that?
“[That] project came to us… They have a robotic dog that does two floors in a warehouse, and they needed coverage for that, and a bid with an outsourcer for Wi-Fi coverage… And I said we could compete with the Wi-Fi bid, and also probably light up the rest of the facility outside. And this facility actually sets-out onto a body of water, where it has some loading [on barges], and had difficulty with traditional networking because of the nuances with the barges. So I said cellular would work well, and, following the site survey, it turned out the bid was very competitive – and [shifted the proposition from] just the warehouse to the whole facility. So they agreed to move forward. We’re in the planning stages now; we’ve kicked off; we are about 30 days in. We hope to roll it out and have it operational in the first quarter of next year. And a lot of people are excited because it… works for the [whole] facility.”
Is that price comparison with Wi-Fi based on just the indoor use cases, and the things Wi-Fi would cover normally, or is it by rolling in a bunch of other use cases for outdoors as well? Where does it deliver the value over and above what Wi-Fi?
“[It is] a number of factors. If you’re just doing indoor, you’re probably looking at about a third or a fourth of the antennas [with 4G/5G]. And that’s being conservative; but let’s say it’s a third. [Which means] it is probably about even money on the antenna costs – because 5G antennas cost more than Wi-Fi. The Wi-Fi stuff’s a third cheaper, maybe, so you can’t make savings there. But the 5G stuff’s going to come down in price, as it gets more popular. But [more than that] the installation charge is almost identical, per mast… They both take the same type of mast to handle-em up. If they’re up in the ceilings, you need a telehandler. But if you’re putting three versus 12 in a warehouse, you make a saving there.
“But we always worry about the total cost of ownership, and fast-forward four or five years to your lifecycle (end-of-life) event, you reap the benefits again. Because you’re saving on the repeat install charges. So you have to factor that in… I think that is overlooked by everybody – because the talk tends to be about the [upfront] equipment charge. A lot of people don’t look at the installation of it. And when you rotate outdoors, that is where 5G really shines – where traditional Wi-Fi can’t compete. Outdoors, the only competitor for private 5G is public 5G, and then the use case spurs back to what connectivity inside the facility looks like.
“And because this is an industrial environment, any device on a public network is not going to be allowed to touch the plant floor. Because that is a security risk. And so those pathways back in are hard to make. But if it’s private outdoors, it can come through a firewall on the network and land right onto the OT network at the facility – so it’s very secure. And that creates an opportunity to have outdoor PLCs, outdoor cameras, whatever we need on the OT network. Which is a big deal. Plus you get this high power, where a whole outdoor facility can light-up with only maybe one or two antennas.”
Is there not an integration / installation premium attached to 5G – insofar as it’s complex? Because it’s a different radio technology for IT departments or outsourced engineers, whatever. And that integration with access and security, with roaming – I mean all of those things sound complicated. Does that stretch the value equation?
“It does, somewhat. I mean there is an intake [with] that. [Plus] there’s some core server equipment at the heart of it, which is the brains of 5G, so to speak, which you have to factor in. And there is an administration cost, as well. Because we are starting with a service model; we’re partnering with outside companies to give us a solution in-a-box, as it were. We want them to come in and provide the private 5G network; we will contract with them [as] the one-throat-to-choke, the one point to [deploy and maintain] the solution…
“A lot of other companies started that way, and some, as they’ve gained more experience, a couple of years in, have pulled that in-house. Some have kept it out. But that’s where we are now – we’re unsure about how it will work, and we want to make sure we get it right. So we are looking at an ’over-the-fence’ provider… Ninety percent of our facilities already have a plant segmentation firewall in place. So we’re looking to leverage that at every plant, as the entry point for 5G. That will be the segmentation piece to bring 5G into the Cargill network at each facility.
You mentioned partners, and people want to know who enterprises are working with; can you say who you are working with variously in the US and Europe? Or can you tell us about your strategy, at least, around supplier partners?
“We’ve been doing open biddings and so I’m just not comfortable saying who we’re with yet. But we have chosen to be broad in that scope. In the US, we started talking with the traditional carriers – Verizon, AT&T, T-Mobile. But we’ve chosen not to go down that route – just because Cargill is global in nature. So we’re looking at vendors that have global experience with private 5G. We have locations in Australia, Africa, Asia Pacific, Europe, Latin America. And we want to know… their experience crosses-over with 80 percent of our [footprint]… That’s a big win for us. Because it means they already know the quirks of spectrum in all those countries. Because 5G is not the same in Canada, Europe, Asia. It’s not the same; everybody thinks it is, but it’s different everywhere you go.”
I spoke with Airbus the other day, and it said all of its industrial workloads at all of its industrial sites will go on private 5G in the next three-to-five years. I think John Deere has talked about putting 80 percent on private 5G. Is that the Cargill view, broadly? How much of your OT workloads will go onto 5G networks, eventually?
“I don’t know if you can put a number on it, but I believe 80 / 100 percent is way too high for Cargill. Most of our companies already have traditional networks; they have fibre already in place. We’re not ripping all of that out… There is still a lot of value in having the core [OT] network connected by wire. There’s a lot of trust in it. And so the heart of our facilities are not going away from a wired connection. If it’s wired today, I really do not see us venturing away from that. If you’re talking about a greenfield site, three years from now, then, yes, we might line up with an 80/10/10 strategy – for 5G, Wi-Fi, ethernet. But man, if you’re talking about going into an existing meat packing facility or a corn milling facility… we’re not walking away from that traditional network. By no means… 5G has to figure out its footprint… This strategy to rip-and-replace [won’t work] – because people will only dig in against it; it will have a hard time to be adopted in industry.”
Just finally, and I’m quoting Appledore Research here, but I like the quote, and want your view: does Cargill have a ‘5G problem’? Is 5G being sold to you in the correct way by the telecoms industry? Does the telecoms industry understand your problems, as a business, or does it believe your problems will all be solved by putting private 5G to work?
“No, we don’t have a 5G problem. And I think people don’t understand Cargill. But I also think we have enough people inside Cargill that know what we need – who are able to [explain] to these ‘experts’ that we’re bringing in… It’s like when you go to the doctor, you need to be your own advocate. The doctor’s going to give you an opinion, but you need to be your own advocate out there. And I feel the same way when you go asking for help, whether it’s with Microsoft or Cisco [or telecoms companies]; you need to be your advocate on that. And that’s what we are doing – to say, ‘this is how 5G fits for us’.”
