Note this article is continued from a previous post, under the header: Private 5G for IoT – a year to get real, two to get really-real, six to be ‘seamless’. To read the previous entry, go here.; to watch the original webinar discussion on this topic (‘plotting timelines, defining applications, making bets’) , go here.
So what about applications? Where and how is private 5G to be used for IoT? We have discussed some of this already – or, at least, we have referenced how private networks are being bankrolled, mostly on LTE and sometimes on 5G. The first aspect is important; LTE is the gateway system for private cellular-based enterprise applications, and will, in time, in most cases presumably, give way to 5G upgrades, affording a range of new business cases. So while the focus of this report is 5G-based private cellular, it is worth first considering where and how LTE is making inroads.
Kathiravan Kandasamy at Syniverse, a carrier services company that has reinvented itself as virtual operator (MVNO) for global IoT connectivity, picks up the point about the loose private-networks timetable, for when 5G outpaces LTE for sales (2024), and when 5G exceeds LTE for total deployments. “Most use cases today could use private LTE… You don’t have to wait for 5G. I don’t discount [some] use cases that require close-to-the-edge latency, and network slicing and so forth, but there are use cases that will work just with [LTE or with] Wi-Fi.”
Actually, allowing for repetition, he makes the point better in a separate conversation. He says: “Most use cases, whatever the circumstances, can be met with a private LTE solution. And we ask [enterprises]: ‘Do you really want to invest so much in 5G? What is the driver, really, to go to 5G?’ And we have not encountered a use case where the workload is so intensive or the latency so acute that you need 5G – or where it is complex enough that you need some sort of slicing. We just aren’t there yet.”
He continues: “Really, in the past 12 months, and maybe even in the next 12 months, the task has been to educate enterprises about the art-of-the-possible with private networks – whether that is about high bandwidth, low latency, high security, high density, precise slicing, secure roaming; all of those things are still being played-out. Because today, the question is whether you really need a 5G core – when a run-of-the-mill 4G core can do what you need.”
But, hang on; is there a danger not-to-say to an enterprise it is future-proofed against this art-of-the-possible with a 5G core? “No, I get it,” responds Kandasamy. “And don’t get me wrong, I want to sell a 5G core. But there is a question, still, at this stage, about whether it is worth it.” Indeed – as discussed, and as of-course – this is the definitive question for any business considering any kind of investment. As previously, one of the few clear cases for a private 5G network – which only 5G serves – is to cover large outdoor spaces.
But Kandasamy, actually selling a “secure” cloud-based MVNO core attached to public-network RAN infrastructure (rather than an all-edge on-prem network system) for national and international mobility cases, suggests otherwise; that LTE will deliver equivalent wide-area coverage, if not for an equivalent density of devices. “One of the areas [we are] getting a lot of traction is with power utilities and energy companies,” he explains, referencing the same verticals Gergs suggested previously as ideal homes for private 5G presently.
He says that 5G has been found out, to an extent, as enterprises have started to look more seriously at the practicalities of private networking. “Eighteen months or two years ago, we were all under the impression that private networks would be driven largely by IoT devices. But the devices on private networks are more or less split evenly between IoT devices with consumer devices – like smartphones, tablets, laptops. Which shows, quite unpredictably, that a lot of these private network use cases have a consumer behind them – a person, and not just a device.”
This, says Kandasamy, represents “an interesting evolution” for the real-world application of (mostly) LTE-driven private networks. This revelation about the popular short-term consumption of private cellular has brought into view the role of traditional public infrastructure to stitch together and annex parochial ‘island’ networks running in enterprise premises. Most of the use cases for this hybrid public/private network crossover have one thing in common: mobility. Somehow unpredictably, but perfectly logically, they put big telcos at the heart of the matter.
Except, maybe they don’t. The likes of Syniverse – piggybacking on national infrastructure in each market, multiplied internationally by hundreds of carrier roaming deals – can offer near-global mobility for IoT devices to roam in and out of private and public networks, and an embedded SIM (eSIM) for IoT vendors to provision devices at the point-of-sale. Syniverse also has a patented mechanic, it says, to prioritise private networks when in-range (inbound roaming is more complex), and a promise of enterprise-grade security when the SIM attaches to a macro network.
Kandasamy comments: “In a private network, you have all of these IoT devices, and all these consumer devices, as well. But not all of them are stationary; lots are moving about, outside the premises, and a private network is not enough… Some of the easy use cases are in transportation and logistics – like a truck or a package, covered in the distribution centre by a private network, and attached to a carrier network when it leaves the site; which then rejoins a private network at its destination. We are seeing increased demand for that kind of mobility.”
There is a distinction to make between how IoT sensors are deployed in industrial processes, in the task of making or moving things, for example, and how they are used to bring environmental context to core business processes. As someone else said, this is the difference between ‘industrial IoT’ (IIoT) and ‘IoT for industry’ (which might be usefully/awkwardly acronymized as ‘IoT4I’). These terms are used interchangeably, but actually describe different applications in the Industry 4.0 space – which, on paper, employ different wireless technologies to deliver them.
For that, read: one might read high-power industrial 5G at the critical IIoT-end of Industry 4.0 and low-power wide-area (LPWA) networking standards like LoRaWAN and Sigfox (or WiSUN), more traditionally, at the familiar IoT4I end. The former, wireless IIoT, is about volume, variety, velocity in big data, which (possibly, eventually) only 5G-IoT will deliver; meanwhile IoT for industry, for discrete IoT monitoring in industrial venues, is well served already by LPWA-style IoT. If that all makes sense; mobility use cases, mostly for asset tracking, are an adjunct of the latter.
But Kandasamy also suggests a global IoT roaming SIM, providing fallback to a macro network in case of outages or failures, affords the kind of comms redundancy (‘ultra-reliability’, effectively) that is, typically, offered as a feature of incoming industrial-grade 5G-IoT networks. In truth, it is not the same (redundancy versus reliability), but the value of public-network failover is proving to be “critical” for certain verticals to load sensitive operational data onto private networks. “It is a redundancy play [not a mobility play],” he says.
“The assumption is private/public network interoperability is for mobility, and that the private network is built already with five-nines availability; but one of the advantages – for airports and utilities, for example, with overlapping macro courage – is that if that private network goes down, for whatever reason, all of the devices automatically attach to the public network and retain connectivity.”
Some other quick points about the kinds of applications that are starting to multiply on private cellular networks. This private IoT4I regime, already well-served by the likes of LoRaWAN and Wi-SUN (and BLE and Wi-Fi), is squarely in the sights of 5G RedCap, along with a bunch of other higher-bandwidth machine applications – according to the 3GPP master plan, at least. Leo Gergs at ABI Research outlines likely RedCap-for-IoT applications: “We’re talking about wearables, industrial sensors, and video surveillance,” he says.
He returns to the point about pricing; that the cost of private 5G networks needs to reduce, and that a cut-down and enhanced new MBB (eMBB) version of them will go some way to do that. “The expectation is that by reducing the capabilities of 5G, down to an eMBB technology, essentially, with the deterministic networking features that licensed spectrum brings, will also bring down the price. And the price has to come down – to at least twice the price of industrial Wi-Fi at the moment. Which would be a considerable reduction, but a significant cost centre still,” he says.
“But a lot has to be done on that front,” he adds. Gergs goes on to ask (out-loud) how urgently industrial cellular, even if it is priced like industrial Wi-Fi, is actually required – if the kinds of cellular-based mobility cases Kandasamy references are also removed from the equation. “There is a question about whether enterprises will switch if they don’t need mobility, and if Wi-Fi works fine. I think they won’t, and I think it is something the telco industry has come to terms with; it has moved away from this idea of a quote/unquote ‘panacea’ where 5G replaces everything.”
Which begs the question: what is private 5G good for, actually, if mobility is provided by public infrastructure, and enterprises will not move away from Wi-Fi et al just for the sake of it. We know the answer, of course, although the counter-response remains that LTE will sometimes suffice: that whether it is initially deployed as an LTE or a 5G system, private cellular is providing valuable in-house mobility for automated guided vehicles (AGVs) and autonomous mobile robots (AMRs), and suchlike, with better bandwidth and cell-handover that available with Wi-Fi.
This is a well-understood private 5G-for-IoT use case; others blur the MBB boundary, but are no less potent. Notably, private LTE/5G (shall we just bundle them together?) is providing the kind of remote coverage, predictable latency, and decent (upload) bandwidth to support certain modish new collaboration tools such as augmented reality for knowledge transfer and remote assistance, where video streams can be shared and annotated to train staff in the office and guide workers at the coal-face.
Clearly, however, there is (at least) a two-speed market developing around private networks, between consumer and machine (IoT) applications in the first instance, relating commonly also to their rollout in white-collar and blue-collar style enterprise environments, and also between IoT-for-industry and industrial IoT scenarios – and the last in each couplet, describing the industrial IoT space, is moving slower. Jagadeesh Dantuluri at Keysight Technologies has a sensible view of it.
“If you sub-divide the use cases in Industry 4.0, there are basic ones, which are increasingly being connected with 5G – things like screwdrivers, and other kinds of industrial torque machines. But those are simple swap-outs of the connectivity module, from WI-Fi to 5G. The other big IoT use cases are to connect AGVs and HMIs (human-machine interfaces; the graphical displays that allow operators to interact with machines and controllers). Those things are being rapidly connected to 5G, wherever private 5G is deployed,” he says.
“But these are all quite different to connecting a PLC (programmable logic controller; the industrial computer running manufacturing processes) to a private 5G network. Which is a more critical application, which controls multi-millions of dollars of production. So yes, if you subdivide Industry 4.0 into less-critical and more-critical use cases, then right now 5G is only really being used to connect the less critical use cases. It will take some time for enterprises to trust 5G with these more critical applications.” Indeed, as per the lasy entry, it will likely take until the late 2020s, into 2030.
