More in

IoT: LPWA technologies’ struggle for supremacy

The hype over the Internet of Things might be fading away, but much effort is under way to roll out low-power wide-area networks. At the same time, some question if the market’s vast potential can be realised while it remains divided between so many technologies. Sam Fenwick reports

The alphabet soup that’s endemic to any conversation about the Internet of Things (IoT) gets particularly thick whenever the subject of low-power wide-area networks (LPWANs) comes up. LoRa, Sigfox, Weightless, Ingenu and Wi-SUN, and their cellular licensed spectrum counterparts, NB-IoT and LTE-M, are all jostling for supremacy.

However, William Webb, telecoms consultant and part-time CEO of Weightless SIG, believes that the success of LPWA “absolutely” hinges on whether the industry can converge towards a single fit-for-purpose standard. “If you look at every other area of wireless communications, we just have one standard. For personal connectivity, it’s Bluetooth, for home connectivity, it’s Wi-Fi, and for wide-area, it’s cellular,” he says. “The dynamics of the market drive the adoption of one standard unless there’s a very wide range of different needs and one standard can’t support all of those. Those dynamics are a mix of economies of scale, the cost of a nationwide network meaning just one is likely, and the interoperability benefits of everyone using the same standard.”

Webb adds that the counterargument is that “in the Internet of Things, there’s such a diversity of need that you couldn’t possibly make do with one standard”, and recognises that “there’s certainly a completely different use-case for those things that need very short-range communication, typically in the home, office or factory, where Wi-Fi or Bluetooth is acceptable”. He therefore doesn’t include those requirements in his “one standard” and thinks that “we’ll eventually get to a single LPWA standard, potentially via a couple of interim standards, one of which will be a licensed cellular technology and one of which will be an unlicensed technology”.

Webb says: “At the moment, there are three technologies that have some degree of deployment: Sigfox, LoRa and Ingenu, the US company. Of those three, I can’t see Sigfox being successful in the long term because it’s just too compromised in not having any real downlink capability and its business model is too close to that of the mobile operators. If you have to choose between signing up to Sigfox who may or may not be here in 10 years’ time, or to Vodafone who you’re pretty certain will, I think by and large you’re going to go with Vodafone…

“LoRa is doing much better and has a wide following with the LoRaWan group, and if you look at most trials in the UK or indeed elsewhere, most of them use LoRa and I think that’s because, by and large, it’s the only one you can really use if you want to go and buy some kit and test it out; LoRa is your only option.

“You can’t do that with Sigfox because Sigfox controls the base stations, so you can only buy the terminals; but if you want to do your own little trial, you can do that with LoRa, so that’s got a lot going for it.”

“I do like LoRa’s approach, it gives you the best of both worlds,” says Digital Jersey’s CEO, Tony Moretta. “LoRa seems to be a good balance. You’ve got someone driving the development but happy to put it out there, loyalty-free access, so you get more control around implementation – there’s more freedom, the prices come down but we know someone is driving the standard. And what I’ve seen so far, that seems to be the best bet. For the moment, our focus is on LoRa.”

However, Webb notes that “LoRa is not an open standard; although parts of it are open through the LoRaWan Alliance, the physical layer, the chip itself, is not; it’s proprietary to Semtech. They have licensed the right to manufacture [it to] one other company, but it’s not an open standard – a different silicon vendor couldn’t just turn up, pay all the right loyalties to Semtech, make it and say ‘everything will be fine’.”

Webb adds that there is therefore “an open question around LoRa as to whether it addresses that, in which case it could well become the long-term system, or whether it’s going to remain in its current business model, in which case I think it will be eventually sidelined by a developing standard that comes along.

“Ingenu seems to be a mostly US operation and [recently] lost its CEO… It has a very similar model to LoRa, without the following of LoRa, so I imagine that they’re finding it difficult to differentiate themselves.”

Naturally as the CEO of Weightless SIG, Webb has something to say about Weightless, another LPWA technology, which like LoRa and Sigfox is designed to work in unlicensed spectrum. “It is a fully open standard and Weightless companies are now just starting to shift software development kit, particularly in the Far East. It has a long way to catch up with LoRa, but it is also at the stage where you could buy Weightless kit if you wanted to test it, and the early feedback we’ve got from companies that have done so is that they found it to be much better, much more fit for purpose than LoRa or other technologies. That’s promising, but there’s still a long way to go for an ecosystem to form around that.”

Webb highlights the fact that cellular LPWA technologies such as NB-IoT or LTE-M “can be deployed very quickly at full coverage or at least good coverage”, but also notes that at the moment, they are “quite expensive, quite power-hungry and there’s still some uncertainty as to whether it’s going to be NB-IoT or LTE-M”.

Despite the difficulty in grappling with new technology and the uncertainty over which one may prevail, Patrice Slupowski, Orange’s VP, digital innovation, says one of the pieces of advice he gives to his customers is: “Don’t wait. If it is about saving money or creating new businesses, don’t wait to do that.”

Speaking of saving money, Neal Forse, CEO and co-founder of WND-UK, a Sigfox network operator, tells me that his company is working with a major housing association to combine Sigfox connectivity with their facility management system and that this has reduced their insurance premiums by 18 per cent, saving millions of pounds. He adds that in this example, “the cost of capital to install customer solutions is covered in the first year”.

He gives large goods vehicle trailers as another example of a use-case where LPWA technology can quickly deliver great value, explaining that refrigerated trailers can be worth up to £160,000 and that their owners struggle to know their locations. With LPWA devices, it is possible to track their positions and also the temperature of their cargo, allowing owners to spot assets that aren’t being properly utilised and providing customers with greater visibility.

While the choice of LPWA technology is an important one, Digital Jersey’s Moretta says the most time-consuming part of an IoT project can be getting the various parties to share their data and then managing and analysing it. “[Compared with that,] getting the sensors out there is a doddle.”

In an article on LPWA, Michele Mackenzie, principal analyst at Analysys Mason, and Tom Rebbeck, its research director, enterprise and IoT, note that while the appeal of a wait-and-see strategy for network operators is clear, adopting it would mean “foregoing a valuable learning experience in the early LPWA market and the opportunity to capture some of the early demand”.

Orange is looking to launch its LTE-M networks in 2018 and has (or soon will) launched a number of field trials of the technology. Adam Armer, Vodafone’s IoT global business development and innovation manager, says the company has deployed NB-IoT in Spain and Ireland and is hoping to have it deployed in 10-15 markets in 2018. While Vodafone’s focus is on NB-IoT, Armer says: “We have not released our plans for LTE-M as yet but the intention is there. I can’t give any more specifics because the budget and the timelines aren’t finalised, but it is part of our plans.”

Over in the US, AT&T has rolled out LTE-M nationwide, while T-Mobile is looking to roll out nationwide support for NB-IoT in 2018 and will also support LTE-M.

Mackenzie says that in most cases, enabling cellular networks to support NB-IoT is a “fairly standard upgrade, but it becomes less straightforward if a network operator’s infrastructure is from multiple vendors”. She expects that in future, there will be a greater need for national LPWA coverage to underpin asset-tracking use-cases.

During her presentation at the recent Smart Summit, Mackenzie said that some analysis and forecasts done by her firm expect the market to be worth $72bn in 2025 (not accounting for systems integration), with connectivity accounting for eight per cent, or $6bn. This suggests that network operators especially will “have to look at other areas where they can maximise their LPWA revenue, and not all of them are well-equipped at this point in time to do that”. She adds that “we shouldn’t underestimate the benefits of providing a portfolio of LPWA-ready devices and hardware”, noting that in a new market it is very difficult for enterprises to source all the required technology and they “may need help with that”.

“What’s important for us, it’s not LoRa, it’s not Sigfox, it’s the stuff that wraps round the [LPWAN technology] and what you can do with it,” says Dan Clarke, smart cities programme manager (for the Smart Cambridge Programme), Cambridgeshire County Council. He explains that Cambridge has a LoRa network for experiments and trials to see how services could be delivered differently and at lower cost.

“If there isn’t an ecosystem there that allows us to change the way we deliver services, or enhance the city, we’re not interested.” He adds: “There needs to be huge organisational changes within local authorities to really make the most of this technology. We don’t have the data skills to make the most of it, we don’t have the people who understand the technology and we need to fundamentally transform the way that local authorities are structured. They’re structured to deliver services in a 1970s, 1980s way, they haven’t changed and they need to.”

Webb believes MNOs will struggle to move up the value chain, as they have not succeeded in doing so in other areas, and because a company could develop an application and the underlying analytics for a particular use-case and then make it available to every potential customer globally. “It makes no sense for every mobile operator to try and develop their own analytics solution in that particular space, and they wouldn’t have the bandwidth to do it.” On the data storage side, he notes that that space is already well covered by the likes of Amazon and Microsoft. “It just seems utterly implausible to me that [MNOs] would be able to [go up the IoT value chain] at scale.”

The refrain that when it comes to IoT, ‘the value is in the data’ might have interesting consequences as it could lead to the creation of natural monopolies. After all, if that’s the case then the company with the most data can provide the best service. Webb agrees with this line of reasoning, though he notes that as far as IoT is concerned, “there’s probably thousands of verticals, which equates to thousands of monopolies”.

This sort of network effect may well be the factor that’s making it so hard to predict which LPWA technology will prevail. Just as VHS prevailed over Betamax for reasons that had little to do with the underlying technology, so it might be that ease of use and adoption by large-anchor customers might be just as important if not more than superior battery life. No doubt we’ll have more clarity once NB-IoT and LTE-M networks are fully deployed and enterprises increasingly turn to LPWA to support their business-critical operations, but for now, the crystal ball is decidedly opaque.