• IoT specialist Semtech recently acquired Sierra Wireless, a rival from the other side of the LPWA/cellular divide
• That move could mark the beginning of a new phase in the internet of everything, as the IoT sector is now optimistically being called, where the competitive focus on differentiated IoT networks is replaced by a more horizontal view

Internet of things (IoT) networks may be converging along similar lines to the trend in broadband access, where – as we reported in April – there’s an emerging consensus that a fusion of network technologies will prove to be the best way forward, for both corporate networks and the telecom service providers that support them.

So out goes the ingrained idea of competing network types, such as 5G and Wi-Fi, and in comes a horses-for-courses approach and a push to make the networks play nicely with one another.

IoT appears to be heading in the same direction.

In August, Semtech, the founding member of the LoRaWAN and LoRa, and one of the lead players in the unlicensed, non-cellular low-power wide-area (LPWA) IoT field (that’s a lot of acronyms to cram into one description), announced its purchase of Canadian cellular IoT player, Sierra Wireless, for $1.2bn.

Semtech says the merger will allow it to establish what it describes as a comprehensive IoT platform, stirring in the ultra low-powered and low-cost advantages of LoRa together with the higher bandwidth, licensed, cellular approach of Sierra to create a cross-platform solution for application developers (mostly).

The resulting “chip-to-cloud” platform might arguably serve as a useful course correction for the industry as a whole, as it takes the marketing stress off network type (where the industry has traditionally been battling it out) and puts it squarely where it belongs: on application development rather than network building.

The Semtech move is one step towards that more unified environment, ultimately making it easier for developers to code across network platforms; and, therefore, for users to bring projects in on time and below budget; for platform providers to develop better and slicker onboarding processes and a host of other improvements to make IoT adoption less fraught and more remunerative.

Shaking out the hype

Since IoT found the telco spotlight about 10 years ago, it’s been hampered by over-exuberance on the one hand, while simultaneously being hobbled by network standards fragmentation on the other. It was supposed to do great things: to explore new worlds; to go where no semiconductor had gone before; most importantly, perhaps, to help foster networked applications that could ratchet modern information economies up a gear; and to gather and process data about the way the world worked and how the machines in it were performing. This capability would allow near-real time control in industrial settings and birth clever business models on the path to greater industrial and service automation.

IoT deployment was even seen by some as a potential counter-balance to commercial and governmental surveillance and control. Hopes were high for a sort of bottom-up, data-driven insurgency.

For instance, IoT applications might sniff out air and water quality problems and provide proof of illegal or damaging pollution. They might measure the extent of particulate concentrations in urban environments, for instance, providing evidence to bolster the case for measures, such as car-restricted zones.

But today, 10 to 15 years later, while telco IoT can be presented as a relatively positive growth story and supporters can point to myriad successful IoT projects, it was supposed to do much better than ‘fairly OK’. IoT was supposed to end up in the telco service lineup, alongside mobile and broadband as one of the big telco growth areas boosted by services built on 4G LTE, which would naturally migrate and be enhanced by 5G. The internet of everything may arrive, carried by 5G, but it hasn’t done so yet. 

The root of the problem

Circa 2013/14, I went with colleagues to film and interview participants at an IoT event where young hopefuls presented their IoT demos and business plans. Winners were to be announced, various forms of support could be extended (or not)… you know the sort of thing. It was all very interesting and we were particularly wowed by the enthusiasm of the young developers who, in some cases, had taken expensive air journeys to have their shot at tech start-up stardom.

It wasn’t until I got home and consulted my notes that I realised none of the demos I had witnessed were likely to result in a commercial ‘hit'. There was only one pitch that stayed in my memory – a remote food preparation gizmo – and that was simply because of its weirdness.

I could have gone out on a limb and written a scathing put-down of the whole day and its eager participants. On the other hand, that seemed unfair, if only because those looking to invest are often more interested in the potential of the inventor rather than the invention itself. What’s more, I might have been very wrong. What if the world had been waiting like a coiled spring for the arrival of remote cooking? No one wants to risk being likened to Dave Dexter, the talent scout for Capitol Records who thought the Beatles weren’t "suitable for the American market." Best, I thought, just to mention the event if necessary and leave it at that.

As it turned out I needn’t have worried about going in a Dexterly direction. No IoT unicorns were likely to be sired by that event as far as I could tell and, for me, the experience became emblematic of much of the early IoT enthusiasm and activity, where IoT take-up fell short of expectations, with many of the resulting ideas – if they went anywhere at all – ending up in the novelty gift clearance bin accompanied by a big consumer yawn.

That little story exposes the key question for any IoT application: Will it be worth the expense, trouble and risk involved for its intended market? For many early IoT implementations, the answer had to be ‘no’, but the telecoms industry’s enthusiasm had been piqued by key technology developments that delivered improving costs, which promised to make many otherwise marginal applications viable. 

Sensors with radio connectivity were being manufactured cheaply, and could be rocket-boosted by scale economics which, we were told, would see prices spiral down to something like 50 cents per unit. Likewise, battery technology for the remote sensors had also progressed significantly (thanks to battery-dependent digital technologies, like smartphones). 

Then, on the radio side, connectivity standards for long-distance IoT (as opposed to short distance or indoor standards, which could piggy-back on Wi-Fi or Bluetooth) essentially bifurcated. Licensed service providers could opt to take slivers of their existing spectrum holdings to support very slow-speed, narrow-band data transmission, while non-telcos used the sub-Gigahertz ISM (industrial, scientific and medical) band. Like public spectrum-using Wi-Fi, the big benefit was cost and ease of deployment; as with unlicensed Wi-Fi, however, the downside could be congestion and unreliability.

That crucial “is it actually worth it” issue also reared its head with the emergence of the smart home where telco attempts to win consumers have so far been met with limited success. Again, telcos appeared to overestimate the importance of the ‘network’ and, therefore, the primacy of the network operator in the emerging market for networked home gadgetry. It is still emerging but is usually serviced by indoor networks, while the smart gadgetry is often provided by the likes of Google.

Still, if the convergence of Wi-Fi and 4G/5G continues, telcos who support Wi-Fi in the home might have another option for IoT with the apparent rise of Wi-Fi HaLo, a low-power, long-range (up to a mile or so) version of Wi-Fi, which operates at sub 1-GHz and can properly penetrate walls and operate on wearable devices like smart watches. The standard – set in 2016 – is reckoned to have potential to play significant roles in smart homes, enabling the usual applications, such as battery-powered cameras, video baby monitors and so on, and better yet is able to operate ‘in the presence’ of other Wi-Fi standards, so may end up being a standard feature on home Wi-Fi gateways.

There’s surely much more defragmentation to come as 5G gets built out and adopted for IoT connectivity, bringing telcos property into the IoT game.