There’s a standards problem in IoT, here’s how to solve it
IoB Insiders: AI, big data and the cloud - how do they impact IoT?
IoB Insiders: AI, big data and the cloud - how do they impact IoT?

There’s a standards problem in IoT, here’s how to solve it

IoB Insiders Professor William Webb, CEO of the Weightless SIG, discusses the current standards problem facing the Internet of Things (IoT) market, and how he thinks it can be solved.

Imagine that there were three different varieties of Wi-Fi style connectivity. Whenever you ordered a new Wi-Fi device you would have to select the right option – which might not always be available. Only some hotspots would be accessible to you – those with the same technology as in your laptop. Some places, like planes, might offer multiple technologies but because of lack of spectrum this reduced the data rate for each. Others chose not to deploy Wi-Fi until the confusion over the preferred technology was over. Such a world would be much more annoying and much less productive than the one we currently inhabit.

This is exactly the world of unlicensed IoT – the machine equivalent of Wi-Fi connectivity. Technologies include Sigfox, LoRa, Ingenu, Telensa, Weightless and others. Devices from one are incompatible with others and they mostly compete for the same spectrum, often interfering with each other. Why do we put up with it?

Some claim that these technologies target different market segments – for example Sigfox targets the lowest-cost devices that only transmit data, while LoRa is for more complex devices that need to receive as well. But Wi-Fi covers a wide breadth of use cases with a single standard, as does cellular.

We do not have one cellular technology for the high-bandwidth gamer and a different one for the low-usage pensioner. It is invariably less expensive to have a single standard, able to meet the needs of most users, which can deliver economies of scale and encourage widespread deployment of infrastructure.

Others claim that the market is large enough for multiple technologies – but it is much smaller by value than cellular. Or that competition is needed between different technologies at this early stage of its evolution – but most standards evolve as lessons are learnt rather than compete.

Related: NB-IoT market to grow exponentially in the next five years

Standards wars

These are arguments designed to perpetuate the status quo, rather than ones with substance. A quick glance at the current wireless connectivity we use throughout our lives shows that there are no exceptions to the rule that (1) we have a single preferred connectivity solution in each different space – eg Bluetooth, Wi-Fi and cellular – and (2) only open standards succeed. There are no reasons why IoT should be any different. And more fundamentally, there will not be widespread success of IoT connectivity until these conditions are met.

For our cell-phones we are used to wide-area connectivity being provided by mobile network operators (MNOs) using 3GPP cellular standards and local connectivity self-provided through Wi-Fi using IEEE standards under the auspices of the Wi-Fi Alliance. It seems likely that a very similar outcome will transpire for IoT.

Wide-area connectivity will come from MNOs deploying 3GPP standards, most likely NB-IoT. Local connectivity will come from an ETSI standard under the auspices of the Weightless SIG. Some devices will have dual-purpose chipsets, others will have perhaps only Weightless-certified connectivity – in the same way that some devices have dual cellular-Wi-Fi chipsets and others Wi-Fi only (but it is rare to have a cellular chipset that does not also have Wi-Fi connectivity).

While few would disagree with the prognosis of NB-IoT deployment, many might question the prediction of unlicensed deployment. After all, isn’t it the case that Sigfox and LoRa have significant deployments already? The answer is emphatically “no”. If we are to reach the predicted 50 billion devices in, say, a decade, then we need to deploy 13 million per day.

Sigfox have around 10 million users, less than a day’s worth of deployment. These are more akin to early trials than mass deployment. And this is exactly what would be predicted based on the observation that markets only succeed when there is a clear single open standard. (Sigfox are part of a process to develop a standard within ETSI so it is possible that they will find a route to become that open standard).

The situation of competing proprietary technologies can only be resolved through the wider industry getting together and collectively putting its weight behind a single unlicensed standard. The Weightless SIG is providing such a forum and welcomes membership from all those who wish to see this untenable situation resolved quickly in the interests of all who want to see IoT succeed. Or would you prefer the world of multiple incompatible Wi-Fi variants described above?

Related: 400 platforms and one big fat reality check for IoT