The IoT lacks a commonly adopted standard for wide area narrowband connectivity, creating adoption delays. That’s one of the key findings from ‘The Internet of Things Myth’, from Matt Hatton and William Webb, published in April 2020. An earlier article on IoT Business News looked at the predictions of 50 billion devices and how everyone got carried away with the idea of connecting everything. In this article we dig into the technology and what was lacking.
The early signs for IoT technologies in the 2010s were good. A diverse set of IoT platforms had arrived (or been adapted) from the likes of Jasper Technologies, SAP, and Thingworx making it much simpler to develop applications and manage devices. Mobile Network Operators (MNOs) were becoming much more interested in IoT, resulting in IoT-friendly pricing models. We also had a new set of Low Power Wide Area (LPWA) network technologies arrive in the form of LoRa, Sigfox, Weightless, RPMA and eventually a mobile operator equivalent in the form of NB-IoT.
Fast forward to the end of the decade and we have yet to see the convergence on a single standard for narrowband wide area connectivity, i.e. the area addressed by the LPWA technologies. For ten years we have had established standards for in-building and personal area networks in the form of 802.15.4, Wi-Fi and Bluetooth, with the specific choice being well defined depending on requirements for reliability, range, bandwidth and so forth. For broadband wide area connectivity the choice is even more straightforward: LTE, and eventually enhanced by 5G, deployed as public networks by MNOs. The existence of a few default technologies across these fields made technology deployment decisions much simpler: smart watches used blue tooth, smart speakers used WiFi, smart cars used 4G/LTE, and so forth.
Referring to LTE as the default for broadband wide area connectivity belies some of the problems associated with technology choices when using mobile networks. During the 2010s mobile operators around the world announced, and in many cases proceeded with, switching off either 2G or 3G networks, or both. This left a patchwork of technologies available, creating a headache for anyone deploying devices with a long lifespan across multiple territories: which technologies do you pick, and how do you avoid inflating the bill of materials (BOM) cost unnecessarily? Typically today LTE coverage is 99% in developed markets and module costs have come down a lot, although higher than 2G. Back in 2010 LTE was a premium product and deployments were more limited. Opting for a multi-mode 2G/3G/4G device would inflate the price significantly compared to picking a single technology, preferably 2G. This fragmentation caused additional cost, uncertainty and delay in deploying.
For narrowband wide area connectivity, which promised billions of trackers, monitors and controllers across innumerable use cases, that type of fragmentation persists. Numerous candidates emerged during the 2010s. Some were based on self-deployed networks using unlicensed spectrum, such as LoRa and Weightless (although some MNOs showed interest in deploying LoRa as an overlay public network), some were based on a network operator model, most famously Sigfox and OnRamp Wireless (now Ingenu). Eventually we also saw the mobile industry develop its own variant for use in licensed spectrum in the form of Narrowband IoT (NB-IoT).
What is required is the emergence of a single technology, or perhaps just two consisting of one technology for public networks using licensed spectrum and another for private networks using unlicensed. For that to happen, history suggests two things are required: standardisation and widespread deployment. Most of these were not full standards. Weightless was, but eventually shut up shop, although its core technology was used to underpin NB-IoT. The technologies are also not widely enough deployed. Sigfox’s network operator model has been deployed in 70 countries but with good coverage only in Europe, South Africa, Japan and a few other countries. LoRa is based on private deployments which are inherently patchy in terms of coverage, although could be perfect for a single organisation’s geographically limited need.
Even NB-IoT, which should have the critical mass to win out, has not really been widely adopted due to limited roll outs. It is a standard. It uses licensed spectrum, which should provide a more reliable network. It is backed by network operators who know a thing or two about deploying national interoperable networks. Nevertheless, it has faced teething problems. Cheap and simple upgrades to LTE networks have proved to be more expensive and more difficult than anticipated. Experience with battery life, a critical aspect for LPWA technologies, has not been as advertised. Also, the expectations for average revenue is now so low, at perhaps USD1-2 per year, that hundreds of millions, perhaps billions, of connections are needed to make it an interesting technology. That has not materialised yet. In the face of this, NTT Docomo recently announced that it was closing down its NB-IoT network in Japan. However, all is not lost. Just a few days later Deutsche Telekom, Swisscom, Telia and Vodafone announced an NB-IoT roaming agreement, indicating a commitment to making the technology more mature.