There was a hint of schadenfreude among mobile operators last week when it emerged that Vodafone, probably the loudest exponent of the NB-IoT low power LTE standard, had missed some of its deployment deadlines for the technology, and now risks being overtaken by Deutsche Telekom in the important market of Germany. More serious than MNO build-out races, however, is the question of why Vodafone’s progress has been slower than expected – and more broadly, how many MNOs will actually manage to find a profitable business case for IoT networks.
The operator’s own comments suggest it may not be seeing sufficient demand for low power wide area network (LPWAN) services such as wireless smart metering or smart city applications, to justify a more rapid schedule. A Vodafone spokesperson told LightReading that its Spanish NB-IoT commercial roll-outs had started in the first quarter of this year, as promised, and was now live in six cities. But the UK and Ireland will be delayed until the summer, because those markets are “not ready” yet, while there was silence about progress in Germany, the fourth country where NB-IoT had been pledged by March 2017.
“The launch of NB-IoT in each market is as much about when our customers are ready as when the network is available,” said the spokesperson. There was no comment on whether the timetable for other launches, beyond the first four markets, would be affected. Vodafone has promised full coverage of its global network by 2020.
Deutsche Telekom, however, says it has customers “lined up” in the Netherlands, where it is deploying a network which is scheduled to be national by year end. The Germany telco will introduce services in Germany during the second quarter, possibly edging ahead of Vodafone, and is also rolling out NB-IoT in Austria, Croatia, Greece, Hungary, Poland and Slovakia.
Meanwhile, other European MNOs are using alternatives to NB-IoT, in unlicensed spectrum, either as an interim solution before implementing the LTE-based technology, or as a full solution. Telefonica has a major deal with Sigfox, in which it invests, while Orange, Bouygues and KPN are using LoRa.
The LPWAN market is certainly not looking like a slam dunk for NB-IoT as Vodafone, in particular, has been inclined to argue. Even though the LTE-based technology has got to market quickly and without significant technical hitches, there will clearly be a need for an unlicensed spectrum solution, for non-MNOs, and for mobile operators which want a dual-network strategy, as they have with WiFi. ZTE revealed its IoT network game plan last week, with an array of announcements at the CeBIT tradeshow in Hanover, and LoRaWAN was part of the strategy, alongside NB-IoT.
ZTE has been a strong proponent of LoRa and so its commitment to support the technology in network infrastructure was unsurprising, but still a welcome boost to the technology, since the top network equipment vendors have, in general, shown little interest in unlicensed LPWAN options (though Nokia recently extended its cloud packet core and IoT managed services offerings to support LoRa and Sigfox).
The Chinese company has been heavily involved in the LoRa ecosystem, initiating the China LoRa Application Alliance (CLAA), which now lists 500 participants. Now it seems ZTE is making a bid to seize leadership of the specialist area of LoRaWAN network technology – a space currently populated by smaller companies. In addition, ZTE will be partnering with Semtech, the main developer of LoRa technology, on the development of GPS-free geolocation functionality (LoRa geolocation).
Though GPS provides extremely accurate positional readings, the trade-offs include the cost of the device hardware and higher power consumption. Typically, GPS is paired with a GSMA standard to provide backhaul, which then requires a monthly subscription to an MNO – meaning GPS is an expensive choice for IoT devices.
A pure LoRa approach could provide a far cheaper alternative, with the main trade-off likely being the degree of accuracy. However, cheaper devices with longer battery lives are a big draw for many adopters which do not need pinpoint accuracy.
More important than the connectivity itself will be the business models, enabling operators or others to monetize all those links. ZTE and the CLAA have made some progress in this area, and unveiled a ‘Smart Street 2.0’ offering in Hanover. This is the second generation of ZTE’s city management solution, which combines IoT, cloud and big data technology in a centralized “street command and control center”. This collects data from IoT city infrastructure and then uses an embedded analytics engine to interact with city residents and vehicles, providing notifications, navigation and smart routing information.
Like Nokia, ZTE knows the real value for a large vendor will lie in managed services, tying together different types of connectivity and devices, and enabling security and applications. It is now testing a new “business and cooperation model” harnessing the unified CLAA LoRa gateways, which enable certified hardware from different vendors to access the network. Using algorithms in the cloud-based core network, network resources can be shared among users on-demand, a first step towards network slicing and an echo of Nokia’s new IoT integration platforms. ZTE provides the operation and maintenance services and integrates the numerous dispersed small and medium-sized LPWAN operators so that they can achieve the benefits of scale, geographically and in terms of their cost base and services reach.
When it comes to the operators, however, it remains unclear how much LPWAN connectivity will add to their revenues. Most operators are very clear that connectivity alone will not justify large-scale LPWAN deployments. NB-IoT may be low cost to implement, since it is available as a software upgrade to many LTE base stations (80% of Vodafone’s, says the operator). But it brings its own O&M challenges, not to mention stringent reliability and availability requirements for many services. It will be critical for operators to secure a healthy share of the real IoT goldmines – big data analysis and professional services.
According to one set of IoT predictions, from ABI Research, the Industrial IoT market will add 13m wireline and wireless connections worldwide this year, to a current base of over 53m. That is a healthy growth rate, but the connectivity fees across wireless, wireline and satellite will generate just $138m in the industrial space, globally. That is in the most potentially lucrative segment of IoT, where large organizations are investing heavily in digital transformation and big data strategies, underpinned by connectivity to every object from engines to robots to components. By contrast, a report from Research & Markets looks ahead to an IoT professional services market already worth over $57bn and growing to $158bn by 2021. The largest segments in terms of this spending are the Industrial IoT, cities and transport, and most vertical markets are looking for professional services partners to customize their applications, provide business transformation consultancy around IoT, design and integrate the systems, and help crunch the data.
The big issue for telcos is whether they can wrest enough of that high value business away from other suppliers, some of them with deeper connections within enterprises. Integrators, management consultancies and IT giants like IBM are all eyeing the IoT space, and many operators are, beneath their big statements about a multilayer IoT model, actually focusing mainly on connectivity, device management and telematics. All useful, but these telcos risk being stuck in a new, machine-to-machine version of the bitpipe.
It is very important, for operators which regard IoT revenues as a key source of future growth, to form the right partnerships with companies which have a better pedigree in enterprise, vertical and big data services. A few are doing this, including AT&T, which has recently become very open to partnering with major cloud providers, rather than trying to do everything itself, as it builds one of the world’s more advanced enterprise and IoT services strategies.
It has already formed a strategic alliance with Amazon AWS for enterprise cloud services, and has now followed that with an IoT-specific partnership with IBM, focused on analytics. The deal extends the two firms’ existing cooperation and is a good, if rare, example of how IT and telecoms majors can work together rather than competing head-to-head.
The resulting solutions combine AT&T’s IoT analytics offering and the IBM Cloud to help enterprise customers to analyze industrial IoT data more quickly in order to improve their business processes or customer solutions. The system will be supported by the AI capabilities of IBM’s Watson platform.
An integration deal between the two firms was first outlined last year and the joint capabilities are now available to customers in the pilot phase as of this week. AT&T says its M2X, Flow Designer and Control Center products monitor, collect, manage, and store data, while IBM’s Watson IoT portfolio, including the Watson Data Platform and IBM Machine Learning Service, handles the more intensive processes from which the valuable insights are provided.
Watson, at its most basic level, provides a platform for connecting, building, launching and managing IoT apps and devices. Spearheading this is the Watson Data Platform, a data ingestion engine with cognitive decision making, from where IBM says these business insights are churned out – giving the example of combining a customer’s inventory data with public information, such as weather and road conditions, in order to generate a value greater than the direct sum of the two separate data sources.
As well as the automotive sector, AT&T’s customers in the oil and gas industries have been highlighted as potential beneficiaries. AT&T says its IoT network and the IBM Watson Data Platform can ingest data from hundreds of wells, and then build a model with machine learning libraries and open source technology to help predict possible failures or machine malfunctions. “The company will be able to detect anomalies in less time and with more accuracy,” stated AT&T.
Both companies are part of the Industrial Internet Consortium (IIC), an open organization that aims to improve integration between the physical and digital world and products to facilitate quicker IoT adoption. Board members also include Cisco, Intel, Bosch, Huawei, SAP, and founding father GE – with membership at around 50 names today.