Internet of Things: Birth of the galactic network
18 March 2015 | Guy Matthews
If carriers are to be an integral part of the technology stack for internet of things (IoT), they need to develop so-far underdeveloped abilities.
In 1962 JCR Licklider, then working at the US Department of Defense’s Advanced Research Products Agency (ARPA), described to his colleagues his vision of a “galactic network” in which everything could talk to everything. His “intentionally grandiloquent” phase, as he described it, was a vision that machines might improve the experience of the world for us, without our having to get involved.
“Lick” lived until 1990, and saw his vision implemented by colleagues as ARPAnet, which became the internet. But he never saw his “galactic network” become reality. We may finally see it in the form of the much-hyped Internet of things (IoT). Gartner has forecast that the IoT could add $14.4 (£8.5) trillion to the global economy by 2023.
And, for those in the communications industry who have been working on machine-to-machine (M2M) projects for years, it’s about time the world caught up with Lick’s vision. In November 2014, the Harvard Business Review published an article by Michael Porter (Harvard Business School professor, and the world’s most cited business academic) working with James Heppelmann (president and CEO of software developer PTC), examining the new skills that would be required.
“Smart, connected products raise a new set of strategic choices related to how value is created and captured, how the prodigious amount of new (and sensitive) data they generate is utilised and managed, how relationships with traditional business partners such as channels are redefined, and what role companies should play as industry boundaries are expanded,” Porter writes.
“For many firms, smart, connected products will force the fundamental question: ‘What business am I in?’” Companies are asking this question now for two important reasons: costs for building smart products and services are practical, and the rewards for success are real.
First, costs: “We’re seeing a point where the access technology price has come down to a point where it enables the next wave of applications,” says Phil Skipper, head of M2M business development at Vodafone, who jokes that he can tell people at parties what his job is now and they understand what he is talking about.
“In previous jobs, I was doing this sort of project for large networks initially because the access cost was pretty high,” he adds. “We then moved down and we did instrumentation for buildings and manufacturing equipment. Then we got down into metering. What we’re seeing now is that combination of scale and access point, which has just really opened up a new sector.”
Second, rewards: “This is all about the data,” says Richard Cornish, head of IoT at Xchanging, and a 20-year veteran of M2M applications, first in the military, and then for several service providers. “Every report you read about the IoT focusses on the number of devices, but that value is at the other end – managing the data, obtaining value from it. The machine is 1% of the value of the proposition.”
IoT VALUE CHAIN
What will be the role of the communication service provider in the IoT’s value chain? It has little control over the endpoints, and little influence over how the data is used.
So is it just a message-passer? Not so, says Skipper. “I think the IoT promotes the carrier, which has the ability to impact multiple parts of that value chain. There’s obviously going to be a large expansion in the number of connected devices. What’s interesting in that connected device space is that they are going to be very different connections to the type that carriers are used to, because these are going to be much lower cost and are going to have a much lower data intensity.”
In Vodafone’s experience (it has been involved in M2M for 20 years), the carrier has two primary roles that map to the two business drivers. The first is managing network complexity efficiently and reliably, when millions of devices are connected.
The second is acting as a many-to-many hub for collating, identifying and redistributing data between the authorised parties on the customer side. This might mean redistribution outside a closed network.
“If I’m doing something like billing for water, there’s also a lot of valuable information there that could be used for leakage detection. The carrier is actually in a very good position as an aggregator for both the devices coming in across a range of different communications media and also the management and distribution of that data to other parties,” Skipper says.
While Vodafone has a global presence and experience in this sector, for most carriers the IoT is creating opportunities for innovative partnerships, in which specialist providers will enhance carrier networks to solve these two problems.
For example, CENX announced in November that it has been selected by a Tier-1 North American mobile service provider to automate service assurance for its machine-to-machine (M2M) offering.
Nan Chen, co-founder of CENX, believes that investing in orchestration is a must for carriers who want to participate in the IoT. “Two things are lacking in the IoT: quality of service (QoS) and encryption of the data,” he says.
Carriers are caught between two standards: carrier Ethernet offers guaranteed QoS, but takes weeks to deploy. Internet connection is quick and flexible, but offers no guarantees of security, reliability or performance. This leaves space, Chen believes, for services like CENX’s Cortx Service Orchestrator.
One of the most complex problems which service orchestration can attempt to solve will be the distributed definition of identity for the IoT.
Devices, many of them extremely basic, need to identify themselves to each other to set up a secure connection. The problems of doing this profitably multiply with the scale of the implementation.
This threatens profitability. Thorsten Trapp, the CTO of Tyntec, believes that many communications service providers will be unable to participate in the IoT and generate the returns that investors expect.
“If you talk about the mobile variant of IoT, it’s difficult to be profitable: the entire network was designed for a higher ARPU,” he says. “If your ARPU from an IoT SIM is 50c, you can’t use a network that costs $1 per user per month.”
Tyntec is partnering with carriers who want to create IoT applications, but don’t have the cost structure to do it. It has built a cloud-based infrastructure for low-cost messaging, and is now applying that to M2M applications provided by carriers.
“We want to bridge the internet world with the telco world. For years we have had customers sending GPS co-ordinates from SIM cards via our APIs – that’s a classic definition of a small IoT implementation,” he says, “There won’t be one infrastructure that does it all. The way to an efficient infrastructure is that operators think of their network in terms of elasticity. They are better off partnering with someone who can bring their costs down.”
This won’t be true of all IoT implementations – Trapp draws a distinction between a sensor that reports once a month, or only when it is outside its normal range, and the type of applications that are constantly streaming video, for example.
The former types of applications might be closed in the short term to carriers with the “wrong” business model. “If someone says, can you manage 10 million SIMS for me with an ARPU of 10c? I can say yes. The cost base we have could go down to 1c per SIM for us.
Telcos are not used to thinking in that way: their analysts do not see declining ARPU as a good thing. In large carriers, if someone says, ‘let’s do IoT! It will cut our ARPU in half!” he’s not the most popular person in the meeting. So those changes will happen way quicker in the smaller operators,” Trapp says.
Yet, if carriers can offset declining ARPU with a growth business in analytics, they create a new business model that can be an effective source of long-term revenue.
At Xchanging, Cornish leads a team of analysts who work alongside carrier analysts to determine and exploit value in M2M and IoT data. He prefers to call the IoT the “digitisation of business”, to focus on the process rather than the device.
He warns carriers that they need to let go of some of their traditional practices, because time to value is important: so you can’t control the service end-to-end, and you can’t rely on acquisitions to own the entire value chain. “Carriers still want to test every device connected to the network. Can you test a billion devices? No. And the time to react to the IoT market is much faster than service providers are accustomed to.”
He helps carriers to build applications to exploit data: to create “small data”, which is a minimum data set which gives useful information about how well an IoT infrastructure functions, and to respond to simple alarms efficiently. This sits alongside Big Data applications, in which data can be hubbed and brokered. But the progress towards that may be slow, as the resources (and the data) for large-scale analytics of this type are often distributed piecemeal throughout the carrier’s network and organisation.
Cornish warns, though, that this capability will be the arbiter of success for carriers in IoT: “You can’t lock in the connectivity in this market. You need to be able to offer a complete value proposition to the customer.”
In 1963, Licklider sent a memo addressed jokingly to “Members and Affiliates of the Intergalactic Computer Network”. Carriers have, from the start, been members of the network. Now the promise of the IoT for carriers depends on working constructively with the affiliates in this emerging intergalactic network – the specialists who can help build the communication and data-sharing business models on which successful IoT applications will depend.