Rise of the machines (again)

The concept and practice of using cellular networks to connect machines - from remote sensors and cars to vending machines sending out refill orders by SMS - has been around since at least the late 1990s. The premise sounded like a no-brainer as cellular operators moved towards the data-centric world of 3G and sought out new revenue opportunities that would help make up for the inevitable decline in voice revenues.

Obviously, it never quite worked out that way. While a number of cellcos today do offer M2M services in some form or fashion, it's still very much a niche app, and a costly, low-margin one at that.

However, in the past year cellular M2M has been experiencing a Renaissance of hype as cellcos start revisiting the idea. This time last year, KORE Telematics, Wireless Maingate, Mach Communications and Vianet pulled together to form the M2M Global Alliance "to help global M2M application providers more quickly and cost effectively expand their services overseas" via overriding protocols and interoperability mechanisms to unify disparate networks regardless of underlying technology or carrier.

A few months after that, the GSM Association started a new initiative focusing on embedded modules to establish guidelines for M2M deployments on GSM networks and a certification process for devices and apps, claiming that 50 billion devices will be cellular-enabled by 2025. This year, the GSMA followed that up with a competition to stimulate M2M apps and solutions development.

Indeed, 2009 has seen a number of high-profile moves in the cellular M2M space. In May, AT&T teamed up with global M2M player Jasper Wireless to push its M2M strategy. In July, Vodafone launched a global M2M platform, and added a green-tech spin by publishing a study with Accenture showing how M2M could help reduce carbon emissions (see sidebar, "How M2M could save the planet", p. 15). The same month, Verizon Wireless and Qualcomm - a long-time player in the M2M space - formed a joint venture dedicated to building cellular M2M solutions. Meanwhile, Berg Insight forecasts almost 9 billion euros of M2M revenue by 2012.

What's going on? Experts chalk much of the revived interest and activity to a confluence of various factors, not least of which is the push towards broadband systems like LTE and, to a lesser extent, Wimax (although Wimax identified M2M early on as a potential differentiator - KDDI listed M2M as a potential service for its Wimax offshoot, UQ Communications, as early as February 2008), which promise the most M2M-friendly incarnation of cellular yet.

However, like everything else in the wireless business, where there's unbridled enthusiasm, there's also eyebrow-furrowed skepticism. Even M2M specialists are warning that while there's plenty of reason to be optimistic about cellular M2M in the long term, the short-term picture could be a case of carriers getting in over their heads as they try to tackle a potentially labyrinthine nightmare of device management in exchange for low ARPUs.

Kindles count as M2M

It's worth mentioning at this point that one thing driving the enthusiasm over M2M is that the definition of what constitutes M2M has evolved. Ten years ago, when people thought of M2M modules, they thought of utility meters and Coke machines. Now it includes things like telematics, health monitoring devices and even consumer electronics products like Amazon's Kindle ebook reader.

"In the consumer electronics space, there's a lot of things you can connect," says Al St George, VP of international markets for the newly formed Qualcomm/Verizon JV. "E-readers are getting a lot of press, but there's a big opportunity there. The Kindle is just the tip of the iceberg."

Even smartphones could qualify as M2M terminals for certain types of business apps, says Francesco Masetti-Placci, head of end-to-end LTE and 4G wireless networks at Alcatel-Lucent APAC.

"You can imagine uploading information and documents from the head office to a mobile workforce, for example," he says.

Digital signage is another segment that comes up in the M2M discussion nowadays, Masetti-Placci adds. "That definitely requires more bandwidth to be more appealing. Today, the interactivity is limited to SMS, but if the signs are enabled to stream video to your iPhone, for example, then there is more value."

Even more traditional categories of M2M like meter reading for power companies are expanding to include more sophisticated applications, says St George.

"Utility companies are building out infrastructure to meet peak demand. So they're looking at M2M solutions connecting houses and appliances where you can not only monitor them but control them remotely," he says. "So on a hot day, they can do rolling blackouts in your house, and shut off the air conditioners street to street for five minutes at a time. That brings peak demand down. So there's a lot of activity going on right now to get on that curve."

Price compression

Another factor driving renewed interest in M2M is that the hardware prices have come down significantly, says Alex Brisbourne, CEO of KORE Telematics.

"We've seen a lot of price compression in the last 18 months. A GSM module now costs less than $20, and in the next couple of years it will drop to the low teens before it flattens out," he says. "So the actual finished device can be priced at under $100, and that will go down to between $50 and $70, which will open up the market significantly. If you just need simple connectivity, of course, it will be even lower."

Similarly, bandwidth pricing has changed as cellcos revamp their mobile data plans from expensive and confusing per-megabyte charges to flexible flat rates, says St George. "Because carriers are more interested now in M2M, we've seen a lot of them modifying their prices and deals to go after that market."

The 4G factor

The move towards LTE and Wimax is also a significant factor in the M2M revival, despite the fact that current M2M apps use so little data that they can purr along smoothly even on legacy GPRS or CDMA 1x networks.

"Smart meters don't need a lot of bandwidth. But as more and more devices come onto the network, operators need to be able to manage them better, and the extra capacity helps," says Doug Chambers, director of market development at the GSM Association. "The other thing to consider is the consumer electronics sector - not just Kindles and smart books, but digital photo frames and cameras will have wireless connectivity, and you're going to need big pipes for that."

St George of the Qualcomm/Verizon JV lists several "bandwidth hog" apps coming to M2M: "consumer devices with larger screens displaying video, camcorders uploading directly over the air while the consumer is shooting video, interactive gaming and video monitoring for security."

He adds that while current 3G networks can handle such apps in terms of throughput and latency, "the critical element for these apps to scale is network capacity. One person streaming HD video won't crash a carrier's network, but millions doing it would with today's networks."

That said, the even bigger attraction of LTE for M2M is the all-IP architecture sitting behind the fat access pipes, says Masetti-Placci of Alcatel-Lucent. "The evolution towards a full IP-based infrastructure simplifies the deployment and implementation in an open way of those types of services."

That will also pave the way for IPv6, which is widely seen as a necessary component of an IP-based wireless broadband future where almost every electronic device will have an Internet port, he adds.

"That means carriers will not only be able to manage a large number of devices simultaneously, but also assign IP addresses for each of them, which would simplify tracking, application processing and so on," Masetti-Placci says.

Chambers of the GSMA adds that LTE's efficiencies could be maximized for M2M apps in which the terminals are stationary, like meters or digital signage.

"If it doesn't move, you don't need to support the mobility protocols or manage things like handover, so you need to look at ways to reduce the signaling overhead," he says.

Complexity shielding

With all that, it's little wonder carriers are keen to get all these machines on their networks, if only to take advantage of their upgraded network capacity and rake in whatever revenue it generates.

The question, as always, is figuring out a bankable business model. That's likely to vary from app to app. For example, the Kindle angle suggests a model where the content provider subsidizes the wireless data charges, says St George. "Amazon did a great job with the Kindle in terms of dealing with what the customer pays for. The wireless charges are completely invisible to the user. They just pay for the books or magazines they download, and the device itself."

However, carriers that have already ventured into M2M see it more as a managed services play for enterprises that simplifies an otherwise complex service and delivers it over a nationwide or even international footprint.

When Vodafone launched its M2M platform, its global enterprise chief, Nick Jeffery said, "service providers like Vodafone need to provide customers with managed connectivity that is reliable, cost effective and universal" in order for M2M to reach its true potential."

SK Telecom has a similar strategy behind the various M2M services it's been offering since 2002, from remote meter reading and street lamp control to a managed M2M platform, says SKT solution business team manager Jung-Hyun Kim.

"Mobile carriers can develop a new market with B2B business models by increasing the usage of existing networks," Kim says. "By using solutions provided by the mobile carriers, corporate customers that do not own networks will benefit from being able to manage their operational resources in the most efficient manner, and the overall efficiency of the industry will be enhanced as well."

The Qualcomm/Verizon venture is also focused on shielding enterprises from the complexity of M2M, says St George.

"A lot of companies underestimate what needs to be done for a successful M2M business - how to manage devices, add and drop them, data mining," he says. "We can make it easy for people who don't understand it and just want a portal to manage devices and make it easy. It's like you see a duck on the surface of the water and we're underneath doing all the paddling."

However, while M2M sounds more appealing than ever as a B2B play for cellcos and has still come a long way in overcoming the hurdles, it hasn't cleared all of them. Not yet.

Defragging M2M

For a start, fragmentation is still an issue, driven by the fact that M2M apps and modules tend to be highly diverse and specialized.

The GSMA's embedded module initiative that hopes to mitigate the issue, and the GSMA is tackling the fragmentation issue on three fronts, according to Chambers: certification of devices, provisioning and efficient management of devices and capacity.

"If you're going to have that many devices, you can't just treat the network as an open pipe," he says.

The GSMA intends to publish guidelines for M2M modules - the first of which are due by the end of this year - that will pave the way for certifying devices, which in turn will lower costs and boost economies of scale. Chambers also cites the GSMA's work on embedding HSPA in laptops and netbooks as a model for developing M2M certification.

"Our laptop initiative helped operators understand the laptop/netbook ecosystem, the value chain and the operators' role in that," Chambers says.

Another significant - and familiar - barrier to M2M adoption is the upfront cost, which can still look intimidating, says St George. "You can get the business model to work in the long term, but the upfront cost is hard for companies to swallow."

Disappointing ARPUs

However, perhaps the biggest challenge for carriers may be looking past the current hype regarding the potential M2M payoff - which, says Brisbourne of KORE, is going to be disappointing for many operators.

"Carriers are being dazzled by the hype over the 'billions and billions of machines' that will need cellular connectivity, which is an interesting stretch, when considering that today in the world, we have about 55 million true cellular M2M connections," says Brisbourne. "To move from there to billions is a pretty steep hockey stick. I think it's going to take a longer runway than people may think."

Brisbourne offers KORE's own M2M residential power meter business in North America, which connects 1.1 million homes, as an example. "It took five years to get to that number, and in reality represents less than 10,000 actual cellular connections, because the Phase 2 infrastructure in the utility companies is all about ratios of 40:1 to 100:1 per cellular uplink."

Combine that with the ARPU per M2M connection - which Brisbourne puts at $5 a month on the high end, which works out to just $300 off that one connection over the five-year life-cycle of the device - and compare it to $3,000 per subscriber on a three-year iPhone plan, and the economics of M2M don't look nearly as attractive, Brisbourne says.

"How excited can a carrier get about taking five years to add another 10,000 connections at five bucks a month?" he says. "Then the fundamental question becomes at what point are they going to be fed up with machines not making a blip on their revenue screen, and contrast that with the complexity of front-end tools and systems, integration and relatively slow ramp time. That's the hidden curse in this."

That's not to say Brisbourne thinks carriers should stay out of the M2M business. "Carriers are contributing to M2M significantly by building out more capacity, larger footprints and better reliability, which helps M2M tremendously as we move into new areas like e-book readers, PNDs and even smartphones that can be used for industrial apps like updating tech manuals," he says.

But he maintains that carriers need to recognize that M2M is a complex business with very specialized requirements.

"Carriers can best serve customers that need very simple, vanilla connectivity. For apps that require complex business process management, that's where specialized platforms from Jasper Wireless or ourselves play an important role," he says.

Sidebar: How M2M could save the planet

Accenture and Vodafone are touting M2M's green credentials as a way to reduce CO2. But it's no silver bullet, writes Ovum's Jeremy Green

A recent detailed study by Accenture and Vodafone builds on earlier modeling to reveal concrete and credible examples of how mobile communications can help to reduce carbon emissions. The study claims targeted implementation of wireless communications could save 113Mt of CO2e (carbon dioxide equivalent, including other greenhouse gases), reduce spending on energy by E43 billion, and require around a billion connections for mobile operators by 2020 within the EU25 countries.

The study covers 25 EU countries, and also provides some consideration of Australia and India. Eighty percent of the identified carbon savings would be enabled by M2M services. Some 80Mt of CO2e could be saved in India through the application of M2M in the electricity generation, transmission and consumption sector alone, says the report.

The report consciously builds on the Global e-Sustainability Initiative (GeSI) report Smart 2020: enabling the low carbon economy in the information age, which argued that ICT could deliver 7.8Gt of CO2e savings, representing 15% of total CO2e emissions in 2020. However, it differs in that it takes a much more detailed and concrete approach, and focuses exclusively on select mobile technologies rather than ICT as a whole.

It finds that mobile technology could reduce CO2e emissions by 2.4% of EU 2020 totals - a smaller proportion than the GeSI report, but much more closely tied to actual data, and therefore much more credible than the slightly Panglossian tone of the earlier document.

The Accenture-Vodafone analysis breaks down the possible carbon-reduction opportunities to a shortlist of 13, organized into the broad categories of smart logistics, smart grids, smart manufacturing (i.e. real-time monitoring of high-value equipment) smart cities (i.e. advanced traffic management) and dematerialization (remote working schemes, video conferencing and e-commerce).

However, it is clear that neither wireless M2M nor ICT in general can be a silver bullet that can solve the problem of how to reduce emissions without impacting on growth. The study finds savings of up to 2.4% of EU25 emissions in 2020; by way of a comparison, it says that a 'business as usual' scenario will deliver reductions of 8.8% by 2020 against 1990 levels. The 2.4% savings would bring the EU25 somewhat closer to its target of a 20% cut in emissions by 2020, although some other authoritative sources suggest rather higher levels are required.

To the authors' credit, the report contains detailed and strong recommendations as to what should be done to encourage and stimulate the deployment of technology-based approaches to emissions reduction. It rightly gives first place to the need to "deliver an effective price for carbon".

-Jeremy Green is the Mobile Practice leader at Ovum