It is no surprise to anyone who has covered either the computing or telecommunications markets for any length of time that manufacturers' visions of the future are centered around the ubiquity of the products they create. The 2007 vision of ubiquitous communications among carriers revolved around a kind of flip-phone with a detachable antenna you might wear on your head or in your pocket. When the iPhone happened, it was called "disruption," but really in the sense that a bad dream was disrupted by a better reality.
The 1995 vision of ubiquitous computing from Microsoft revolved around a universal acceptance of the role of packaged software; the word "Internet" was surgically inserted into a later draft of Bill Gates' The Road Ahead. So when you look carefully at concepts of an "Internet of Things" (IoT), if you're a veteran, you might want to focus on what these things are supposed to be. KORE Telematics President and COO Alex Brisbourne (whose business is machine-to-machine communication, or M2M) has done precisely that, and shares his thoughts with us in part 2 of his three-part discussion with ReadWriteWeb.
Alex Brisbourne, KORE Telematics: I was on a panel recently with a leading guy from Intel, and he was talking up the fact that the whole industry needed to have Intel Atom processors at its edge points, and lots of capacity in the network, moving lots of data through. And I said, "Look, nobody's going to buy an Intel Atom-based sensor, which all it's doing is trying to determine whether to switch the sprinkler system on, under the sod on a golf course."
Scott M. Fulton, III, ReadWriteWeb: I can see Intel's need to create a market for itself, and Atom being its smallest processor - albeit one that really should go into a smartphone, and really nothing smaller than a wristwatch - but knowing that it's not going to have a tremendously low-power device, and that it's not going to build a 4-bit, 4004 processor again, it's going to use as much publicity as it can to create a market for itself outside the netbook.
AB: We have to be realistic. The fact is that, you go back to the [period of time] shortly post-Andy Grove, but certainly with Pat Gelsinger. With every single slide deck [Intel used] to sell a project [was driven by], "Does it sell processors?" Everything else was subservient to that particular beast. You had to convince people that they had to move up the processor tree.
I thoroughly buy the notion that that's what they [Intel] said. But I don't entirely agree that it's a universal message for the trunk of the M2M market.
SF3: Is there a possibility that just the same way Intel needs to sell processors, IBM needs to sell middleware, so it may be taking a package that doesn't quite fit what the real world envisions for the [IoT] application, and trying to shove that hexagonal peg into a round hole?
AB: It's always hard to deduce with IBM whether the engineering or the marketing organizations are at the front of the bus. By and large, it's a marketing organization. And with the whole "Solutions for a Smarter Planet" and other related taglines which they've put together, I think IBM and Google have both done a fantastic job of raising the practical awareness of what could be achieved, particularly if you look at their areas of strength: distribution, logistics, and other areas. You're seeing the initiatives coming out of Qualcomm, the Qualcomm Life program for health management, a significant number of initiatives we're involved with in the health management area as well.
Clearly, every company goes back to its mother lode of where its profitability lies, and builds its strategies around it. But if you look at it in the context of IBM, it really is at the heartland of data management. So this seems to be markedly relevant to them. I think the challenge with IBM, honestly, is that remote telecom-enabled data management, particularly with thin communications - going out to these individual devices - has never in truth been one of their great strengths. I mean, they're fantastic in what they've done in the world of RFID, and now what they're doing in NFC and financial services. But how to marry that into the thin-route, M2M stuff, still remains challenging. At the end of the day, it's all about selling the data management solution.
Scott M. Fulton, III, ReadWriteWeb: About four months ago, the CEO of Salesforce.com [Marc Benioff] was giving a conference where he wanted to make a point about the ubiquity of communication, and the need to bring things and products, non-sentient beings, into the discussion. So he stood next to a Coke machine, and said we're missing a very valuable opportunity here, because it could have detected, by the virtue of the fact that he was wearing an iPhone, that he was standing within three feet of the Coke machine. And if the Coke machine were endowed with certain logic, it could have sent him a coupon, saying, "Here, press this button, have a Coke."
He tried to make the point that we need products to be more intelligent so that they, individually, can engage with their customers. And he painted, literally, a picture of walking through a grocery store, and having the goods off the shelf communicate with your phone, and say, "Hi, there. I'm a can of beans. And if you pick me up within the next 15 seconds, I'll give you 30 cents off." Is this anything that you see real-world customers - people who actually engage with KORE Telematics - doing, or is this really beyond the edge of the pipe dream?
Alex Brisbourne, KORE Telematics: Currently it's just beyond the edge, but not very much. I suppose we've been talking as an industry about the whole notion of the interconnectivity of location, presence, as components in an enriched marketing world for ten or fifteen years, I suppose. But in reality, we've been talking about it seriously for five or seven, and there have been some very spectacular flameouts of businesses that believed that this could be picked up, run with, and made money out of pretty early on.
I do think that when you look at some of the things like Foursquare, there's some evidence that this is starting to creep in. Actually, this is an area where there are a lot of concerns with regard to security and privacy, and the whole mechanisms around opt-in services, but for location as well as for things like couponing. But...the take-up level, I have to say, remains extremely low in that segment. It's gone down. Going back to my days ten years ago as a general manager for one of the carriers up here in Canada, we were playing around with some of this technology then, and were truly astonished by the level - or more accurately, the lack of level of interest in the adoption of this technology. There's been limited technology [tests] playing around the edges, but most of them seem to be science projects, honestly, more than broad adoption.
SF3: But when you do play around with technologies with your customers in a nearer timeframe than the five- to ten-year one, to what extent do your customers start asking for the need to have more ubiquitous access to machine data from devices that they can carry in their pocket, from smartphones or tablets? Do they really need their pipelines to communicate their status to them in an app? Or can it, or should it be done better, though a remotely managed resource?
AB: There's quite a lot of difference between what people say they want and what people actually end up using. And I'm going to sound a bit like a Luddite on this one, for which I apologize, because I say generally I'm relatively forward-looking, but also I'm old enough to have become somewhat pragmatic as well.
We break out the machine-to-machine market, the Internet of Things, into two quite distinct groups, for our analysis of the industry. One is the true consumer-centric demand applications. This is where there is typically a one-to-one relationship of a consumer with a supplier, in some form. Let's call it the "connected iPad," and all it is, is the connectivity relationship. Then there's the true business-to-business application, which may have a consumer touchpoint. With the B2B one, the connectivity of the M2M device is built in as a result of an integrated, and frankly, quite narrowcast application.
Take GM/OnStar as being a relatively good example. We want to know where the airbags go off. By and large, people aren't doing too much with the other features of GM/OnStar - occasionally we have to remotely unlock their doors, things like that, when they leave the keys in. But it's a pretty narrowcast application.
The same takes place for a majority of others. I've just finished a meeting with a large customer of ours that's in the medical diagnostics industry, and they manage things like real-time glucose monitoring of diabetics to improve quality of life, and reduce healthcare costs. It's an extraordinarily narrow, vertically integrated application, for the device to the settings to the device to the user experience to the back end. And it's very specific to you, Scott, as a user of that particular device or service.
Where I think you reach a little bit of commonality is, to whom do you want to convey that information? There is certainly a growing trend [for] thin, alerting [smartphone] applications, using text messaging in some cases, when one of these alerts takes place. Perhaps a slightly more useful example would be, say, a home alarm system, which used to simply get a call from a call center. Today, if an intruder breaks in and you've got alarms from certain manufacturers, it will pop up immediately and alert your iPhone, it'll show you maybe a picture of your home, or at least where the sensors are, and it'll tell you that the back door's been opened or a window's been opened, or maybe the cat's knocked over the flower pot, or whatever.
There is a point where you can embellish applications, make them richer - frankly get people to pay more - as a result of delivering smart functions to these new, smart devices. But it's really an adjunct to the mainstream Internet of Things service. It has in effect, when you think about it, become a third level of the application, which is essentially a browser access-type application.
In Part 3: Should everything in an Internet of Things have an API?