Making Sense of the IoT Market
This is the first of several pieces I intend to be contributing over the next few months on the subject of Internet of Things (IoT). While network gurus today spend their time wresting with MPLS backbones, Internet services and SIP trunks -- or, more likely, combining them in an SD-WAN or other next-generation solution -- IoT is going to be the next big opportunity for the IT industry and IT professionals.
As I pointed out in a piece a few weeks back, by definition, IoT applications will depend on networks, and those networks will have vastly different characteristics and requirements than what we typically deal with today. To address those requirements, virtually every segment of the wireless industry is offering up solutions.
As has happened more times than I can count, I'm lucky enough to be getting involved just as the ball starts rolling. I'm currently working on a project where part of our responsibility is to do an objective analysis of the various IoT transmission solutions that are being brought to the market. This deals not only with fundamental issues like coverage, data rates, and latency, but also how much these services are going to cost, how the pricing mechanism will work (e.g. per device, per GB, per transaction), how the service selection will impact battery life, what type of network management systems we will need to monitor/maintain/troubleshoot/update/etc. all of these devices, and the list goes on.
Organizing the Market
In order to bring some organization to the analysis, we're having to take a cut at categorizing all of the various types of applications that fall under the amorphous heading of "IoT." Before we even got into categorizing though, we first had to come to an agreement as to what constitutes an IoT application.
My working definition of IoT (and it is admittedly a work in progress) is:
The idea of building a network of physical objects ("things") embedded with electronics, software, sensors, actuators, and network connectivity, enabling them to collect and exchange data and potentially respond to commands from an application.
More simply put, it's the idea of applying information technology to the real world as opposed to the "virtual world." Given the breadth of IoT applications that have been discussed, I prefer to refer to it as a design "idea" rather than a specific class of applications. The important thing is what differentiates IoT applications from all of the other stuff we do on networks.
The fundamental difference I see is that in IoT applications we are linking machines to machines rather than connecting people to applications. People are involved in IoT applications, but not directly. Waze (which I count as the single biggest IoT app today) provides great driving directions to people, but what makes that happen is an application in that person's smartphone providing speed, direction, and location information to a server that analyzes that along with information from thousands of other users and computes optimal routes for all of them based on mapping technology -- crazy!
While all IoT applications incorporate those characteristics, the components, expertise, business models, marketing requirements, and purchase justifications will all be different. How do you go about organizing that diverse a set of capabilities into meaningful categories? For the moment, I propose four major categories. The more precise delineation of these categories and their challenges and relative stages of development will be the topics of future pieces.
For now, the four IoT categories I discern are:
1. Smartphone-Based Applications: Those who have started their investigation of IoT with their heads in the clouds, often fail to recognize that the smartphone, or more specifically a smartphone with an IoT-enabling application, is the most widely deployed IoT endpoint today. The majority of Americans have smartphones, and with their built-in accelerometers, GPS-based and other location technologies, and soon altimeters, they are highly functional endpoints.
Smartphone-based IoT applications take two primary forms: neutral and vendor-affiliated. I put things like Waze in the "neutral" category, because Waze isn't trying to sell us anything (though its sponsors most certainly are). Amazon and the various loyalty apps represent the biggest segment of the vendor-affiliated selections, though UC and team collaboration mobile apps and fixed mobile convergence solutions fit into this category as well. The difference between success and failure with these is first marketing to increase awareness and then offering a user benefit that outweighs the aggravation of dealing with a separate app. In that regard, the UC vendors have consistently failed on both counts.
2. Specialized (Often Embedded) Endpoint: Applications that depend on specialized endpoints are the model of IoT that most people recognize. GM embeds its OnStar terminal in GM vehicles (and other auto manufacturers with similar offerings do the same). Progressive Insurance's Snapshot and many fleet management solutions utilize a device that connects to the vehicle's On-Board Diagnostics (OBD) port. Medical devices, like infusion pumps, might have built-in IoT modules. In this category of IoT apps, the vendor is not selling IoT per se, but a vastly improved product whose benefits are linked to IoT technology.
3. Industrial IoT (I-IoT): I separate industrial IoT applications primarily because they are so much farther along in terms of development. While the idea of telemetry might be new to some, we've been doing things like that in industrial applications for decades. Many of the case studies of IoT applications I have reviewed have merely taken the idea of "telemetry" and rechristened it "IoT." These are the same things we did with SCADA systems back in the 1970s.
Industrial engineers have gone to great lengths to refine these ideas and incorporate technology into their operations. As a result, factory and warehouse automation systems have already gone through several generations. Unlike consumer-oriented applications which have to work their way into people's lives, I-IoT adoption is based on hard-dollar analysis, and these people have gone through that analysis thousands of times already. For my money, I-IoT is a sure bet, and its development will be largely independent of IoT adoption in the consumer space.
4. Grand Visionary Ideas: I was at a loss in coming up with a name for this last category, though I do love the promise it holds. This segment would be comprised of applications that do things like monitor precipitation and weather conditions over entire watersheds and automatically control distribution systems to optimize water usage. Or systems that monitor sea and weather conditions and control the operation of massive sea gates to control flood waters. We could have systems that monitor air quality and post public alerts or take steps to reduce vehicle traffic into certain key areas.
That stuff is great, but it is clearly going to take us a lot longer to get there. The challenge won't be the availability of tools to build the solution, but the expertise to design, engineer, secure, troubleshoot, and manage the solution, let alone finding the money to pay for it. While I'd love to work on those projects, for the near term my focus will be on the first three categories.
UC's Place with IoT
I have no doubt that IoT will represent a major market force and a major revenue opportunity for somebody going forward. Certainly, the mobile operators are getting in line to provide the connectivity. We also have dozens of device manufacturers and mobile app developers chomping at the bit to deliver the endpoints once someone has spec'd the application.
The question is, will any of the UC (now "team collaboration") vendors have a place in this rapidly evolving market? The marketing folks have latched onto the term "digital transformation," much in the way they latched onto "enterprise mobility" some years back. All of that mobility talk produced an embarrassing string of failures, so if history is any guide...
Success in the IoT space is going to hinge on user adoption, and that will be a major challenge for enterprise-oriented UC vendors. Most haven't yet come to the realization that their ability to build something has absolutely nothing to do with whether anyone needs or wants it -- no matter what buzzword you use to promote it.
Scoring points in the IoT game will require the ability to recognize and meet user needs and deliver functional and appealing solutions. So we'll have to see which of Forrest Gump's immortal maxims they adopt. Let's hope it's, "Life is like a box of chocolates," rather than, "Stupid is as stupid does."