The Comcast/Icontrol Deal is Another Sign of IoT Reality

Comcast’s announcement on buying a big chunk of the software assets of Icontrol (already used by Comcast in Xfinity’s home control) is another possible sign that the IoT space is getting sensible.  In fact, it’s two signs in one deal, and these could reinforce the trend toward a logical model of IoT that I blogged on earlier.  If that happens, we could finally see strong IoT progress.

We have to start with some facts.  There are over 600 million prospects for home and office security and control technology globally.  Today, those who have monitoring pay about $300 per year, which would mean that the total addressable market (TAM) is almost $200 billion.  It’s harder to say what the other segments of the prospective IoT market size up to be, but there’s not much doubt that security and control represents the largest single segment.  Further, it’s also the one that requires the least push to get going—because it already is.  That’s the problem with it, though.  “News” means “novelty” and the old market can’t make news.

Sad, given that truth doesn’t have much of a shot in the space either.  In my prior IoT blogs I’ve noted that “Internet of Things” has been taken too literally to be logical.  To put a bunch of sensors and controllers directly on the Internet begs all kinds of business model, security, and public policy questions.  A smarter model for the broad, public, IoT is the analytics-and-big-data approach, now being taken by about a half-dozen players of various sizes.

OK, but let’s circle back here.  The database model is great as an architecture for global IoT, but it doesn’t directly address two things.  First, how do you economically connect the literally millions of security and control devices already out there?  Second, how does the home security and control market, a giant piece of populist IoT, evolve?  The Comcast/Icontrol deal might offer insight to both.

Icontrol is one of a number of companies that offer sensor/control solutions based not on IP or the Internet but on lightweight protocols designed for short-range in-home (or business) connections between sensor/controller devices.  There are a number of protocols in this space, ranging from the venerable X.10 to Insteon’s proprietary protocol and to Zigbee and Z Wave, both of which are supported by Icontrol.  The first two of these are typically used over powerline wiring in a home or business and the latter two are wireless.

All these protocols support meshed direct connections, and all can also be used in a star configuration with a central controller, or a tree of controllers and devices for larger spaces.  The controller configurations have become more popular over time because they allow for programming of events based on time and sensor conditions.  They also provide, at least in most cases, for a link between the control network and the Internet, which lets people control things from their phones or tablets.  The sensors and controllers are not directly on the Internet, and if the gateway feature isn’t enabled there’s no Internet connection at all.  Many people use the networks this way for security reasons.

Comcast and a host of other service providers and specialty or security companies; Comcast is the largest of Icontrol’s service customers.  These providers offer a security-and-control-as-a-service model, a kind of PaaS, and the value proposition is simple—there are tens of millions of security systems out there that have minimal features, and probably as many as ten million with home control features that are obsolete.  Installing the complicated high-feature systems requires considerable skill and often even some programming, and finding devices that work with what you have is also complicated.  Companies like Comcast believe that people will pay for a service-centric ecosystem, and that seems to be true.  Remember, most good home alarms are connected to central monitoring and a monthly or annual fee is collected.  Often it’s more than a residential phone system’s fee, or even low-speed broadband.

These services are a kind of back door to IoT in the strict sense, if they’re IoT at all.  As I said, most of them don’t provide Internet access to the control networks at all, and where the access is provided it’s simple to add security/authentication the same way you’d do with an email account or secure website.  The sensors and controllers don’t have to bear the cost of direct Internet connectivity and security either.

These are the two signs of progress I mentioned.  First, Icontrol/Comcast represents a business model to evolve that which is already out there, the home/office control and security model that has worldwide well over a hundred million users already.  Second, it demonstrates a network technology that’s inherently secure and Internet-independent unless you take explicit steps to open it up.  While some of the control protocols (like X.10) could in theory be used from outside the home to trigger something within, others have explicit sensor/controller registration and the central hub would ignore a foreign device.

A service based on a platform like Icontrol could still perform many IoT-ish functions, though.  That’s particularly true if you can incorporate local WiFi devices, and even allow roaming phones and tablets to do things.  For example, a phone or tablet might register with the local hub and then be able to report its location, which could be helpful for parental control.  Similarly, a phone or tablet could be privileged based on its identified user to extract data from the control network or control devices.  And that’s without presuming that any newer sensors/controllers would be added to the mix.  Since Comcast’s rumored reasons for buying the Icontrol software assets include the desire to customize the stuff to incorporate new devices, Comcast might be thinking of expanding the scope of the control network beyond the ordinary gadgets now available and the Comcast-created custom devices and third-party elements it already supports.

Most of the home-network stuff is based on the simpler and shorter-range Zigbee technology, and this is what Comcast acquired.  Z Wave is usually professionally installed, and usually by companies that do home alarm and control systems.  However, it has about three times the range, and it might be that something like Z Wave would be a better option if the Icontrol model were to be applied to things like retail stores and even streets/intersections, things that people often think of when they think “IoT”.

The key point, though, is that the Icontrol model aims explicitly at supporting current sensor/controller technology as an upgrade.  You can add features (and devices) to many current systems or upgrade to something that is able to support that sort of incremental functional growth.  And because the security/control space is typically associated with a recurring payment for a service external to the home/office even today, it lends itself to an implementation based on cloud-hosted technology.

Once you pull security/control even a little bit into the cloud, you can start looking at those expanded IoT-like services.  You can implement the service using database and analytics technology, complex event processing, and all the other good stuff that IoT really has to be based on if it’s to succeed.  In short, Comcast could be on to something.

We already have cloud IoT, of course, as I said in my prior blog, but that was largely in the form of cloud-hosted services to augment IoT, not a cloud form of an IoT application.  Comcast is betting now on just that, and it might mean that other players who now offer only IoT features in the cloud (like Amazon) will start thinking platforms.  Amazon, recall, has a home controller already and hopes to expand its capabilities.  You can, in fact, link it to many popular home-control technologies already.  It would take only a little systemization for this to become a full IoT offering.

This is, to be sure, “IoT lite” in the eyes of those who think the whole world of sensors and controllers are going on the Internet, but that’s never been a practical option.  What we may be developing now is evolutionary not revolutionary, interesting not exciting, but real and not fiction.