Channel's IoT Opportunities Aren't Hard to SeeBy Howard M. Cohen | Print
COMMENTARY: Where the channel opportunities are in the Internet of things is an increasingly popular question, but the answer is really right in front of us.
Everyone we ask about the channel opportunities coming from the Internet of things has a different answer. Some say they haven't shown up yet. Others tell us definitively that it's all in the data and the analytics. Still others feel it's a big infrastructure play while some folks are focusing on carrier opportunities.
Yes. They're all right.
However, to truly focus in on the opportunity, one only has to look at the numbers—in this case, the IP address numbers.
IPv4, the original Internet addressing system, offered up 4.3 billion addresses. Even when it was introduced, many of the engineers involved felt it would be insufficient, so they started work on developing the next version. Expedience prevailed and IPv4 has served the Internet well for almost 50 years.
The conversion to IPv6, the successor to IPv4, was supposed to be completed by 2011. Five years later, we're only a fraction of the way there. Right now, Cisco is predicting that only a third of users and 41 percent of all fixed and mobile devices will be IPv6-capable by 2019.
The clue in this scenario is the sheer number of addresses the Internet Engineering Task Force and others felt we would need in the new addressing scheme, 340 undecillion.
It's difficult to conceptualize a number like 340 undecillion. Here's a visual aid comparing the number of IPv4 addresses with the number of IPv6 addresses:
Implications for the Channel
Here's where this all starts to be real fun for channel executives and technologists. What could we possibly need so many addresses for?
We can start with people. The number of people on the planet continues to increase, and the number of devices each one carries or wears is growing quickly from one to five. Each of those devices on each of those people requires several addresses.
Then, there are the computer devices that don't have any individual people associated with them. It used to be that every server needed an IP address to support its one instance of an operating system. Now, most servers are supporting dozens or hundreds of instances, and each of those instances requires a dozen or more IP addresses.
Finally, we get around to the Internet of things. Start looking at a world where every motor vehicle requires at least one IP address. Also connected to the Internet would be most every home appliance and entertainment system plus speakers and displays, every industrial manufacturing machine, every light switch, every thermostat, every door lock, every smoke detector, every industrial heating and cooling system and all their ducts and louvers, and every light post along every road.
Next, there are sensors. Many manufacturers have announced their intention to blanket the world with sensors that will measure everything and report back so that everything can be accomplished far more effectively everywhere.
Okay, so we need 340 undecillion addresses to identify all these people, computers, things, sensors and more. What else do we need?
All these things need to be invented and built. Then, they need to be installed and integrated, and then secured, monitored and managed. That will require management and security software at a level of sophistication we haven't reached yet.