1 Comment

Summary:

Much of the potential of IOT will exist only in theory if there is no infrastructure to support the coming stampede of data it will generate. Here are a few ideas of what needs to happen next.

(L to R): Peter Bakas, CEO and Co-Founder, DataScale;
Jai Menon — VP and Chief Research Officer, Dell Research, Dell; Tamara Budec, VP, Portfolio Operations, Digital Realty;
Cory von Wallenstein, Chief Technologist, Dyn;
Dave Ohara, Founder, GreenM3 and Analyst, Gigaom Research

The “internet of things” refers to the idea of turning everyday objects around us into responsive data devices. It’s one of the hottest ideas in tech right now — even if we’re totally lacking the resources we need to support it.

Speaking at Gigaom’s Structure event in San Francisco, four experts well-versed in the practical problems of making this happen offered their thoughts about the infrastructure that will be needed to make use of it.

For Tamara Budec, an executive with data services provider Digital Realty, the building of the physical layer for the internet of things has barely begun, and moving forward will require constructing new clusters of data centers and also and freeing up more wireless spectrum.

There are also questions of how to build in global security and privacy controls, according to Jai Menon, VP and chief research officer at Dell Research. The challenge is especially daunting, Menon said, given the sheer scale of data that can now be collected. He pointed out that capturing all the information points generated by New York City’s taxi fleet would produce an unmanageable 26 exabytes a year.

Menon also noted that building an infrastructure for the internet of things will involve constructing not only physical plant, but next-generation analytic tools that are not just predictive, but prescriptive too — for instance, sensors that would not just predict a storm tomorrow, but also be capable of sending emergency responders to meet it.

Others shared this notion that building this infrastructure will require not just investments, but a better philosophy about why we are doing so in the first place.

“We talk about big data, but what we really want is useful data,” said Peter Bakas, CEO of DataScale, adding that building to scale is all well and good, but it’s hard to do so if no-one knows what exactly they’re scaling for.

To go from hype about the internet of things to actually building the platforms to harness it may also require more than just pointing out what will happen if we don’t build it.

Cory von Wallenstein, Chief Technologist at Dyn, pointed out that out everyone in the tech world knew that we’re running out of IPv4 addresses, but that fact hasn’t prompted people to rush out and acquire IPv6 ones. Instead, people are just stretching the existing internet architecture — something that won’t be possible as potentially billions of new devices come online in coming years.

To encourage people and companies to build a new architecture that welcomes these devices will require explaining the opportunity and incredible value that they could bring. Right now, von Wallenstein says, we are mainly looking for gold in “data exhaust” derived from activities that consumers are already performing; but the real value will come when the right infrastructure is in place to get entirely new forms of data.

The panel was hosted by Dave Ohara, the founder of GreenM3 and an analyst with Gigaom Research, and a video embed of the session follows below.

Photo by Jakub Mosur

Structure 2014 ticker

  1. There will be so much evolution in this arena, I don’t think we’ll even recognize the ways in which ‘things’ are connected in a decade.

    Reply Share