Big Data Gets Bigger as Internet of Things Awakens

Business adoption of Internet of Things solutions will be fast — in fact, as I wrote yesterday, it’s already here for some industries. That’s why CIOs and other IT leaders need to gear up for supporting the unique data issues related to this trend.

Let’s look at what makes the Internet of Things data a bit different from other IT data resources.

The Problem: Mega Big Data. One of the main differences will be in the amount of data you’ll need to sort, improve, integrate, analyze and manage. You’ve heard of Big Data? All these devices, constantly chattering updates about moisture, light, movement and whatnot, will create crazy amounts of Big Data.

IT Requirement: A (possibly real-time stream) data analytics platform that can handle Big Data and a scalable infrastructure to support it.

Duke Chung, co-founder and chief marketing officer of customer service support technology company Parature, says a data analytics platform will be a must-have. Chung’s Harvard Business Review blog post actually focuses on how the Internet of Things will affect customer service, but much of it is also applicable to or will require IT.

“The fact that millions of devices will soon be Wi-Fi enabled will cause a flood of user data for companies to sift through,” he writes.  “Businesses can use this data to understand where issues are happening on their products, how frequently, and best resolutions — but only if they have the means to analyze it.”

Already, we’re seeing vendors positioning predictive analytics solutions for the Internet of Things, including IBM, Pentaho/ThinkBig and, of course, industrial players such as GE,  which already has a solution that will handle the unique challenges of Internet of Things Big Data.

“This type of predictive analytics solutions will be the norm, and companies will need to incorporate tools that will inform and improve customer service engagement on all of their devices,” Chung writes.

Before you analyze the data, you will need some way to collect, integrate, aggregate, model and distribute it. The most important criteria here will be what IBM Big Data Evangelist James Kobielus calls “operational economies of scale.” One way to achieve that, of course, will be via the cloud.

“From a planning perspective, you should start with the assumption that all IoT analytic functions benefit from the scale economies of centralization, unless one or both of the following criteria—functional scope and real-time speed—make distributed deployment more appropriate,” Kobielus writes.

He also discusses the issue of real-time interaction speed with the Internet of Things.

Next, I’ll look at the other major data-related challenge with the Internet of Things: APIs.