Cisco Blogs
Share

Why Data Integration Provides the Critical Foundation for Capturing Actionable Insights


December 16, 2014 - 1 Comment

We all know that Big Data is getting bigger. And, the gap between the amount of data with hidden value and the amount of value that is actually extracted keeps widening. In fact, according to IDC, less than 1 percent of the world’s data is currently being analyzed. What good is data if it doesn’t produce actionable insights that generate improved outcomes?

A large portion of the world’s data is produced by the billions of connected objects that make up the Internet of Things (IoT), a critical enabler of the Internet of Everything. In Cisco Consulting Services, we recently conducted a blind global survey to learn more about how organizations are harnessing IoT to transform their businesses — and what they can do to drive more value.

Perhaps the most interesting finding of the study was that IT and operational technology (OT) leaders now perceive IoT as being about much more than just “things.” When Cisco asked them which area (people, process, data, or things) they needed to improve most to make effective use of IoT solutions, the largest number (40 percent) indicated “data,” while “process” (27 percent) ranked second. “People” placed third (20 percent) and “things” finished last (13 percent).

These leaders understand that connecting “things” is but a means to an end. The primary value that IoT creates is a direct result of the data that can be captured from connected things — and the resulting insights that drive business and operational transformation.

To capitalize on the wide range of data IoT generates, organizations must overcome three key challenges identified by our survey respondents:

1. Integrating data from multiple sources
2. Automating the collection of data
3. Analyzing data to effectively identify business-relevant, actionable insights

Only by addressing all three can organizations turn raw data into true value.

Let’s focus on the first of these essential ingredients: data integration. In a future blog, I will address automation (including edge computing) and analytics.

Data Integration and Virtualization
Before data can be processed and analyzed, it needs to be captured and integrated. Data integration is essential across business and IT operations. For example, transportation planning becomes more efficient if internal packing and delivery data is merged with weather and traffic data.

The unparalleled distribution and variety of devices and data today, however, make data integration a bigger hurdle than ever before. Organizations must consider multiple factors, including the physical installation of devices, the best communication standards, how to handle many different types of data created by a specific device (e.g., video, geolocation data), and how to effectively integrate IoT data with data from other sources, such as third-party data providers from the cloud, as well as internal data stores.

Furthermore, because copying all data to one centralized node for integration is no longer the best possible option for a variety of reasons — data volume, velocity, variety, and possible regulatory issues — organizations are now starting to rely upon data virtualization to integrate widely dispersed data. Data virtualization makes a heterogeneous set of data sources look like one logical database to users and applications. These data sources don’t have to be stored locally — they can be anywhere. This is particularly valuable for an IoT application that relies on data from many distributed sources, such as embedded sensors, video cameras, and third-party data sources.

As Rick van der Lans explains in “The Network Is the Database: Integrating Widely Dispersed Big Data with Data Virtualization,” data virtualization provides another powerful advantage: “Data virtualization technology is designed and optimized to integrate data live. There is no need to physically store all the integrated data centrally. It’s only when data from several different sources is requested by users that it’s integrated, but not before that. In other words, data virtualization supports integration on demand. The time is over that we can push the data to a centralized location for integration purposes — we have to push the integration to the data.”

Data integration sets the stage for the next two steps on the road to capturing actionable data insights: automation (including edge computing) and analytics. Stay tuned for an update on those two.

 



In an effort to keep conversations fresh, Cisco Blogs closes comments after 60 days. Please visit the Cisco Blogs hub page for the latest content.

1 Comments

  1. Concur with your three steps but might I suggest a fourth and an alternative approach. IoT is about "smart connected" things that create actionable decision making. If so, then higher level capabilities aren't sensing, monitoring, controlling - but rather - diagnostics, prognostics, forecasting, optimization and autonomics. And increasing the velocity of the OODA loop / SIDR loop. This means some knowledge of where one is headed with decision making and the algorithms to achieve these higher order capabilities is needed - and - the data structure for said algorithms. This makes data architecture and standards a key enabler and or hurdle to successful outcomes for any IoT initiative. Unless that imitative is turning your smart phone into a remote control for your TV, HVAC or home security system. There's a value to autonomic "Things" in IoT http://goo.gl/vCb2KW