Really Thoughtful Data: Is ‘Cognitive Computing’ the Next Big Thing in Analytics?

Really Thoughtful Data: Is ‘Cognitive Computing’ the Next Big Thing in Analytics?

Along with all the excitement over artificial intelligence and machine learning, there’s an acknowledgement that none of this is possible without one essential raw material: data, and lots of it. As data gets rolled into these systems, and these systems increasingly gain the ability to intelligently act on this data without human intervention, we witness the dawn of “cognitive computing.”

Cognitive computing is intended to mimic human thought processes. According to the definition supplied by the Cognitive Computing Consortium, cognitive computing redefines “the nature of the relationship between people and their increasingly pervasive digital environment. They may play the role of assistant or coach for the user, and they may act virtually autonomously in many problem-solving situations. The boundaries of the processes and domains these systems will affect are still elastic and emergent. Their output may be prescriptive, suggestive, instructive, or simply entertaining.”

A cognitive computing system “offers a synthesis not just of information sources but of influences, contexts, and insights. To do this, systems often need to weigh conflicting evidence and suggest an answer that is ‘best’ rather than ‘right.’”

So it all sounds good in the lab. But is it ready for the business world? A recent report out of the IBM Research Institute suggest that cognitive computing is gaining ground across many business applications. In the survey of 500 executives, most expect to double their use of cognitive computing in the next three years, especially for operational opportunities.

Current areas in which cognitive computing is being applied include performance of assets, facilities and energy, employed at 12% of sites, with 16% expecting to adopt cognitive computing in the next three years. Connected transportation visibility is the next most employed areas at 10%, with another 11% anticipating adoption.

The areas likely to see the most growth within the next three years includes product quality monitoring and predicting failures, with 21% intending to adopt (currently, eight percent already use this kind of solution). Another 20% plan for manufacturing plant optimization, with nine percent currently using. Nineteen percent intend to adopt in the next three years for inventory and network optimization, also up from the current nine percent.

What’s notable is all these applications are in-depth, operational uses of data, versus the more visible, glitzier applications such as marketing, customer relationship management, healthcare, or new product development.  The data that may have been flowing through organizations for years, essentially unnoticed, is suddenly part of a new wave of intelligence guiding organizations. As the report’s authors suggest, cognitive computing, as applied to operations, is the backbone of any efforts to move to digital enterprise.

Cognitive computing is readily available to organizations, and “the whole process doesn’t need to be exotic and it doesn’t have to cost very much,” according to Tom Davenport, professor at Babson College and MIT, and author of numerous books on analytics. He calls cognitive computing “the next logical step past analytics.”


The Next Logical Step Past Analytics Is Cognitive Computing


Cognitive computing can be applied to problems that so far have been beyond the reach of traditional analytics, Davenport says. “In the world of big data, for example, the data from sensors, social media, and online applications often flow and accumulate much faster than humans could possibly analyze or act on it. Without machine learning to create the models for such data, it couldn’t be analyzed at all.”

Here are some ways to get started with cognitive computing:

Look to cloud and online resources for help and insights. Davenport recommends machine learning algorithms that are available through leading cloud providers. Such tools “facilitate exploration of deep-learning neural network applications like speech and image recognition,” Davenport says. In addition, skills issues can be addressed through numerous online courses now available through universities and learning sites.

Identify areas of business pain. “Perhaps it’s a knowledge bottleneck – a situation that might benefit from the application of knowledge that previously has been inaccessible,” Davenport suggests. “Or perhaps it’s a situation with so much data that humans couldn’t possibly handle it. Then start your experimentation with cognitive technologies on that problem.”

Look at data in new ways. Integrate the data from IoT-connected devices and cloud-based transactional apps to engage new audiences while creating new channels of product/service distribution,” the IBM Institute report authors recommend. Look to “nontraditional sources of data (such as weather or social trends) into your transparent, operational control tower. Apply supply chain analytics and network optimization in real time, enabling automated decisions and mitigating risk.”

Engage in sensing, integration and advanced analytics through connected devices. The IBM Research Institute report urges that cognitive computing efforts be directed toward improving “return on assets across all asset classes. Predict failures and improve part quality, while foreseeing product availability.”