Pittsburgh Technical Council

Decisions, Decisions: Resilient Cognitive Solutions Clears the Clutter of Data Overload

Decisions, Decisions: Resilient Cognitive Solutions Clears the Clutter of Data Overload

Article Published: October 21, 2014

“Select a numbered tile from a shared pile of nine numbered pieces, alternating choices with your opponent. Be the first to gather three pieces that sum to 15 and you win.” 

This simple game can be devilishly difficult, or intuitive and effortless, depending on how it is presented to you. The player presented with just the data struggles to keep up. The player with the proper decision support, in this case a ‘Magic Square’ of numbers on a Tic-Tac-Toe grid is quick to win.

The real question, if you were to ask Bill Elm, President of South Side-based Resilient Cognitive Solutions (RCS), really comes down to this: are your decisions about math and combinatorics done in real time, or is deciding where to put your next X or O done against a landscape of established winning paths?

You need data, but the best decisions spring from a foundation of knowledge, not data alone. RCS has flourished in serving its clients—found in the worlds of defense, private industry, and additional commercial applications—because it clears away the clutter of data overload to present a wealth of rich, reliable, robust knowledge so critical to good decision-making.

Elm uses the Game of 15 as a standard demonstration to prove the point. Using the tic-tac-toe Magic Square, his opponents often claim he has an unfair advantage over them armed only with raw data. Elm, on the other hand, knows the odds of success or failure change dramatically when effective Decision Support changes that data into a knowledge context. It’s a powerful point made in the simplest of scenarios.

“We are drowning in data overload,” Elm said. “Massive investments by companies into more data have not improved decision-making. Executives say the decisions their organizations make are no better, and are taking longer, even with all of this added data. We have only made the forest more dense. This is not about the needle in a haystack: we need to see the larger crop circle in the entire hayfield. Everyone making critical business decisions under pressure can use this.”

Elm pointed to trendy business software systems that present data in visually stunning ways, but that fail to provide context or analysis of that data.

“These things may look pretty, but it’s hard to see trends or make decisions based on them,” he noted. “It becomes a frame-of-reference problem. Without context, you can’t make a good decision when data is always repackaged to be visually interesting. It might win graphic design awards, but all you’re left with is: what do I do with this data?”

Instead, RCS serves C-level customers with mission-critical jobs who are unhappy and dissatisfied after constructing a big data build.

“I see the need for our services everywhere,” Elm admitted. “But I’ve got to wait for the horse to get thirsty, for the leader to realize he or she has pain. Once that happens, we do a design exquisitely tailored to solve the problem.”

The first step in the RCS process entails figuring out what the client’s fundamental decisions are, the essential cognitive work their brain is struggling with, then designing a solution that fits the decision-makers’ brains. RCS uses the same underlying science as the Artificial Intelligence community, designing to support how the human brain functions, instead of building a box to replace the brain. It may sound complicated and a bit esoteric at first, but Elm can explain.

“The brain is the only truly adaptive decision-making engine in the world,” he said. “We engage the brain with computer-generated information presented in ways that flow naturally along with how the brain processes it and reaches conclusions based on it.

“A lot of what we do, engineers say is too ‘researchy,’ while researchers say it’s too ‘engineer-driven.’ All that proves to us is that we are in exactly the sweet spot we want to be.”

The key to RCS’ success comes in finding what Elm calls “abstractions”—the big picture, how systems work, the crop circle instead of the needle in the haystack. A lot more than just the facts of a situation. Like a grand master in chess, an RCS system is based on those abstractions, along with the knowledge to make the best decision. At the same time, an RCS system presents this information in a format that is visually easy to understand and act up, customized for each application.

Elm offered one example from the military. During training exercises, a missile defense system always chose the wrong incoming missile to target. Officials believed a faulty algorithm caused the problem. In the end, servicemen more often turned the system off and relied on their own experience, rather than the incomplete data-reliant reasoning of the software. RCS identified the real issue as the system inaccurately equating the easiest missile to defend against as the most threatening one and misrepresenting it to the user. With that discovery as the platform, a new decision-centered RCS system replaced the original and performance improved.

The 18 employees at RCS represent a multi-disciplinary team of psychologists, engineers, and designers—an eclectic mix that helps the firm bridge the varied disciplines of brain understanding, software engineering, and appealing visual presentation.

Balanced with Elm’s two degrees in electrical engineering from Carnegie Mellon University, his service with the U.S. Army Signal Corps and Military Intelligence, and his experience with Westinghouse—particularly in the post-Three Mile Island nuclear accident period—RCS continues to blaze a new trail with its clients. Engagements can show breakthrough concepts in as little as 60 days, and longer, as would be expected, for very complex, comprehensive reworks, but the results are unmistakable.

“We are the only shop in the world that does what we do from a blank page to a working system,” Elm said. “We’re inventing things no one’s ever seen before.”

Sign In or Join Up