Navigating Signal-to-Noise Ratio in the 5G Enterprise of the Future

While some companies are just beginning to wrap their heads around 5G and what it means for business, Bloomberg has been actively planning for the coming data deluge and using it as a driver to invest in artificial intelligence, notification systems and other customization tools.

If the buzz at the recent Mobile World Congress in Barcelona is to be believed, then digital Nirvana is on the horizon. 5G, the next fifth generation of mobile networks, is on its way and promises to turbocharge businesses with speeds of up to 1 gigabyte per second – 50 to 100 times faster than what 4G offers today.

But business and 5G may have a few loose connections. Says Phil Miller, Bloomberg’s head of mobile engineering: “5G will blur the line between performance and productivity in the office and on a mobile device.” While fifth-gen networks will sharply increase the amount of data that networks can handle, they will also up the amount of ‘noise’ or streams of information which may or may not be relevant.

There is a lot at stake — as our core business is all about turning billions of bytes of data into actionable information. When thinking about 5G, Phil says these questions must be asked: How can we improve the signal to noise ratio? How do we filter all this information and convey it to the customer? And how can we add intelligent value to all this information?

In preparation, engineering teams across Bloomberg have been developing software and machine learning tools to filter data through the Bloomberg terminal, one of the world’s largest private networks which serves up real-time news, data and financial information to financial professionals around the globe.

Far from being a simple ‘wheat from the chaff’ filtering exercise, the higher aim is to create a framework which  delivers customers personalized, actionable value while maintaining an intuitive user experience.

Artificial Intelligence to the rescue

Given the sheer volume of information, our tools and filters will have to intelligently learn from the user to understand what information is relevant, when. However, “Artificial intelligence, or AI, won’t take over from humans,” says Phil. “At least not anytime in the near future.”

Since humans will have written the rules governing such software, Phil says, most of the time people will trust AI filters to make decisions for them. However, there will be instances where AI will have to seek human input to make particularly complex or difficult decisions.

“In a world where decision making has been taken out of the 9-to-5 environment, the filter has to determine: Do I need to be interrupted?” says Phil.

As growing broadband connections allow more work to be done out of the office, software must also understand how to parse and cull down information so it can be digested quickly by humans on small screens.

These opportunities also present a serious UI challenge: How can a user interface be optimized for the many things a 5G-powered mobile device can show its owner? And, how can the data be filtered in a way so that it is useful at the right time for users?

Mainstream 5G adoption may still be years away, but for Bloomberg , the work to tame the data deluge of tomorrow’s networked society has already started.