A Cloud Falls Over The U.S. Open

September 9, 2013 Jenny Thomas

Tennis fans might think of clouds as being a bad omen. Rain delays can put a quick stop to the action on the court. But clouds mean something different to those of us in computing circles, and fans of this year’s U.S. Open Championships might be surprised to learn that clouds actually had a big hand in creating a better tennis experience.

The USTA has been partnering with IBM for more than 20 years, and in that time the technology has changed a lot. Today’s tech savvy fan, whether part of the 700,000 crowded into the USTA National Tennis Center in Flushing Meadows, New York or one of the millions watching at home, expects to be able to examine each serve, volley, and point throughout every match–and in real time.

Just like the technology has changed through the years, so have the demands of the fans, players, broadcasters, and media who watch and cover the U.S. Open. Itâ€™s not just about watching the action, it’s about the stats, facts, videos, live scoring, and historical and real-time analysis of every game, set, and match. In computing terms, it’s about big data.

To make that happen, IBM employed a series of Power Systems, including a Power 750 server, which is the same system that provides the foundation for IBM’s Watson question-answer system. Running both AIX and Linux, the upgraded IBM infrastructure that supported USOpen.org this year was built to bring analytics, cloud computing, mobile devices, and social media together to deliver real-time insights into match data. And it was all delivered via cloud.

For IT Jungle readers, it is of interest to note that the IBM i was once part of the computing team for the U.S. Open, but in 2008, IBM pulled the i5/OS and its System i iron in favor of six Power6-based Power 550 midrange servers running Linux. AIX has been able to wiggle back onto the system, but IBM i has not.

During recent U.S. Open tournaments, IBM manually monitored and adjusted computing workloads for the USTA. For 2013, IBM upgraded the infrastructure that supports USOpen.org with a combination of predictive analytics and cloud computing technologies to predict, monitor, and continuously adjust the capacity and computing resources. This versatile cloud computing environment, powered by Power-based servers and cloud technologies located in three geographically dispersed data centers that are virtualized and treated as a pool of computing, provides continuous availability and scalability. IBM says the analytics and cloud integration helps the USTA more efficiently accommodate and respond to unpredictable spikes in demand for content and data during the tournament, and provides a more effective means to manage IT costs over the course of the year when the USTA does not require as much infrastructure support.

Managing the big data action at the U.S. Open draws an easy parallel to real world applications for IBM exec Rick Singer, who is vice president of sports sponsorship marketing for IBM.

“Sports fans demand ubiquitous and uninterrupted real-time access to scores, stats and insights from big data, much like clients and C-suite execs in most industries,” said Singer. “The explosion of mobile and social technologies has radically changed the ways companies and people access data, giving them many choices over how, when, and where they tap into information. From watching a game on TV, catching video highlights on a smartphone, to monitoring online sales via an iPad, uninhibited access to big data has become an essential element of business and life.”

In 2012, the U.S. Open cloud handled over 325 million page views, with 117 million generated from mobile devices. To keep those online fans happy and engaged, IBM stepped up the U.S. Open mobile fan experience to include a refreshed, iPad app that offers integrated social features and instant access to live video and in-depth match analysis and statistics. IBM also improved apps it has developed for the USTA for the iPhone and Android platforms, with a scrolling timeline that keeps fans informed of trending social topics and match photos.

A redesigned SlamTracker was introduced at this year’s Open. The SlamTracker, which debuted in 2007, is built on IBM predictive analytics software and analyzes more than 41 million data points from eight years of Grand Slam tennis data. The improved SlamTracker offers a new embedded social sentiment feed that allow fans to follow and contrast how the volume and tone of fan sentiment on Twitter correlates and shifts as player performance changes on the court.

On site at the Tennis Center, IBM brought back the IBM Game Changer Interactive Wall for a second year. The Game Changer wall extends USOpen.org and mobile app features and provides greater insight into the U.S. Open action using analytics. Tennis fans of all ages can interact with the wall and access live match scores, in-depth analysis, and data visualizations.

Looking into the future, the private cloud IBM set up for the USTA can be shared by other organizations with similar demand profiles, which for IBM clients would include the Australian Open, the French Open, Wimbledon, the U.S. Open golf tournament, the Masters, the Tony Awards, and, of course, ibm.com. Sporting events make a nice backdrop for the real story IBM wants to tell: A similar solution could also be implemented by larger companies interested in sharing IT resources internally between divisions, departments, or projects.