IBM Introduces Innovative Serve Quality Feature at U.S. Open
As world number three and the foremost American tennis player, Jessica Pegula, readied herself to serve against Aryna Sabalenka from Belarus during the U.S.
Open semifinals on a recent Thursday evening, she found herself under the gaze not only of spectators, broadcasters, and journalists, but also of an advanced technological system.
Commencing with this year’s tournament, IBM unveiled a groundbreaking feature within the U.S. Open app, in collaboration with the United States Tennis Association (USTA).
This function, known as “serve quality,” harnesses sophisticated camera technology to track over 20 anatomical points on Pegula, as well as all other singles athletes in the competition, encompassing both male and female participants.
The serve quality score, rated on a scale of 100, incorporates critical movements involving the knees, wrists, and elbows, among various other parameters. The intricate data is subsequently analyzed by IBM’s WatsonX, producing the score accessible within the app.
While limb and skeletal tracking have been employed in professional sports such as soccer, this feature marks a historic first for Grand Slam tennis tournaments, signifying a new era enhanced by artificial intelligence.
This initiative forms part of a broader array of enhancements for the 2026 event, which will also include an AI chat function and the ability to highlight “key moments” throughout matches.
According to IBM, the innovative feature has undergone private testing over the past few years, with an anticipated analysis of 1.2 billion joints by the close of this year’s tournament.
Furthermore, the organization predicts that the app will generate around 7 million serve quality insights.
The mechanism behind the serve quality score is straightforward: users can access match summaries via the tournament app, click on “match recap,” and inquire through IBM’s “match chat feature” about how serve quality influenced the match.
The AI-generated insight remarked, “Jessica Pegula surpassed Aryna Sabalenka in serve quality with a score of 72.32% against Sabalenka’s 71.95%.”
Notably, despite Pegula’s defeat, her serve quality score was superior, thanks to IBM’s computation that includes data from ball, racket, and player movement tracking.
“Pegula displayed remarkable accuracy, placing 75.95% of her 79 total serves within the box, maintaining an average distance of 1.555 feet from the ideal serve zone,” the response elaborated.
Serve quality scores are publicly available for all singles matches post-completion, enhancing both viewer engagement and data insight.
IBM has released a video elucidating how the serve quality feature operates:
“It begins with the camera,” remarked Tyler Sidell, IBM’s Technology Program Director for Sports & Entertainment Partnerships, during an interview.
He detailed that twelve cameras are strategically positioned around Arthur Ashe Stadium, functioning in tandem with Hawk-Eye technology to facilitate automatic line-calling.
The limb tracking initiative at tennis tournaments was catalyzed by the introduction of Hawk-Eye’s “SkeleTRACK” product at the 2024 Laver Cup, although this application did not permit fan interaction concerning serve scores.
“The Hawk-Eye cameras are now equipped to capture limb movements; we actively analyze 21 limbs and joints from every single player,” Sidell elaborated, noting the significant innovative acceleration this technology has provided.
“We initiated model training with 2025 data to refine the algorithm for this feature. 2026 marks the inaugural year for its deployment to fans,” he continued.
“The volume of data generated from a tennis match is staggering,” expressed Brian Ryerson, USTA’s Senior Director for Digital Strategy.
“Skeletal data is a novel framework for us at the U.S. Open. Although we have possessed it in recent years, it presents a complex and rich dataset.”
The collaborative effort aimed to impart a “unique perspective” to fans in an accessible format. IBM and USTA placed primary focus on the serve due to its critical importance within tennis matches.
“The serve is the linchpin stroke of a tennis match,” Sidell stated. “We meticulously trained on existing data, incorporating academic resources to strengthen our system.”
Bryan Ryerson articulated that the threshold for success regarding the serve quality score is determined by two chief criteria: “Firstly, we must ensure it is digestible for fans, given its technical nature. The aspiration was to resonate effectively, and we believe we have achieved that.”
“Additionally,” he added, “we aim to enhance our narrative capabilities day-to-day, striving for accuracy at every turn.”
This development represents merely the genesis of limb-tracking capabilities in prestigious tennis tournaments.
Executives from both IBM and USTA have indicated that additional shots, including forehands and backhands, could soon be tracked and shared with app users in subsequent years.
“The potential is vast,” Sidell noted. “Future iterations may include racket insights. We are launching serve quality now, but next year, with comparative analysis alongside serve quality, we could draw pivotal correlations.”
Ryerson concurred, stating, “As we amass more skeletal data, a wealth of insights previously inaccessible will come forth.”
IBM envisions that more tennis tournaments and even other sports could integrate skeletal tracking data available to app users.
“This is merely IBM’s introductory venture into this domain, yet there’s certainly scope to extend it to other sports or even our other Grand Slams,” Sidell suggested. “You may consequently witness similar features at Wimbledon.”
The horizon for limb-tracking technology may extend beyond tennis, with possibilities for application in prestigious golf tournaments.
For instance, IBM’s enduring collaboration with The Masters offers avenues for capturing limb movements.
“If hardware can monitor golfers’ limbs and joints, there is ample opportunity to adopt similar methodologies across diverse sports,” he affirmed.
Source link: Fortune.com.
