I have never been a sports person. Case in point, when I was a kid playing T ball, I was more interested in catching bugs in the outfield, than I was in catching fly balls. But as a grown-up software developer, I am paid to deal with bugs and I jumped at the chance learn about the advanced systems used at the U.S. Open.
For the other non-sports people out there the U.S. Open is a tennis tournament held annually since 1968. The folks over at IBM Research and IBM iX have put together some impressive real time analysis systems for the event.
There are actually two U.S. Open application diagrams I want to share with you. The first is the U.S. Open Tennis Championships app affectionately called the “Slamtracker”. The second is the “AI Highlights” application which drives the production of the tournament’s highlight videos for streaming and television.
“Slamtracker” — The US Open Tennis Championship app system architecture diagram.
The Slamtracker application aggregates the official scores and stats along with analytics from computer vision analysis of the players on the court. The system extracts over a dozen key measures and generates real-time statistical analysis.
The distance run, serve placement, serve speed, return speed & depth of return are just some of the values recorded.
These data points are run through contextual and historical analysis using IBM SPSS and Watson Machine Learning.
Fans can explore higher level match statistics with the new feature “Momentum” that lets you see which player is statistically on a hot streak towards winning the match..
Additionally, the Slamtracker application identifies situations fans will find interesting using Apache Spark. It identifies “pressure situations” by analyzing 12 years of match data (26 million data points).
Did this player historically choke or make a moving comeback when they were loosing a match and down 2 sets? The Slamtracker can share these insights with fans in real time, point by point and play by play.
US Open AI Highlights
The AI Highlights system uses Watson computer vision, audio and analysis to help the U.S. Open video team produce tournament highlight videos. Over the course of the tournament there are over 100 tennis matches across 17 courts and each match is recorded by multiple cameras. This creates thousands of hours of video footage. Normally that footage would have to be review and edited manually.
By analyzing video and other data, the developers have been able to cut the production time for individual highlight videos from nearly an hour to just a few minutes each. Extracted features such as the gestures a player makes after scoring a point or the roar of the crowd are all useful in identifying the exact moments a fan will find meaningful and interesting.
Meet the developers behind the AI Highlights and the
upcoming IBM Fantasy Football application.
If you happen to be in the New York City area on August 30th, 2018 please join us at LMHQ in welcoming Aaron and David from IBM iX. They are key members of the team behind the U.S. Open applications and Fantasy Football.
Description: Cutting edge and production ready Artificial Intelligence (AI) algorithms and services on the IBM Cloud are used throughout professional sports and entertainment events. Within this session, you will learn how to leverage multiple regression, deep learning, distribution functions and support vector machines within real-time applications. Two award winning systems, AI Highlights and the ESPN Fantasy Football, will be demonstrated. Scale and availability considerations will be discussed.
Please register for the event here:
(Please note you must be registered for this event to gain admission.)
IBM Code NYC
LMHQ — A Creative Event & Meeting Space
150 Broadway, 20th Floor · New York, NY
Thursday, August 30, 2018 — 6:30 PM to 9:00 PM
Further details about the technology powering the US Open can be found at https://www.ibm.com/usopen.
It is not all fun and games.
Call for Code : Natural disaster preparedness, relief and recovery.
In addition to the sports and business applications, IBM is also dedicating it’s technologies to a global tech for good initiative: Call for Code.
“For 2018, the Call for Code Global Challenge asks developers to create solutions that significantly improve preparedness for natural disasters and relief when they hit. This competition is the first of its kind at this scale, encouraging developers who want to pay their skills forward for a specific mission to alleviate human suffering.”