Real-time data intelligence for Football

01 apr Real-time data intelligence for Football

The BallJames application

BallJames tracks the movement of all players, the ball, and referees in 3D. The system uses multiple high-resolution cameras and advanced deep learning and computer vision technologies to generate accurate on-pitch data. The system works autonomously, so there is no need for human interaction in the stadium or requiring sensors on the players or the ball.
Insight is a sophisticated platform that quantifies the quality and potential of all professional football players around the world, supporting football experts in expanding their scouting scope. The platform can suggest viable trade (transfer) targets and identifies player talent at a young age by calculating their potential using advanced machine learning algorithms. The goal is to increase the trade (transfer) success rate.
SciSports’ data intelligence services are used by clubs, agents, associations, players, and other football-related parties.
Clients include Royal Belgian Football Association, Olympique Lyonnais and Ajax, as well as other clubs from Europe’s top leagues and the MLS.

The BallJames system works by sending multiple live 4K video streams, recorded at a football match, to SciSport’s proprietary analytics engine (AE) hosted in Google Cloud Platform (GCP). Feedback is produced very quickly and incrementally through its deep learning algorithms. The AE in GCP will start to produce analytical results after 10 seconds of analyzing video streams allowing a coach to dynamically see the results of his tactical moves on the bench, or empowering fans with next generation statistics, or providing visuals displayed on the TV or web.

BallJames will have far reaching applications in the future extending beyond football, including providing advanced statistics for online gambling, as well as revolutionizing the way players are chosen and leveraged for professional sports teams.

Network challenges and requirements

The essence of the BallJames application requires the visibility, control and delivery of high-definition video streams from different cameras to an AE in GCP. The real-time nature of this application puts enormous demands on the network to deliver guaranteed performance without risking the security or integrity of the data.

Some of the network requirements that need to be satisfied for the BallJames application include:

• Performance – the network needs to provide sufficient guaranteed bandwidth, latency, and jitter to meet the prescribed service level agreement (SLA) from each video stream.
• Security – each video stream needs to be encrypted to secure the data from hackers and other threats.
• Reliability – the network has to be architected in a highly available manner so that the application can survive equipment and link failures.
• Visibility and control – SciSports must have full control and visibility of each video stream, in real-time, to measure performance, analyze for research purposes, and to be able to troubleshoot issues.
• Flexibility – as SciSports business expands in both locations and applications, the network must be flexible enough to adapt to different location requirements and cloud capabilities.

Simply put, the critical and time-sensitive requirements of the BallJames application demands the highest performance, reliability, and security, leaving no room for error.

SDNstudio and Nuage Networks

SDNstudio is an SD-WAN managed service provider that operates across 220 countries worldwide. As a key partner of Nuage Networks, SDNstudio leverages Nuage Networks SD-WAN 2.0 technology to deliver managed SD-WAN services. SDNstudio deployed their managed SD-WAN service to meet all of the stringent network requirements of the BallJames application