Other tech

A former pro tennis player has developed an artificial intelligence app to make everyone’s game better

- October 31, 2020 2 MIN READ
A tennis app utilising artificial intelligence is set to revolutionise player analysis and coaching interaction.

Developed former professional tennis player Dr David To in Adelaide, South Australia, the subscription-based Tennis AI app launched last week and is available globally.

The AUD$10 a month subscription allows players to upload videos of their games and training sessions to a site where the footage is automatically processed with AI within 30 minutes, depending on available upload speeds, and then share it with their coach.

Dr To said although this type of analysis has been available at the elite level for years, it has not been available for amateurs because manual labelling of footage made video analytics too slow and the need for high-tech cameras made it too expensive.

To solve this problem, the Tennis AI program processes regular smartphone footage with artificial intelligence analytics, machine learning and computer vision to provide a stroke-by-stroke breakdown of a player’s game almost immediately.

“It helps the (amateur) see themselves and filter to specific parts of their game,” Dr To said.

“This allows the player to improve their game quicker and more efficiently. Most people are visual based learners, so the ability to filter down to a specific stroke type and direction is a major advantage.”

To use the program, a smartphone or GoPro is placed directly behind the player, attached to the fence, on a tripod, or simply with a parent filming behind the court.

Dr To, who has a PhD in Chemical Engineering and an MBA from the University College London, said coaches can use the program for free.

“From the coaches point of view, they can expand their business globally,” Dr To says

“They can coach players not only at their club, but they can coach others around the world, (helping) to market their business online.”

Data available to coaches and players include rally length, a heat map of where the player is positioned on the court, as well as filtering serve direction and where the player is making their errors.

The program was tested and delivered in real-time at the inaugural ATP Cup in January this year.

The app was developed in conjunction with the Innovation and Collaboration Centre at the University of South Australia by Dr To, Dr Mark McDonnell and Dr Guy Gallasch, with the help of Senior Data Engineer James McNeill.

The team formed startup Athletes AI in July 2017 to scale their idea across multiple sports globally, and Dr To says the vision of the company is to put video analytics in the hands of all athletes.

To do so, Dr To is looking to raise AUD$1.2 million in capital for market and product development for the company that has a pre-money valuation of $5 million.

He expects to have around 23,000 customers by 2022.

The capital would go towards linking with key influencers in the sports industry as well as further refining the product to include cricket and baseball.

So far they have formed partnerships with Cricket Australia and the Adelaide Football Club.