AI in the swing of things




ACE INVENTOR—Above, Kevin Ma, 15, tries out his robotic shot sensor with a ping pong paddle. Above right, his technology pairs with a tennis racket. At right, the teen takes a swing on the tennis court.

ACE INVENTOR—Above, Kevin Ma, 15, tries out his robotic shot sensor with a ping pong paddle. Above right, his technology pairs with a tennis racket. At right, the teen takes a swing on the tennis court.

Practice makes perfect, but a little computer assistance never hurt.

Kevin Ma is a 15-year-old tennis player from Oak Park who built a robot to help him perfect his game. Then he built a second one, to help him play ping pong better.

The teenager published an academic paper on his creation last fall and hopes to have a second published soon.

The robots are small sensors that can be affixed to the handles of the tennis racket or ping pong paddle. They read the player’s swing, speed and accuracy to provide feedback on where they can improve their game.

“I took a class at UCLA in 2018 that was all about artificial intelligence. It was fun,” Kevin said. “When quarantine started, there was no one to practice tennis with, so I decided to make something where I could practice with myself, something that would tell me how to practice.”

The class Kevin took taught him the basics of building an artificial intelligence system, and another class a year later taught him basic engineering and how to use a 3D printer.

Courtesy photos

Courtesy photos

When he started building his device he bought factory-made sensors online, wired them and used a 3D printer to create the handle-mounts in his garage. Then he set about programming it to recognize the ball, the swing and the other data he would need for useful feedback.

“It’s a lot of trial and error. It’s really annoying because when you start there’s this thing called the bathtub curve. When you first do something there’s a ton of errors at the beginning, nothing works, (you fix it), but then over time things break down and there’s more errors,” he said.

“You take these classes to build up your knowledge, but you have to really work on doing stuff, researching on your own when you engage in your own projects.”

Kevin finished his first prototype last May, but he’s still improving it.

The teen, who plays varsity tennis at Oak Park High School, spent hours on the community tennis courts, practicing the same swing over and over and over so he could train the AI to read the movement of the racket.

 

 

Once the tennis device worked, he decided to build another one to read ping pong swings. Kevin plays the game with his dad sometimes.

“If you already have this paddle sport technology you can just go wild,” Kevin said. “I might expand into badminton in the future.”

Once his robots were finished, the youth decided he wanted to show them off, so he looked into publishing academic papers. He found the Institute of Electrical and Electronics Engineers, and from there it was like any other school paper.

“I’d never written an academic paper before, but once you know the format, it’s just an abstract, which is an introduction, and then you basically regurgitate everything you’ve done with it,” Kevin said. “Just throw in a bunch of fancy language and make it sound cool.”

His paper about his ping pong sensor was published in September. He said he expects his paper about the tennis sensor to be published soon.

Kevin said he enjoys all the work: building and testing his devices, and writing about them, too. It’s a challenge, and it’s kept him busy while he’s been locked at home during quarantine.

“Part of it is that you want to share what you did. I built this thing, it’s kind of cool, I want to show it off,” he said. “The other part is I’m still in quarantine. I don’t have much to do. It’s like, I already built this thing, let’s see how far we can go with it.”

Follow Ian Bradley on Twitter @Ian_ reports.