Author丨Linfeng
Editor丨Sea waist
Title Picture丨Picture Worm Creativity
At last year's Asian Games, a domestic AI "sparring" serving robot imitating Xu Xin played against Xu Xin himself, reporting the batting score in real time. According to reports, Xu Xin said after the game that the AI robot almost "exactly matched" his own hitting action.
AI+ sports is not a new application, and intelligent means have long been reflected in athletes' training and competitions. VAR (Video Assistant Referee) was used in the 2018 World Cup, and since 2017, Second Spectrum has used AI to provide real-time data and analytics services for the NBA.
The American start-up volley.ai also aims to create AI coach-empowered racquet sports, the volley.ai trainer consists of multiple computers and cameras, is a portable machine that can track the path of the ball, players and field positioning, can record video while simulating professional matches, hit high balls, putts and special pitches, suitable for tennis, cricket, platform tennis, padel (Padel), softball and other racket sports.
Source: volley.ai
Volley.ai was founded in 2021 by two brothers, John and Dan Weinlader, both of whom have engineering backgrounds, with Dan receiving a Ph.D. in electrical engineering from Stanford University. The Weinlader brothers are serial entrepreneurs who have worked together for more than 20 years and have run 3 companies together. They operated Superior Tech for 13 years for the distribution of John Deere equipment, and they founded volley.ai out of their love of platform tennis. The team is made up of ball enthusiasts and engineering scholars, including former professional tennis players.
Last September, volley.ai announced a new partnership with the American Platform Tennis Association to provide them with exclusive training content.
All along, volley.ai was self-funded, until recently in the midst of its first round of funding.
The ball machine which can see the player
Amin Khadduri, CEO of APTA, said: "After meeting John's team and seeing volley.ai trainer in action, we immediately realised that volley.ai would create opportunities for clubs, professionals and players. "In this partnership, the volley.ai trainer will follow in more than 270 annual tournaments, including 47 national championships, during which players will have access to a database of training, batting and players.
According to reports, the volley.ai's trainer, that is, the AI trainer, can replicate the dynamic serve and simulate the game scene, and the skills include serve, putt, high lob, barrier ball and special ball skills. It can also record live coaching sessions for off-field learning and provide players with personal, real-time service stats that can be controlled via a smartphone.
Loading...
Video source: YouTube
Through the volley.ai app, players have access to detailed data, including footage and statistics of the players' practices. Players can control the position of the ball from their mobile phones, set up the machine digitally, and the trainer will keep the player's historical data, so it can ensure the continuity of training. Volley.ai said that the features in its app are constantly being updated. At the same time, volley.ai is rolling out gesture features that allow players to control the trainer without using a mobile device.
Source: volley.ai
The Volley.ai trainer comes with 3 cameras for person tracking, video recording, and requesting help.
In terms of details, it comes with speakers and LED plates for communication and security, and four swivel wheels on the foot, which can be adapted to the lowest temperatures in Chicago and the hottest in Miami. The trainer is 87 inches tall, can be tilted up 56°, tilted down -38°, moved left and right by 34°, and has 3 independent hinges for multi-dimensional rotation. Inside the machine is the NVIDIA vision system, which is capable of 3.5 hours of battery life and holds 70 balls.
Source: volley.ai
Unlike ordinary ball machines, volley.ai's AI trainer builds a large model that analyzes court data, people's positions, and hitting paths.
It is claimed that the volley.ai's trainer system uses vision algorithms and machine learning to simulate practice procedures, and after the vision system tracks the ball, the machine returns the ball at a time that precisely matches the opponent's. This allows the trainer and the user to play the full ball, i.e. from the serve until one side wins or misses, forming the complete course of the game, rather than the impact of a single ball.
Even a smart ball machine can't fully simulate the court. For example, if a smart football ball machine serves at a fixed angle, the player's reaction in training will be slower than that of the ball machine, and it is difficult to train the goalkeeper's ability to predict and make remedies in combination with the timing.
John understands this, "Our goal is to make training aids and make racquet sports better, not how to help players win games. ”
He said professionals are the first partners of volley.ai, but he stressed that volley.ai's business doesn't include providing guidance on how to make technology better, but building communities where professionals have a place to spread their knowledge. For example, within the community, members can receive comments and guidance from other professional users after uploading a replay video in the app.
Volley.ai give the machine to the club for free, charging a monthly rental fee of $1,500 to $3,000. The way to make money is to pay for subscriptions, and after members create an account, they incur a fee for subscriptions, and they can use and share user data to receive guidance from professionals. Users pay $30 for a one-time use of the system, or they can purchase it for $40 per month for an annual fee of $300.
According to reports, after three years of development, the volley.ai trainer was released to the market in September 2023 after eight prototypes, and 110 trainers were distributed to about 45 platform tennis and padel clubs in the northeastern United States and Chicago area, including eight states including Connecticut and Florida, and were distributed in less than four months.
Homemade ball machine
Despite being in business for 27 years, John was not involved in sports until he founded volley.ai in 2021.
It is reported that after graduating from the University of Pennsylvania in the United States, John started his entrepreneurial career after working as a senior software engineer at Motorola for 5 years. He is a successful serial entrepreneur who started with a Windows client application for music in 1997, before computers were fully ubiquitous, and gained 50,000 users.
Later, he ventured into Internet product marketing, and both of his companies were acquired by well-known American companies.
At the beginning of the 21st century, John Deere, an American manufacturer of agricultural and construction machinery, had established a worldwide network of distributors and agents. John, whose family has a background in engineering, joined the family business Superior Tech in 2003 to help his father distribute John Deere's high-quality equipment sprayers, where he grew significantly and ran the business for 13 years before selling the company to Agri-Fab in 2021.
As a distributor of John Deere, John has been exposed to a variety of mechanical and electrical products from major manufacturers, and his modified sprayers have also been approved by the original manufacturer and have been put on sale.
"And I'm going to let everyone see our new products on the touchscreen or on the display panel. ”
So around the same time, John also founded VS Networks in Chicago, a large-scale IoT package for the design, manufacture and transportation of multimillion-dollar devices, which was acquired by rival Spectrio LLC in August 2019.
Since 2010, his brother Dan Weinlader, a technician with a master's degree in electrical engineering from Stanford University, has been working with him since 2010. Dan serves as CTO of VS Network and Director of Superior Tech. Dan has been instrumental in the deployment of program frameworks and operating systems on the software side, as well as the actual mechanical design of the hardware.
Over the years, the brothers have developed a strong engineering background and machine learning knowledge that made it possible for them to invent and modify AI tennis ball machines.
Platform tennis was a hobby that John had cultivated for nearly 10 years, and it wasn't until he was 40 that he had time to pick up a racket to socialise and compete. "I regret that my parents didn't teach me ball games when I was a child, otherwise I could have saved a lot of time for the establishment of the volley.ai. ”
According to reports, platform tennis is different from volleyball and tennis, the court is relatively mini, the ball is very dense and the racket is thinner, and professional players will not have an overwhelming advantage over ordinary players, because it requires enough patience and skill. This aroused John's enthusiasm and interest.
After becoming obsessed with platform tennis, John dragged the traditional tennis ball machine directly to the training ground, but it was of little effect, and the design inertia of the engineering product made him wonder: "Why can't the ball machine locate itself? Why can't it figure out where to throw the ball? Why can't it see where I am?"
He made a makeshift ball machine at the club to improve his training, and unexpectedly, no matter the skill level, people always came to ask for the link to buy his ball machine. This question made John, a veteran of more than ten years of marketing experience, see the market opportunity.
In 2021, brothers John and Dan Weinlader founded volley.ai with the goal of changing the way racquet sports are trained. In addition, John assembled a team of Ph.D. engineers and tennis enthusiasts to run the volley.ai, including PhDs from the University of Michigan, former professional tennis players, and coaches.
Source: volley.ai
John, who prefers platform tennis, chose to start the sport in this game, which is played by about 500 clubs in the United States, including the American Platform Tennis Association (APTA) with 32,000 members. In September 2023, Volley.ai announced a partnership with APTA as its official sponsor.
In his view, the platform tennis market is small, but there is no substitute, and it is an opportunity to fill the gap in the market. Moreover, players have a high tolerance for this type of AI serving robot, which will give them ample opportunity for trial and error.
It can be used by both novices and professionals
In the early days, they bought Ace Attack, a traditional ball machine company that made high-end ball machines designed to run at 120 mph and was not suitable for sale to the general public.
Ball professionals have suggested that the ball machine is effective for repetitive practice movements, but it does not necessarily improve or stimulate the player's batting skills, that is, if a machine keeps feeding the ball to the player at the wrong angle and height, the training effect will be greatly reduced, and the teaching cannot be done without the human coach by the side.
In response to this, there have been technological innovations in the field of racquet sports, such as Slinger's portable ball machine with a mobile app that can record the number of shots and give teaching tips, and the tennis ball machine Proton has also released a smart ball machine that can determine the speed and direction of rotation of the ball. Swing Vision is similar to volley.ai's camera system, producing a bracket that is attached to a fence and can be uploaded to its app to watch where the ball is shot, how fast and how long it will be.
In contrast, volley.ai is unique in its combination of video recording and smart serves, and its serve opportunities are designed to play at different speeds for different types of players, creating different learning curves to help with training and hobbies. "That's where AI comes in to the ball machine system."
A simple ball machine wasn't enough to enter the market, so John and Dan set out to make the original tennis ball machine "smarter" and a trainer. Human tennis coaches can also rent volley.ai trainers to assist in course teaching and improve lesson efficiency.
Source: volley.ai
They distinguish between models based on different levels of players. "If you're a beginner, it's impossible to get a high-speed serve (above 100 mph) in training. "For professional players, volley.ai trainers can be used as a tool for challenging training, as well as a content library to provide customized training and professional analysis based on videos after training.
In the face of professionals, volley.ai trainers know the position and level of the players, and can use the camera system to learn the game and replicate the player's service path during the game to simulate the real game.
John mentioned that he will set up a community in volley.ai to bring together professionals to share and create workouts to build a well-engaged ecosystem. "We have built-in instructional videos so professionals can set up and use the trainer in their own way. When people see the value of its use, the panic about the machine is eliminated. ”
For novice players, the volley.ai trainer does not play with them fiercely, but feeds the ball. The machine will throw the ball in the same position, and novices can see "Success" on the screen for a very short time after hitting the ball to inspire interest in the sport.
In terms of price, John does not plan to sell more than $70,000 to the professional crowd, but intends to work with the team/club for a long time. "If I'm just a dealer, I don't see how often and how the ball machine is used, and how satisfied it is. My goal is that every time it sends a ball, I know what kind of ball it is being used for, what it is used for, and make sure that the ball machine is being used correctly. ”
Source: volley.ai
After making the AI trainer for platform tennis, John expanded the scope to similar racket sports, such as pickleball, padel and more. And, with the increase in the number of clubs and the increase in the number of teaching professional components, they have access to the university market or elite training centers.
In the next step, he gamified volley.ai trainer to create a single-player racquet sports experience.