Who’s your favorite baseball pitcher? Shane McClanahan? Sandy Alcantara? Justin Verlander? Whoever you said, two of the top sports-tech companies in the U.S. — Rapsodo and Trajekt Sports — have teamed up to build a robot version of them, and the results are reportedly uncannily accurate.
Okay, so we’re not talking about walking-talking-pitching standalone robots, as great a sci-fi-tinged MLB ad as that would be. However, Rapsodo and Trajekt have combined their considerable powers to throw a slew of different technologies at the problem of building a machine that’s able to accurately simulate the pitching style of whichever player you want to practice batting against — and they may just have pulled it off, too.
Their solution brings together the Trajekt Arc, a pitching robot that utilizes super-precise motor control, image processing, and machine learning to toss out perfect consistent pitches and Rapsodo’s PRO 3.0, a ground-based radar and image-based monitoring system designed to accurately measure hitting and pitching data.
Heck, the setup even features a video projection system that lets you see life-sized footage of the player pitching right at you. (This offers stock pitcher footage as standard, although teams can also upload their own video data for added customization.) Consider it the future of batting practice as we know it.
“We are trying to be the leader in building baseball pitching machines that replicate and can control all the degrees of freedom necessary to replicate a human pitcher,” Joshua Pope, CEO and co-founder of Trajekt, told Digital Trends.
“It’s really designed and geared toward enhancing the player’s ability to perform on the field,” added Seth Daniels, director of product at Rapsodo.
The first baseball pitching machine was invented back in 1897 by Princeton professor Charles Howard “Bull” Hinton. Somewhat ominously hailed as the “baseball gun” and blasting out baseballs with the aid of gunpowder, it was, by turns, reportedly inaccurate and dangerous. Fortunately, the technology developed from there.
Over the 20th century, a variety of pitching machines were invented, with one notable creation being Paul Givagnoli’s arm-style pitching machine in 1952. This machine launched the ball toward the plate with a pitching style that was intended to mimic that of a real pitcher’s arm-throwing action. For the first time, a player in a batting cage could carry out endless practice against a machine able to recreate what it was like to go up against a human pitcher.
For most of us, the evolution of pitching machines could probably stop right around there. Add in some incremental tweaks and modifications, and a machine that can pitch somewhat like a human is enough for us to work with to improve our batting average. If you really need to visualize hitting against a superstar pitcher, well, just do what most of us do when faced with a basketball hoop in the backyard: Use a bit of imagination.
But things are different when you get to the pros. At that level, the differences that separate an MLB home run leader from an also-ran are almost imperceptible. For these elite athletes seeking the slightest advantage over their opponents, the idea of being able to train against a pitcher’s exact arsenal and pitch characteristics prior to setting foot on the field could make all the difference. That’s where the combination of Trajekt Arc and Rapsodo’s PRO 3.0 enters the picture.
“The way it works is that any tracking technology that is used to measure the player’s pitch can be used as the data source that gets inputted to Trajekt,” said Joshua Pope.
In many training cases, this involves getting your pitcher to throw sample pitches which are then used to teach the device, courtesy of its ball-tracking technology. Throwing out a few sample pitches is enough to help teach the Trajekt Arc the necessary pitching metrics, such as speed, spin, movement, and strike zone location. Once this data has been captured, that pitch will be added to the device’s system so as to be made available for future training sessions. Think of it like the Netflix of pitches.
However, what if you want to model a particular opposing player’s pitch — say, the Los Angeles Dodgers’ Tony Gonsolin — on who you’ll be going up against in an upcoming game? A call to Gonsolin (“Hey, Tony. We’re playing you next week, and we just wondered…”) is probably out of question, but you’re not out of luck. It’s also possible to train the Trajekt Arc using pitch-tracking data from a source like Hawk-Eye, which tracks the pitches for MLB games, thereby still allowing you to create a convincing replica of how a rival pitcher performs on the field.
“There’s something called the nine-parameter fit, which fully defines the trajectory of the ball, the break parameters, the speed spin,” said Pope. “What Trajekt does is we control the release condition with high precision. As long as you have measured those release positions using any tracking system, you can use that as an upload. This data is ubiquitously available at all levels of baseball.”
Spoiler alert: Sports are competitive. That competition doesn’t just take place on the pitch, the court, the track, or the field either. Teams compete over the most promising players, and yes, the best new technologies to help support them.
For that reason, neither Daniels nor Pope were able to reveal which MLB teams are currently using their pitching tech, although they noted that it is currently being used in the training grounds of seven such teams. (Previous Rapsodo tech is being used by all 30 MLB teams.)
“No, we aren’t able to disclose [that information],” Pope said. “I think a lot of folks see it as a competitive edge they want to keep close. We’re respecting that, and [therefore] keeping it confidential.”
Should it be proven to deliver on its promise, however, you can bet your bottom dollar that technology like this will start filtering out to other teams — and maybe beyond the MLB, too.
“Our flagship product … is the most accurate replication of a pitcher that we’ve ever encountered with any pitching machine,” Pope said. “That will always be like our premium offering, and I think the highest value use-case will be with those MLB teams who can upload video data, who can upload all the pitch data they have from games for premium training. But we are working on lighter versions that are more portable, that retain some of the key features like the video projection of the pitcher wind-up [and] the ball orientation controller.”
Coming soon to a batting cage near you? Hey, there are worse ways to improve your game than squaring off against MLB all-stars — virtual or not.
- Fighting football injuries with 3D-printed, hyper-personalized pads
- Meet the robotic pioneers that will help humanity colonize Mars
- SmokeBot robot can see things human firefighters can’t in smoke-filled rooms
- Before robot soccer players can beat humans, they’ll need to stop toppling over
- These ridiculous robots can match Japan’s best volleyball players