“That’s about it,” she said this week.
Carissa Moore, who on Tuesday faced off with Buitendag for the sport’s first-ever Olympic gold medal, takes a different approach. She loads up on performance analytics, wave pools and science. The American, who beatBuitendag by nearly 6.5 points to win the gold medal on Tuesday,has competed on artificial wavesand uses technology such as a wearable ring that tracks her sleep and other vitals to help her coaches fine-tune her training and recovery.
Their different approaches go to the heart of a long-running tension in surfing: dueling images of the spiritual, naturalist wave rider versus the modern, techie athlete.
“There’s this illusion that you’re trying to sustain, even if you’re aware of all the stuff that’s gone into [surfing],” said Peter Westwick, a University of Southern California surfing historian. He’s talking about the use of advanced polymer chemistry-enabled products in surfboards and wetsuits and complex weather modeling that helps govern where and how competitions like this Olympic event are held. The tech has roots in military research and development, he said.
“It’s now the basis of this billion-dollar industry,” Westwick said.
The latest iteration of that loaded conflict involves software that’s invisible but powerful, like the wind that helps propel the waves the sport relies on. Machine learning algorithms could further shape surfing in years to come, helping to improve wave forecasting, and making inroads into training, injury prevention, and recruitment of top athletes, according to researchers and coaches.
“We’ve been really trying to figure out ways to get our athletes to perform. There’s so many variables you can’t control, like wind and tides,” said Kevyn Dean, USA Surfing’s medical director. “Taking a deeper dive into analytics and data was our roadmap…We really want to follow the data.”
USA Surfing took cues from other sports that have leveraged data analytics to improve athletic performance and output, including basketball, baseball, soccer and football, he said. Among the metrics they’ve looked at: cardiovascular output, sleep patterns, heart-rate variability, and the frequency of certain injuries. They’ve also examined peer-reviewed research from around the globe to figure out what physiological factors could affect performance.
One study, for instance, showed that a surfer’s ability to land “an air,” an aerial maneuver during which an athlete takes the board above the lip of the wave and lands on the frothy, white water, was affected by a lack of ankle mobility and hip stability.
“That’s kind of obvious, but the question is: how much ankle mobility and hip stability do you need?” Dean said. “What position do they get into that allows you to land an air?”
The answer requires data.
USA Surfing experimented in 2019 with biomechanics data acquired with the help of motion-capture cameras and force sensors. They quantified different jumping and landing mechanics. They were also able to get data on sway—slight movements made to maintain balance, as well as the force applied by the athlete’s legs to a plate that stands in for the surfboard.
“There tended to be a lot of imbalances between each limb, and that imbalance may be that critical factor that makes an athlete superhuman” or likely to get injured, said Tracy Axel, USA Surfing’s manager for data and analytics. “There’s a critical breaking point.”
With the help of an engineering team, the organization is developing a machine-learning system that could analyze some of the same information, but using imagery taken as a surfer rides actual waves, she said. In January, they developed a proof-of-concept version capable of identifying basic maneuvers, as well as the dominant stance a surfer takes on their board, she said. With further development, the hope is that artificial intelligence could help teams with talent recruitment, injury prevention and training by sifting through tons of videos and identifying patterns of interest, she said.
Wave forecasting is among the biggest applications machine learning has had in surfing to date. The use of wave-forecasting technology has a long history that precedes surfing, including in coastal engineering, shore protection and combat-planning for World War II, according to Westwick, the historian.
Wave forecasting has more recently benefited from some of the same technological advances that have enabled other commercial machine-learning technologies, like image and voice recognition: namely the large-scale availability of data and computing power, said Ning Li, an ocean wave models system specialist at the University of Hawaii at Manoa’s Pacific Islands Ocean Observing System.
In part, that’s what enabled a team from Surfline Inc. to help identify Tsurigasaki Surfing Beach here as the venue for surfing’s inaugural Olympic ride, despite Japan not being known for its big surf. Using decades of climate data, the team forecast that tropical storms could help feed the waves.
On the first day of competition Sunday, several surfers, including Moore, said the waves were small or hard to ride. But as the week went on, the surfers rejoiced at the development of a typhoon that helped create bigger waves for the finals. For Moore and Buitendag’s gold-medal matchup, wave heights reached 2.5 meters, or just over 8 feet.
Surfline’s software ingests and analyzes multiple data sources, including satellite imagery, polar-ice cover, seafloor shape, wind patterns and buoy measurements, according to Kevin Wallis, Surfline’s director of forecasting.
“We’ve got the computer crunching all of those numbers to guide us in making calls,” he said. For the Olympic competition, Surfline models helped make the call on which days athletes would compete.
The company’s forecast also leverages feedback from human surfers, plus inputs from a network of roughly 800 cameras that allow the company’s small army of forecasters to analyze wave patterns remotely. That information is used, in part, to tweak models to improve their forecasts. Sometimes, company staff will delay meetings to go test the accuracy of their algorithms and provide their models feedback in the name of R&D, he and his colleagues said.
For instance, when Southern California gets a mix of swells from the south and the northwest, that tends to create bigger waves than the models predict, Wallis said. The models can be off by a couple of feet, which makes a huge difference in wave quality and rideability, he said. Human feedback makes a difference.
“It’s a work in progress,” he said. But “to actually go and ride the waves you’ve predicted is really helpful. That’s another awesome benefit of the job.”
This story has been published from a wire agency feed without modifications to the text
Subscribe to Mint Newsletters * Enter a valid email * Thank you for subscribing to our newsletter.
Never miss a story! Stay connected and informed with Mint.
Download
our App Now!!
Source: https://www.livemint.com/sports/olympics-news/how-olympic-surfing-is-trying-to-ride-the-machine-learning-wave-11627388786162.html
