Artificial intelligence (AI) in sport may sound like something from a science fiction movie.
But unbelievable footage has captured the moment an AI pilot overthrows three human champions in a fierce drone race.
Drone racing world champion, Alex Vanover, was among the trio of professionals thrashed by AI as it won an impressive 15 of 25 trials in a Swiss-based study.
The so-called ‘Swift’ technology mastered speeds of over 62mph (100kmh), with a reaction time that was 120 milliseconds faster than humans on average.
Its impressive track record is a milestone for scientists at the University of Zurich, who sought to move this pioneering technology beyond board games.
Drone racing world champion, Alex Vanover, was among the trio of professionals thrashed by AI as it won an impressive 15 of 25 trials in a Swiss-based study
‘Deep RL (deep reinforcement learning) has enabled some recent advances in artificial intelligence,’ scientists say.
‘Policies trained with deep RL have outperformed humans in complex competitive games, including Atari, Go, chess, StarCraft, Dota and Gran Turismo.
‘Swift competed against three human champions, including the world champions of two international leagues, in real-world head-to-head races.
‘This work may inspire the deployment of hybrid learning-based solutions in other physical systems, such as autonomous ground vehicles, aircraft and personal robots, across a broad range of applications.’
The Swift technology uses a complex combination of learning-based algorithms to map sensory information of a drone while controlling its commands.
As part of this study, experts put Swift head-to-head with Vanover as well as the two-time MultiGP International Open World Cup champion, Thomas Bitmatta, and 16-year-old Swiss national champion Marvin Schaepper.
The study track was set up with seven square gates that players had to pass through in order to complete each 246ft-long lap.
To win a race, a competitor was required to complete three consecutive laps before its opponent in 25 rounds.
Each pilot wore a headset that showed a video stream in real-time aboard their aircraft, providing a similarly immersive first-person experience.
To win a race in the 25-round test, a competitor was required to complete three consecutive laps before its opponent
The fierce race came as part of a study developed by Switzerland’s University of Zurich
Thomas Bitmatta, Marvin Schaepper and Alex Vanover (left to right) racing against AI
Footage shows Swift – illuminated with blue lights – often rocketing past red, human-piloted drones through each gate.
The AI-powered drone outperformed all three champion flyers – scoring five, four and six wins against Vanover, Bitmatta and Schaepper respectively.
Swift’s completion of three laps also took around 16.98 seconds on average – 0.4 seconds faster than runner-up Vanover at 17.38 seconds.
Scientists attribute Swift’s speed to its longer-term thinking style, helping it make sharper turns and more efficient moves.
On the flip side, it’s claimed that human pilots only plan their course of action up to one gate into the future.
Despite this, Swift does have its disadvantages as scientists believe it isn’t as resilient as humans.
‘Human pilots are impressively robust: they can crash at full speed, and—if the hardware still functions—carry on flying and complete the track,’ scientists say.
‘Swift was not trained to recover after a crash. Human pilots are also robust to changes in environmental conditions, such as illumination, which can markedly alter the appearance of the track.’
Even still, Swift is a milestone in robotics, and experts hope it will inspire other technologies in future.
‘Swift won several races against each of the human champions and demonstrated the fastest recorded race time,’ they wrote.
‘This work represents a milestone for mobile robotics and machine intelligence, which may inspire the deployment of hybrid learning-based solutions in other physical systems.’
Read the full article here
Discussion about this post