Here’s what happened when AI drones raced against a world-class human pilot

Artificial intelligence systems have been shown to be better than humans at many things – from spotting cancer on tissue slides to mastering the ancient Chinese game Go – but when it comes to drone racing, humans still rule. For now.

In a race funded by Google and organised by Nasa, world-class drone pilot Ken Loo beat artificial intelligence (AI) drones in speed and skilled manoeuvres.

Scientists at Nasa’s Jet Propulsion Laboratory (JPL) in the US built three custom drones – called Batman, Joker and Nightwing – and developed the algorithms to enable the drones to fly at high speeds while avoiding obstacles.

When pitted against Loo, the AI-powered drones were slower, averaging 13.9 seconds compared with Loo’s 11.1 seconds. They were also unable to mimic Loo’s impressive aerial corkscrews.

(Nasa JPL)

However, the AI drones were more consistent in their performance and flew more cautiously compared with Loo.

Rob Reid, projects task manager at the JPL, said: “We pitted our algorithms against a human, who flies a lot more by feel.

“You can actually see that the AI flies the drone smoothly around the course, whereas human pilots tend to accelerate aggressively, so their path is jerkier.”

The researchers say their AI drones are a work in progress.

The drones were built to go as fast as 129 kilometres per hour in a straight line but on the obstacle course set up in a JPL warehouse, the teams could only fly at 48 to 64 kilometres per hour before applying the brakes.

While Loo was more creative and nimble in his performance he, unlike his competitors, was limited by exhaustion.

He said: “This is definitely the densest track I’ve ever flown. One of my faults as a pilot is I get tired easily.

“When I get mentally fatigued, I start to get lost, even if I’ve flown the course 10 times.”


Most Read in World