It’s been two years since Yamaha Motobot challenged MotoGP GOAT Valentino Rossi to a lap time duel. And now, the race has been run, at California’s Thunderhill raceway.
Humans have been getting their tails handed to them by computers. A.I. has been regularly trouncing humanity in classic games like chess, Go, and Mortal Kombat for years now. But there’s one area where we still reign supreme (for now): motorcycle racing.
The Yamaha MOTOBOT project began in 2015 with the goal of an A.I. driving a motorcycle “with no modifications made to the motorcycle itself.” MOTOBOT controls six actuators on the bike: the steering wheel, throttle, front brake, rear brake, clutch and gearshift pedal. As a test for the ambitious project, Yamaha and partners SRI International gave themselves an equally ambitious goal: MOTOBOT beating MotoGP racer Valentino Rossi.
Rossi, a coworkers of the researchers while racing for Movistar Yamaha MotoGP, has won a remarkable 89 out of the 301 races he has been in professionally. Make that 90 as he smoked Yamaha MOTOBOT.
Check Out: Auto Expo 2018 to be held between February 9-14.
Of course, Yamaha MOTOBOT is actually performing quite well. The computer, after three years of trials, has a much greater understanding of motorcycle racing than an amateur would. Yamaha hopes that by 2020, they’ll have gleaned enough from MOTOBOT to “deliver new value from Yamaha to our customers,” according to their website. While Rossi sounds eager for a rematch, it appears that Yamaha will settle for second place as long as the average, non-racing customer can glean some real driving benefits from MOTOBOT, in which case we all win.
And while Rossi beat the metal pants off a 2 year old Motobot, how much money would you put on the fleshy GOAT in five years’s time? What about 10? Food for thought. Enjoy an explanation of some of the tech inside Motobot and the thinking behind the project in a second, longer video below:
