Can A.I. Beat a Human in a Drone Race?
For the last two years, Google has been funding research into drone autonomy. Last week, JPL engineers put this research to good use and tested their AI piloted drone against world-class drone pilot, Ken Loo aka. FlyingBear.
While Loo used his superhuman reflexes and skill to navigate the course, the A.I. drone used a set of cameras and sensors to track its position relative to a preprogrammed map.
These sensors allowed the drone to successfully complete the course with a lap average of just 13.9 seconds. Sadly (for the A.I drone) this was not fast enough to best Loo who averaged 11.1 seconds/lap.
The good news for all current and aspiring drone pilots is that it appears JPL/Google still have some work to do before their drones can compete in the DRL or otherwise take your jobs. Still, it is quite impressive to see how far they have come and equally exciting to think about where they will go in the years to come.