[ad_1]
An autonomous car used artificial intelligence to improve its driving skills in 20 minutes.
Engineers at Cambridge University have developed what is called "reinforcement learning" that allows the vehicle to improve its ability to stay in the race.
The Wayve team showed how a Renault equipped with the technology used by trial and error to properly track a track, without the use of sensors and engineering rules as do it. other autonomous cars.
First, the car deviates to the side, requiring regular safety driver intervention.
With each correction, the AI makes fewer mistakes, finally learning how to stay properly in the lane without help.
"The missing piece of the puzzle is smart algorithms, no more sensors, rules and maps" "Humans have a fascinating ability to perform complex tasks in the real world, because our brains allow us to to learn quickly and transfer knowledge through our many experiments.We want to give our vehicles better brains, not more equipment. "
Wayve provides that reinforcement learning IA will be used in the future to handle everything from traffic lights to intersections and intersections.
19659010]
[ad_2]
Source link
