uses three main pillars. one is mapping so uses three main pillars. one is mapping. so for uses three main pillars. one is mapping. so for the uses three main pillars. one is mapping. so for the system i uses three main pillars. one is mapping. so for the system to| mapping. so for the system to work mapping. so for the system to work you mapping. so for the system to work you send mapping. so for the system to work you send vehicles, - mapping. so for the system to work you send vehicles, much| work you send vehicles, much like google work you send vehicles, much like google sends work you send vehicles, much like google sends vehicles - like google sends vehicles around like google sends vehicles around the like google sends vehicles around the streets - like google sends vehicles around the streets to - like google sends vehicles around the streets to mapj like google sends vehicles - around the streets to map out street around the streets to map out street view
solvable problem, but very, l very expensive to do - and even worse to maintain. there s another limitation - with current approaches in terms of the sensors that get used. they re very expensive sensors, something called lidar, - which is a way of sending out beams of light, laser light and measuring the time for that light to reflect back. and that gives you a sense of how far things are away. that s a wonderful sensor. i love depth information. i used it in my own previous work, but it s very expensive, hard to integrate in the vehicles. so that s the second one. and the third one is| how the intelligence for these is set up. and today, most approaches - to autonomous driving are based on a very modular system where you sort of first try| and detect all the cars and the pedestrians . and the traffic lights and whatever else you think you need. -
use an vehicles much like google since vehicles around the streets to map out street view images so we can navigate ourselves. but for vehicles, autonomous vehicles, this needs to be done is a much higher percent and so we have sent me to level detailed maps of entire cities and which is actually a remarkable solvable problem but very, very expensive to do an even worse to maintain. there is another limitation with current approaches in terms of the sensors that get used which are very expensive which our way of sending out beams of light, laser light in determining the amount of time that latex to give back and determine how far things are away. that s a wonderful centre and i love that type of information i used it in my previous work and it s hard to integrate into vehicles. in the third one is,