The robotaxi reality check
The sensory showdown
The brutal edge cases
The legal, ethical, and social road
audi Match sensors to tricky weather to figure out why self-driving cars still get confused.
audi End-to-End Neural Networks
No more lines of code, the network takes the pixels and decides how to steer.
audi Eight external cameras and artificial intelligence mimic human vision to navigate roads, allowing Tesla to avoid using expensive LiDAR hardware.
Which vehicle manufacturer is famous for using a vision-only camera approach, rejecting LiDAR and radar?
How many external cameras does Tesla's Full Self-Driving (FSD) hardware suite rely on to navigate the road?
audi Heavy steam and dust can confuse the expensive laser shields used by autonomous cars.
Which environmental factor is known to occasionally confuse LiDAR sensors by reflecting light unpredictably?
LiDAR in the dark: Because LiDAR sensors use invisible lasers to map their surroundings, a self-driving car can navigate a winding road in pitch-black darkness with its headlights completely turned off.
What type of map is required by Level 4 robotaxis to know the exact position of curbs and lane lines down to the centimeter?
Why are geofenced robotaxis like Waymo unable to instantly operate in a newly built neighborhood?
A detection system that works on the principle of radar, but uses light from a laser. By 2026, Waymo paired these laser sensors with cameras to achieve redundant, all-weather vehicle perception.