Self-propelled vehicles are "trained" to be overly cautious, but there may be situations where they need to make high-speed maneuvers to avoid a collision. Will such vehicles, equipped with high-tech sensors costing tens of thousands of dollars and programmed to move at slow speeds, be able to do it in a split second like a person?
This question is going to be solved by the experts of Stanford University. They created a neural network that allows self-propelled cars to perform high-speed maneuvers with low safety intervention, just like race car drivers do.
When self-driving cars eventually enter mass production, they are expected to be capable of far beyond human capabilities, with 94% of accidents attributed to human error. Therefore, the researchers consider this project an important step in improving the ability of autonomous vehicles to avoid accidents.
Source: 3dnews.ru