Driverless cars can't reach the height of human brain in "visual ability", and their sensors complete the collection of road environment through infrared cameras and ordinary cameras. When the vehicle is in a densely populated building area, an accident area or other people command the vehicle to pass through this area through general gesture signals, unmanned vehicles will encounter difficult situations to judge. In addition, there are old and deformed signal signs on the road, and unmanned vehicles may misunderstand or miss them, causing unnecessary accidents.
2. The degree of humanistic acceptance.
There are still many questions about driverless cars in society. For example, when a driverless car is driving in this densely populated world and it is found that an accident is inevitable, should the intelligent computer choose pedestrians rushing to the road or directly hit oncoming vehicles? Can you still drive at full load after being attacked by an external virtual network? How to exercise the road not fully drawn by Google or GPS? Driverless cars also have great challenges in laws and regulations. Such as product liability, legislation and multiple jurisdiction. Accident liability judgment between unmanned vehicles and manned vehicles and accident liability judgment between unmanned vehicles.
3. Low security and defensiveness
JonathanPetit, chief scientist of SecurityInnovation, a software security company, said that most lidar systems used to detect obstacles in driverless cars can be cracked with a device that costs less than $60. Petit said that through this device, hackers can set up cars, pedestrians or walls that don't actually exist anywhere, which will slow down the driving speed of driverless cars and even make them difficult to move. Its related papers have been published at the European Black Hat Safety Conference? .
Million car purchase subsidy