The implementation of autonomous driving requires verification of the functions, performance, safety, stability, and robustness of products and systems.
Functional test: The main functional indicators include whether it can correctly respond to various road traffic facilities, whether it can abide by traffic rules, and whether it can correctly respond to traffic participants such as vehicles, bicycles, and pedestrians on the road according to the design indicators of the automatic driving function. Whether it can normally exit outside the set ODD and prompt the driver to take over, and whether it can correctly complete other automatic driving functions planned during functional design.
Performance test: The main indicators include various vehicle movement data (such as speed, acceleration, driving route), the recognition accuracy rate of traffic participants, response speed, recognition range, adaptability to various lighting and climate environments, drivers and the subjective experience of the occupants (such as whether they feel confused, nervous, and uneasy, whether the driving operations of the vehicle are comfortable and natural during automatic driving).
Safety testing: Safety indicators include functional failure probability, passing of functional safety scenarios and passing of expected functional safety scenarios. Stability test: mainly to verify whether the function and performance can run stably.
Robustness test: Robust is the transliteration of Rubust, which is the key to system survival in abnormal and dangerous situations. It mainly tests the system's anti-strike capability. It mainly verifies whether the system can recover in time and reduce the severity when problems are encountered in complex scenarios. For example, whether computer software will not freeze or crash the system in the event of input errors, disk failures, network overloads, or intentional attacks.