AV-FUZZER: An automated fuzzing framework for safety testing of autonomous vehicles

Abstract

Autonomous vehicles (AVs) are getting increasing attention, since people believe that AVs have signi cant potential to improve the productivity of transportation and quality of life. However, there is a public perception problem with regard to the safety of AVs, so ensuring AV safety is among the top issues regarding the development of AVs, and it is critical to the success of AVs in the marketplace. However, the question is how to test out the AVs system whether it is as safe as expectations. In this thesis, we will discuss the history and background of the AV's development. In the meanwhile, the focus of this thesis is on testing methodologies, AV-Fuzzer and the other two existed testing frameworks.