A Survey on Automated Driving System Testing: Landscapes and Trends

Automated Driving Systems ( ADS ) have made great achievements in recent years thanks to the efforts from both academia and industry. A typical is composed of multiple modules, including sensing, perception, planning, control, which brings together latest advances different domains. Despite these achievements, safety assurance significance, since unsafe behavior can bring catastrophic consequences. Testing has been recognized as an important system validation approach that aims expose behavior; however, context ADS, it extremely challenging devise effective testing techniques, due high complexity multidisciplinarity systems. There much literature focuses on a number surveys also emerged summarize technical advances. Most focus system-level performed within software simulators, they thereby ignore distinct features modules. In this article, we provide comprehensive survey existing literature, takes into account module-level testing. Specifically, make following contributions: (1) We techniques for highlight differences affected by modules; (2) with empirical studies issues occurring development or deployment, problems collaborations between gap simulators real world; (3) identify challenges opportunities testing, pave path future research field.