Verification and Validation (V&V) for Safe Deployment of Automated Driving Systems – In Depth Evaluation of State-of-the-Art V&V Methods in the Automotive Sector
Sebastian Siegl, Tobias Düser
Abstract
The number of Advanced Driver Assistance Systems (ADAS) for SAE L2 and higher automation levels has increased dramatically in recent years. ADAS encompasses a wide range of technologies like Adaptive Cruise Control (ACC), Lane Departure Warning (LDW) or Automatic Emergency Braking Systems (AEBS). For higher automation levels (SAE L4), robotaxi and roboshuttle services are emerging significantly since 2018. Integration of the technology for SAE L2 to L4 and offering on-demand services are aimed at augmenting traffic safety and minimizing the risk of accidents through human failure. Robotaxi service providers (SAE L4) are live with services in US American and Chinese cities. These technological advancements require the development and application of testing methods to prove the safety, quality and reliability of automated driving systems (ADS) before deploying them. This publication presents an overview of state-of-the-art testing methods used in the automotive industry for the verification and validation (V&V) of ADS with SAE L2 to L4. The three methods shadow mode, functional real-world testing and simulation are identified as the most relevant and are evaluated using specific categories.