Test Framework for Automatic Test Case Generation and Execution Aimed at Developing Trustworthy AVs from Both Verifiability and Certifiability Aspects

Abstract

Recent developments in the testing and safety assurances of Automated Driving Systems (ADSs) have shifted from a traditional distance-based approach (i.e. vehicle miles travelled) to scenario-based approach. Various studies have been conducted on different aspects of the scenario-based testing workflow such as scenario generation, scenario description language, and scenario analysis. This study intends to build on top of the individual functional modules introduced for scenario-based testing and contributes towards a common test framework, based on the experience gained from the Innovate UK's OmniCAV project. The Test Framework introduced in this paper consists of: Description of Test Scenarios, Different Types of Testing and Allocation of Tests, Generation of Test Cases, Data Collection & Analysis, Correlation of obtained Results and Evidence Gathering for functional and behavioral verification. The framework and processes defined here have relevance to real-world practice and contribute towards both verifiability and certifiability of ADSs.