Automated Function Assessment in Driving Scenarios

Abstract

In recent years, numerous innovations in the automotive industry have addressed the field of driver assistance systems and automated driving. Therefore additional required sensors, as well as the need for digital maps and online services, lead to an ever-increasing system space, which must be covered. Established test approaches in the area of Hardware-in-the-Loop (HiL) use predefined and structured test cases to test the systems on the basis of requirements. In the approach of systematic testing, an evaluation is only carried out for a specific test case respectively the duration of a test step. This paper presents a concept for an automated quality assessment of driving scenarios or digital test drives. The aim is the analysis and subsequent evaluation of continuous function behavior during a realistic test drive within a simulated environment. Compared to conventional systematic test approaches, the presented concept allows a continuous evaluation of the test drive, whereby multiple evaluations of systems in similar scenarios with deviating boundary conditions is possible. For the first time, this enables a functional evaluation of a complete test drive comprising numerous scenarios and situations. The presented approach was prototypically implemented and demonstrated on a Hardware-in-the-Loop (HiL) test bench evaluating an adaptive cruise control (ACC) system.