[VDZ22] Cost-effective Simulation-based Test Selection in Self-driving Cars SW by Christian Birchler

Self-Driving Cars (SDCs) are the future of mobility. They are meant to drive better and more conscious than humans. Unfortunately, in the recent past, there have been several life-critical incidents with SDCs, which hinder the widespread usage of SCDs in real society. One major limiting factor is that testing complex systems such as SDCs are costly and time-consuming. Therefore, to enhance the safety aspect of SDCs and lower the costs necessary to test these systems, a proper cost-effective testing approach is required. Hence, we proposed SDC-Scissor, a framework that instantiates a Machine Learning pipeline to identify (simulation-based) tests that are unlikely to detect faults in the SDC software under test and skip them before their execution, and this is to increase the fault detection ability of tests while reducing their cost. By filtering out those tests, SDC-Scissor reduces the number of long-running simulations to execute and drastically increases the cost-effectiveness of simulation-based testing of SDCs software. More information at https://github.com/ChristianBirchler/sdc-scissor