Browse Publications Technical Papers 2021-01-0870

Automated Test Case Generation and Virtual Assessment Framework for UN Regulation on Automated Lane Keeping Systems 2021-01-0870

Validation of highly automated or autonomous vehicles is nowadays still a major challenge for the automotive industry. Furthermore, the homologation of ADAS/AD vehicles according to global regulations is getting more essential for their safe development and deployment around the world. In order to assure that the autonomous driving function is able to cope with the huge number of possible situations during operation, comprehensive testing of the functions is required. However, conventional testing approaches such as driving distance-based validation approach in the real world, can be time- and cost-consuming. Therefore, a scenario-based virtual validation and testing method is considered to be a proper solution.
In this paper, we propose a virtual assessment framework using a fully automated test case generation method. This framework is embedded into the continuous development and validation process. We conduct a detailed case study on an automated lane keep system (ALKS) function according to the UNECE regulation, hereby presenting the underlying framework and method. We utilize AVL ADAS Load Matrix to automatically generate test cases guided by a coverage metric and optimize the critical-case-finding process within a large parameter space to identify functional boundaries. We show the results of testing an ALKS system where the quality of testing is assured by the given quality metrics, while the effort of executing the test scenarios is reduced. Results obtained from initial test runs together with identified functional boundaries, form the basis for further development and validation steps. For the latter, the generated test scenarios can be re-used in a Hardware-in-the-Loop (HiL) environment, to allow for efficient performance, safety and regression testing throughout the whole development process. In terms of homologation, the presented testing process furthermore offers the opportunity to test and release the developed systems in a virtual world, before final homologation on proving ground and public road is conducted.


Subscribers can view annotate, and download all of SAE's content. Learn More »


Members save up to 16% off list price.
Login to see discount.
Special Offer: Download multiple Technical Papers each year? TechSelect is a cost-effective subscription option to select and download 12-100 full-text Technical Papers per year. Find more information here.
We also recommend:

Effect of Adherent Rain on Vision-Based Object Detection Algorithms


View Details


Unsettled Topics Concerning Coating Detection by LiDAR in Autonomous Vehicles


View Details


Toward an Automated Scenario-Based X-in-the-Loop Testing Framework for Connected and Automated Vehicles


View Details