Home // VALID 2016, The Eighth International Conference on Advances in System Testing and Validation Lifecycle // View article

Automatic Test Evaluation for Driving Scenarios Using Abstraction Level Constraints

Authors:
Steffen Wittel
Daniel Ulmer
Oliver Bühler

Keywords: Automotive Testing; Test Generation; Test Evaluation; Test Automation

Abstract:
Sophisticated Driver Assistance Systems (DASs) on the way to highly automated driving require extensive testing activities to verify the functionality and the safety of the developed software. With each step towards autonomous driving, the automobile manufacturers are increasingly taking on responsibility for driving maneuvers automatically performed by such systems in unknown environmental situations. Whereas recent DASs use the driver as fallback, in the future this fallback will be only available after a legally prescribed period of time since the driver might be distracted by other activities. To take account of this, robustness testing becomes more and more important to ensure a safe operation at different environments. This paper presents a constraint based approach that applies automatic testing to evaluate DASs. Thereby, the focus is set on the determination of the expected responses that are the basis for the automatic evaluation of the generated test scenarios. The introduced approach is working on different levels of abstraction in combination with an analysis of the observed behavior to classify individual situations of the scenarios after the test execution. The approach enables a full automation of the evaluation, which is the bottleneck of current state-of-the-art scenario generators.

Pages: 14 to 19

Copyright: Copyright (c) IARIA, 2016

Publication date: August 21, 2016

Published in: conference

ISSN: 2308-4316

ISBN: 978-1-61208-500-5

Location: Rome, Italy

Dates: from August 21, 2016 to August 25, 2016