Exploring the Assessment List for Trustworthy AI in the Context of Advanced Driver-Assistance Systems

Borg, Markus; Bronson, J. C.; Christensson, L.; Olsson, Fredrik; Lennartsson, Olof; Sonnsjö, Elias; Ebabi, Hamid; Karsberg, Martin

doi:10.48550/arxiv.2103.09051

Cited by 1 publication

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Machine Learning (ML) nowadays is used in a wide range of application areas such as automotive [1], [2], health care [3] and manufacturing [4]. Many of the ML-driven systems in these domains present a high level of autonomy [5] and meanwhile are subject to rigorous safety requirements [6].…”

Section: Introductionmentioning

confidence: 99%

Machine Learning Testing in an ADAS Case Study Using Simulation-Integrated Bio-Inspired Search-Based Testing

Moghadam¹,

Borg²,

Saadatmand³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

This paper presents an extended version of Deeper, a search-based simulation-integrated test solution that generates failure-revealing test scenarios for testing a deep neural network-based lane-keeping system. In the newly proposed version, we utilize a new set of bio-inspired search algorithms, genetic algorithm (GA), (µ + λ) and (µ, λ) evolution strategies (ES), and particle swarm optimization (PSO), that leverage a quality population seed and domain-specific cross-over and mutation operations tailored for the presentation model used for modeling the test scenarios. In order to demonstrate the capabilities of the new test generators within Deeper, we carry out an empirical evaluation and comparison with regard to the results of five participating tools in the cyber-physical systems testing competition at SBST 2021. Our evaluation shows the newly proposed test generators in Deeper not only represent a considerable improvement on the previous version but also prove to be effective and efficient in provoking a considerable number of diverse failure-revealing test scenarios for testing an ML-driven lane-keeping system. They can trigger several failures while promoting test scenario diversity, under a limited test time budget, high target failure severity, and strict speed limit constraints.

show abstract