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

TL;DR

This paper enhances a simulation-integrated testing tool for neural network-based lane-keeping systems by incorporating bio-inspired search algorithms, demonstrating improved effectiveness and diversity in failure detection through empirical evaluation.

Contribution

Introduction of new bio-inspired search algorithms into the Deeper testing framework, improving failure scenario generation for ML-based ADAS testing.

Findings

New algorithms outperform previous versions in failure detection.

Test generators produce diverse failure scenarios efficiently.

Effective under limited testing time and strict constraints.

Abstract

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), $({\mu}+{\lambda})$ and $({\mu},{\lambda})$ 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.