Simulation-based testing of cyber-physical systems (CPS) is costly due to the time-consuming execution of CPS simulators. In addition, CPS simulators may be flaky, leading to inconsistent test outcomes and requiring repeated test reexecution for reliable test verdicts. Many test inputs within the input space of CPS may not effectively exercise the behaviour of the system under test (SUT) – for instance, those that violate system preconditions, exceed operational design domain (ODD) limits, or represent inherently safe scenarios. In this article, we propose to use test validators to filter out such test inputs before execution. We describe two methods for generating test validators: one using genetic programming (GP) that employs well-known spectrumbased fault localization (SBFL) ranking formulas, namely Ochiai, Tarantula, and Naish, as fitness functions; and the other using decision trees (DT) and decision rules (DR). We evaluate our test validators through case studies in the domains of aerospace, networking and autonomous driving.We show that test validators generated using GP with Ochiai are significantly more accurate than those generated using GP with Tarantula and Naish or using DT or DR. Moreover, this accuracy advantage remains even when accounting for the flakiness of the simulator. We further show that our test validators generated by GP with Ochiai are robust against flakiness with only 4% average variation in their accuracy results across four different network and autonomousdriving systems with flaky behaviours. Finally, we show that, on average, 88.7% of the assertions inferred by our approach align with or overlap with requirements precondition violations, ODD-limit violations, and nominal safe conditions extracted from technical standards and empirical results in the literature. Our full replication package is available online [1].

Traditional software testing techniques focus on discovering inputs that reveal failures in the system under test. However, such techniques often fall short of providing explanations for the underlying reasons of system failures. I have made efforts to employ simulation-based testing to provide interpretable feedback to the developers regarding the circum-stances of system failures. To achieve this, I first explore machine learning (ML)-assisted test generation techniques for building failure models. These techniques leverage ML models to enhance the effectiveness and efficiency of testing, either by predicting the test outputs or providing guidance through a reduced search space. Subsequently, I build failure models using interpretable machine learning models based on tests generated by the ML-assisted test generation algorithms. Such failure models generate a set of rules that developers can easily interpret. The systems for which I build failure models belong to the cyber-physical and network domains. In this doctoral symposium, I present my research and outline plans for the remainder of my Ph.D. study.

Test inputs fail not only when the system under test is faulty but also when the inputs are invalid or unrealistic. Failures resulting from invalid or unrealistic test inputs are spurious. Avoiding spurious failures improves the effectiveness of testing in exercising the main functions of a system, particularly for compute-intensive (CI) systems where a single test execution takes significant time. In this article, we propose to build failure models for inferring interpretable rules on test inputs that cause spurious failures. We examine two alternative strategies for building failure models: (1) machine learning (ML)-guided test generation and (2) surrogate-assisted test generation. ML-guided test generation infers boundary regions that separate passing and failing test inputs and samples test inputs from those regions. Surrogate-assisted test generation relies on surrogate models to predict labels for test inputs instead of exercising all the inputs. We propose a novel surrogate-assisted algorithm that uses multiple surrogate models simultaneously, and dynamically selects the prediction from the most accurate model. We empirically evaluate the accuracy of failure models inferred based on surrogate-assisted and ML-guided test generation algorithms. Using case studies from the domains of cyber-physical systems and networks, we show that our proposed surrogate-assisted approach generates failure models with an average accuracy of 83%, significantly outperforming ML-guided test generation and two baselines. Further, our approach learns failure-inducing rules that identify genuine spurious failures as validated against domain knowledge.

An input to a system reveals a non-robust behaviour when, by making a small change in the input, the output of the system changes from acceptable (passing) to unacceptable (failing) or vice versa. Identifying inputs that lead to non-robust behaviours is important for many types of systems, e.g., cyber-physical and network systems, whose inputs are prone to perturbations. In this paper, we propose an approach that combines simulation-based testing with regression tree models to generate value ranges for inputs in response to which a system is likely to exhibit non-robust behaviours. We apply our approach to a network traffic-shaping system (NTSS) – a novel case study from the network domain. In this case study, developed and conducted in collaboration with a network solutions provider, RabbitRun Technologies, input ranges that lead to non-robustness are of interest as a way to identify and mitigate network quality-of-service issues. We demonstrate that our approach accurately characterizes non-robust test inputs of NTSS by achieving a precision of 84% and a recall of 100%, significantly outperforming a standard baseline. In addition, we show that there is no statistically significant difference between the results obtained from our simulated testbed and a hardware testbed with identical configurations. Finally we describe lessons learned from our industrial collaboration, offering insights about how simulation helps discover unknown and undocumented behaviours as well as a new perspective on using non-robustness as a measure for system re-configuration.