As a methodology for system design and testing, use cases are well-known and widely used. While current active machine learning (ML) algorithms can effectively automate unit testing, they do not scale up to use case testing of complex systems in an efficient way. We present a new parallel distributed processing (PDP) architecture for a constrained active machine learning (CAML) approach to use case testing. To exploit CAML we introduce a use case modeling language with: (i) compile-time constraints on query generation, and (ii) run-time constraints using dynamic constraint checking. We evaluate this approach by applying a prototype implementation of CAML to use case testing of simulated multi-vehicle autonomous driving scenarios.

We explore the application of graph database technology to spatio-temporal model checking of cooperating cyber-physical systems-of-systems such as vehicle platoons. We present a translation of spatio-temporal automata(STA) and the spatio-temporal logic STAL to se-mantically equivalent property graphs and graph queries respectively. We prove a sound reduction of the spatio-temporal verification problem tograph database query solving. The practicability and efficiency of thisapproach is evaluated by introducing NeoMC, a prototype implementation of our explicit model checking approach based on Neo4j. To evaluate NeoMC we consider case studies of verifying vehicle platooning models. Our evaluation demonstrates the effectiveness of our approach in terms of execution time and counterexample detection.

Cooperating cyber-physical systems-of-systems (CO-CPS) such as vehicle platoons, robot teams or drone swarms usually have strict safety requirements on both spatial and temporal behavior. Learning-based testing is a combination of machine learning and model checking that has been successfully used for black-box requirements testing of cyber-physical systems-of-systems. We present an overview of research in progress to apply learning-based testing to evaluate spatio-temporal requirements on autonomous systems-of-systems through modeling and simulation.

We introduce a new methodology for virtualized fault injection testing of safety critical embedded systems. This approach fully automates the key steps of test case generation, fault injection and verdict construction. We use machine learning to reverse engineer models of the system under test. We use model checking to generate test verdicts with respect to safety requirements formalised in temporal logic. We exemplify our approach by implementing a tool chain based on integrating the QEMU hardware emulator, the GNU debugger GDB and the LBTest requirements testing tool. This tool chain is then evaluated on two industrial safety critical applications from the automotive sector.

Learning-based testing (LBT) is an emerging paradigm for fully automated requirements testing. This approach combines machine learning and model-checking techniques for test case generation and verdict construction. LBT is well suited to requirements testing of low-latency safety critical embedded systems, such as can be found in the automotive sector. We evaluate the feasibility and effectiveness of applying LBT to two safety critical industrial automotive applications. We also benchmark our LBT tool against an existing industrial test tool that executes manually written test cases.