UncertainTree: Analyzing Multi-Model Consistency under Uncertainty

Model-driven development of complex software-intensive or cyber-physical systems involves the creation of heterogeneous artifacts and models. Although most modeling tools seem to convey certainty for each value or model state, the information they represent often relies on assumptions and so, by necessity, include a degree of uncertainty. When specifying and managing such uncertainty within a single model, it is hard to determine its effect on the set of modeling artifacts and as well, on model consistency and correctness. The potential effects of uncertainty can be unclear, especially when models are automatically kept consistent. Effects become even more unclear when consistency and correctness are determined not solely by constraints but for a given model state, also by simulation or other black-box mechanisms. In this paper, we present UncertainTree, an end-to-end process that makes uncertainty a first-class concern of multi-model consistency analysis. The approach's analysis includes interacting uncertainties from various sources and models. Engineers annotate uncertainty in the multi-model environment. The process explores the induced possible models, propagates them, and evaluates their consistency. UncertainTree aggregates the heterogeneous verdicts in a central reporting metamodel and communicates the outcome through stakeholder-specific uncertainty views. To improve explainability, we enrich the views with explanations provided by large language models, accessing the evidence provided by previous analysis steps. To make the analysis scale, we envision interchangeable execution modes, from state-space exploration to symbolic execution of consistency-preservation rules. As evidence of feasibility, we report a pre-study, conducted as a small-scale evaluation with a working prototype on two subject systems, as evidence of feasibility. Further, we lay out a detailed plan to evaluate the end-to-end vision.