Abstract: Model-based approaches to the verification of non-terminating Cyber-Physical Systems (CPSs) usually rely on numerical simulation of the System Under Verification (SUV) model under input scenarios of possibly varying duration, chosen among those satisfying given constraints. Such constraints typically stem from requirements (or assumptions) on the SUV inputs and its operational environment as well as from the enforcement of additional conditions aiming at, e.g., prioritising the (often extremely long) verification activity, by, e.g., focusing on scenarios explicitly exercising selected requirements, or avoiding </i>vacuity</i> in their satisfaction. In this setting, the possibility to efficiently sample at random (with a known distribution, e.g., uniformly) within, or to efficiently enumerate (possibly in a uniformly random order) scenarios among those satisfying all the given constraints is a key enabler for the practical viability of the verification process, e.g., via simulation-based statistical model checking. Unfortunately, in case of non-trivial combinations of constraints, iterative approaches like Markovian random walks in the space of sequences of inputs in general fail in extracting scenarios according to a given distribution (e.g., uniformly), and can be very inefficient to produce at all scenarios that are both legal (with respect to SUV assumptions) and of interest (with respect to the additional constraints). For example, in our case studies, up to 91% of the scenarios generated using such iterative approaches would need to be neglected. In this article, we show how, given a set of constraints on the input scenarios succinctly defined by multiple finite memory monitors, a data structure (scenario generator) can be synthesised, from which any-horizon scenarios satisfying the input constraints can be efficiently extracted by (possibly uniform) random sampling or (randomised) enumeration. Our approach enables seamless support to virtually all simulation-based approaches to CPS verification, ranging from simple random testing to statistical model checking and formal (i.e., exhaustive) verification, when a suitable bound on the horizon or an iterative horizon enlargement strategy is defined, as in the spirit of bounded model checking.

Abstract: A Cyber-Physical System (CPS) comprises physical as well as software subsystems. Simulation-based approaches are typically used to support design and Verification and Validation (V&V) of CPSs in several domains such as: aerospace, defence, automotive, smart grid and healthcare. Accordingly, many simulation-based tools are available to support CPS design. This, on one side, enables designers to choose the toolchain that best suits their needs, on the other side poses huge interoperability challenges when one needs to simulate CPSs whose subsystems have been designed and modelled using different toolchains. To overcome such an interoperability problem, in 2010 the Functional Mock-up Interface (FMI) has been proposed as an open standard to support both Model Exchange (ME) and Co-Simulation (CS) of simulation models created with different toolchains. FMI has been adopted by several modelling and simulation environments. Models adhering to such a standard are called Functional Mock-up Units (FMUs). Indeed FMUs play an essential role in defining complex CPSs through, e.g., the System Structure and Parametrization (SSP) standard. Simulation-based V&V of CPSs typically requires exploring different simulation scenarios (i.e., exogenous input sequences to the CPS under design). Many such scenarios have a shared prefix. Accordingly, to avoid simulating many times such shared prefixes, the simulator state at the end of a shared prefix is saved and then restored and used as a start state for the simulation of the next scenario. In this context, an important FMI feature is the capability to save and restore the internal FMU state on demand. This is crucial to increase efficiency of simulation-based V&V. Unfortunately, the implementation of this feature is not mandatory and it is available only within some commercial software. As a result, the interoperability enabled by the FMI standard cannot be fully exploited for V&V when using open-source simulation environments. This motivates developing such a feature for open-source CPS simulation environments. Accordingly, in this paper, we focus on JModelica, an open-source modelling and simulation environment for CPSs based on an open standard modelling language, namely Modelica. We describe how we have endowed JModelica with our open-source implementation of the FMI 2.0 functions needed to save and restore internal states of FMUs for ME. Furthermore, we present experimental results evaluating, through 934 benchmark models, correctness and efficiency of our extended JModelica. Our experimental results show that simulation-based V&V is, on average, 22 times faster with our get/set functionality than without it.

Abstract: Model-based approaches to safety and efficacy assessment of pharmacological drugs, treatment strategies, or medical devices (In Silico Clinical Trial, ISCT) aim to decrease time and cost for the needed experimentations, reduce animal and human testing, and enable precision medicine. Unfortunately, in presence of non-identifiable models (e.g., reaction networks), parameter estimation is not enough to generate complete populations of Virtual Patient (VPs), i.e., populations guaranteed to show the entire spectrum of model behaviours (phenotypes), thus ensuring representativeness of the trial.We present methods and software based on global search driven by statistical model checking that, starting from a (non-identifiable) quantitative model of the human physiology (plus drugs PK/PD) and suitable biological and medical knowledge elicited from experts, compute a population of VPs whose behaviours are representative of the whole spectrum of phenotypes entailed by the model (completeness) and pairwise distinguishable according to user-provided criteria. This enables full granularity control on the size of the population to employ in an ISCT, guaranteeing representativeness while avoiding over-representation of behaviours.We proved the effectiveness of our algorithm on a non-identifiable ODE-based model of the female Hypothalamic-Pituitary-Gonadal axis, by generating a population of 4 830 264 VPs stratified into 7 levels (at different granularity of behaviours), and assessed its representativeness against 86 retrospective health records from Pfizer, Hannover Medical School and University Hospital of Lausanne. The datasets are respectively covered by our VPs within Average Normalised Mean Absolute Error of 15%, 20%, and 35% (90% of the latter dataset is covered within 20% error).