Tommaso Dreossi

Efficient Guiding Strategies for Testing of Temporal Properties of Hybrid Systems

Techniques for testing cyberphysical systems (CPS) currently use a combination of automatic directed test generation and random testing to find undesirable behaviors. Existing techniques can fail to efficiently identify bugs because they do not adequately explore the space of system behaviors. In this paper, we present an approach that uses the rapidly exploring random trees (RRT) technique to explore the state-space of a CPS. Given a Signal Temporal Logic (STL) requirement, the RRT algorithm uses two quantities to guide the search: The first is a robustness metric that quantifies the degree of satisfaction of the STL requirement by simulation traces. The second is a metric for measuring coverage for a dense state-space, known as the star discrepancy measure. We show that our approach scales to industrial-scale CPSs by demonstrating its efficacy on an automotive powertrain control system.

Parameter synthesis for polynomial biological models

Parameter determination is an important task in the development of biological models. In this paper we consider parametric polynomial dynamical systems and address the following parameter synthesis problem: find a set of parameter values so that the resulting system satisfies a desired property. Our synthesis technique exploits the Bernstein polynomial representation to solve the synthesis problem using linear programming. We apply our framework to two case studies involving epidemic models.

Compositional Falsification of Cyber-Physical Systems with Machine Learning Components

Cyber-physical systems (CPS), such as automotive systems, are starting to include sophisticated machine learning (ML) components. Their correctness, therefore, depends on properties of the inner ML modules. While learning algorithms aim to generalize from examples, they are only as good as the examples provided, and recent efforts have shown that they can produce inconsistent output under small adversarial perturbations. This raises the question: can the output from learning components can lead to a failure of the entire CPS? In this work, we address this question by formulating it as a problem of falsifying signal temporal logic (STL) specifications for CPS with ML components. We propose a compositional falsification framework where a temporal logic falsifier and a machine learning analyzer cooperate with the aim of finding falsifying executions of the considered model. The efficacy of the proposed technique is shown on an automatic emergency braking system model with a perception component based on deep neural networks.

Parallelotope Bundles for Polynomial Reachability

In this work we present parallelotope bundles, i.e., sets of parallelotopes for a symbolic representation of polytopes. We define a compact representation of these objects and show that any polytope can be canonically expressed by a bundle. We propose efficient algorithms for the manipulation of bundles. Among these, we define techniques for computing tight over-approximations of polynomial transformations. We apply our framework, in combination with the Bernstein technique, to the reachability problem for polynomial dynamical systems. The accuracy and scalability of our approach are validated on a number of case studies.

Parameter Synthesis Through Temporal Logic Specifications

Parameters are often used to tune mathematical models and capture nondeterminism and uncertainty in physical and engineering systems. This paper is concerned with parametric nonlinear dynamical systems and the problem of determining the parameter values that are consistent with some expected properties. In our previous works, we proposed a parameter synthesis algorithm limited to safety properties and demonstrated its applications for biological systems. Here we consider more general properties specified by a fragment of STL (Signal Temporal Logic), which allows us to deal with complex behavioral patterns that biological processes exhibit. We propose an algorithm for parameter synthesis w.r.t. a property specified using the considered logic. It exploits reachable set computations and forward refinements. We instantiate our algorithm in the case of polynomial dynamical systems exploiting Bernstein coefficients and we illustrate it on an epidemic model.

Combining Model Checking and Runtime Verification for Safe Robotics

A major challenge towards large scale deployment of autonomous mobile robots is to program them with formal guarantees and high assurance of correct operation. To this end, we present a framework for building safe robots. Our approach for validating the end-to-end correctness of robotics system consists of two parts: (1) a high-level programming language for implementing and systematically testing the reactive robotics software via model checking; (2) a signal temporal logic (STL) based online monitoring system to ensure that the assumptions about the low-level controllers (discrete models) used during model checking hold at runtime. Combining model checking with runtime verification helps us bridge the gap between software verification (discrete) that makes assumptions about the low-level controllers and the physical world, and the actual execution of the software on a real robotic platform in the physical world. To demonstrate the efficacy of our approach, we build a safe adaptive surveillance system and present software-in-the-loop simulations of the application.

Falsifying Oscillation Properties of Parametric Biological Models

We propose an approach to falsification of oscillation properties of parametric biological models, based on the recently developed techniques for testing continuous and hybrid systems. In this approach, an oscillation property can be specified using a hybrid automaton, which is then used to guide the exploration in the state and input spaces to search for the behaviors that do not satisfy the property. We illustrate the approach with the Laub-Loomis model for spontaneous oscillations during the aggregation stage of Dictyostelium.

Hybrid Automata and e-Analysis on a Neural Oscillator

In this paper we propose a hybrid model of a neural oscillator, obtained by partially discretizing a well-known continuous model. Our construction points out that in this case the standard techniques, based on replacing sigmoids with step functions, is not satisfactory. Then, we study the hybrid model through both symbolic methods and approximation techniques. This last analysis, in particular, allows us to show the differences between the considered approximation approaches. Finally, we focus on approximations via e-semantics, proving how these can be computed in practice.

Parameter Synthesis Using Parallelotopic Enclosure and Applications to Epidemic Models

We consider the problem of refining a parameter set to ensure that the behaviors of a dynamical system satisfy a given property. The dynamics are defined through parametric polynomial difference equations and their Bernstein representations are exploited to enclose reachable sets into parallelotopes. This allows us to achieve more accurate reachable set approximations with respect to previous works based on axis-aligned boxes. Moreover, we introduce a symbolical precomputation that leads to a significant improvement on time performances. Finally, we apply our framework to some epidemic models verifying the strength of the proposed method.

Sapo: Reachability Computation and Parameter Synthesis of Polynomial Dynamical Systems

Sapo is a tool for the formal analysis of polynomial dynamical systems. Its main features are 1) Reachability computation, i.e., the calculation of the set of states reachable from a set of initial conditions, and 2) Parameter synthesis, i.e., the refinement of a set of parameters so that the system satisfies a given specification. Sapo can represent reachable sets as unions of boxes, parallelotopes, or parallelotope bundles (symbolic representation of polytopes). Sets of parameters are represented with polytopes while specifications are formalized as Signal Temporal Logic (STL) formulas.