Automated bicycle-sharing systems (bss) are a prominent example of reconfigurable cyber-physical systems for which the locality and connectivity of their elements are central to the way in which they operate. These features motivate us to study bss from the perspective of Actor-Network Theory-a framework for modelling cyber-physicalsystem protocols in which systems are networks of actors that are no longer limited to programs but can also include humans and physical artefacts. In order to support logical reasoning about information-flow properties that occur in bss, we use a logical framework that we have recently developed for actor networks, which results from a two-stage hybridization process. The first stage corresponds to a logic that captures the locality and connectivity of actors in a given configuration of the network; the second stage corresponds to a logic of possible interactions between actors, which captures the dynamics of the system in terms of network reconfigurations. To illustrate the properties that can be checked using this framework, we provide an actor-network specification of a particular bss, and use a recently developed tool for hybridized logics to highlight and correct an information-flow vulnerability of the system.

No abstract

Modern software systems are increasingly exhibiting dynamicreconfiguration features analogous to naturally occurring phenomena where the architecture of a complex changes dynamically, at run time, on account of interactions between its components. This has led to a renewed interest in modal logics for formal system development, building on the intuitive idea that system configurations can be regarded as local models of a Kripke structure, while reconfigurations are captured by accessibility relations. We contribute to this line of research by advancing a modal logic with varying quantification domains that employs typed modalities and dedicated modal operators to specify and reason about a new generation of Kripke structures, called dynamic networks of interactions, that account for the context of a system's dynamics, identifying which actants have triggered a reconfiguration and what are its outcomes. To illustrate the expressiveness of the formalism, we provide a specification of the biological process of membrane budding, which we then analyse using a sound and complete proof-by-translation method that links dynamic networks of interactions with partial first-order logic.