UTP Semantics for Shared-State, Concurrent, Context-Sensitive Process Models

Abstract

Abstract-Process Modelling Language (PML) is a notation for describing software development and business processes. It takes the form of a shared-state concurrent imperative language describing tasks as activities that require resources to start and provide resources when they complete. Its syntax covers sequential composition, parallelism, iteration and choice, but without explicit iteration and choice conditions. It is intended to support a range of context-sensitive interpretations, from a rough guide for intended behaviour, to being very prescriptive about the order in which tasks must occur. We are using Unifying Theories of Programming (UTP) to model this range of semantic interpretations, with formal links between them, typically of the nature of a refinement. We address a number of challenges that arise when trying to develop a compositional semantics for PML and its shared-state concurrent underpinnings, most notably in how UTP observations need to distinguish between dynamic state-changes and static context parameters. The formal semantics are intended as the basis for tool support for process analysis, with applications in the healthcare domain, covering such areas as healthcare pathways and software development and certification processes for medical device software. c IEEE 2016 http://doi.ieeecomputersociety

UTPCalc — A Calculator for UTP Predicates

“…So our UTCP theory is based on a non-homogeneous relation with alphabet s, s , ls, ls , g, in, and out. See [6] for details.…”

UTCP: Compositional Semantics for Shared-Variable Concurrency

“…In [7] we had the explicit run form, and this reduced the semantics of a to a, that of a + b to a ∨ b and a b to ab ∨ ba. While this looks cleaner, it has lost too much information, and we cannot compose these further to get correct answers.…”

“…Actions can be combined using sequencing, selection, branching and iteration. We published initial work on a UTP semantics for PML [7], noting that it is essentially the same as the Command language. The semantics we gave in [7] was not compositional, however, and finding a fully compositional semantics was noted for future work.…”