Non-Monotonic Reasoning For Requirements Engineering: State Diagrams driven by Plausible Logic

Abstract

We extend the state diagrams used for dynamic modelling in object-oriented analysis and design. We suggest that the events which label the state transitions be replaced with plausible logic expressions. The result is a very effective descriptive and declarative mechanism for specifying requirements that can be applied to requirements engineering of robotic and embedded systems. The declarative model can automatically be translated and requirements are traceable to implementation and validation, minimising faults from the perspective of software engineering. We compare our approach with Petri Nets and Behavior Trees using the well-known example of the one-minute microwave oven.