A Product Feature Evolution Validation Model for Engineering Change Management

Abstract

Product design integrates several disciplines in a concurrent engineering (CE) environment. Each one of these disciplines has a specific point of view on the product being developed. While each discipline exerts its own expertise and methods on the definition of the product and its related processes, information must remain consistent for all disciplines and through the evolution of the product definition. This paper proposes a product feature evolution validation (PFEV) model that aims at controlling the information flow needed to support a product definition evolution (PDE) while insuring its validation by all disciplines involved. The model applies both to the product design and modification phases, i.e., before and after releasing its definition. The PFEV model thus supports CE and enables managing the product feature evolution throughout the product life cycle. The PFEV model defines an exchange protocol between the disciplines in order to preserve the consistency of the numerical model, which includes the complete numerical information characterizing the product. The model addresses two qualities of an information system: dispatching relevant PDE information to appropriate disciplines and providing this information according to specific views. This is achieved by centralizing the product numerical model and by exploiting the product’s features rather than managing product model as black boxes. Links between features are formalized in a shared product features table that is used to dynamically identify all disciplines impacted by a product feature evolution (PFE). A PFE is also characterized by its potential impact, detrimental or beneficial, on every discipline previously identified as impacted. In the case of a detrimental impact, the discipline is asked to validate the evolution. If the impact is beneficial, the discipline is simply notified about the evolution. Specific views are generated for the impacted disciplines based on feature filtering and adaptation mechanisms.