Information modelling for the manufacturing system life cycle

Abstract: This thesis deals with information modelling within the scope of the manufacturing system life cycle, i.e. the development phase and the operation phase. Information modelling defines and structures information that needs to be managed, and is thereby an important step towards realising efficient information management throughout the manufacturing system life cycle.The research goal of this work was to find a modelling approach that simplifies management and integration of manufacturing system information, both within and between the development and operation phases. The starting point was an assumption that information integration requires a common information model for the manufacturing system life cycle. The approach was to evaluate the usefulness of the STEP standards AP214 and AP239 (PLCS) regarding how they meet the information requirements. Case studies within the automotive industry were carried out for gathering test data.Modelling experience showed that PLCS has the most suitable scope since it can represent the manufacturing system from a life cycle perspective. However, the generic character of PLCS introduced other issues, such as how to ensure consistent instantiation. Further guidance is needed regarding how to use PLCS for representing domain-specific objects such as machining centres.As a response to the inconsistency issue, a concept model of a machining centre was developed to guide the instantiation of PLCS. However, it was found that there are multiple ways to translate the concept model to PLCS depending on viewpoint. Moreover, the characteristics of information management within the operation phase were found to be notably different compared to characteristics of the development phase. For these reasons, it is discussed whether or not a common modelling format for the whole manufacturing system life cycle is appropriate or even realisable.From a practical viewpoint, it is concluded to be both inevitable and necessary to find appropriate delimitations and interfaces between complementary information models. A promising step towards information integration is to classify the information concepts of different models according to terms defined in concept models.