Towards a reference ontology for functional knowledge interoperability

Functionality is one of the key aspects of artifact models for design. A function of a device, however, can be captured in different ways in different domains or by different model-authors. Much research on functions has been conducted in the areas of engineering design, functional representation and philosophy, although there are several definitions and notions of functions. We view conceptualization of function is multiplicative in nature: different functions can be captured simultaneously from an objective behavior of an artifact under different teleological contexts of users/designers, or from different viewpoints (perspectives) of a model-author. Such differences become problematic for sharing functional knowledge among engineers. In this article, we attempt to clarify the differences of such perspectives for capturing functions on the basis of the ontological engineering. On the basis of a generalized model of the standard input-output model in the well-known systematic design methodology, we show descriptive categorization of some upper-level types (classes) of functions with references to some definitions of functions in the literature. Such upper-level ontological categories of functions are intended to be used as a reference ontology for functional knowledge interoperability. One of the two usages here is to convert functional models between different functional taxonomies. A functional term in a taxonomy is (ideally) categorized into a generic type defined in the reference ontology. It is widely recognized in the literature that such an upper-level ontology helps automatic `mapping discovery' which is to find similarities between two ontologies and determine which concepts represent similar notion. The reference ontology of function might have such an effect. Another usage of the reference ontology is to integrate fault knowledge into functional knowledge and automatic transformation of FMEA sheets. The designer can describe an integrated model of both functional knowledge and fault knowledge. Based on ontology mappings, automatic transformations of FMEA sheets can be realized. In this article, we discuss the detail of the definitions of the upper-level categories of functions ontologically. Then, we give an overview of usages and effects of the upper-level categories as a reference ontology for functional knowledge interoperability.