An Ontology-Based Approach for Closed-Loop Product Lifecycle Management

The main goal of the Product Lifecycle Management (PLM) is the management of all the data associated to a product during its lifecycle. Lifecycle data is being generated by events and actions (of various lifecycle agents which are humans and/or software systems) and it is distributed along the product's lifecycle phases: Beginning of Life (BOL) including design and manufacturing, Middle of Life (MOL) including usage and maintenance and End of Life (EOL) including recycling, disposal or other options. Closed-Loop PLM extends the meaning of PLM in order to close the loop of the information among the different lifecycle phases. The idea is that information of MOL could be used at the EOL stage to support deciding the most appropriate EOL option (especially to make decision for re-manufacturing and re-use) and combined with the EOL information it could be used as feedback in the BOL for improving the new generations of the product. Several PLM models have been developed utilising various technologies and methods towards providing aspects of the Closed-Loop PLM concept. Ontologies are rapidly becoming popular in various research fields. There is a tendency both in converting existing models into ontology-based models, and in creating new ontology-based models from scratch. The aim of this dissertation is to include the advantages and features provided by the ontologies into PLM models towards achieving Closed-Loop PLM. Hence, an ontology model of a Product Data and Knowledge Management Semantic Object Model for PLM has been developed. The transformation process of the model into an ontology-based one, using Web Ontology Language-Description Logic (OWL-DL), is described in detail. The background and the motives for converting existing PLM models to ontologies are also provided. The new model facilitates several of the OWL-DL capabilities, while maintaining previously achieved characteristics. Furthermore, case studies based on various application scenarios, are presented. These case studies deal with data integration and interoperability problems, in which a significant number of reasoning capabilities is implemented, and highlight the utilisation of the developed model. Moreover, in this work, a generic concept has been developed, tackling the time treatment in PLM models. Time is the only fundamental dimension which exists along the entire life of an artefact and it affects all artefacts and their qualities. Most commonly in PLM models, time is an attribute in parts such as "activities" and "events" or is a separate part of the model ("four dimensional models"). In this work the concept is that time should not be one part of the model, but it should be the basis of the model, and all other elements should be parts of it. Thus, we introduce the "Duration of Time concept". According to this concept all aspects and elements of a model are parts of time. Case studies demonstrate the applicability and the advantages of the concept in comparison to existing methodologies.