Thomas Paviot

Improving the interoperability of industrial information systems with description logic-based models-The state of the art

Semantic technologies that have arisen with web development have brought out new tools, concepts, and methodologies which are increasingly employed in Product Lifecycle Management (PLM) applications. This paper proposes a literature review of papers related to ontologies in the area of product lifecycle management. However, it only focuses on inference ontologies, i.e. ontologies that enable reasoning, for instance, models expressed in the Web Ontology Language (OWL). The goals of this paper are to explore the field of such applications, to figure out the advantages of inference ontologies in a PLM context and to synthesize major existing inference models in terms of methodology and structuration. Finally, this paper proposes several research perspectives.

A PLCS framework for PDM / ERP interoperability

Wide diffusion of methodologies and software relevant to product lifecycle management (PLM) in industrial companies faces heterogeneity of information technology (IT) systems. Especially, the lack of interoperability between product data management (PDM) systems, that drive virtual product development, and enterprise resource planning (ERP), which manages real product, cannot lead to a global description of the product development process. We demonstrate that a mediator approach is pertinent for the coordination of these two systems. The use of open standards, and more precisely STEP Application Protocol 239, known as product lifecycle support (PLCS), allows overcoming issues related to semantic part of this interoperability. The last part of the study focuses on implementation issues, in order to develop an interoperability framework based upon a consistent product data model, whether the viewpoint is from the design or logistics businesses.

SWRL as a rule language for ontology-based models in power plant design

Current standard models have improved interoperability in Product Lifecycle Management. But the increase of constrains surrounding information management in an industrial context, and notably for power plant design, like the increase of data, actors, systems and the need of automation and collaboration leads to an increasing use of ontology-based models. However, the design of power plants presents a specific issue, which is rules expression and validation. This need of rules is generated by safety, interoperability or collaboration constrains. The Semantic Web Rule Language (SWRL) is an ontological language based on OWL-DL that allows in addition rules expression. This paper presents SWRL rules and their industrial applications but also limits of this language in PLM applications. It proposes in consequence other rules languages.

5 root concepts for a meta-ontology to model product along its whole lifecycle*

PLM approaches are based on the definition of a common and shared meta-model. Defining these meta-model, especially in the case of ontology-based models, is often a manual a difficult task. A framework for modeling meta-ontology. The framework is based on five root concepts : product, process, resource, rule and business. These concepts are disjoints, high level, meaningful, and generic. We compare them to concepts of existing models and propose an application scenario to validate the proposed framework.

A generic multiCAD/multiPDM interoperability framework

Many Small and Medium Enterprises (SME) work as subcontractors (or co-contractors) for several clients for the design of mechanical components. During the design process, they must use a variety of Computer Aided Design (CAD) softwares and connect all the Product Data Management (PDM) systems of their customers. After defining the specific needs of these companies, we show that the available commercial CAD/PDM integrations, as well as the current literature, are inappropriate for a multiCAD/multiPDM collaborative design. We first define the few simple processes required to ensure an efficient collaboration. Then, the instantiation of these processes in our CAD/PDM integration can be split into four points: the general definition of a CAD product structure tree and its associated model, the conversion algorithm of the product structure to an Engineering Bill of Material (EBOM), the creation of an Unified Modelling Language (UML) data model, an implementation based upon Component Object Model (COM) and Service Oriented Architectures (SOA) technologies. We conclude by presenting the results obtained from the demonstrator developed.

Expressing formal rules within ontology-based models using SWRL: an application to the nuclear industry

Ontology-based models propose a new modelling paradigm for the industry. In this study, we explore the maturity of ontology-based models for industrial applications from a special point of view: the execution of business rules. Indeed, expressing and implementing rules nowadays is a crucial issue. We therefore propose an overview of the use of OWL-DL/SWRL approaches in the industry, as well as a critical overview of the execution of SWRL rules on an industrial use case in the nuclear industry.

A transformation model to express business rules from natural language to formal execution: an application to nuclear power plant.

Abstract Expressing and implementing business rules is a way to enhance product or business models efficiency. However, business rules are made of, by and for the business. Because information system experts do not master business issues and business experts do not master formal languages and computational issues, expressing business rules within an IS system is a challenge. To overcome the semantic part of this challenge, we propose a semantic transformation framework in four steps, from natural language assertions to formal rules. This transformation model uses semantic web technologies. The application domain is the nuclear industry.

A concept-based approach to modelling shared ontology-based models for industrial applications.

Abstract Ontologies have emerged as a new modelling paradigm in industry during the last decade. They propose new design paradigms for information modelling that engineers must be aware of and manage, whenever they aim to design consistent models. In this paper, guidelines are proposed as well as an industrial illustration in order to help modellers to anticipate and well manage the specificities of ontology modelling. An illustration based on family modelling and an industrial example about classification of parts based on their basic shapes are proposed.

A Modeling Language for 3D Process Plant Layout Representation, Exchange and Visualization

In the nuclear industry, achieving Long Term Data Preservation is a requirement for nuclear power plants to be safely built, operated over five or six decades and retired. Among them, CAD data suffers from some strong dependencies on the software vendors and its data model thus leading to a possible weakness in the preservation workflow. This paper presents a modeling language, suitable for the 3D representation of a process plant layout, based upon a procedural Constructive Solide Geometry (CSG) approach. The language execution, as well as the layout rendering and exchange, are experimented using a platform independent implementation, based on free software and open standards.

Automated business rules and requirements to enrich product-centric information

Abstract Current PLM or BIM based information systems suffer from a lack of checking components for business rules. One reason is the misunderstanding of the role and nature of business rules, and how they should be treated in a product-centric information system. This paper intends to provide both a process and a related model to build such a component and enrich future systems. Rules and requirements process management enables the unambiguous formalization of implicit knowledge contained in business rules, generally expressed in easily understandable language, and leads to the formal expression of requirements. In this paper, the requirements are considered a consequence of the application of a business rule. A conceptual model is then introduced, called DALTON (DAta Linked Through Occurrences Network), which supports this process. In this ontology, concepts and product data, coming for instance from an existing product database, are represented using instances and occurrences, connected together with triples built from business rules and requirements according to previous management processes. An experiment involving a set of SWRL rules is conducted in the Protege environment that validates the model and the process.