Article Preview
Top1. Introduction
Today's business must meet growing demands regarding quality and quantity of products and services, responsiveness, and cost reduction. To cope with these requirements, the company must have a reliable production system, well maintained by an efficient and inexpensive maintenance system. An efficient and well-organized maintenance service contributes to the consistency of the production system (Brumbach & Clade, 2013). In addition, it will extend the life of industrial equipment and therefore obtain the best overall performance of the company. This maintenance need concerns any business, industry, or service provider.
Since the 1980s, a phase of structuring and standardization of maintenance services has been implemented (Tomlingson, 1998). However, the evolution of markets, globalization, and the emphasis on profit and the company's competitiveness leads to developing new production and maintenance organization concepts. At the same time, the quality aspect plays an essential role in companies' reliability and, precisely, the maintenance function. New information and communication technologies (ICTs) have helped establish and evolve these roles. Thanks to ICTs and the Internet, maintenance and monitoring services can be realized automatically, remotely, and through various distributed information systems. Hence, the emergence of services offered through maintenance architectures, ranging from stand-alone systems to integrated systems where collaboration is essential to any operation (Karray et al., 2010; Emmanouilidis et al., 2011).
Thus, the maintenance stakeholders need to have “the right information in the right format so that the right people are doing the right things at the right time” (Ruschel et al., 2017). Therefore, it has become necessary to integrate all maintenance support systems into a global maintenance management platform.
Integration of applications is not enough to provide maintenance actors with the correct information, which can be used at the right time. Therefore, maintenance platforms must also strengthen the use of knowledge in maintenance by developing the standardization of information and knowledge in terms of understanding, interpretation, and sharing, thus improving semantic interoperability.
Ontological engineering seems to be one of the best ways to respond to these problems since ontologies have well-defined terminologies whose semantics are unambiguous (Guarino, 1998; Zemmouchi-Ghomari, 2021) due to their formal and explicit representation of a shared understanding of domain concepts and relationships.
The most common definition of ontology is Gruber's definition, which describes ontology as a specification of a conceptualization (Gruber, 1993). In more precise terms, an ontology is a hierarchical structure describing the knowledge of a particular domain. It contains domain concepts described by attributes and relationships between them. The interest in ontologies stems from their ability to structure the studied domain's explicit knowledge and deduce other interesting implicit knowledge from them. Ontologies are designed to formally represent a domain of knowledge to gain consensus from the community of domain experts. A domain ontology consists of concepts, interpretation constraints, and procedures and rules required for different use cases.
On the other hand, the industry demands a standardized description of industrial maintenance resources. The objective is to avoid ambiguities and facilitate communication between the various stakeholders, especially in incidents or breakdowns. Indeed, integrating industrial data can be facilitated by core ontologies. Through core ontologies, different datasets can be combined and merged, allowing for the expansion of information clouds without requiring the adaptation of new ontologies. Only the core ontology needs to be modified to model the new data structures according to their semantics. Companies can use industry-specific vocabulary without using a global model covering all industries. Additionally, reasoning on ontological data allows managers to infer new knowledge to assist in decision-making. It is then possible to query the developed ontology automatically; in the same way, databases are queried.