Graph-based reasoning in collaborative knowledge management for industrial maintenance

Capitalization and sharing of lessons learned play an essential role in managing the activities of industrial systems. This is particularly the case for the maintenance management, especially for distributed systems often associated with collaborative decision-making systems. Our contribution focuses on the formalization of the expert knowledge required for maintenance actors that will easily engage support tools to accomplish their missions in collaborative frameworks. To do this, we use the conceptual graphs formalism with their reasoning operations for the comparison and integration of several conceptual graph rules corresponding to different viewpoint of experts. The proposed approach is applied to a case study focusing on the maintenance management of a rotary machinery system. * Corresponding author. Tel.: +33 624302337/562442718. E-mail addresses: [email protected] (B. Kamsu-Foguem), [email protected] (D. Noyes). improvement that emphasizes the ongoing monitoring and verification of the root causes of problems in the monitored system to eliminate repetitive failures and recurring problems [14]. The approach proposed in this work is based on the EF exploitation in collaborative decision-making situations. The goal is to improve collaboration among maintenance actors by the deployment of knowledge engineering tools to effectively share experiences, generating knowledge to better resolve problems. The approach of dynamic capitalization of knowledge supports knowledge validation process among collaborating actors, hence increasing the reliability of capitalized knowledge [15]. Problem solving will include knowledge reasoning based on the conceptual graphs (CG) formalism [16]. The choice of knowledge representation by CG should enable a better understanding of critical situations and provide assistance to the appropriate decision-making in order to anticipate them [17]. The expected result is a better use of knowledge and skills distributed among different experts: (1) strengthening the collective knowledge with the promotion of access by the collaborative actors to relevant information and the enhancement of plans for the maintenance of the target system, (2) facilitating the sustainable management of experience learning and providing support to the modeling and improvement of the quality of knowledge sharing within the collaborative organization. The paper is structured as followed. Section 2 exposes the proposed methodology and its main components concerning knowledge engineering. Section 3 presents the cognitive experience feedback approach applied to industrial maintenance management. Section 4 presents the conceptual graphs operations used to implement the modeling of expert rules in collaborative decision-making processes. An illustrative application example for the maintenance management of a rotary machinery system from the railway field is exposed in Section 5. Finally, Section 6 concludes and discusses future challenges. 2. Collaboration in maintenance activities: situation and