Title: QoS of cloud prognostic system: application to aircraft engines fleet

Authors: Zohra Bouzidi; Labib Sadek Terrissa; Noureddine Zerhouni; Soheyb Ayad

Addresses: Smart Computer Sciences Laboratory, Computer Sciences Department, University of Biskra, P.O. Box 145 RP – Biskra 07000, Algeria ' Smart Computer Sciences Laboratory, Computer Sciences Department, University of Biskra, P.O. Box 145 RP – Biskra 07000, Algeria ' AS2M Department, Femto-ST Institute, 15B Avenue of the Montboucons, 25030 Besançon cedex, France ' Smart Computer Sciences Laboratory, Computer Sciences Department, University of Biskra, P.O. Box 145 RP – Biskra 07000, Algeria

Abstract: Recently, prognostics and health management (PHM) solutions are increasingly implemented in order to complete maintenance activities. The prognostic process in industrial maintenance is the main step to predict failures before they occur by determining the remaining useful life (RUL) of the equipment. However, it also poses challenges such as reliability, availability, infrastructure and physics servers. To address these challenges, this paper investigates a cloud-based prognostic system of an aircraft engine based on artificial intelligence methods. We design and implement an architecture that defines an approach that is prognostic as a service (Prognostic aaS) using a data-driven approach. This approach will provide a suitable and efficient PHM solution as a service via internet, on the demand of a client, in accordance with a service level agreement (SLA) contract drawn up in advance to ensure a better quality of service and pay this service per use (pay as you go). We estimated the RUL of aircraft engines fleet by implementing three techniques. Next, we studied the performance of this system; the efficient method was concluded. In addition, we discussed the quality of service (QoS) for the cloud prognostic application according to the factors of quality. [Received: 19 May 2018; Revised: 10 August 2018; Revised: 31 August 2018; Revised: 21 March 2019; Accepted: 28 March 2019]

Keywords: prognostics and health management; PHM; remaining useful life; RUL; prognostic as a service; Prognostic aaS; cloud computing; artificial intelligence; measure performance; quality of service; QoS.

DOI: 10.1504/EJIE.2020.105080

European Journal of Industrial Engineering, 2020 Vol.14 No.1, pp.34 - 57

Published online: 12 Feb 2020 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article