Collective approach for modeling complex system failures

Abstract

This paper defines the functional requirements of a computerized decision support system (DSS) necessary to serve as a working tool in assisting the reliability engineer in equipment repair/replacement management of material handling equipment. It incorporates the knowledge required to build analytical models of complex material handling systems using known techniques of system modeling. The system can be used to guide the data analysis and statistical modeling for both repairable and nonrepairable systems. This project aims at bridging the existing gap between the reliability theory and practical applications in the field of reliability engineering. A prototype of this computerized system is being developed (to be tested at research sponsors locations). The nonhomogenous Poisson process (NHPP), renewal process, Prentice, Williams and Peterson (PWP) model and proportional hazards models are implemented accordingly in this project. The complete knowledge base addresses both parametric and nonparametric methods of reliability estimation. An additional focus of this project involves research on the left-truncated power law intensity model for cases where the early lifetime data are missing. The software is being developed to support each of these models. Case studies are conducted based upon the failure data of industrial forklift trucks. These analyses are being utilized to verify the design of the DSS. In the presentation, we present the organization of the DSS and present one of the case studies.