A comparative study on data-driven prognostic approaches using fleet knowledge

Recently, Prognostics and Heath Management techniques have been deeply investigated with the aim to reduce life-cycle cost of products and systems. The increasing availability of condition monitoring data in substantial quantities for multitudes of homogeneous products and the need for generic algorithms that are applicable to complex systems motivates the development of new data-driven prognostic approaches. In this paper, two data-driven algorithms, one based on a statistical approach and another based on Neural Network, are discussed and tested for an application case. The analysis of the results has shown that both the considered approaches are characterized by reliable prediction performances on Remaining Useful Life calculation, thus resulting as potential tools for the application of a Condition-Based Maintenance strategy.