Fleet Knowledge for Prognostics and Health Management – Identifying Fleet Dimensions and Characteristics for the Categorization of Fleets

Carolin Wagner and Bernd Hellingrath
Submission Type: 
Full Paper
phmc_17_019.pdf292.89 KBSeptember 6, 2017 - 2:59am

Current prognostics and health management approaches are often not able to meet expectations due to their limited ability to accurately detect abnormal machine conditions, identify failures and estimate the remaining useful life. This is in many cases attributed to the lack of real data and knowledge about the component or machine under consideration. Instead, experimental data is often used for algorithm training, which is not able to reflect the complexity of real-world systems. To improve prognostics and health management approaches condition data from fleets of machines rather than single units can be taken into consideration. Therefor machine conditions are assessed against situations encountered by machines in the same fleet and knowledge is transferred to allow algorithms to intelligently learn and improve their capabilities.
Several approaches have recently been presented in the literature, which make use of the fleet knowledge for condition-based maintenance. These approaches are designed for specific fleet compositions and characteristics. Therefore, in order to incorporate fleet knowledge into diagnostic and prognostic approaches the fleet under consideration and resulting requirements have to be analyzed. With this information, it is possible to determine whether fleet-based approaches are applicable in general to the specific case as well as facilitate the selection of a suitable fleet-based approach. Three types of fleets are distinguished in the literature, namely identical, homogeneous and heterogeneous fleets. This distinction makes reference to the structural dimension of fleets. For fleet-based approaches, however additional dimensions should be taken into account. These include among others the operating condition in the fleet (e.g. identical, different, or dynamically changing) and the type of available data (e.g. sensor reading, context data, textual description). This paper aims at identifying and analyzing different dimensions and respective characteristics of fleets to be considered in the context of prognostics and health management. The results are synthesized in a classification structure to support the categorization of fleets.

Publication Year: 
Publication Volume: 
Publication Control Number: 
Page Count: 
Submission Topic Areas: 
CBM and informed logistics
Data-driven methods for fault detection, diagnosis, and prognosis
Submitted by: 

follow us

PHM Society on Facebook Follow PHM Society on Twitter PHM Society on LinkedIn PHM Society RSS News Feed