Railway rolling stock fleet predictive maintenance data analytics

Research output: Contribution to journalArticlepeer-review

830 Downloads (Pure)


In this paper predictive maintenance data analytics model is presented within the framework of stochastics point process. The concept of predictive maintenance is supported by data analytics model algorithms development The generalised proportional intensity model (GPIM) is proposed. The model employs analytics, methods and techniques that use asset data, such as condition and loading data or experience, to detect or predict changes in the physical condition of equipment. Condition monitoring of train doors as a critical system is considered. The data arising from monitoring the state of the door systems occur randomly. Numerical example presented of modelling process considers the way in which intermittent failure data of the door is censored by preventive maintenance. Simulation of the GPIM model is used to predict the time to failure of the door system in terms of the
expected cost per unit time. The cost model is inspired by a study of the practices in the railway industry. The use of the GPIM in this paper contributes to a wider shift towards integrating predictive maintenance and service operations in the railway industry.
Original languageEnglish
Number of pages9
JournalInternational Journal of Railway Technology
Publication statusAccepted/In press - 3 Mar 2019


Dive into the research topics of 'Railway rolling stock fleet predictive maintenance data analytics'. Together they form a unique fingerprint.

Cite this