The purpose of this paper is to model the occurrences of successive failure types and times to failure of the two repairable machine systems.
Historical data on failure types and time to failures of the given machine systems (4 nos) were gainfully used. Second order time homogeneous Markov Chain models were used to characterize the occurrences of the two broad failure type, namely, mechanical and electrical, after having found that the occurrences of failure types were dependent. Second order time homogeneous Markov Chain with Bivariate Distribution function (M2BVD) was used to model the times between successive failures {Tn, n≥1} for each machine system.
It is possible to apply the theoretical framework of Markov chain models to the accumulated data on failure types and failure times of any repairable system, which provide a wealth of information on the systems and are often left unused.
The framework used in the study can be improved to accommodate multiple failure types and failure times of any repairable system to the extent that a more accurate prediction of these two variables and a better estimate of the system reliability are available.
The models for failure types and failure times of the given machine systems would be of immense use to the maintenance crew for predicting the future failure types and failure times of any given system and subsequently organizing and fine‐tuning their state of preparedness.
