Integrating Predictive Analytics with RAMS Analysis to Quantify Performance Improvement

Traditional Reliability, Availability, Maintainability and Supportability (RAMS) analyses rely on a mix of work history and the assessment of experienced personnel to estimate probability of failure of components within systems. In this paper we present an improved approach to process delay data from equipment downtime logs into probability density functions that are used in analyses leading to quantification of potential savings. The analytics include data conditioning to classify types of faults and support standard Weibull calculations to classify failure modes to specific blocks representing components within the system. The approach relies on a data model that handles a wide range of codification schemas with plain English interpreters to convert commentary to failure modes which can then be allocated to the blocks. The outcome of this work is the automated assessment of the many problems impacting complex systems of equipment, rolling up component-level issues to the system level and advising on reliability improvements.