Big Data Driven Reliability Growth for Healthcare Systems

SUMMARY & CONCLUSIONSThis paper presents a repairable Health Care systems development approach that integrates system Reliability Growth with the structured automated collection and manipulation of 'big data'. The approach focuses on the collection of Critical Parameters generated during endurance runs and translating these into system robustness numbers like (Cp/Cpk), system time-to-failure data and failure mode information and Crow-AMSAA (NHPP) graphs. The robustness numbers are used for a daily management dashboard and used as input for actions to drive continuous improvement of the system dependability.During the development process of a medical device, endurance runs were executed on early prototypes in order to optimize the design and to further improve the Reliability aspects of the system prior to shipment to the customer.The chosen approach ensures an automated, repeatable and reproducible data collection, calculation and visualization of big data sets and connecting them to well-known Reliability Engineering strategies like Reliability Growth Analysis. Using open source software for Statistical Computing we proved to be able to bridge the gap between data generated from the endurance runs and data analysis, while preventing manual collecting and handling of big data sets.