Root cause analysis-based approach for improving preventive/corrective maintenance of an automated prescription-filling system

Pharmacy automation is a new, but rapidly developing industry. Even though the products and services developed are relatively new, it is extremely important that the reliability and availability of the products is high. The heterogeneous nature of customers and their requirements makes the machines at these pharmacies, different entities. This heterogeneity makes way for different forms of faults that essentially lead to a common error. In order to improve the reliability of a sub-system, thereby a system, it is necessary to identify the root causes of these problems precisely so that the corrective action taken can be effective. In the absence of clearly defined field feedback data, estimation of the importance of the root causes becomes difficult. This research endeavor is aimed at identifying the usage of simpler approaches to estimate the importance of root causes that can lead to improved preventive/corrective maintenance.;As technology increases at a rapid pace, new products are developed and are sent to market at the same rate. The lure of increased profits and market share make it necessary for technology firms to develop products and sell them before their competitors. Most technological firms do this at the expense of increased and thorough testing. Decreased testing releases relatively premature products into the market which need constant upgrades to stay in-line with the currently developed products.;Given that the products developed are new, the data collection process is not exactly a refined one. The depth of usability of field feedback data depends on the depth to which data is collected. Collection of statistical failure data, although useful, does not provide a clear insight to the various errors modes. It now becomes necessary to know how this data is distributed among the various failure modes so that erroneous corrective maintenance can be avoided. Failure to do so, leads to degraded system reliability and thereby, low customer satisfaction levels.;In the absence of complex and detailed statistical and knowledge data, other measures have to be used to estimate the critical modes of failure and the corresponding root causes. This research work aims at minimizing the errors of a particular sub-system in a automated prescription-filling system. The main errors influencing system performance are Vial Feed errors, Partial Ejected Vial., Vial Double Feed, Barcode Misreads, Gantry errors. Of these, "Barcode Misreads" accounted for approximately 40.6% of all errors in the system. The components of the sub-system affected by Misread errors were identified and a fault tree was constructed. Since the probability of each root cause could not be determined sufficiently using the available field-feedback data, the concept of Failure Mode Effect and Criticality Analysis (FMECA) was used to determine the initial criticality of the factors. The criticality for factors was then confirmed by means of an experiment. The root causes were then prioritized and a dynamic Diagnostic Decision Tree (DDT) was constructed which was then used for preventive/corrective maintenance plans.