Análise da confiabilidade de uma ilha robotizada operando em linha de montagem automobilística

Abstract

This paper presents an analysis of the failure modes of a robotized island of an automobile company located in the metropolitan region of Curitiba. We collected data from a supervisory system at the company in early 2017, in order to obtain the failure times of the assets that were part of that island. From this data, a survival analysis was performed and found the parametric models that best fit the analyzed failure data sets of each asset and the system. As a result, we found, for the 5 categories employed and for the whole system, the 3-parameter Weibull model, showing that the predominant failure mode of the system and its component assets was that of wear or aging. However, for the ‘Tucker Welding Robots’ category, it was observed that its failure behavior did not follow the same pattern as the others, presenting premature failure mode or infant mortality.