Estimating the warranty returns and proving root causes using statistical analysis of archived parameters measurements for an automotive mechatronic device

Abstract

In the mass production of the automotive industry, there are many cases in which a problem is not detected in the production processes of the supplier, but after delivery, in the customer's assembly line. As a result, it is necessary to know both root causes to prevent them, but especially, if it cannot be detected even in the customer's processes, to estimate the probability of occurrence in the warranty period. The paper's aim is to present the analysis of the measurements of the mechatronic devices indirect parameters linked with the effect of the failure and by statistical correlation to prove the root causes. A study case will be used to understand the failure first occurrence and the duration of its effect, as the symptom description as well, in that way it will support the estimation of the warranty return number of parts, eventually, the accounting department can establish the provisions based on the output data.