Early detection of reliability related problems from two-dimensional warranty data considering labour code priority index

Abstract

Early detection of reliability related problems through the use of sensitive statistical methods, allowing early actions to mitigate potential reliability problems could save a considerable amount of money and product goodwill for large manufacturing companies. The items which are covered under two-dimensional warranty, both age and usage are important and apart from the number of failures, the cost of repair needs to be considered to adequately tackle the problem. In the present paper, we have addressed the above issues and developed an index, known as Labour Code Priority Index (LCPI), based on number of failure at different age-usage windows as well as the cost of repair of the failed items. We have explored various desirable and interesting properties of the index and borrowed the strength of the CUSUM technique applied on the sequentially generated index values to identify early reliability related issues both for a simulated as well as, a real-life synthetic data.