Background and Problem Description Sample Clauses

Background and Problem Description. Within the manufacturing industry high product quality is an essential part of the brand. Warranty payments are used as a common way of assessing product quality and the automotive industry spends huge amount of money on warranty claims. For Scania itself, the product obligations, mostly related to vehicle warranties, amounted to ~SEK 1480 m in the year 2013 [5]. Therefore, precise warranty forecasts are of immense importance for the company management. After the delivery of the vehicle, its warranty period starts and with the passage of time, warranty claims are gradually reported. Warranty claims are the claims from customers for replacement or repair of, or money compensation for malfunctioning or non-functioning of any part of the vehicle, as specified and agreed in the warranty guidelines. Companies usually maintain a warranty database to collect the details of all the claims reported during the warranty period of the vehicle. There is a vast literature available on modeling of warranty claims. Empirical models obtained by the analysis of past warranty details from technologically similar products usually form the basis of commonly employed warranty forecasting techniques. These models usually incorporate (or allow for) modifications in the methodology to adjust according to the design and technology changes in the product. The originative contribution in this regard is the publication by Xxxxxxxxxxx et al. [2] where they have proposed a Poisson distribution model for prediction of warranty claims in the time domain. Since, then several papers have been published to predict the warranty claims for the remaining period based on the claims reported so far and the empirical model based on past warranty claims of similar products. Majeske et al. [4] proposes a mixture model for warranty data including the parameters for product field quality performance, manufacturing and assembly process, and dealer preparation process. Xxxxxxx et al. [3] proposes a model of two dimensional automotive warranty characterized by time in service and vehicle usage mileage. A related aspect here is the warranty forecasting for recently launched products that have very small history of claim records of their own. Since the technology is constantly evolving and also because of new environmental and safety regulations, it is inevitable to face frequent launches of new products. And it becomes vitally important as well for the manufacturers to make good forecast for the newly lau...