The role and importance of warranty has changed significantly over the last two decades. Currently all most all products are sold with some form of warranty attached. Warranty is of importance to both manufacturers and consumers. From the consumers point of view, warranty provides some information about product reliability and quality and acts as an insurance in the event of early failure of the product. From the manufacturer's view point, it protects the manufacturer from unreasonable claims from the consumers and has been used as an effective signal to communicate product reliability for the marketing of the product.
Selling a product with warranty results in additional costs to the manufacturer due to the servicing of warranty. This cost can be reduced by improving the reliability (through better design) and quality (through better manufacturing actions -- quality control activities).
In general, items are multi-components where one or more components are critical in the sense that the item fails whenever one of these components fails. One way of improving reliability of the item is through redundancy where a module of replicated critical components is used instead of a single critical component. Three different types of redundancy have been used in this context. They are -- Hot, Cold and Warm standby. The improvement in the reliability is achieved at the expense of increased manufacturing cost per unit due to the replication. Only if the increase in the manufacturing cost per unit is less than the resulting savings in the expected warranty cost per unit, building in redundancy is worthwhile. We deal with this problem when all components are statistically similar and conform to the design specification and carry out a comparative study between no redundancy and the three different types of redundancy (Hot, Cold and Warm standby). We consider both the Free Replacement Warranty and the Pro-rata Warranty.
In real life, due to manufacturing variability, a fraction of the components fail to meet the design specification. As a result, the reliability improvement through redundancy is affected by this quality uncertainty. In this situation, it is not clear whether improving reliability through redundancy is worthwhile or not. Testing to weed out nonconforming components (or modules, if redundancy is used) adds an extra dimension to the decision problem as there is the option of testing at either component or module level. An additional decision problem relates to the duration of the testing and this depends on the nature of nonconformance. We classify two types (Type I and Type II) of nonconformance. A Type I nonconforming item is one which is not functional when it is put into use. This type of item can be easily identified through testing for a very short time. A Type II nonconforming item is functional when it is put into use but its performance is inferior to that of a conforming item. In other words, its mean time to failure (MTTF) is shorter than that of a conforming item. This type of item can not be easily identified and can be weeded out through life testing. We first deal with redundancy decisions where the nonconforming components are of Type I and examine testing at component and module levels. We consider 11 different options including the option of no testing. Later on, we deal with redundancy decisions where nonconforming components are of Type II and life testing is used to weed out such components. ……………………………