After new products are launched to the market, many companies will continue the effort to improve product reliability and reduce warranty and life time ownership cost. Vehicles in automotive industry are the typical examples. First, same model of the vehicle will be produced for many years, and cust...
After new products are launched to the market, many companies will continue the effort to improve product reliability and reduce warranty and life time ownership cost. Vehicles in automotive industry are the typical examples. First, same model of the vehicle will be produced for many years, and customers expect that the issues found in the field will be fixed or mitigated in new model years. Meeting this customer expectation is important for vehicle's long term success. Secondly, the new vehicle models are being developed in a much faster pace today than ever before. It is an increasing challenge to sufficiently test and validate the reliability of new vehicle models and their components, especially when the reliability requirements are extremely high. However, timely releasing new products is crucial for companies staying competitive in the market. On the other hand, technology has advanced to allow automotive companies to collect field data much easier. The field data reflects the product reliability performance in the real use and environment. The use of this data makes the effort of continuing reliability improvement necessary and possible. A reliability growth model reflecting field reliability performance and the effect of improvement can systematically and effectively facilitate this effort. Many reliability growth models have been proposed and applied using testing time as an indicator to reflect reliability growth in product development cycle, such as Duane model [1] and Crow-AMSAA model [2]. These models are very popular in application. Sun, Luck, and Mizumachi [3] incorporated the allowable calendar time between tests in their dual-time domain growth model to reflect the impact of allowable time to the reliability growth during product development. The “testing time” in these models generally is not directly applicable to the field reliability growth. Even though the time of customers using their vehicles allows more chances to expose the weakness of the product, the calendar time is more relevant. After product is released to the market, its reliability will manifest itself with customer's use of the product. The longer the customers drive their vehicles or use their products, the easier the age related issues would show up. Once the lessons are learned from the field, companies would need time to report issues from field operation, collect data, investigate the root cause for the failure mode, figure out and validate the fixes, and implement solutions. The improved reliability will be reflected in the vehicles and other products built afterwards. Therefore, the calendar time is a better, and also a simple indicator, of both chance to expose field reliability issues and potential opportunity to improve reliability. In this paper, a new field reliability growth model is proposed along with a modified Crow-AMSAA model. They are applied to a number of vehicle brands and models, and the results are discussed and compared. It shows the new model generally fits the field data very well while modified Crow-AMSAA model can be useful for special applications. Two models are compared with each other. The parameter estimate and statistical inference are presented with an example. The management aspect of the field reliability growth is very important and is therefore also discussed.
After new products are launched to the market, many companies will continue the effort to improve product reliability and reduce warranty and life time ownership cost. Vehicles in automotive industry are the typical examples. First, same model of the vehicle will be produced for many years, and customers expect that the issues found in the field will be fixed or mitigated in new model years. Meeting this customer expectation is important for vehicle's long term success. Secondly, the new vehicle models are being developed in a much faster pace today than ever before. It is an increasing challenge to sufficiently test and validate the reliability of new vehicle models and their components, especially when the reliability requirements are extremely high. However, timely releasing new products is crucial for companies staying competitive in the market. On the other hand, technology has advanced to allow automotive companies to collect field data much easier. The field data reflects the product reliability performance in the real use and environment. The use of this data makes the effort of continuing reliability improvement necessary and possible. A reliability growth model reflecting field reliability performance and the effect of improvement can systematically and effectively facilitate this effort. Many reliability growth models have been proposed and applied using testing time as an indicator to reflect reliability growth in product development cycle, such as Duane model [1] and Crow-AMSAA model [2]. These models are very popular in application. Sun, Luck, and Mizumachi [3] incorporated the allowable calendar time between tests in their dual-time domain growth model to reflect the impact of allowable time to the reliability growth during product development. The “testing time” in these models generally is not directly applicable to the field reliability growth. Even though the time of customers using their vehicles allows more chances to expose the weakness of the product, the calendar time is more relevant. After product is released to the market, its reliability will manifest itself with customer's use of the product. The longer the customers drive their vehicles or use their products, the easier the age related issues would show up. Once the lessons are learned from the field, companies would need time to report issues from field operation, collect data, investigate the root cause for the failure mode, figure out and validate the fixes, and implement solutions. The improved reliability will be reflected in the vehicles and other products built afterwards. Therefore, the calendar time is a better, and also a simple indicator, of both chance to expose field reliability issues and potential opportunity to improve reliability. In this paper, a new field reliability growth model is proposed along with a modified Crow-AMSAA model. They are applied to a number of vehicle brands and models, and the results are discussed and compared. It shows the new model generally fits the field data very well while modified Crow-AMSAA model can be useful for special applications. Two models are compared with each other. The parameter estimate and statistical inference are presented with an example. The management aspect of the field reliability growth is very important and is therefore also discussed.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.