Many quality-related issues – especially with regard to the Warranty Management – directly rely on reliability prognoses, such as the determination of expected failures in field operation or expected warranty costs. Additional areas of application comprise the determination of the quantity of spare parts and concerning an End-of-Life-Inventory, additionally the assessment of different concepts, constructs and technologies for longer operating times and further the estimation of potential risks.
The following explanations refer to the automotive sector, but can generally be transferred to all fields of technological products and processes.
Field trials, endurance testing and accelerated tests are used for the general validation of automotive components regarding their functionality and robustness. Due to the small sample sizes they are not suited to answer the above mentioned questions sufficiently. In addition, the correlation between testing and the real field operation remains unknown. On the contrary, field data reflect the actual conditions for operation as well as all relevant negative factors, which make them a solid basis for the examination of failure behavior. In general, concerning large serial deliveries there are sufficient failures available that guarantee a high statistical significance. However, the complete recording of field failures takes place only during the warranty period, which makes the data timely limited and incomplete.
The time difference between the vehicle’s registration and the failure of the component under investigation is not suitable for measuring its load. For this purpose, the pure operating time would be necessary, which, in general, is not recorded and thus remains unknown. As confirmed by many extensive investigations, the distance travelled up to the failure may serve as an adequate substitute. However, the driving behavior whose monitoring is restricted to the warranty period shows to be highly diverse (from a few thousand km to over 100.000 km). Accordingly, this fact needs to be accounted for when investigating the reliability.
The after-image depicts a process of a reliability prognosis.
Reliability Prognosis Models offer (amongst others) the following possible uses:
– Calculation of future guarantee and warranty costs
– Risk management concerning the extension of the warranty period
– Calculation of required serial spare parts or End-of-Life-Inventory in spare parts management
– Detection of warranty fraud
– Possibility of supplier assessment/ – monitoring
– Statistical analysis of registration’s delay (optimizing Supply-Chain-Management)
– Statistical analysis of reporting delays (optimizing flow of information between supplier and purchaser or intern)
Functional Safety (ISO 26262, IEC 61508)
– Generation of failure rates in consideration of the specific strains of components
– Possibility to prove operational dependability under real conditions – Proven-in-Use according ISO 26262
– Proving standard conformity
Informations Research and Development
– Assessment of system modifications
– Aiding component selection
– Deployment of components proven in use – optimization of test and validation planning (actual exposure in field operation)
– Feedback for research and development
– Statistical analysis of company intern field data via proven model
– Extensive experience over years with major OEM and suppliers
– Pure data analysis as well as interpretation by experts possible
– Continuing development (state of the art)
Download model description
You can download a summary of the Reliability Prognosis Model of Wuppertal here:
German version: IQZ_Zuverlässigkeitsprognosemodell
English version: IQZ_Reliablity-Prognosis-Modell