Development of a multivariate reliability model for mapping the degradation behavior, determination of the remaining useful life and prediction of damage cases (DEREPRO) of technically complex products with the aid of multivariate usage profiles

• Applicant: Professor Dr. Stefan Bracke
• Subject Area: Production Systems, Operations Management, Quality Management and Factory Planning
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 557903399

The demands placed on consumer goods, particularly in terms of functionality, reliability, and second-life applications, are undergoing a period of rapid growth. This is accompanied by an increase in product complexity, which is characterised by a larger number of components and the integration of different technical disciplines. This complexity presents novel challenges in reliability analysis, as damage causalities can only be recorded through the use of multivariate approaches. Concurrently, the accelerated pace of development due to intense competitive pressure increases the likelihood of incomplete identification of potential damage causes during the testing phase. The objective of this research project is to develop a multivariate reliability model for mapping complex degradation processes, determining the remaining useful life (RUL), and predicting instances of damage in technically complex products. The univariate methods currently employed consider a single variable for damage prediction, which precludes the consideration of complex multivariate relationships. It is therefore necessary to incorporate multivariate usage data into the model in order to enable a more precise prediction of failure behavior. This model will be developed as an exemplar in the field of automotive engineering, with a particular focus on the analysis of customer-specific usage profiles in order to identify significant differences between products that are susceptible to failure and those that are not. The objective is to define a multivariate limit range that predicts the time of failure. Furthermore, the model will be implemented in a software prototype and validated through case studies in order to demonstrate its effectiveness and efficiency in comparison to univariate approaches.The development of the model comprises several steps: 1. Identification of influencing parameters: investigation of relevant variables that have a significant influence on the occurrence of claims. 2. Modeling of the product fleets: creation of distributions of the influencing variables and mapping of typical usage profiles. 3. Development of the multivariate model: mapping of degradation behavior and prediction of future damage cases based on Monte Carlo simulations. 4. Implementation of a software prototype: Integration of the methodology into a prototype software solution for practical application and validation. The overall objective is to use the multivariate reliability model to gain a deeper understanding of the interactions between usage data and their influence on product reliability. This helps to improve the accuracy of predictions and prevent product failures, which minimizes economic risks for companies.

• DFG Programme: Research Grants