Concept for machine learning and field data driven adjustment of testing conditions of technical prototypes

The present dissertation deals with the adjustment of test scenarios of technical prototypes. The overall objective of this research work is the development of a concept for the detection of differences between defect and non-defect products concerning the redesign of test procedures and application of the obtained knowledge during the test phase of the next generation products. An essential part of the presented concept is the application of various machine learning algorithms as well as goal-oriented interpretation of the results and deduction of corresponding conclusions. The proposed concept has a generic character and is suitable for all products in the consumer and investment goods sector, in which the possibility of the field data gathering is given. The application of the concept is performed based on various examples from the automotive industry. Über den Autor Marcin Hinz has studied mechanical engineering (BSc) and Computational Mechanical Engineering (MSc) at the University of Wuppertal, Germany. He worked as a CFD engineer and as a scientific assistant at the Chair of Reliability Engineering and Risk Analytics at the University of Wuppertal. He was a lecturer for Statistical Optimization at the University of Applied Sciences in Cologne, Germany. He received his doctoral degree in the field of reliability engineering at the University of Wuppertal. Currently, he is a postdoctoral researcher at the Chair of Reliability Engineering and Risk Analytics at the University of Wuppertal. The focus of his research work is machine learning, accelerated life tests, condition monitoring systems and optimization of test procedures as well as analysis of field data.