researchXchange: Statistical Process Optimisation 17. Mai'19
17.05.2019, 12:00-12:45 Uhr – Raum E 013, Pestalozzistrasse 20, Burgdorf
DoE: An Example of Statistical Process Optimisation
The purpose of the industry project was to optimise a new production technology. Concretely, we wanted to analyse whether it is possible to reach the same output accuracy as the gold standard technology without compromising on speed. For a competitive advantage, it will be critical to maintain the quality, and subsequently maximise the speed. For this purpose, we estimated the effect of four key explanatory variables (such as speed) on the output quality. Those variables can easily be adjusted in the production technology (machine). The common approach of changing one factor at a time would lead to wrong conclusions, as we expected the factors to interact strongly. Therefore, reaching the optimal setting via optimising one factor after the other was not possible and a Design of Experiment (DoE) approach was needed. Using a simple, but flexible low-order polynomial regression model, we were able to find a setting which results in high quality products at a very high speed.
Dr. Jasmin Wandel, Head of Institute/ Professor of Mathematics and Statistics, BFH, Institute for Optimisation and Data Analysis IODA
Jasmin Wandel studied statistics at the University of Bern and completed a doctorate in extreme value theory. She has extensive experience in statistical consulting, data analysis and modelling. For several years she was responsible for the statistical analyses in research projects of the School of Dental Medicine (ZMK) of the University of Bern. In 2013 she joined the BFH as Professor of Mathematics and Statistics. She teaches mainly in the Department of Electrical Engineering and Information Technology, as well as in Biomedical Engineering – a Master program jointly conducted by the BFH and the University of Bern. Jasmin Wandel has been head of the Institute for Optimisation and Data Analysis (IODA) since the beginning of the year. As a statistician she is involved in several research projects with partners from the industry and federal offices, and also supports internal projects of the BFH.
Please register for the Seminar. A sandwich and drinks are provided for participants registering until Thursday 16.05.2019 12:00h.
- Startdatum 17.05.2019, 12:00-12:45 Uhr
- Ort Raum E 013, Pestalozzistrasse 20, Burgdorf