newAIDE – Enhancing vehicle design processes by AI
divis participates in the research project newAIDE: new Artificial Intelligence based Design in Engineering. The aim is to use artificial intelligence methods for the optimization and automatization of vehicle design processes. The project will develop AI methods for learning, supporting, and partially replacing decision making processes based on construction and [...]
The latest updates to our software
ClearVu Solution Spaces (CVSS) in Ultra-HD and 4K Our new version of CVSS is now based on the latest Java version. This ensures a better scaling of the user interface on high definition screens. Evaluation of Prediction Confidence How reliable is the prediction of a metamodel? To answer this [...]
New publication: “Towards an adaptable quality monitoring process for self-piercing riveting”
With his co-authors Vanessa Noller, Dr.-Ing. Ulrich Walther (both Mercedes-Benz AG Sindelfingen) and Prof. Dr.-Ing. Gerson Meschut (University of Paderborn) Thomas Bäck has published an article about the quality monitoring process of self-piercing riveting. The article introduces machine learning methods for predicting and avoiding faulty joints. The results showed a [...]
Research project EnXylaScope
divis will be participating in the new European research project EnXylaScope, starting this month. The project focus is the development of new enzyme-based methods for debranching xylan, an innovative ingredient of many consumer goods and food. Due to the good chemical and physical characteristics, debranched xylan is meant to replace [...]
Optimization with Python
The new version of our ClearVu Python Package now provides optimization methods. The methods which are used for hyperparameter optimization of the models can also be used for the optimization of target functions in python projects. Like for the modelling, the data exchange uses the panda package. The API-documentation was [...]
Thomas Bäck was nominated as a member of the Royal Netherlands Academy of Arts and Sciences
Thomas Bäck was honored in a very by special way. He was chosen as a member of the Royal Netherlands Academy of Arts and Sciences. This institution chooses its members based on their achievements. The academy represents a wide range of academic fields and, furthermore, functions as an advisory body [...]
Machine Learning for Vehicle Development
Jana Büttner, Stefan Schwarz, Axel Schumacher and Thomas Bäck published their results on Machine Learning for vehicle development in the ATZ – Automobiltechnische Zeitschrift. In their article “Automated Machine Learning for Supporting Vehicle Design: The Case of a Global Sensitivity Matrix” the authors describe how relations between design variables and [...]
Machine Learning for Vehicle Development
On February 25, Jana Büttner (Porsche AG) and Prof. Thomas Bäck will give a presentation at the ASCS online event “Machine Learning for Virtual Vehicle Development”. The presentation “Automated Machine Learning for Supporting Vehicle Design: The Case of a Global Sensitivity Matrix” focuses on using machine learning for analyzing relations [...]