At this edition of the event, we will focus on the Cluster of Excellence: Integrative Computational Design and Construction for Architecture (IntCDC) of the University of Stuttgart as well as on research groups at the University of Tübingen focusing on geo sciences, climate and sustainability research.
Due to COVID-19 restrictions, the workshop will be held online (Zoom). If you would like to participate, please register here.
|14:05 – 14:15||
Prof. Achim Menges
Institute for Computational Design and Construction, University of Stuttgart, IntCDC
Title: Introduction to IntCDC
Abstract: The vision of the Cluster of Excellence Integrative Computational Design and Construction for Architecture (IntCDC) is to harness the full potential of digital technologies in order to rethink design, fabrication and construction based on integration and interdisciplinarity, with the goal of enabling game-changing innovation in the building sector as it can only occur through highly integrative fundamental research in an interdisciplinary, large-scale research undertaking.
|14:15 – 14:35||
Prof. Dr.-Ing Cristina Tarín
Institute for System Dynamics, University of Stuttgart, IntCDC
Title: Co-Design for Architecture, Engineering and Construction using Bipartite Graphs
Abstract: The architecture, engineering and construction (AEC) industry consists of several players from different domains. They usually work in a sequential process, which is slow, costly and problems are identified in late stages. Co-design helps to identify and predict problems in early stages throughout different domains such that the design, engineering, manufacturing and construction process can be improved. We present an approach to enable co-design using bipartite graphs, which are capable to integrate expert knowledge, physical models and machine learning models from different domains. The different models form a computational network, which allows e.g. the architect to automatically analyze the feasibility of manufacturing a design.
|14:35 – 14:55||
Prof. Dr.-Ing. Peter Middendorf
Institute of Aircraft Design, University of Stuttgart, IntCDC
Title: Data-driven approaches for fiber composites technologies and simulation
Abstract: Fibre composites for lightweight and wide-span architecture applications are one research area within IntCDC. The technology focus is laid here on freeform winding using collaborative robotics and the respective winding path is generated in a co-design approach. Whilst this is a collaborative effort from a group of IntCDC PIs, there is also a growing activity of supporting data-driven approaches at the Institute of Aircraft Design to make use of various sensor data along the process chain of a composite structure and thus to overcome limitations of classical modelling and simulation techniques.
|14:55 – 15:15||
Tenure-Track Prof. Dr. Thomas Wortmann
Chair for Computing in Architecture, Institute for Computational Design and Construction, University of Stuttgart, IntCDC
Title: AI for AEC
Abstract: Tenure-Track Prof. Wortmann will present examples of the rich opportunities to apply AI in the Architecture, Engineering and Construction (AEC) sector, including architectural design optimization, modelling design and construction processes, predicting wind flow around buildings, predicting the behavior of heterogenous materials, and adaptive robotic control. Ultimately, these applications contribute to a reimagining of how our built environment is planned and constructed, which is an urgent need in light of the impending climate catastrophe.
|15:15 – 15:35||
Dr. Nicole Ludwig
Machine Learning in Sustainable Energy Systems, Cluster of Excellence Machine Learning: New Perspectives for Science, University of Tübingen
Title: Forecasting Sustainable Energy under Uncertain Weather
Abstract: The increasing share of sustainable energy generation in the electricity system comes with significant challenges. One of the main challenges is the volatility of sustainable energy generation, due to their dependency on the weather. To efficiently integrate sustainable energy into the grid, the expected electricity supply and demand have to be forecast accurately. In this talk I will introduce the challenges we face when forecasting weather-dependent energy supply and demand and how probabilistic machine learning can help alleviate some of these challenges.
15:35 – 15:55
Dr. Bedartha Goswami
Machine Learning in Climate Science, Cluster of Excellence Machine Learning: New Perspectives for Science, University of Tübingen
Title: Uncovering hidden structure in climate data
Abstract: The last decades have witnessed an unprecedented increase in data related to climatic processes all around the globe, which contain till yet undiscovered mechanisms of how the climate functions and how it is currently changing. We present the idea of "climate networks," which allow us to reveal non-trivial patterns in databases of climate data, and reveal potentially new climatic mechanisms and connections that have not been studied till date. We will look at how climate networks are estimated, what they can do for us, and three examples of how they are used in practice. Towards the end, we will also talk briefly about other projects that are ongoing in the "Machine Learning for Climate Science" research group, in the hope that we find common interests for cross-pollination of ideas and efforts.
|15:55 – 16:15||
Prof. Dr. Thomas Scholten
Chair of Soil Science and Geomorphology, Department of Geosciences, University of Tübingen
Title: Machine Learning in Soil Science
Abstract: Soils are the most important basis for our nutrition. Their quality determines the food security of the world. Machine learning methods help to determine the underlying soil properties and to predict their spatial distribution on Earth. The presentation will explain the underlying theoretical approach of state factors and address the concept of relevant range of scales for multi-scale contextual spatial modeling of soil properties. Further, I will show results from my research of different parts of the world and demonstrate the potential of data-driven models for geo- and environmental science.
|16:15 – 16:30||Wrap-up|