Building Bridges across Disciplines
The research at the IWR is dedicated to the development of mathematical and computational methods for science, technology and humanities, requiring formal mathematical research as well as the development of efficient, tailored software to utilize high-performance computing.
Following its philosophy, the IWR has become one of the most integrative research centers of Heidelberg University with at present 45 active members and twelve junior research group from eight different faculties, i.e. mathematics, informatics, physics, chemistry, earth sciences, biological sciences, medicine, philosophy. As one consequence, scientific computing has become one pillar of the strategy of the Ruprecht-Karls University as a comprehensive university. However, the IWR is not only the key player in scientific computing here in Heidelberg, but served also as blueprint for other national and international institutes and research centers, like the Interdisciplinary Center for Mathematical and Computational Modeling in Warsaw.
The IWR has built up a unique expertise in all aspects of scientific computing with the core competence in innovative methods for modeling, simulation, optimization, and image processing. Key methodological research fields of the IWR members comprise:
- Multi-Scale Modeling and Simulation: A great challenge for modeling and simulation is the building of bridges from microscopic structures and models across the disparate scales of space, time and organization to macroscopic systems.
- Optimization and Optimization-Based Simulation: Optimization-based methods play an ever-increasing role in scientific computing, because most applications require the solution of complex optimization problems like parameter estimation, optimum experimental design, optimal design, optimal control or optimization-based control. Optimization-based methods, nowadays, also play a crucial role in modeling processes.
- Model and Mesh Adaptivity: Discretization of continuum models leads to sequences of nonlinear and then linear systems of algebraic equations. Since the accuracy and computational effort of the solution is directly linked to the discretization level, i. e. the density of the mesh, adaptive discretization leads to efficient and accurate computational methods.
- Multilevel and Parallel Methods: The simulation of complex multiscale, multi-physics problems including optimization and uncertainties requires enormous computational resources despite all algorithmic improvements. Parallel algorithms once being an esoteric area of research has now become mainstream as even desktop computers only get faster by adding more compute cores.
- Image Processing: Imaging technologies rapidly advance in areas like medicine and driver assistance systems, for example. The research aims at providing cutting-edge solutions to basic image analysis problems for applications in science, engineering and the humanities.
- Data Analysis and Statistical Modeling: Research in statistics and data analysis addresses pressing needs in the analysis and modeling of complex systems, high dimensions and massive data sets. These developments are driven by qualitatively new demands in a very broad range of scientific, business and societal applications.
Examples of areas in which challenging problems of scientific computing, which have been successfully addressed through the research at the IWR, demanding optimal computational tools in modeling, simulation and optimization are:
- Computational Chemistry: Breakthroughs have been achieved in the simulation and modeling of reactive flows, the theoretical description of time-dependent chemical phenomena like the intermolecular coulombic decay and nuclear motion, for which efficient computer programs have been developed and disseminated into the scientific community.
- Computational Physics: Algorithms and computer programs could be devised to model, for instance, three-dimensional radiation transport in astrophysics and to simulate the fate of pesticides in ground water in collaboration with environmental physicists.
- Industrial Applications of Scientific Computing: New methods and algorithms have been tailored for real-time optimization and control of industrial processes, for example, reactive distillation columns (BASF) or heavy-duty trucks (Daimler). The Heidelberg Collaboratory for Image Processing (HCI) as a new form of private-public partnerships serves as a role model for “Industry-on-Campus” ventures.
- Computational Biosciences: Efficient algorithms and methods have been devised to allow for high-performance computer simulations of biological macromolecules, in systems biology, in cell biology and in plant physiology, for example.
Tackling Future Research Challenges
The interdisciplinary approach of application-driven development of new mathematical methods and computational tools is unique to the IWR from a national as well as international point of view. It is clear that further strengthening the existing expertise is a key ingredient for an ongoing success. These areas comprise:
- Multiscale, multiphysics, multidimensional modeling
- Adaptive numerical simulations
- Model-based optimization and optimal control
- Image processing and static analyses, big data
- Cutting-edge software development and high-performance computing
In addition to the already ongoing and very successful research directions at the IWR, it is planned to reinvigorate application fields, which had already been very actively pursued. These fields are in particular:
- Computational astrophysics
- Computational environmental physics
Through the generational change among the researchers at the IWR and the respective institutes new scientific interaction points have become apparent, which will be promoted in the near future.
Most important for a striving institute like the IWR, however, is the discovery and exploration of totally new areas of application for scientific computing. Emerging from the research of the last years and in line with the institutional strategy of Heidelberg University, the IWR has, does and intends to further build bridges between research areas by promoting mathematical and computational methods to scientific areas, in which computation plays no or only a minor role so far. It is important to stress that the researchers of the IWR will focus on methodological developments thereby promoting the computational science in the corresponding individual fields, which are in particular:
- Computational Humanities: Several interdisciplinary, already running projects are a unique selling point of the IWR. The appointment of a W3-Professor for “Scientific Computing in the Humanities” and the establishment of the junior research group “Forensic Computational Geometry Laboratory” are first successes.
- Computational Molecular Material Science: The scientific orientation of the science campus of Heidelberg University moved recently towards organic electronics, with a special emphasis on fundamental research. Together with the great expertise of the IWR in modeling, simulation and optimization of complex processes, this impetus led to the successful application for the Center of Advanced Materials (CAM), which will be harbored in an own building on the Neuenheimer Feld science campus. The appointment of a W3-Professor for “Theoretical and Computational Chemistry” was a first step into this direction.
- Computational Social and Behavioral Sciences: Several ongoing Ph.D. projects in this new area are already running, mostly funded through the HGS MathComp. In these highly interdisciplinary projects, problems arising in vision, cognition and psychology, for example, are tackled.
- Computational Medicine and Health Care: In this highly innovative and widely unexplored branch of computational science, first interdisciplinary projects are successfully running addressing problems in cardiology, sepsis, epidemiology and tumor recognition, to name just a few.
The means by which the IWR strengthens these research areas follow its interdisciplinary, open-minded philosophy. “Challenge Workshops” and mini-symposia will be organized, to which leading experts as well as promising junior researchers in the field of application are invited, and hot problems calling for computational support are to be identified. These interactions usually readily lead to initial grant proposals of individual researchers of the IWR together with the partner of the other discipline, as they are currently already under way. Another promising model for the establishment of a research field within the IWR is to implement junior research groups, as it has already been successfully done in the past for computational humanities. Fostering these initial steps and expansion to related problems lead to smaller and larger collaborative centers, as for example, graduate colleges, collaborative research centers, like the HGS MathComp and the HCI.
Last Update: 25.08.2016 - 16:29