SimTech Doktorandenwochenende 2011
Söllerhaus in Hirschegg, Kleinwalsertal, Austria
8 - 10 July, 2011
"The whole is greater than the sum of its parts." (Aristoteles)
The SimTech Cluster of Excellence at the University of Stuttgart provides a unique environment for interdisciplinary co-operations
of its members. New research ideas and future visions are being developed due to the sum of ideas of each member and the efficient exchange of knowledge within the project networks as well as between them.
Organizers of the Ph.D. weekend: David Molnar, Institute for Materials Testing, Materials Science and Strength of Materials Judith Rommel, Institute of Theoretical Chemistry Patrick Weber, Institute for Systems Theory and Automatic Control
1. Motivation: Simulating SimTech
A key component of the Graduate School SimTech is the interdisciplinary collaboration of its members. Cooperations are aspired in this context within the project networks as well as across the project networks within the entire SimTech Cluster. Important prerequisites for the development of successful collaboration are the visibility of possible tie points and the possibility to identify potential cooperation partners. Therefore, the goal of the SimTech PhD weekend 2011 (SimTech PhD WE 2011) was to
- improve the visibility of possible tie points for new cooperations.
- provide an integrative discussion forum to identify new potential cooperation partners.
These goals were reached by the conception Simulating SimTech which will be explained in the following. In advance to the SimTech PhD WE 2011 the participants were divided into interdisciplinary teams (4-6 people per group) for a joint project work. In these groups the participants worked on solutions for fictive major projects chosen by their teams. A major project is distinguished by innovation, interdisciplinarity as well as scientific, economical and societal relevance. During the project work in teams, potential tie points were worked out as a possible bases for cooperations after the PhD WE 2011. By the active and creative team work of the participants from different project networks and research areas, a crucial contribution to realize the goals stated above, was achieved.
2. Course of Events
Preparation - Before the PhD Weekend
All participants were asked to send brief information about their project in the following form to the PhD speakers:
- Name, institute
- Project network, project title, project description (max. 1200 characters)
- Five keywords describing the project
- Five keywords about used methods and software
- Suggestions for possible major projects.
After registration, the PhD speakers formed Inter PN (Project Network) groups consisting of 4 to 6 members from different project networks. Each team member received the previously collected brief information from all other team members. The teams were asked to organize a first meeting, where a project moderator was elected. The moderator acted as the contact person for the PhD speakers. In the first meeting also a decision upon which major project the group wanted to work on had to be made. Optionally, the selection could be made from the list of submitted proposal projects of the group members.
Phase 1: Intra PN
Each participant discussed within his Intra PN group about the used methods, software and established expertises within the PN. For a better orientation the groups have been equipped with various material about their PN. The goal of this phase was to update each participant about the ongoing research within his PN. The result of this phase was a short method catalogues of the PNs. At the end of phase 1, each PN method catalogue was presented in plenum.
Phase 2: Inter PN
In this phase the participants were reunited in the predefined groups from the preparation phase. Each participant acted as if he would have the background and knowhow of his whole PN collected in phase 1 at his disposal. In these interdisciplinary teams, solution approaches for the major projects were developed. The following points had to be included:
- Motivation, research goals and entrepreneurial goals within the major project.
- Development of interfaces with and intersections within SimTech, e.g. via connection graphs.
- Selection and assignment of the needed Know-How to reach the goals.
- Definition of milestones within the project.
- Discussion about the benefit for science, economy and society.
At the end of this phase, the re¬sults were presented in a plenary meeting. The individual points above were discussed with all participants after the talks. The results were additionally summa¬rized by the groups in reports (for two shortened reports: see Chapter 3). The timetable (Figure 1) shows the arrangement of the separate phases. Besides the working phases, group activities were organized in order to strengthen the team spirit within the SimTech Cluster.
3. Inter Project Network Phase
In the following two of the five major projects are shortly presented.
3.1 Investigation of a Novel Treatment Approach for Gene Therapy of Lung Disease
This project is concerned with simulation of treatment approaches for gene-therapy of lung diseases. The therapy bases on inhalation of an aerosol, containing modified adeno-viruses suspended in small fluid droplets. Then, the viruses interact with the affected cells in order to treat the disease. The simulation aims at the detection of optimal treatment parameters, i.e. reactant concentration and inhalation duration. We propose to simulate this process by a twophase flow model and local transport-diffusion equations for reactant uptake in cells. Due to the resulting large-scale systems model reduction techniques are employed in order to enable parameter studies. Adequate volume visualization methods finally aid the interpretation of the results by medical staff and biologists.
Cystic fibrosis is a common recessive genetic disease affecting the entire body, but primarily the bronchial and lungs, causing processive disability and often early death. Breathing difficulties are the most serious symptoms which result from frequent lung infections. These infections can be treated, yet not cured by application of antibiotics, inhalation and other medications. Several other symptoms, e.g. sinus infections and poor growth result from the effects of cystic fibrosis on other parts of the body. Cystic fibrosis is caused by a mutation in a single gene, coding for the cystic fibrosis trans-membrane conductance regulator (CFTCR), a chloride-channel. This gene is required to regulate the components of mucus, but also e.g. sweat. A potential novel treatment approach, mainly for the bronchial symptoms of this disease, is gene therapy. This approach would imply the use of recombinant adenoviruses to deliver a correct version of the gene to the affected epithelial cells in the bronchia. These viruses would be administered in an aqueous solution using a vaporizer or nebulizor, comparable to an asthma spray. A requirement for successful treatment is to obtain a homogeneous distribution of the reactant to the cells of the mucous layer in the bronchia. As possible side effects of this treatment are only poorly understood and should be kept to a minimum in potential later studies in the clinic, as much previous knowledge as possible is needed. These information would best be obtained from a modeling approach.
Intersections within SimTech
A project of the size and complexity like the one we propose here would be suited ideally to be worked on within the framework of SimTech. Efforts from various project networks would be needed in order to tackle the demanding tasks which are connected to model development, reduction, simulation and visualization. In addition to the contributions of the already involved Project networks (PN) 5, which accounts mainly for simulation and visualization of the obtained data, PN 6, responsible for model reduction and PN 7 which would aid both in (experimental) verification of the model as well as in providing parameters necessary for model development, interconnections with several further PNs would be of significant benefit. For example, connections to PN 4 (Coupled problems in biomechanics and systems biology) are obvious. Their expertise in the field of biomechanics is of considerable use in order to obtain an appropriate description of the movement, i.e. contraction and expansion, of the bronchia and lung alveolae during breathing. Due to the high complexity of the models, high computational power is required in order to obtain solutions in a timely fashion, hence an involvement of PN 2 (Molecular simulations with high-performance and many-core computer systems) is required. Ultimately, a strong participation of PN 9 (Foundations for the integrative reflection and evaluation of simulations) would be important to discuss potential ethical implications, should the results of our modelling approach support an application of this treatment in clinical studies (see Figure 2).
Benefit for Science, Economy or Society The project is aimed at helping people suffering from cystic fibrosis. Not only those people will benefit from research, but also their descendants, who may carry the disease. The treatment will certainly ease the burden on the healthy family members, since they are often affected (indirectly) by the disease. The project will bring an answer to the question as to how much reactant must be absorbed by a patient so that the treatment is efficient and potential side effects are minimized. Since the production of the reactant is rather expensive, an accurate estimation of its concentration in a dose may significantly reduce the production costs and thus increase the number of CF patients that may be effectively treated.
3.2 Simulation Quality – Methods of Evaluation
Simulation is used to support decision making processes of high responsibility today. E.g. simulations are used to prototype expensive machinery. Simulation errors may lead to high costs or even endangerment of humans and ecology. Therefore it is very important to track the propagation of errors and uncertainty in simulation technology. The more complex simulations get, the more important it is to have good control of the individual parts of the simulation process.
Motivation and benefit for Science, Economy or Society
The benefit of a major project which at first defines and later on monitors and ensures simulation quality is various. In the field of science it could help by providing standards and supporting a common language. In economy it would support incorporating simulation quality in auditing processes, more reliable risk and product quality assessment. The benefit for society would be more transparent decision making by communicating uncertainty of actions that are motivated by simulation predictions.
Intersections within SimTech
At first the dependencies of a interdisciplinary simulation have to be identified (Figure 3).
- PN2 can deliver methods to quantify numerical errors and uncertainties. (e.g. Jacobian determinant analysis)
- PN7 can support methods to analyse the accuracy of data. E.g. Bayesian data analysis or error model estimation.
- PN8 can deliver methods for data management, if e.g. data from various sources are combined and redundancy is present. For example sensor grids methods.
- Data can be enriched with meta data to assign e.g. timestamps. PN8 can contribute tagging methods. These methods are also usefully to analyze timeliness.
- Volatility and consistency has to be analyzed by the individual experts of the physics and background of a simulation. Here we see contributions from all PNs.
- PN2 has competences and can contribute methods like “core discrete stochastic arithmetics" in this field to validate hardware errors
- PN1 can contribute in the section of software quality management with e.g. the expertise gained from the Espresso project.
- PN5 can contribute stability criteria methods to propose solution methods for model classes.
The PhD students have been asked some questions after the PhD weekend in form of a questionnaire. The main outcome of this evaluation is shown in this chapter. Figure 4 shows a general shift towards a better overview of the applied methods within SimTech (left figure). A shift is also observed towards a better understanding of how the individual PhD student's project is embedded within SimTech after the PhD weekend (right figure).
It is also interesting how the PhD students experienced the work in the Intra-PN-Group in comparison to the Inter-PN-Group (Figure 5). While the difficulty of gathering the parts was the same in both configurations, discussions were slightly more difficult in the interdisciplinary Inter-PN-Group. Surprisingly the PhD students managed the elaboration much better in the Inter-PN-Group. This outcome motivates a more intensive Inter PN work as it is experienced being effective by the PhD students. The overall shift towards a better overview over SimTech strongly motivates a repetition of the PhD weekend.
The interdisciplinary results of the PhD weekend, both from the Intra PN Phase and the Inter PN Phase made the SimTech PhD WE 2011 a successful event. The work within the individual project networks provided an update for 2nd and 3rd year members of the Graduate School SimTech and a very good first insight for the new members into the current research topics of SimTech.
The good overview gained during the Intra PN Phase lead to a very productive Inter PN Phase which in total is indicated by the quantity and quality of the submitted project elaborations. Especially the connection graphs obtained during the Inter PN Phases show the competence of the interdisciplinary SimTech environment to solve problems. Their development was one of the main goals of the SimTech PhD WE 2011. Furthermore, the SimTech PhD WE 2011 provided an integrative discussion forum and supported the process of identifying potential cooperation partners within the SimTech research cluster. The participants enjoyed the canyoning tour and the mountain adventure, respectively, which were organized as an additional team building activity. The positive feedback during the weekend, the versatile and interdisciplinary outcome as well as the results of the evaluation strongly motivate the repetition of the PhD weekend.
The organizers of the Ph.D. weekend thank the Executive Board of Directors of SimTech for the opportunity of organizing this weekend. Furthermore, we thank the German Research Foundation (DFG) for the financial support of the SimTech PhD WE 2011 within the context of Cluster of Excellence in Simulation Technology (EXC 310/1) at the University of Stuttgart.
We also thank the SimTech Management Team, in particular Birgit Pferdmenges who helped organizing the weekend and contributed to the discussions before and during the weekend.
Finally, we would like to thank all participants for having a great SimTech PhD WE 2011 with fruitful discussions and newly formed connections. We hope, you enjoyed it as much as we did.