Configurable nD-visualization for complex Building Information Models Helga Tauscher Schriftenreihe des Instituts f¨urBauinformatik Herausgeber: Univ.-Prof. Dr.-Ing. R. J. Scherer c Institut f¨urBauinformatik, Fakult¨atBauingenieurwesen TU Dresden, 2017 ISBN: 978-3-86780-526-1 Institut f¨urBauinformatik, TU Dresden Postanschrift: Besuchsanschrift: Technische Universit¨atDresden N¨urnberger Str. 31a 01062 Dresden 2. OG, Raum Nr. 204 01187 Dresden Tel.: +49 351 / 463-32966 Fax: +49 351 / 463-33975 Email: [email protected] www: http://cib.bau.tu-dresden.de Configurable nD-visualization for complex Building Information Models Konfigurierbare nD-Visualisierung f¨urkomplexe Building Information Models An der Fakult¨atBauingenieurwesen der Technischen Universit¨atDresden zur Erlangung des akademischen Grades einer Doktor-Ingenieurin (Dr.-Ing.) genehmigte Dissertation vorgelegt von Dipl.-Ing. Helga Tauscher geboren am 05. August 1976 in Rostock Erster Gutachter: Prof. Dr.-Ing. Raimar J. Scherer (TU Dresden) Zweiter Gutachter: Ao. Univ. Prof. Dr. techn. Bob Martens (TU Wien) Dritter Gutachter: Prof. Dr.-Ing. Frank Petzold (TU M¨unchen) Tag der Einreichung: 23. Februar 2016 Tag der Verteidigung: 18. Oktober 2016 Preface When I applied to the Institute of Construction Informatics in 2010, I did so with a research proposal that I had developed earlier at Bauhausuniversit¨atWeimar. I wanted to explore tool sets and workflows for energy{efficient building design and construction. However, Professor Scherer suggested a different research area to me, that was at that time part of the mefisto research project and to be included in other projects as well: that of visualization in the context of construction project management. Looking back I am surprised how well this suggestion allowed me to isolate a topic that would foster my curiosity, creativity, and commitment over a period of several years. This would not have been possible without the freedom to identify and follow my very own path into scientific waters. This freedom is a gift, but also poses the liability upon every PhD candidate, to assume responsibility for their own work without relying on detailed external guidance. I want to thank Professor Scherer who allowed for and called for this kind of early scientific independence and provided the environment, the resources and just the right impulses for this work to evolve. Frank Petzold from TU Munich and Bob Martens from TU Wien agreed to examine my work. Thank you for the patience to follow my arguments. Thinking and writing takes not a tremendous amount of resources, but it can't be carried out without basic needs covered first. I am happy to have been part of two projects under the auspices of the German Ministry of Education and Research and gratefully acknowledge their support. During the finalization stage I decided to fully commit to the concentrated work at the thesis and to refrain from further project obligations. During this time I received a short-time scholarship from TU Dresden and two travel grants from the German Academic Exchange Service (DAAD). These institutions facilitated the completion of this work, I appreciate their support. A good part of research is carried out in happy solitude, but at the same time heavily relies on scientific exchange and communication, be it during lunch breaks, at confer- ences, or through written publications. I want to thank my colleagues at CIB for the fruitful discussions and the cooperative attitude. Special thanks go to Romy Guruz, who shared a tiny room with me for the first year, reminded me to type quietly while main- taining speed, and became a close companion and friend later on. Sven-Eric Schapke as mefisto project leader introduced me to the intricacies of the project, but he also had a fair share in finding my research topic. I enjoyed working with Martin Voigt from the Institute for Multimedia Technology at the Faculty of Computer Science, who worked on a complementing aspect of visualizations in the mefisto project and was an inspir- v ing partner. I name these three colleagues to represent many persons I met during the years as colleagues at CIB, as project partners, or in research communities dedicated to computation in architecture and construction, who spured my thoughts through their comments and discussions. During the final stages of the write-up the whole text had to be read and revised several times. I want to thank Lene and Lars for reading parts of the thesis in advance and commenting with fresh eyes. Ursula, Joachim, and Traudl provided me their most beautiful forest view and seaside rooms to immerse myself into writing. This work would not exist without friends and family who supported me and my adventurous undertaking with many more or less noticeable, but in any single case notable contributions. In particular, I want to thank J¨orn,Claudia, Anne and Henning for listening to my stories, sharing my concerns, encouraging my decisions, for just backing me up. This book is dedicated to my parents and grandparents who laid the foundations for what I am. I wish they could see what they brought up. When I left the software development team at conject AG to pursue a PhD, one of my former colleagues asked me why I would want to miss out on the development part in Research and Development and do only research that would be discontinued half-way through, long before it could be brought to any productive use. To me, it was a great pleasure to intensely focus on a question and to thoroughly think it through without the pressure of immediate commercial exploitability. At the same time I always had practical applications in mind and verified applicability through prototypical software implementations. I hope that some of the ideas developed in this thesis will be picked up in the future and get incorporated into productive software. Singapore, August 2017 Helga Tauscher vi Abstract With the ongoing development of building information modelling (BIM) towards a com- prehensive coverage of all construction project information in a semantically explicit way, visual representations became decoupled from the building information models. While traditional construction drawings implicitly contained the visual representation besides the information, nowadays they are generated on the fly, hard-coded in software applications dedicated to other tasks such as analysis, simulation, structural design or communication. Due to the abstract nature of information models and the increasing amount of digital information captured during construction projects, visual representations are essential for humans in order to access the information, to understand it, and to engage with it. At the same time digital media open up the new field of interactive visualizations. The full potential of BIM can only be unlocked with customized task-specific visual- izations, with engineers and architects actively involved in the design and development process of these visualizations. The visualizations must be reusable and reliably repro- ducible during communication processes. Further, to support creative problem solving, it must be possible to modify and refine them. This thesis aims at reconnecting building information models and their visual representations: on a theoretic level, on the level of methods and in terms of tool support. First, the research seeks to improve the knowledge about visualization generation in conjunction with current BIM developments such as the multimodel. The approach is based on the reference model of the visualization pipeline and addresses structural as well as quantitative aspects of the visualization generation. Second, based on the theoretic foundation, a method is derived to construct visual representations from given visualization specifications. To this end, the idea of a domain-specific language (DSL) is employed. Finally, a software prototype proofs the concept. Using the visualization framework, visual representations can be generated from a specific building information model and a specific visualization description. vii viii Kurzfassung Mit der fortschreitenden Entwicklung des Building Information Modelling (BIM) hin zu einer umfassenden Erfassung aller Bauprojektinformationen in einer semantisch expliziten Weise werden Visualisierungen von den Geb¨audeinformationen entkoppelt. W¨ahrendtraditionelle Architektur- und Bauzeichnungen die visuellen Repr¨asentationen implizit als Tr¨agerder Informationen enthalten, werden sie heute on-the-fly generiert. Die Details ihrer Generierung sind festgeschrieben in Softwareanwendungen, welche eigentlich f¨urandere Aufgaben wie Analyse, Simulation, Entwurf oder Kommunikation ausgelegt sind. Angesichts der abstrakten Natur von Informationsmodellen und der steigenden Menge digitaler Informationen, die im Verlauf von Bauprojekten erfasst werden, sind visuelle Repr¨asentationen essentiell, um sich die Information erschließen, sie verstehen, durch- dringen und mit ihnen arbeiten zu k¨onnen.Gleichzeitig entwickelt sich durch die digi- talen Medien eine neues Feld der interaktiven Visualisierungen. Das volle Potential von BIM kann nur mit angepassten aufgabenspezifischen Visual- isierungen erschlossen werden, bei denen Ingenieur*innen und Architekt*innen aktiv in den Entwurf und die Entwicklung dieser Visualisierungen einbezogen werden. Die Visualisierungen m¨ussen wiederverwendbar sein und in Kommunikationsprozessen zuverl¨assigreproduziert werden k¨onnen. Außerdem muss es m¨oglich sein, Visual- isierungen zu modifizieren und neu zu definieren, um das