For full functionality of this site it is necessary to enable JavaScript. Here are the instructions how to enable JavaScript in your web browser.

Position within the page tree

Home
Team
Renner, Markus

Markus Renner

M.Sc.

Research Associate
Institute for Computational Design and Construction (ICD)
2024 - Current

Contact

+49 711 685 81949
Email

Keplerstraße 11
70174 Stuttgart
Deutschland

Publications

Publications:

2025
1. Renner, Markus, Matthias Hornung, Diellza Elshani, Mathias Niepert und Thomas Wortmann. 2025. Training and evaluating a Co-pilot tool using graph neural networks for generating non-orthogonal building typologies in architectural autocompletion. Hg. von Markus Renner und Matthias Hornung. International Journal of Architectural Computing. August. doi:10.1177/14780771251354914, https://doi.org/10.1177/14780771251354.
  - BibTeX
  BibTeX
  @mastersthesis{doi:10.1177/14780771251354914, abstract = {This study introduces a context-sensitive generative artificial intelligence (AI) co-pilot tool designed to assist architectural design by generating predictive suggestions based on Building Information Modeling (BIM) and Industry Foundation Classes (IFC) data. The approach adopted in this study involves the utilization of Graph Neural Networks (GNNs) and Deep Generative Models of Graphs (DGMG) to enhance the synthesis of 3D architectural spatial typologies. To capture spatial and relational patterns between building elements, various GNN architectures where employed, including Graph Convolutional Networks (GCNs), GraphSAGE, and Graph Attention Networks (GATs). The custom training dataset comprised 180,000 subgraphs derived from real-world BIM models, with IFC files converted into heterogeneous graph representations. A combination of multi-label classification and regression techniques was applied to address the complexities of architectural design predictions. The developed co-pilot tool integrates these models within an interactive human-AI workflow, offering users different levels of control. A System Usability Scale (SUS) assessment was conducted for each control mode, demonstrating the potential of the tool to generate context-sensitive design recommendations. Visual evaluation findings highlight the co-pilot’s ability to learn spatial relationships between building elements in 3D and predict contextually appropriate architectural components. }, author = {Renner, Markus and Hornung, Matthias and Elshani, Diellza and Niepert, Mathias and Wortmann, Thomas}, doi = {10.1177/14780771251354914}, editor = {Renner, Markus and Hornung, Matthias}, eprint = {https://doi.org/10.1177/14780771251354914}, journal = {International Journal of Architectural Computing}, month = {08}, number = 0, pages = 14780771251354914, title = {Training and evaluating a Co-pilot tool using graph neural networks for generating non-orthogonal building typologies in architectural autocompletion}, url = {https://doi.org/10.1177/14780771251354 }, volume = 0, year = 2025 }
2. Hornung, Matthias, Markus Renner, Diellza Elshani, Mathias Niepert und Thomas Wortmann. 2025. Augmenting IFC-Based BIM Models for Graph Neural Network Training of Non-Orthogonal Spatial Typologies: A Context-Informed Design Methodology. Hg. von Matthias Hornung und Markus Renner. CAAD Futures 2025 – Catalytic Interfaces: Conference Proceedings. Conference paper, September. doi:10.25442/hku.29365832.v1, https://datahub.hku.hk/articles/conference_contribution/49_Augmenting_IFC-Based_BIM_Models_for_Graph_Neural_Network_Training_of_Non-Orthogonal_Spatial_Typologies_A_Context-Informed_Design_Methodology/29365832.
  - BibTeX
  BibTeX
  @mastersthesis{Hornung2025, abstract = {Building Information Modeling (BIM) and Industry Foundation Classes (IFC) data are valuable resources in architecture, offering detailed and structured information on building elements. However, using such resources to develop artificial intelligence (AI) models that are capable of learning and generating spatial and geometrical suggestions remains a challenging problem. This study presents a systematic method for converting real-world BIM/IFC projects into graph representations, suitable for training Graph Neural Networks (GNNs), which overcome the limitations inherent to conventional AI techniques based on pixel and 3D voxels in architecture. First, a custom schema was defined to encode each building element and spatial relationship as triplets, allowing for the generation of heterogeneous graphs that capture geometry, topology and semantic labels. Using models provided by multiple architecture offices, we extracted over 180,000 subgraphs that represent diverse non-orthogonal spatial typologies. Next, each subgraph was annotated with multi-label classifications (element type) and continuous attributes (e.g. dimensions, orientation) to support both classification and regression tasks. Finally, to demonstrate the utility of the dataset, we trained a custom generative neural network (NN) to complete partial spatial graphs, producing non-orthogonal room layouts based on the existing building context. The custom NN suggests that the dataset has the capacity to support advanced 3D generative tasks in architecture and could lay the groundwork for future studies on automated design synthesis.}, author = {Hornung, Matthias and Renner, Markus and Elshani, Diellza and Niepert, Mathias and Wortmann, Thomas}, booktitle = {CAAD Futures 2025 – Catalytic Interfaces: Conference Proceedings}, day = 12, doi = {10.25442/hku.29365832.v1}, editor = {Hornung, Matthias and Renner, Markus}, key = {Hornung2025}, month = {09}, number = 49, pages = {869-883}, title = {Augmenting IFC-Based BIM Models for Graph Neural Network Training of Non-Orthogonal Spatial Typologies: A Context-Informed Design Methodology}, type = {Conference paper}, url = {https://datahub.hku.hk/articles/conference_contribution/49_Augmenting_IFC-Based_BIM_Models_for_Graph_Neural_Network_Training_of_Non-Orthogonal_Spatial_Typologies_A_Context-Informed_Design_Methodology/29365832}, volume = {II}, year = 2025 }
3. Ploennigs, Joern, Markus Berger, Thomas Wortmann, Jakob Kirchner, Jakob Beetz, Alina Roitberg, Karsten Menzel und Björn Ommer. 2025. Building Foundation Models - Potentials, Challenges and Research Directions for Using LLM and LVM in AEC. In: Proceedings of the 2025 European Conference on Computing in Construction, hg. von Marijana Sreckovic, Pedro Meda, Ranjith Soman, und Ekaterina Petrova, 6: Computing in Construction. Porto, Portugal: European Council on Computing in Construction, Juli. doi:10.35490/EC3.2025.268, https://ec-3.org/publications/conference/paper/?id=EC32025_268.
  - BibTeX
  BibTeX
  @inproceedings{EC32025_268, abstract = {Foundation Models like Large Language Models (LLM) and Large Vision Models (LVM) show huge potential in automating processes in building design, construction, and operation. However, there are concerns about their capability to create coherent, usable, sustainable, and safe architecture. This paper analyzes the challenges and potentials of LLM and LVM in the AEC sector. We discuss a collection of relevant AEC modalities like CAD models, semantic graphs, and time series and how they could integrate into the current landscape of models. From the analysis we derive research directions toward the development of domain specific â€œBuilding Foundation Modelsâ€.}, address = {Porto, Portugal}, author = {Ploennigs, Joern and Berger, Markus and Wortmann, Thomas and Kirchner, Jakob and Beetz, Jakob and Roitberg, Alina and Menzel, Karsten and Ommer, Björn}, booktitle = {Proceedings of the 2025 European Conference on Computing in Construction}, doi = {10.35490/EC3.2025.268}, editor = {Sreckovic, Marijana and Meda, Pedro and Soman, Ranjith and Petrova, Ekaterina}, isbn = {978-9-083451-31-2}, issn = {2684-1150}, language = {en-GB}, month = {07}, organisation = {European Conference on Computing in Construction}, publisher = {European Council on Computing in Construction}, series = {Computing in Construction}, title = {Building Foundation Models - Potentials, Challenges and Research Directions for Using LLM and LVM in AEC}, url = {https://ec-3.org/publications/conference/paper/?id=EC32025_268}, volume = 6, year = 2025 }
2024
1. Renner, Markus, Evgenia Spyridonos und Hanaa Dahy. 2024. Tensegrity FlaxSeat: Exploring the Application of Unidirectional Natural Fiber Biocomposite Profiles in a Tensegrity Configuration as a Concept for Architectural Applications. Hg. von Markus Renner und Evgenia Spyridonos. Buildings 14 (August). doi:10.3390/buildings14082490, https://www.mdpi.com/2075-5309/14/8/2490.
  - BibTeX
  BibTeX
  @article{buildings14082490, abstract = {Material selection is crucial for advancing sustainability in the building sector. While composites have become popular, biocomposites play a pivotal role in raising awareness of materials deriving from biomass resources. This study presents a new linear biocomposite profile, fabricated using pultrusion technology, a continuous process for producing endless fiber-reinforced composites with consistent cross-sections. The developed profiles are made from flax fibers and a plant-based resin. This paper focuses on the application of these profiles in tensegrity systems, which combine compression and tension elements to achieve equilibrium. In this study, the biocomposite profiles were used as compression elements, leveraging their properties. The methods include geometrical development using physical and digital models to optimize the geometry based on material properties and dimensions. A parametric algorithm including physics simulations was developed for this purpose. Further investigations explore material options for tension members and connections, as well as assembly processes. The results include several prototypes on different scales. Initially, the basic tensegrity principle was built and explored. The lessons learned were applied in a final prototype of 1.5 m on a furniture scale, specifically a chair, integrating a hanging membrane serving as a seat. This structure validates the developed system, proving the feasibility of employing biocomposite profiles in tensegrity configurations. Furthermore, considerations for scaling up the systems to an architectural level are discussed, highlighting the potential to enhance sustainability through the use of renewable and eco-friendly building materials, while promoting tensegrity design applications.}, article-number = {2490}, author = {Renner, Markus and Spyridonos, Evgenia and Dahy, Hanaa}, doi = {10.3390/buildings14082490}, editor = {Renner, Markus and Spyridonos, Evgenia}, issn = {2075-5309}, journal = {Buildings}, month = {08}, number = 8, title = {Tensegrity FlaxSeat: Exploring the Application of Unidirectional Natural Fiber Biocomposite Profiles in a Tensegrity Configuration as a Concept for Architectural Applications}, url = {https://www.mdpi.com/2075-5309/14/8/2490}, volume = 14, year = 2024 }

Teaching

Seminar

Computational Explorations 2 - WS 2025
Computational Explorations - SS 2024/25

Markus Renner is a research associate at the Institute for Computational Design (ICD/CA) at the University of Stuttgart. His current research project focuses on the implementation of surrogate models for design space exploration with an emphasis on user-friendly autocompletion interactions.

Prior to joining ICD, Markus worked as a computational designer for str.ucture GmbH. He holds a Bachelor of Science and a Master of Science in Architecture and Urbanism at the University of Stuttgart, where he acquired extensive expertise in architecture, programming, robotics, and building construction.

Together with Matthias Hornung, their master’s thesis “Generating Non-Orthogonal Spatial Building Typologies Using Graph Neural Networks: An AI-based Co-Pilot for Context-Informed Architectural Design”, demonstrates a state-of-the-art machine learning approach for the generation of context-informed building elements in 3D, utilizing Graph Neural Networks (GNNs). The study introduces an autocompletion Co-Pilot tool designed to support planners in the Architecture, Engineering, and Construction (AEC) industry to streamline previously unachievable design processes.

His research interests include developing GNN models with graph-based building information data to infer missing node features, and to enable context-informed generation of building elements, especially their geometry in three-dimensional space.

Markus Renner

Contact

Publications

2025

BibTeX

BibTeX

BibTeX

2024

BibTeX

Teaching

Seminar

Vita

Audience

Formalities

Services

Organization

Markus Renner

Contact

Publications

2025

BibTeX

BibTeX

BibTeX

2024

BibTeX

Teaching

Seminar

Vita

Here you can reach us

Audience

Formalities

Services

Organization