MAS 2022/23 Thesis Projects
Nubian Vaulted Slabs: Design and fabrication using 3D concrete printing
Concrete, the most extensively used material in construction, produces significant carbon emissions and formwork waste, especially slab elements.
To address these concerns, this thesis explores a horizontal 3D concrete printing method. The proposed fabrication method, inspired by Nubian vaults, will focus on compression-only structures, slicing with inclined planes and using an angled end effector.
The experimental process consists of five steps: (1) geometric exploration, (2) structural simulation, (3) parameters optimization, (3) testing the printability of the geometry, and (5) 1:1 fabrication. After a series of experiments, the goal of this thesis is to understand the geometric parameters, the printability strategies and explore the further potential of this method in architectural practice.
Students: Huang Su, Yo Cheng Lee
Thesis Supervisors: Dr. Ana Anton dbt, Dr. Andrei Jipa dbt, Dr. Lukas Gebhard Chair of Concrete Structures and Bridge Design
Automated Finishing Strategies for Impact Printed Structures
This research explores Automated Finishing Strategies (AFS) for Impact Printed Structures (IPS). Previous research into IPS demonstrates form-work-free, rapid deposition for efficient earth-based fabrication at low resolution. This exposes new research challenges of surface treatment and finishing operations for market viability, customisation, aesthetic potential and environmental resistance while maintaining the ecological benefits of the developed method. A secondary surface treatment method compatible with IPS should be efficient, robust, and accommodate deep surface texture. Furthermore, additional materials for earth-based construction should consider circularity. To address these considerations, a 6 DoF robot is used to investigate scanning and adaptive robotic surface finishing techniques. Experiments inspired by traditional and state-of-the-art surfacing techniques are computationally generated, resulting in a range of low- to high-resolution finishes.
Student: Carl P-Conquilla, Zac Zhuo Zhang
Thesis Supervisors: Kunaljit Chadha GKR, Dr. Lauren Vasey GKR
Clay Acoustic Patterns
This research explores the utilization of clay 3D printing technology in the area of acoustic design. Through a systematic approach and a timeline of design iterations, fabrication experiments, and acoustic evaluations and measurements, we aim to develop a design-to-fabrication workflow specific to clay 3D printing in acoustic applications. Customized acoustic clay blocks are the product of traditional brickmaking practices combined with advanced computational design and robotic fabrication techniques. The research focuses on three key areas: design strategies for interior freestanding walls, suitable 3D printing and robotic non-planar motion planning techniques, and integration of acoustic evaluation methods into the design process. The outcomes include clay acoustic prototypes, a sound-aware design workflow, and insights into the creative and practical applications of clay enabled by multi-axis robotic motion. This research has the potential to impact architectural designs in sound engineering, enhancing the quality of existing spaces and transforming interior spaces through innovative clay-based acoustic solutions that offer superior sound quality and widen the palette of available acoustic solutions for interiors.
Students: Ana Ascic, Ramon Lopez
Thesis Supervisors: Maria Smigielska dbt, Dr. Achilleas Xydis GKR, Ioanna Mitropoulo dbt
Collaborative Augmented Assembly
In the field of architecture, engineering, and construction (AEC), significant progress has been witnessed in computational design, robotic fabrication, and digital fabrication. Even though these advancements bring new advantages to AEC they still lack adaptability, operate in a linear manner, and lack the implementation of human-in-the-loop processes. Digital fabrication processes especially lack flexible task distribution setups, enabling humans to continuously make decisions throughout fabrication. To support such task distribution setups an interactive user interface is necessary, to allow humans to take over tasks, distribute them to a robot and receive fabrication information on-the-fly.
This project presents Collaborative Augmented Assemblies (CAA), a phone-based augmented reality (AR) application that enables multiple users and robots to collaboratively assemble a complex wooden structure.
Students: Eleni Vasiliki Alexi, Joseph Clair Kenny
Thesis Supervisors: Dr. Daniela Mitterberger GKR, Gonzalo Casas GKR
Exploring Architectural Expressions through Discrete Design and Reinforcement Learning
The general problem addressed in this thesis lies in the limited exploration of aesthetic and design potentials enabled by Reinforcement Learning (RL) within the fields of architecture, engineering, and digital fabrication. The specific research question focuses on how the fusion of Discrete Design methodologies with Reinforcement Learning can inspire innovative architectural expressions that prioritize unconventional spatial relationships and human-centric living spaces. To answer this question, the research introduces a novel human-machine collaborative design workflow using a Hierarchical Reinforcement Learning approach with multiple RL-based agents trained to optimize and innovate within design spaces based on designer input and reward signals. The research demonstrates that integrating Discrete Design with RL leads to unconventional architectural designs emphasizing interconnected functional spaces, voids, and unique human-centric qualities. This answer is particularly important for architects, designers, and urban planners looking to advance creative solutions that optimize space and enhance the quality of life for occupants. Finally, the findings inspire future research in refining RL-based design workflows and facilitating interdisciplinary collaborations for large-scale urban planning.
Student: Ahmed Elmaraghy
Thesis Supervisors: Dr. Anton Savov dbt, Hang Zhang dbt, Mark Tam BRG
Designing with Deep Reinforcement Learning
As the growing interest in artificial intelligence and digital fabrication has risen, utilizing AI to generate habitability-centered and fabrication-aware designs with discrete approaches remains an open question.
The focus of this research is to incorporate deep reinforcement learning and voxelized space generation to generate novel design solutions that comply with spatial affordance.In this model, a given size of space is voxelized to allow the RL-agent to carve out spatial configurations consisting of primitive actions from studies of human-affordance. Composite neural networks are used to learn the task with the objective to maximize the cumulative reward given by satisfying spatial affordance in such an environment.This research contributes to the ongoing discourse on integrating AI in architecture, digital fabrication, and construction. We recognize the potential of voxelization to incorporate complex spatial configuring tasks and the combination of DRL actions with architectural anthropometric studies.
Students : (Hill) Yixiao Huang
Thesis Supervisors: Dr. Anton Savov dbt, Hang Zhang dbt, Mark Tam BRG
Digital Fabrication of Biologically Active Cement Spatial Structures
This research explores the integration of microbial biocementation and robotic granular 3D printing techniques to advance the creation of cementitious structures with intricate geometries for architectural applications.
The main goal is to investigate bio-fabrication methodologies that can accommodate, and eventually enhance, bacteria activity by exploring diverse geometry configurations facilitated by robotic fabrication and computational design approaches. First, the research analyzes the microbial bio-cementation procedure, considering species selection and environmental conditions. Secondly, the study investigates fabrication process aspects such as material deposition, tool and robotic path planning. Lastly, extraction and curing methods for post-fabrication process were explored. With these investigations, we aim to contribute to the field of bio-fabrication in architecture by exploring the potential of integrating microbial bio-cementation, computational design, and robotic fabrication.
Student: Nijat Mahamaliyev
Thesis Supervisors: Andrea Ling dbt, Karen Antorveza dbt
Technical staff: Tobias Hartmann dbt; project intern: Georg Bauer TU Wien
Brick Design 2.0
In the past fifteen years a growing interest in digital bricks has risen as a consequence of infrastructure evolution in both design and fabrication. Nevertheless, despite the development of numerous tools and plug-ins within this context, the emphasis has predominantly been on research, with limited practical applications having been realized. The focus of this research is to cater to practical requirements and integrate relevant state-of-the art academic research to deliver an advanced tool for practitioners.
The primary objective of this plug-in is to provide a parametric design environment, enabling a robust yet flexible build-up of brick facade geometries. The plug-in will be Python-based and leverage the COMPAS infrastructure by creating a dedicated brick library, enhancing the plugin's maintainability, reusability, and potential for future development. By incorporating relevant data for robotically fabricated brick wall, the geometrical model is therefore turned into an informed model, thus making it ready for IFC interoperability and exchange.
Students: Adam Anouar
Thesis Supervisors: Chen Kasirer GKR, Gonzalo Casas GKR, Dr. Petrus Aejmelaeus-Lindström GKR
This thesis explores the immersive exploration of numerous architectural designs using VR and the potential of motion tracking technologies. Initially the research aimed to research the possibility of enabling two users (an architect and a client) to simultaneously and intuitively explore multiple design alternatives and investigate techniques by remapping the visualized space to the physical boundaries of the Immersive Design Lab. The potential of OptiTrack and VR backpack for architectural visualization and client feedback recording has been evaluated. The research focuses on exploring the transition from 2D design space exploration (optioneering methods) to immersive 3D design space exploration. Prototype scenarios for design review and client feedback are developed using VR technology. The objective is to experiment with existing or novel user interfaces in order to enhance the immersive VR experience and provide suitable user interfaces for architectural visualization and client feedback recording.
Student: Yuki Xue Chen, Etienne Pavoncello
Thesis Supervisors: Dr. Anton Savov dbt, Wenqian Yang, dbt
Circular Material Inventory for Circular Design Generation
In the realm of circular construction practices, a notable gap exists, particularly regarding the limited timeframe for design generations with reclaimed materials and the focus on facade element reuse. The challenge lies in the absence of a comprehensive system and 3D data that can facilitate seamless reuse within existing digital workflows, especially the early design phase and the 3D design processes. To address this, the thesis proposes a solution through research that explores the use of an enhanced material database and machine learning tools. The resulting platform serves as a user-friendly tool for informed decision-making on designing with reclaimed materials, thereby contributing to sustainable design practices. This study also sets the stage for future investigations into different material genres and generative design methods, ultimately promoting the circular design platforms within the construction industry.
Student: Jan Law, Kevin Chang
Thesis Supervisors: Beril Önalan CEA, Simon Griffioen dbt, Dominik Reisach CHP / dbt, Yael Ifrah dbt
Computational Design and Assembly of Infinitely Reusable Kit of parts
This project explores methods to build fully reversible and multipurpose space frames by combining standardized components (1 m wooden sticks and couplers) with computational design tools and Augmented Reality fabrication tools. The computational design tool is based on the “Frame X” algorithms, which automatically optimize the composition of couplers and bars.The geometry is computationally generated then the algorithm is recalculating the structure with offsetting the bars and coupler rotating. For the on site assembly process, the geometry is exported to HoloLens which is assisting user locating parts.
Both chairs and full scale pavilions have been realized with the system which demonstrates both the flexibility of the design and assembly approach as well as the robustness of the FrameX algorithm and shows that this system enables circular construction.
Students: Issue Yi Hsiu Hung, Chenming Jiang
Thesis Supervisors: Dr. Ziqi Wang CRL, Dr. Yijiang Huang CRL, Aurèle Gheyselinck GKR, Dr. Petrus Aejmelaeus-Lindström GKR
Fungal Flexible Formworks
There is an increasing need for bio-based solutions in architecture and the need to find alternative materials due to the construction industry’s heavy reliance on finite resources. Among all current practices, mycelium based composites are a promising alternative. When replacing conventional building materials, there are 2 ways of doing this, complete material substitution, or material integration. Mycelium has the capacity to bind agricultural and industrial waste and is biodegradable, however their use is limited to non-structural application and with standardised techniques such as moulding. This research proposes integration of fungi-based bio-composites with textile formworks. We aim to produce 2 full-scale prototypes demonstrating the concepts of substitution and integration with concrete. We hope to expand the design space and propose a novel fabrication methodology for laminated fungal composites.
Students : Abhipsa Pal, Karunadhipathi Lihin Weera
Tutors: Tiziano Derme dbt, Francesco Ranaudo BRG
This research investigates on printing a series of facade panels by using ABB gantry robot arms, which are integrated with thermal functionality and aesthetic. In this project, several combinations of different subdivisions of patterns and inner cavities will be applied to examine the U-value of each panel. The optimal solution is developed in a final demonstrator, consisting of a multi-panels, 1:1 facade mock-up of dimensions roughly around 2m x 2m. The objective of this research is to conceive and fabricate a 3D-printed facade panel with thermoplastic polyester that seamlessly integrates thermal function and aesthetic considerations into an unified mono-material composition.
Students: Brian Chen, Katie Gallagher
Thesis Supervisors: Francesco Milano GKR, Valeria Piccioni NCCR Digital Fabrication