Skip to main content
main-content
Top

About this book

The book presents research from Rob|Arch 2018, the fourth international conference on robotic fabrication in architecture, art, and design. In capturing the myriad of scientific advances in robotics fabrication that are currently underway – such as collaborative design tools, computerised materials, adaptive sensing and actuation, advanced construction, on-site and cooperative robotics, machine-learning, human-machine interaction, large-scale fabrication and networked workflows, to name but a few – this compendium reveals how robotic fabrication is becoming a driver of scientific innovation, cross-disciplinary fertilization and creative capacity of an unprecedented kind.

Table of Contents

Frontmatter

Design and Simulation

Frontmatter

Image Classification for Robotic Plastering with Convolutional Neural Network

Inspecting robotically fabricated objects to detect and classify discrepancies between virtual target models and as-built realities is one of the challenges that faces robotic fabrication. Industrial-grade computer vision methods have been widely used to detect manufacturing flaws in mass production lines. However, in mass-customization, a versatile and robust method should be flexible enough to ignore construction tolerances while detecting specified flaws in varied parts. This study aims to leverage recent developments in machine learning and convolutional neural networks to improve the resiliency and accuracy of surface inspections in architectural robotics. Under a supervised learning scenario, the authors compared two approaches: (1) transfer learning on a general purpose Convolutional Neural Network (CNN) image classifier, and (2) design and train a CNN from scratch to detect and categorize flaws in a robotic plastering workflow. Both CNNs were combined with conventional search methods to improve the accuracy and efficiency of the system. A web-based graphical user interface and a real-time video projection method were also developed to facilitate user interactions and control over the workflow.

Joshua Bard, Ardavan Bidgoli, Wei Wei Chi

Designing Natural Wood Log Structures with Stochastic Assembly and Deep Learning

Advances in contemporary 3D scanning and bespoke robotic technologies have enabled architectural structures to be directly constructed from naturally grown wood. However, design precedents using natural wood logs are still dominated by design approaches using predefined geometric models. The limit of this approach lies in the necessity to model the form of every structural member based on the captured geometries of all the materials before design begins. Moreover, human designed rules for joining irregular components are limited and solutions are prone to be limited by empirical knowledge. In this paper, we introduce a method for assembling natural wood log structures with higher goals autonomously using robotic stochastic assembly and deep learning. The novelty of this method is that the design of structures does not rely on prior-knowledge of the to-be-assembled materials but is generated by assembling materials iteratively. A vision system with a position suggestion network based on convolutional neural networks (CNNs) was implemented and trained to drive an industrial robotic arm for negotiating between the topological changes from potential connections and the local assembly constraints of the log. A robotic hand-eye coordination database recording the assembly of birch logs has been established and small-scale wood structures were built by the trained robot. Results show that our robot can find desired structural configurations autonomously and can assemble unfamiliar batches of wood logs. The cost and gain of using stochastic assembly and deep learning as a design strategy are discussed and future research on using different learning strategies and large-scale implementations are laid out.

Kaicong Wu, Axel Kilian

Mockup Method: Heuristic Architectural Fragments as Central Models in Architectural Design

Standardization and increased specialization have slowly begun to separate the means and methods of making from the process of architecture. The introduction of digital tools towards the latter half of the century have functioned to further this divide, removing any remaining traces of materiality and scale. Accordingly, architectural design exploration primarily resides in the creation and modification of digital objects, which must then be translated into the physical world. This positions built architecture in a curious position of constant catch up, chasing the impossible ideal of its digital counterpart. However, the tools predominant in architectural design and fabrication today (CAD, CAM) may be appropriated, along with sensory feedback, towards the development of a new material workflow. This paper presents a prototypical workflow which combines computational methods and robotic fabrication techniques with the spontaneity of the human and the messiness and contingency of material. The workflow is tested through the design of 1:1 heuristic architectural fragments.

Kevin Pazik

Haptic Programming

Current industrial robotics focuses on the utilization within clearly defined and structured production environments. However due to increasing product variety, a paradigm shift away from repetition of static task towards dynamic human-robot collaboration is noticeable. Due to the fact that static automation can only be achieved at a prefabrication level within the construction industry, this shift towards adaptable robotics can be utilized for new concepts for on-site robotic assistance. We extensively illustrate our approach towards robotics that adapts to changing environmental conditions and material features, while retaining a degree of predictability necessary for effective collaboration. Furthermore, by integrating human-robot collaboration with parametric modelling a feedback to design is established. The term haptic programming is coined in order to illustrate the direct interconnection between parametric model and human-robot collaboration. First application examples are shown to illustrate the use of a priori knowledge from the design phase in combination with haptic interaction primitives to enable intuitive human-robot collaboration. Haptic programming allows the exchange of knowledge between the user and a robot on a physical level.

Sven Stumm, Sigrid Brell-Çokcan

Towards Automatic Path Planning for Robotically Assembled Spatial Structures

This paper discusses the integration of automatic robot path planning into the computational design environment. A path planning software interface is presented that allows to support fabrication-aware design of robotically assembled structures with discrete elements. Using the large-scale Robotic Fabrication Laboratory (RFL) as test-bed, the software interface is validated through three experiments, in which building members need to be guided around obstacles and which are fabricated using two cooperative robotic arms. Specific focus of this paper is the investigation of strategies to narrow the path search by adjusting design and path planning parameters in order to achieve a calculation time that is suitable for design applications. A close integration of automatic path planning and design is presented, which does not only enable the negotiation between design intention and fabrication feasibility, but allows for an understanding of the constraints present in robotically fabricated spatial structures. Thus, this research contributes to expand these structures’ design and fabrication space.

Augusto Gandia, Stefana Parascho, Romana Rust, Gonzalo Casas, Fabio Gramazio, Matthias Kohler

Communication Landscapes

In this paper we present an installation which explores a robotic fabrication process through real time human-machine interaction using natural interfaces. Aiming to contribute to the debate regarding automatization processes described by the industry 4.0, we suggest a future collaborative approach to distributed and participatory design. The installation invites participants to shape a physical object through a robot that interprets the acoustic signal of their voices. The object is then fabricated in near real-time through robotic hotwire cutting. The visitor is gifted with the result of its exploration to take home as a souvenir of a possible future. The negative traces of voices are aggregated through an algorithm in a sculptural wall.

Giovanni Betti, Saqib Aziz, Andrea Rossi, Oliver Tessmann

Towards Visual Feedback Loops for Robot-Controlled Additive Manufacturing

Robotic additive manufacturing methods have enabled the design and fabrication of novel forms and material systems that represent an important step forward for architectural fabrication. However, a common problem in additive manufacturing is to predict and incorporate the dynamic behavior of the material that is the result of the complex confluence of forces and material properties that occur during fabrication. While there have been some approaches towards verification systems, to date most robotic additive manufacturing processes lack verification to ensure deposition accuracy. Inaccuracies, or in some instances critical errors, can occur due to robot dynamics, material self-deflection, material coiling, or timing shifts in the case of multi-material prints. This paper addresses that gap by presenting an approach that uses vision-based sensing systems to assist robotic additive manufacturing processes. Using online image analysis techniques, occupancy maps can be created and updated during the fabrication process to document the actual position of the previously deposited material. This development is an intermediary step towards closed-loop robotic control systems that combine workspace sensing capabilities with decision-making algorithms to adjust toolpaths to correct for errors or inaccuracies if necessary. The occupancy grid map provides a complete representation of the print that can be analyzed to determine various key aspects, such as, print quality, extrusion diameter, adhesion between printed parts, and intersections within the meshes. This valuable quantitative information regarding system robustness can be used to influence the system’s future actions. This approach will help ensure consistent print quality and sound tectonics in robotic additive manufacturing processes, improving on current techniques and extending the possibilities of robotic fabrication in architecture.

Sheila Sutjipto, Daniel Tish, Gavin Paul, Teresa Vidal-Calleja, Tim Schork

Function Representation for Robotic 3D Printed Concrete

The use of Function Representation (FRep) to synthesise and specify geometries for 3D printing is finding renewed interest. The usefulness and extension of this representation in the synthesis and analysis of geometries for the process of large-scale, layered concrete 3D printing has been previously articulated by the authors. This paper fully extends the implicit representation used previously in shape-design to fabrication-related processing of compressive skeletal structures for realisation by robotic 3D printing of concrete. In particular, we use an initial value formulation of a propagating front to process the nodes and bars of a given funicular spatial structure.

Shajay Bhooshan, Johannes Ladinig, Tom Van Mele, Philippe Block

Material and Processes

Frontmatter

Thermally Informed Robotic Topologies: Profile-3D-Printing for the Robotic Construction of Concrete Panels, Thermally Tuned Through High Resolution Surface Geometry

This paper explores the thermal design and robotic construction of high-performance building components. The complex surface geometry of these components actuate specific thermal behavior in passive building systems through implementing the principles of convection in thermal mass. Our seamless design-to-fabrication workflow uses optimization methods that combine measured thermal data and simulation feedback with advanced modeling and emerging robotic manufacturing techniques. Bridging an understanding of thermal performance, geometry, and manufacturing we suggest direct formal relationships between the behavior of airflow to tool-path planning for a robotic arm. This paper will focus on describing an experimental process we term Profile-3D-Printing that demonstrates a novel approach to the construction of concrete panels with complex surface geometries. This hybrid construction method combines material deposition with tooled post-processing to achieve high-resolution surface definition. The process entails automated delivery of material for selective deposition of panel geometry, and tooled shaping of rough and finish layers for the physical production of computationally generated forms.

Joshua Bard, Dana Cupkova, Newell Washburn, Garth Zeglin

Hold Up: Machine Delay in Architectural Design

This paper introduces an architectural design approach that is founded on working with digital fabrication machines, materials, and time. This approach is called Machine Delay Fabrication. Machine Delay Fabrication is contextualized within the lineage of productive delays using two examples from other creative disciplines. It is then contrasted with the “real-time” approach to digital fabrication research and practice. The concept of machine delay is outlined along with the three types of machine delay that are pertinent to digital fabrication. A new concrete 3D printing method called pointillistic time-based deposition, or dripping, is introduced as an example of Machine Delay Fabrication. The setup and variables of dripping are discussed, along with some experimental findings. Finally, dripping is used to demonstrate the constructive and aesthetic possibilities of Machine Delay Fabrication in architectural design.

Zach Cohen

Concrete Fabrication by Digitally Controlled Injection

This paper explores the volumetric modification of material properties through precise injection of chemical agents as a new method of concrete fabrication. We inject concrete while in its slurry state with a hydrated aluminium solution, triggering an effervescent reaction. This reaction causes changes in material density which affect the structural integrity and visual characteristics of the end-product once it is set. The process is administered by a purpose-built end-effector attached to an industrial robotic system and guided by a parametric design system. Being able to produce localized differentiation, suggests for an innovative approach to a fabrication technique of texturing, 2D and 3D forming of concrete.

Ryan Wei Shen Chee, Wei Lin Tan, Wei Hern Goh, Felix Amtsberg, Stylianos Dritsas

Towards the Development of Fabrication Machine Species for Filament Materials

The research presented in this paper explores the concept of deploying collaborative heterogeneous robot systems where machines are working together towards a common fabrication goal. Augmenting or replacing existing industrial-robot fabrication processes with task-specific architectural construction machines has the potential to expand the design space of digital fabrication methods beyond the limitations of previously existing strategies. The proposed system implies the development of a library of hardware solutions as well as a digital control tool to enable successful execution of fabrication tasks.This research is focusing on heterogeneous mobile robotic fabrication strategies specific to filament materials. Deploying smaller robots for manipulation of lightweight thread-like materials allows building significantly larger structures. Multiple task-specific machines developed in this research are designed to carry, manipulate, anchor and pass filament materials in an on-site architectural environment of interior space. This paper presents the current state of the catalogue of robot species developed in this research as well as the experiments and demonstrators performed to evaluate them. Ultimately this research aims to create a larger toolbox of hardware and software tools and methods for heterogeneous teams of custom single-task fabrication and construction robots.

Maria Yablonina, Achim Menges

Spatial Print Trajectory

Controlling Material Behavior with Print Speed, Feed Rate, and Complex Print Path

Current digital clay fabrication techniques comply with the innate material behavior of clay by extruding in two-dimensional layers. This method inevitably uses an excess amount of material and is a time-consuming process that does not take advantage of the viscous properties of clay. However, by utilizing spatial print trajectories with embedded print parameters (e.g. print speed and extrusion rate), the extrusion behavior of the material can be controlled via simulating actions like anchor, drag, and pull of the clay at the nozzle tip. The aforementioned spatial print trajectory can then form a voxel that can be heterogeneously controlled in order to quickly form self-supporting complex geometries with different density, macro-porosity, and structural rigidity. The print path can also be scaled up to exploit the potential of digital fabrication at the construction scale.

Sulaiman AlOthman, Hyeonji Claire Im, Francisco Jung, Martin Bechthold

An Additive and Subtractive Process for Manufacturing with Natural Composites

We present research work on a manufacturing process deploying natural composite materials. The objective of the project is to create a sustainable manufacturing process integrating materials, hardware, software and fabrication logic from the ground up. We deploy a bioinspired natural composite comprised by renewable, widely available, biodegradable and low-cost natural components. Material properties closely resemble those of high-density foams or low-density timbers and it is produced without any petrochemical or harmful solvents associated with adverse environmental effects. We designed a mobile material deposition system using the Direct Ink Writing method, with work envelope of over 3 m vertically and indefinite horizontal range, comprised of industrial robotic hardware and purpose-built mechanical mobile platform. We performed testing in characterizing material properties with and without the introduction of the printing process, tightly integrated material behavior with manufacturing and developed design software for direct transition from design to production. To address scaling, we approached the fabrication process from the perspective of fusing the best principles from both additive and subtractive manufacturing, offering geometric freedom and material efficiency of additive manufacturing while targeting production and quality efficiencies of subtractive and forming processes. We believe this process has the potential of significant impact on general manufacturing as well as the building industry.

Stylianos Dritsas, Yadunund Vijay, Marina Dimopoulou, Naresh Sanadiya, Javier G. Fernandez

Hard + Soft: Robotic Needle Felting for Nonwoven Textiles

This project explores the development of an additive manufacturing technique for nonwoven textiles. Nonwoven textiles, based on natural materials, synthetic polymers, or blends of the two, have numerous performative aspects, including excellent acoustic absorption, thermal insulation, and tactile characteristics. Felt is a typical example of a nonwoven material, and can be manufactured by both wet or dry processes. One example of a dry process involves needle felting, whereby fibers of the textile are meshed and entangled when punched together. This process binds the material together seamlessly without the addition of sewn thread or adhesives. Needle felting can range in scale from hand craft techniques with a single needle to large scale web processing. Integration into a robotic process not only enables precision and speed in manufacturing but also extends needle felting as a three-dimensional process, allowing for local differentiation of stiffness and other properties across a homogeneous solid. Through a customized digital workflow, formal and material properties can be varied at local level within a component. By developing a fully integrated design to production methodology for influencing these properties, this research opens a wide range of potentials for nonwoven textiles in architectural applications. The research involves three areas of development; the process tooling for robotic felting, the digital workflow that enables the formal and material properties to be specified computationally and embedded into the machine code, and prototypes of architectural elements such as acoustic panels and furniture demonstrating different techniques and processes.

Wes McGee, Tsz Yan Ng, Asa Peller

Construction and Structure

Frontmatter

SCRIM – Sparse Concrete Reinforcement in Meshworks

This paper introduces a novel hybrid construction concept, namely Sparse Concrete Reinforcement In Meshworks (SCRIM), that intersects robot-based 3D Concrete Printing (3DCP) and textile reinforcement meshes to produce lightweight elements. In contrast to existing 3DCP approaches, which often stack material vertically, the SCRIM approach permits full exploitation of 6-axis robotic control by utilising supportive meshes to define 3D surfaces onto which concrete is selectively deposited at various orientation angles. Also, instead of fully encapsulating the textile in a cementitious matrix using formworks or spraying concrete, SCRIM relies on sparsely depositing concrete to achieve structural, tectonic and aesthetic design goals, minimising material use. The motivation behind this novel concept is to fully engage the 3D control capabilities of conventional robotics in concrete use, offering an enriched spatial potential extending beyond extruded geometries prevalent in 3DCP, and diversifying the existing spectrum of digital construction approaches. The SCRIM concept is demonstrated through a small-scale proof-of-concept and a larger-scale experiment, described in this paper. Based on the results, we draw a critical review on the limitations and potentials of the approach.

Phil Ayres, Wilson Ricardo Leal da Silva, Paul Nicholas, Thomas Juul Andersen, Johannes Portielje Rauff Greisen

Versatile Robotic Wood Processing Based on Analysis of Parts Processing of Japanese Traditional Wooden Buildings

Processing parts of wooden buildings by robots and computerized numerical control (CNC) machines has been attempted in recent years. However, the shapes of the processed parts are constrained by the machines and attached tools. Modified parts cannot be applied to traditional wooden buildings, as the shapes of the parts are an aspect of the tradition. This paper introduces a method of versatile processing of wooden building parts using an articulated robot, based on the analysis of part shape and carpentry tools used in Japanese wooden building construction. The goal is to process the original part shape of the construction method as is, without optimizing it for the robot and machining. We analyze Japanese carpentry tools and processing methods and propose a method to flexibly process wooden building parts using a circular saw, square chisel, vibration chisel, and router. Then, using robots employing these tools, we process parts of a five-storied pagoda of a Japanese traditional wooden building. We consider flexible processing of the wooden building parts.

Hiroki Takabayashi, Keita Kado, Gakuhito Hirasawa

Form Finding of Nexorades Using the Translations Method

The aim of this paper is to discuss the dialectic form-finding of a complex timber structure based on an innovative structural system: shell-nexorade hybrids. Nexorades, also known as reciprocal frames are elegant structures that suffer from a relatively poor structural behavior due to in-plane shear and bending of the members. Introducing plates as bracing elements significantly improves their performance, but increases the manufacturing complexity and sets high tolerance constraints. We present the fabrication and assembly of a 50 m2 timber pavilion with 6-axis robotized milling. The use of a mobile robot and fixed machining stations is explored to allow for maximal flexibility of iterations between design and fabrication.

Tristan Gobin, Romain Mesnil, Cyril Douthe, Pierre Margerit, Nicolas Ducoulombier, Leo Demont, Hocine Delmi, Jean-François Caron

Sub-Additive 3D Printing of Optimized Double Curved Concrete Lattice Structures

The research presented in this paper investigates architectural-scale concrete 3D printing for the fabrication of rapidly-constructed, structurally-optimized concrete lattice structures. Sub-Additive Manufacturing utilizes a three-dimensional tool path for deposition of material over a mechanically-shaped substructure of reusable aggregate. This process expedites the production of doubly-curved concrete form by replacing traditional formwork casting or horizontal corbeling with spatial concrete arching. Creating robust non-zero Gaussian curvature in concrete, this method increases speed over typical concrete fabrication practices. Utilizing robotics to integrate a streamlined workflow from digital design to physical fabrication, Sub-Additive leverages digital workflows to produce structurally, materially, and spatially optimized building components while dramatically reducing waste material. Addressing digital form finding and optimization, material behaviors (both concrete and supportive aggregate), nozzle design and novel utilization of robotic fabrication, this paper introduces a series of key concepts for Sub-Additive Manufacturing, radically advancing concrete 3D printing at full scale.

Christopher A. Battaglia, Martin Fields Miller, Sasa Zivkovic

Investigations on Potentials of Robotic Band-Saw Cutting in Complex Wood Structures

In the field of wood manufacturing, CNC milling seems to be the only way to deal with geometrically complex wood components. Robotic band-saw, the combination of robotic fabrication and band-saw cutting provides a feasible solutions for producing curved wood beams without the immense time consumption of CNC milling method. This paper explores potentials of robotic band-saw cutting technology in the construction of complex wood structures. The research processes of two large-scale research pavilions, from structure design and optimization, Glulam production, robotic fabrication to site assembly, are presented in this paper to explore an integrated workflow of complex wood structures construction based on robotic band-saw. As the two pavilions show, robotic band-saw technology has the ability to perform high efficiency and accuracy both in the plane and space curved wood elements, and shows the potential to take over part of the work of CNC in wood construction.

Hua Chai, Philip F. Yuan

Direct Deposition of Jammed Architectural Structures

The research presented in this paper investigates novel techniques and tools for robotic fabrication of fibre reinforced granular structures built without any type of formwork. Combining granular jamming with strategically placed, continuous reinforcement, allows for precise fabrication of Jammed Architectural Structures (JAS) out of crushed rock and string that are fully recyclable. By further combining the material system with computational design and robotic fabrication enable to place and compact materials directly where needed, without formwork, allows building bespoke architectural structures in an additive, fully reversible and waste-free manner. The paper describes techniques and challenges for implementing a robotic direct deposition of JAS for fabrication of building-scale and loadbearing artefacts, and highlights the prospects of in-situ fabrication that is exemplified by two one-story tall prototypes. The robot direct depositions the material system in space, without formwork: first, a robot places string in loop-based pattern; second, it places small piles of aggregates inside the string loops, on top of the previously aggregated material; third, it carefully packs the piles and activates the string trough tension.

Petrus Aejmelaeus-Lindström, Gergana Rusenova, Ammar Mirjan, Fabio Gramazio, Matthias Kohler

Control and Fabrication

Frontmatter

FIBERBOTS: Design and Digital Fabrication of Tubular Structures Using Robot Swarms

Rapidly increasing demands to enhance speed and safety for on-site and site-specific construction drive the need to develop collaborative and autonomous systems. Such systems generally involve the use of retrofitted gantry and robot arm systems, which have size constraints and are computationally complex to use in collaboration with other machines. This paper describes an alternative multi-robot system built from the ground up to enable collaborative and site-specific construction. The strategy simplifies design workflows while simultaneously maintaining structural, environmental, and robot dynamic constraints. This system of ‘swarm fabricators’ enables robotic agents to operate in parallel, digitally fabricating independent tubular forms. Each robot controls its position allowing the system to effectively ‘grow’ a large-scale woven architecture. The robots fabricate by pulling fiber and resin from ground-based storage and winding a composite around their own bodies. Additional sections of a composite tube are appended upon each other, starting from a base. The system relies on an environmentally informed flocking-based strategy to design the structure and inform the robots’ trajectories.

Markus Kayser, Levi Cai, Christoph Bader, Sara Falcone, Nassia Inglessis, Barrak Darweesh, João Costa, Neri Oxman

InFormed Ceramics: Multi-axis Clay 3D Printing on Freeform Molds

This paper addresses a novel technique of 3D clay printing using a robotic arm along with the architectural applicability of the proposed fabrication technique. In the initial phase of research, we selected the clay material to be used for printing in consideration of its viscosity and developed the clay extruder customized for the material. In the late phase, to verify how effective the robotic fabrication technique we propose is, we manufactured the freeform foam mold and tested multi-axis clay 3D printing on it by using three end-effectors consecutively: hotwire cutter, spindle, and clay extruder. Also, applying the proposed technique to a larger-scale ceramic structure consisting of nineteen panels, we checked the feasibility of fabricating actual ceramic building components with this technique. If this robotic clay 3D printing technique is improved more and fulfills the demands of the industry in future, it will enable cost-effective ceramic modular fabrication for a customized design in architecture.

Minjae Ko, Donghan Shin, Hyunguk Ahn, Hyungwoo Park

Altered Behaviour: The Performative Nature of Manufacture Chainsaw Choreographies + Bandsaw Manoeuvres

This paper explores the ventures of the Robotic Fabrications AA Visiting School. Cultivating and implementing radical and innovative modes of architectural fabrication, we borrow from neighbouring creative fields such as choreography, performance and martial arts - efficient code being infected with an exciting spatial complexity and positing an artistic engagement within the world of physical production.

Emmanuel Vercruysse, Zachary Mollica, Pradeep Devadass

Cyber Physical Macro Material as a UAV [re]Configurable Architectural System

The research presented in this paper describes a new strategy for deploying a cyber-physical macro material combined with aerial construction robots as a reconfigurable architecture. The resulting architectural system is capable of autonomous rearrangement and stagnant operation driven by behavioral design patterns rather than a singular robotic assembly or construction method. The central part of the paper describes the functional framework for applying the system to a self-supporting roof canopy where the material is comprised of discrete, lightweight, polyhedron shaped, carbon fiber units with onboard sensing, processing, and communication. The material is autonomously reconfigured by calling and coupling with task-specific UAV builders. Programming and hardware for intra- and inter-material, and builder communication and user interaction are described. The system’s live decision making, interaction, and continuous reconfiguration based on programmatic input are outlined. The cyber-physical framework is produced and shown in 1:1 functional prototype.Three potential behavioral scenarios applying the system as a new type of architecture are outlined: adaptive behavior, interactive behavior, and learning behavior. Further, the programming of the existing framework using machine learning and artificial intelligence is proposed as a future development. In conclusion, developments of the system and the impact and disruption of architectural design, planning, and construction processes are proposed.

Dylan Wood, Maria Yablonina, Miguel Aflalo, Jingcheng Chen, Behrooz Tahanzadeh, Achim Menges

Adaptive Robotic Carving

Training Methods for the Integration of Material Performances in Timber Manufacturing

The paper presents the developments of a series of methods to train a fabrication system for the integration of material performances in timber manufacturing processes, combining robotic fabrication together with different sensing strategies and machine learning techniques, and their further application within a prototypical design to manufacturing workflow. The training cycle, spanning from the recording of skilled human experts to autonomous robotic explorations, aims to encapsulate different layers of instrumental knowledge into a design interface, giving designers the opportunity to engage with material and tool affordances as process driver. The training methods are evaluated in a series of experiments and design iterations, proving their potential in the development of customized design to manufacturing workflows and integration of material performances, with a specific focus on timber.

Giulio Brugnaro, Sean Hanna

Multimode Robotic Materialization

Design to Robotic Fabrication Method of Integrating Subtractively Produced Hard Components and Additively Deposited Soft Silicone

This paper explores and discusses the implementation of multimode, subtractive-additive, robotic production methods. The research addresses the opportunities and challenges in the integration of subtractive and additive robotic manufacturing routines for architectural scale applications. It specifically presents a customized computational design to robotic fabrication workflow of a material system composed of subtractively produced expanded polystyrene and additively manufactured silicone. The integration of these two sets of production methods allows for a symbiosis of hard and soft materials. Through initial case studies and a pilot subtractive-additive prototype, this paper discusses how multimode robotic production methods, with the goal of purposefully expanding the potential of isolated fabrication techniques, results in multiscale and efficient design materialization. Furthermore, this paper demonstrates the design, computation, materialization, and prototyping process of a chair project with multiple materials in more detail.

Sina Mostafavi, Benjamin N. Kemper, Daniel L. Fischer

Digital Composites: Robotic 3D Printing of Continuous Carbon Fiber-Reinforced Plastics for Functionally-Graded Building Components

Carbon fiber-reinforced thermoplastic (CFRP) can impact the built environment significantly. In addition to its extraordinary weight-to-strength ratio, CFRP provides numerous benefits in architecture, including its plasticity and aesthetics. In theory, CFRP can take any shape, but the fabrication constraints of mold making currently limit the design. The mold is the most resource-intensive element of CFRP fabrication, and it is particularly challenging and costly for the bespoke non-standard shapes, which are often found in architecture. This research presents a novel approach to utilizing robotic three-dimensional printing (3DP) to make CFRP materials accessible in architecture by combining layered 3DP of neat polymers with add-on 3DP of CFRP materials. This paper begins by providing background information and an overview of state-of-the-art CFRP fabrication. It then outlines the challenges related to the fabrication process and design of 3DP CFRP building components. Taking into account the design parameters, material properties and fabrication constraints, the paper describes the fabrication process development, computational design tool development and demonstrative physical experiment. The presented research is the first part of an interdisciplinary research involving architecture and mechanical engineering in order to investigate the potential for large-scale mold-less fabrication of CFRP components.

Hyunchul Kwon, Martin Eichenhofer, Thodoris Kyttas, Benjamin Dillenburger

Robotic Extrusion of Architectural Structures with Nonstandard Topology

This paper presents a fast and flexible method for robotic extrusion (or spatial 3D printing) of designs made of linear elements that are connected in nonstandard, irregular, and complex topologies. Nonstandard topology has considerable potential in design, both for visual effect and material efficiency, but usually presents serious challenges for robotic assembly since repeating motions cannot be used. Powered by a new automatic motion planning framework called Choreo, this paper’s robotic extrusion process avoids human intervention for steps that are typically arduous and tedious in architectural robotics projects. Specifically, the assembly sequence, end-effector pose, joint configuration, and transition trajectory are all generated automatically using state-of-the-art, open-source planning algorithms developed in the broader robotics community. Three case studies with topologies produced by structural optimization and generative design techniques are presented to demonstrate the potential of this approach.

Yijiang Huang, Josephine Carstensen, Lavender Tessmer, Caitlin Mueller

On-Site Robotics for Sustainable Construction

Although additive and robotic manufacturing, is considered a technology with lots of potentials in the construction industry, its deployment has not yet reached wide applications for on site construction of sustainable architectural structures. This paper focuses on the deployment of a 3D printing technology that combines robotics with natural materials for the construction of environmentally performing small scale buildings. Cable robots are explored for 3D printing with adobe, while drones are explored for real-time monitoring technologies of the construction process. A full scale prototype of the technology has been deployed during 15 days at an international construction fair, demonstrating its potential by producing live a 20 m2 pavilion. The paper describes and discusses the experience, feasibility and limitations of the technology operating on site and in a direct collaboration with human operators and craftsmen. The prototype demonstration presented in the paper has led to the conclusion that there is a significant potential of using the technology for large scale sustainable architectural constructions on-site.

Alexandre Dubor, Jean-Baptiste Izard, Edouard Cabay, Aldo Sollazzo, Areti Markopoulou, Mariola Rodriguez

Application and Practice

Frontmatter

Tailored Structures, Robotic Sewing of Wooden Shells

This paper investigates the use of robotics with sensing mechanisms in combination with industrial sewing techniques to explore new strategies for the fabrication of thin wooden shells.The investigation is characterized by a parallel theoretical and prototype-based methodology, the latter serving as a vehicle to further the technical development, which could ultimately enable novel architectural qualities.The development unfolds in four interdependent avenues: (1) The transfer of textile patterning techniques used in garment production to inform the design of flexible 3 mm beech plywood segments; (2) The capacity of wood to be elastically bent and connected into geometrically stable structures; (3) the use of sewing as a new construction joint for thin material; and (4) The integration of sewing into an automated and adaptive robotic fabrication workflow enabled by sensing and scanning with the capacity to join complex three-dimensional curved structures at an architectural scale.

Martin E. Alvarez, Erik E. Martínez-Parachini, Ehsan Baharlou, Oliver David Krieg, Tobias Schwinn, Lauren Vasey, Chai Hua, Achim Menges, Philip F. Yuan

Dynamic Robotic Slip-Form Casting and Eco-Friendly Building Façade Design

Robot arm technology has been increasingly considered a future solution for the building industry, which is keen to develop eco-friendly and labor-efficient processes. The automobile industry provides a powerful reference for such a revolution. However, in contrast to the automobile industry, which mostly uses standard industrial materials, such as metal, the building industry still relies on a variety of traditional materials, including concrete cement, brick and clay. Although precast building components have been highly promoted, they present innate constraints, including limitations associated with transportation, storage, and installation capacity, which greatly impact the overall budget. In this paper, research on a dynamic slip-form concrete extrusion method based on a robot arm technique will be presented within the scope of laboratory experiments, on-site fabrication and a design-oriented installation that carefully considers energy efficiency and lighting optimization.

Lei Yu, Dan Luo, Weiguo Xu

Ceramic Constellation | Robotically Printed Brick Specials

The fired clay brick, a mainstay of Chinese construction, has for many years languished. Rough cast and hidden behind layers of stucco, it plays a supporting role to reinforced concrete and steel - rarely however in a literal sense. This has not always been the case - highly decorative, geometrically complex, perforated or materially modified in response to specific structural or environmental conditions, the brick has played an instrumental and expressive role in the built history of China. This paper describes a means by which a traditional craft and its inherent material intelligence may be transformed into a flexible and performative system of construction through the integration of parametric design software and 3D robotic clay printing technology, ultimately offering a culturally rooted yet modern material solution for a modernizing China.

Christian J. Lange, Donn Holohan, Holger Kehne

Robotic Fabrication of Bespoke Timber Frame Modules

This paper presents methods and techniques to robotically prefabricate timber frame modules. The key challenge of this research lies in enabling the digitally informed and fabricated spatial assembly of timber beams into prefabricated timber frame modules. The project combines the fabrication and the spatial assembly of timber beams into one fully integrated robotic fabrication process. A cooperative robotic construction procedure that minimises the need for scaffolding and allows for the informed assembly of spatial structures with non-planar geometries was developed. This required the examination of suitable timber joining methods, assembly sequencing, as well developing appropriate and novel strategies to register and handle material deviations and construction tolerances. The physical implementation of the research in multiple experiments and finally, a full-scale building project validates the approach.

Andreas Thoma, Arash Adel, Matthias Helmreich, Thomas Wehrle, Fabio Gramazio, Matthias Kohler

Large-Scale Additive Manufacturing of Ultra-High-Performance Concrete of Integrated Formwork for Truss-Shaped Pillars

In the present paper a new additive manufacturing processing route is introduced to produce ultra-high-performance concrete complex architectonic elements, by printing integrated formwork. Interdisciplinary work involving material science, computation, robotics, architecture and design resulted in the development of an innovative way of 3D printing cementitious materials. The 3D printing process involved is based on a FDM-like technique, in the sense that a material is deposited layer by layer through an extrusion printhead mounted on a 6-axis robotic arm. An architectural application is used as a case-study to demonstrate the potentialities of the technology. Along with the detailed description of the design and construction process, a description of the responsibilities and their distribution amongst the stakeholders involved in the project is given. The steps taken to include the 3D printed element in an authorized regulatory context are presented as well. The structural elements produced constitute some of the largest 3D printed concrete parts available until now. Multi-functionality was enabled for structural elements by taking advantage of the complex geometry which can be achieved using our technology for large-scale additive manufacturing. The proposed process succeeds in solving several of the current issues problems that can be found in the production of 3D printed architectural features for an AEC industrial context and therefore suggests an immediately viable route for industry assimilation.

Nadja Gaudillière, Romain Duballet, Charles Bouyssou, Alban Mallet, Philippe Roux, Mahriz Zakeri, Justin Dirrenberger

Realization of Topology Optimized Concrete Structures Using Robotic Abrasive Wire-Cutting of Expanded Polystyrene Formwork

This paper presents a new method for cost-effective realization of topology optimized structures using robotic abrasive wire-cutting of expanded polystyrene formwork. Topology optimization has shown potential for generating material efficient designs and increasing performance in architectural structures. However, the method results in complex, structural morphologies which frustrate efficient construction of said structures. To overcome this, and make the realization of the potential of topology optimization feasible in general construction, new approaches are needed. We propose an integrated method of ruled surface rationalization and robotically controlled abrasive wire-cutting of formwork parts in Expanded Polystyrene. The method is demonstrated through robotic production of EPS formwork using a pilot abrasive wire-cutting end-effector on a containerized robotic work cell with an ABB IRB 6700 industrial manipulator, extended with external rotary axis. The usability of the formwork is demonstrated through the construction of a 21 m, prefabricated demonstrator structure using Ultra High Performance Concrete.

Asbjørn Søndergaard, Jelle Feringa, Florin Stan, Dana Maier

The Brick Labyrinth

This paper presents a project developed within the Master of Advanced Studies in Architecture and Digital Fabrication programme at ETH Zurich. The Brick Labyrinth is the first large-scale construction built in the Robotic Fabrication Laboratory, a unique multi-robotic setup for automated prefabrication at architectural scale. The project continues the tradition of robotic brick laying started at Gramazio Kohler Research but increases the significance of computational design and robotic control by solely working with a dry-stacked construction method. The paper introduces the design methodology, the computational framework and the robotic fabrication setup and addresses the structural challenges of the constructive system. It introduces strategies for an automated multi-robotic brick laying process at large-scale including custom-made robotic end-effectors to increase the speed of the process. The unique setup of the project and its material system demonstrate a fully reversible construction process at architectural scale, suggesting a new approach to physical prototyping, which could fundamentally change the way we design buildings. While this paper highlights the design explorations leading towards the final structure – featuring the development of a flexible dry-stacked brick bond and its potential to create unique spatial sequences – it also provides an outlook on how the integration of computational tools into automated fabrication processes can lead to new design typologies.

Luka Piškorec, David Jenny, Stefana Parascho, Hannes Mayer, Fabio Gramazio, Matthias Kohler

Backmatter

Additional information