Cumincad : CUMINCAD Papers : Search Results

CumInCAD is a Cumulative Index about publications in Computer Aided Architectural Design
supported by the sibling associations ACADIA, CAADRIA, eCAADe, SIGraDi, ASCAAD and CAAD futures

PDF papers
References

Hits 1 to 20 of 627

Reformat results as: short short into frame detailed detailed into frame

_id	ijac201816406
id	ijac201816406
authors	As, Imdat; Siddharth Pal and Prithwish Basu
year	2018
title	Artificial intelligence in architecture: Generating conceptual design via deep learning
source	International Journal of Architectural Computing vol. 16 - no. 4, 306-327
summary	Artificial intelligence, and in particular machine learning, is a fast-emerging field. Research on artificial intelligence focuses mainly on image-, text- and voice-based applications, leading to breakthrough developments in self-driving cars, voice recognition algorithms and recommendation systems. In this article, we present the research of an alternative graph- based machine learning system that deals with three-dimensional space, which is more structured and combinatorial than images, text or voice. Specifically, we present a function-driven deep learning approach to generate conceptual design. We trained and used deep neural networks to evaluate existing designs encoded as graphs, extract significant building blocks as subgraphs and merge them into new compositions. Finally, we explored the application of generative adversarial networks to generate entirely new and unique designs.
keywords	Architectural design, conceptual design, deep learning, artificial intelligence, generative design
series	journal
email
last changed	2019/08/07 14:04

_id	ecaade2018_111
id	ecaade2018_111
authors	Khean, Nariddh, Fabbri, Alessandra and Haeusler, M. Hank
year	2018
title	Learning Machine Learning as an Architect, How to? - Presenting and evaluating a Grasshopper based platform to teach architecture students machine learning
doi	https://doi.org/10.52842/conf.ecaade.2018.1.095
source	Kepczynska-Walczak, A, Bialkowski, S (eds.), Computing for a better tomorrow - Proceedings of the 36th eCAADe Conference - Volume 1, Lodz University of Technology, Lodz, Poland, 19-21 September 2018, pp. 95-102
summary	Machine learning algorithms have become widely embedded in many aspects of modern society. They have come to enhance systems, such as individualised marketing, social media services, and search engines. However, contrasting its growing ubiquity, the architectural industry has been comparatively resistant in its adoption; objectively one of the slowest industries to integrate with machine learning. Machine learning expertise can be separate from professionals in other fields; however, this separation can be a major hinderance in architecture, where interaction between the designer and the design facilitates the production of favourable outcomes. To bridge this knowledge gap, this research suggests that the solution lies with architectural education. Through the development of a novel educative framework, the research aims to teach architecture students how to implement machine learning. Exploration of student-centred pedagogical strategies was used to inform the conceptualisation of the educative module, which was subsequently implemented into an undergraduate computational design studio, and finally evaluated on its ability to effectively teach designers machine learning. The developed educative module represents a step towards greater technological adoption in the architecture industry.
keywords	Artificial Intelligence; Machine Learning; Neural Networks; Student-Centred Learning; Educative Framework
series	eCAADe
email
last changed	2022/06/07 07:52

_id	ecaade2018_w12
id	ecaade2018_w12
authors	Rahbar, Morteza
year	2018
title	Application of Artificial Intelligence in Architectural Generative Design
doi	https://doi.org/10.52842/conf.ecaade.2018.1.071
source	Kepczynska-Walczak, A, Bialkowski, S (eds.), Computing for a better tomorrow - Proceedings of the 36th eCAADe Conference - Volume 1, Lodz University of Technology, Lodz, Poland, 19-21 September 2018, pp. 71-72
summary	In this workshop, data-driven models will be discussed and how they could change the way architects think, design and analyse. Both supervised and unsupervised learning models will be discussed and different projects will be referred as examples. Deep learning models are the third part of the workshop and more specifically, Generative Adversarial Networks will be mentioned in more detail. The GAN's open a new field of generative models in design which is based on data-driven process and we will go into detail with GANs, their branches and how we could test a sample architecture generative problem with GANs.
keywords	Artificial Intelligence; Machine Learning; Generative Design; Knowledge based Design; GAN
series	eCAADe
email
last changed	2022/06/07 08:00

_id	ecaade2018_301
id	ecaade2018_301
authors	Cocho-Bermejo, Ana, Birgonul, Zeynep and Navarro-Mateu, Diego
year	2018
title	Adaptive & Morphogenetic City Research Laboratory
doi	https://doi.org/10.52842/conf.ecaade.2018.2.659
source	Kepczynska-Walczak, A, Bialkowski, S (eds.), Computing for a better tomorrow - Proceedings of the 36th eCAADe Conference - Volume 2, Lodz University of Technology, Lodz, Poland, 19-21 September 2018, pp. 659-668
summary	"Smart City" business model is guiding the development of future metropolises. Software industry sales to town halls for city management services efficiency improvement are, these days, a very pro?table business. Being the model decided by the industry, it can develop into a dangerous situation in which the basis of the new city design methodologies is decided by agents outside academia expertise. Drawing on complex science, social physics, urban economics, transportation theory, regional science and urban geography, the Lab is dedicated to the systematic analysis of, and theoretical speculation on, the recently coined "Science of Cities" discipline. On the research agenda there are questions arising from the synthesis of architecture, urban design, computer science and sociology. Collaboration with citizens through inclusion and empowerment, and, relationships "City-Data-Planner-Citizen" and "Citizen-Design-Science", configure Lab's methodology provoking a dynamic responsive process of design that is yet missing on the path towards the real responsive city.
keywords	Smart City; Morphogenetic Urban Design; Internet of Things; Building Information Modelling; Evolutionary Algorithms; Machine Learning & Artificial Intelligence
series	eCAADe
email
last changed	2022/06/07 07:56

_id	ijac201816401
id	ijac201816401
authors	Doyle, Shelby and Nick Senske
year	2018
title	Digital provenance and material metadata: Attribution and co-authorship in the age of artificial intelligence
source	International Journal of Architectural Computing vol. 16 - no. 4, 271-280
summary	This speculative essay examines a single drawing, produced in a collaboration between the authors and a Turtle robot, in a search for methods to evaluate and document provenance in artificial intelligence and robotic design. Reflecting upon the layers of authorship in our case study reveals the complex relationship that already exists between human and machine collaborators. In response to this unseen provenance, we propose new modes to document the full range of creative contribution to the design and production of artifacts from intellectual inputs to digital representations to physical labor. A more comprehensive system for artificial intelligence/robotic attribution could produce counter- narratives to technological development which more fully acknowledge the contributions of both humans and machines. As artificially intelligent design technologies distinguish themselves with distinct capabilities and eventual autonomy, a system of embedded attribution becomes the basis for human–machine collaboration, indeterminacy, and unexpected new applications for existing tools and methods.
keywords	Artificial intelligence, robotics, metadata, attribution, co-authorship, ethics
series	journal
email
last changed	2019/08/07 14:04

_id	caadria2018_126
id	caadria2018_126
authors	Khean, Nariddh, Kim, Lucas, Martinez, Jorge, Doherty, Ben, Fabbri, Alessandra, Gardner, Nicole and Haeusler, M. Hank
year	2018
title	The Introspection of Deep Neural Networks - Towards Illuminating the Black Box - Training Architects Machine Learning via Grasshopper Definitions
doi	https://doi.org/10.52842/conf.caadria.2018.2.237
source	T. Fukuda, W. Huang, P. Janssen, K. Crolla, S. Alhadidi (eds.), Learning, Adapting and Prototyping - Proceedings of the 23rd CAADRIA Conference - Volume 2, Tsinghua University, Beijing, China, 17-19 May 2018, pp. 237-246
summary	Machine learning is yet to make a significant impact in the field of architecture and design. However, with the combination of artificial neural networks, a biologically inspired machine learning paradigm, and deep learning, a hierarchical subsystem of machine learning, the predictive capabilities of machine learning processes could prove a valuable tool for designers. Yet, the inherent knowledge gap between the fields of architecture and computer science has meant the complexity of machine learning, and thus its potential value and applications in the design of the built environment remain little understood. To bridge this knowledge gap, this paper describes the development of a learning tool directed at architects and designers to better understand the inner workings of machine learning. Within the parametric modelling environment of Grasshopper, this research develops a framework to express the mathematic and programmatic operations of neural networks in a visual scripting language. This offers a way to segment and parametrise each neural network operation into a basic expression. Unpacking the complexities of machine learning in an intermediary software environment such as Grasshopper intends to foster the broader adoption of artificial intelligence in architecture.
keywords	machine learning; neural network; action research; supervised learning; education
series	CAADRIA
email
last changed	2022/06/07 07:52

_id	acadia18_166
id	acadia18_166
authors	Kvochick, Tyler
year	2018
title	Sneaky Spatial Segmentation. Reading Architectural Drawings with Deep Neural Networks and Without Labeling Data
doi	https://doi.org/10.52842/conf.acadia.2018.166
source	ACADIA // 2018: Recalibration. On imprecisionand infidelity. [Proceedings of the 38th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA) ISBN 978-0-692-17729-7] Mexico City, Mexico 18-20 October, 2018, pp. 166-175
summary	Currently, it is nearly impossible for an artificial neural network to generalize a task from very few examples. Humans, however, excel at this. For instance, it is not necessary for a designer to see thousands or millions of unique examples of how to place a given drawing symbol in a way that meets the economic, aesthetic, and performative goals of the project. In fact, the goals can be (and usually are) communicated abstractly in natural language. Machine learning (ML) models, however, do need numerous examples. The methods that we explore here are an attempt to circumvent this in order to make ML models more immediately useful. In this work, we present progress on the application of contemporary ML techniques to the design process in the architecture, engineering, and construction (AEC) industry. We introduce a technique to partially circumvent the data hungriness of neural networks, which is a significant impediment to their application outside of the ML research community. We also show results on the applicability of this technique to real-world drawings and present research that addresses how some fundamental attributes of drawings as images affect the way they are interpreted in deep neural networks. Our primary contribution is a technique to train a neural network to segment real-world architectural drawings after using only generated pseudodrawings.
keywords	full paper, representation + perception, computation, ai & machine learning
series	ACADIA
type	paper
email
last changed	2022/06/07 07:51

_id	ijac201816407
id	ijac201816407
authors	Mahankali, Ranjeeth; Brian R. Johnson and Alex T. Anderson
year	2018
title	Deep learning in design workflows: The elusive design pixel
source	International Journal of Architectural Computing vol. 16 - no. 4, 328-340
summary	The recent wave of developments and research in the field of deep learning and artificial intelligence is causing the border between the intuitive and deterministic domains to be redrawn, especially in computer vision and natural language processing. As designers frequently invoke vision and language in the context of design, this article takes a step back to ask if deep learning’s capabilities might be applied to design workflows, especially in architecture. In addition to addressing this general question, the article discusses one of several prototypes, BIMToVec, developed to examine the use of deep learning in design. It employs techniques like those used in natural language processing to interpret building information models. The article also proposes a homogeneous data format, provisionally called a design pixel, which can store design information as spatial-semantic maps. This would make designers’ intuitive thoughts more accessible to deep learning algorithms while also allowing designers to communicate abstractly with design software.
keywords	Associative logic, creative processes, deep learning, embedding vectors, BIMToVec, homogeneous design data format, design pixel, idea persistence
series	journal
email
last changed	2019/08/07 14:04

_id	caadria2018_290
id	caadria2018_290
authors	Wang, Zhenyu, Shi, Jia, Yu, Chuanfei and Gao, Guoyuan
year	2018
title	Automatic Design of Main Pedestrian Entrance of Building Site Based on Machine Learning - A Case Study of Museums in China's Urban Environment
doi	https://doi.org/10.52842/conf.caadria.2018.2.227
source	T. Fukuda, W. Huang, P. Janssen, K. Crolla, S. Alhadidi (eds.), Learning, Adapting and Prototyping - Proceedings of the 23rd CAADRIA Conference - Volume 2, Tsinghua University, Beijing, China, 17-19 May 2018, pp. 227-235
summary	The main pedestrian entrance of the building site has a direct influence on the use of the buildings, so the selection of the main pedestrian entrance is very important in the process of architectural design. The correct selection of the main pedestrian entrance of building site depends on the experience of designers and environment data collected by designers, the process is time consuming and inefficient, especially when the building site located in complex urban environment. In order to improve the efficiency of design process, we used online map to collect museums information in China as training samples, and constructing artificial neural networks to predict the direction of the main pedestrian entrance. After the training, we get the prediction model with 79% prediction accuracy. Although the accuracy still need to be improved, it creates a new approach to analysis the main pedestrian entrance of the site and worth further researching.
keywords	Artificial Neural Network (ANN); Main Pedestrian Entrance of Building Site; Automatic Design
series	CAADRIA
email
last changed	2022/06/07 07:58

_id	ecaade2018_139
id	ecaade2018_139
authors	Cudzik, Jan and Radziszewski, Kacper
year	2018
title	Artificial Intelligence Aided Architectural Design
doi	https://doi.org/10.52842/conf.ecaade.2018.1.077
source	Kepczynska-Walczak, A, Bialkowski, S (eds.), Computing for a better tomorrow - Proceedings of the 36th eCAADe Conference - Volume 1, Lodz University of Technology, Lodz, Poland, 19-21 September 2018, pp. 77-84
summary	Tools and methods used by architects always had an impact on the way building were designed. With the change in design methods and new approaches towards creation process, they became more than ever before crucial elements of the creation process. The automation of architects work has started with computational functions that were introduced to traditional computer-aided design tools. Nowadays architects tend to use specified tools that suit their specific needs. In some cases, they use artificial intelligence. Despite many similarities, they have different advantages and disadvantages. Therefore the change in the design process is more visible and unseen before solution are brought in the discipline. The article presents methods of applying the selected artificial intelligence algorithms: swarm intelligence, neural networks and evolutionary algorithms in the architectural practice by authors. Additionally research shows the methods of analogue data input and output approaches, based on vision and robotics, which in future combined with intelligence based algorithms, might simplify architects everyday practice. Presented techniques allow new spatial solutions to emerge with relatively simple intelligent based algorithms, from which many could be only accomplished with dedicated software. Popularization of the following methods among architects, will result in more intuitive, general use design tools.
keywords	computer aideed design; artificial intelligence,; evolutionary algorithms; swarm behaviour; optimization; parametric design
series	eCAADe
email
last changed	2022/06/07 07:56

_id	acadia18_226
id	acadia18_226
authors	Glynn, Ruairi; Abramovic, Vasilija; Overvelde, Johannes T. B.
year	2018
title	Edge of Chaos. Towards intelligent architecture through distributed control systems based on Cellular Automata.
doi	https://doi.org/10.52842/conf.acadia.2018.226
source	ACADIA // 2018: Recalibration. On imprecisionand infidelity. [Proceedings of the 38th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA) ISBN 978-0-692-17729-7] Mexico City, Mexico 18-20 October, 2018, pp. 226-231
summary	From the “Edge of Chaos”, a mathematical space discovered by computer scientist Christopher Langton (1997), compelling behaviors originate that exhibit both degrees of organization and instability creating a continuous dance between order and chaos. This paper presents a project intended to make this complex theory tangible through an interactive installation based on metamaterial research which demonstrates emergent behavior using Cellular Automata (CA) techniques, illustrated through sound, light and motion. We present a multi-sensory narrative approach that encourages playful exploration and contemplation on perhaps the biggest questions of how life could emerge from the disorder of the universe. We argue a way of creating intelligent architecture, not through classical Artificial Intelligence (AI), but rather through Artificial Life (ALife), embracing the aesthetic emergent possibilities that can spontaneously arise from this approach. In order to make these ideas of emergent life more tangible we present this paper in four integrated parts, namely: narrative, material, hardware and computation. The Edge of Chaos installation is an explicit realization of creating emergent systems and translating them into an architectural design. Our results demonstrate the effectiveness of a custom CA for maximizing aesthetic impact while minimizing the live time of architectural kinetic elements.
keywords	work in progress, complexity, responsive architecture, distributed computing, emergence, installation, interactive architecture, cellular automata
series	ACADIA
type	paper
email
last changed	2022/06/07 07:51

_id	sigradi2018_1693
id	sigradi2018_1693
authors	Granero, Adriana Edith
year	2018
title	The Inclusion of decentralized and self-organized system in the process of construction of design thinking
source	SIGraDi 2018 [Proceedings of the 22nd Conference of the Iberoamerican Society of Digital Graphics - ISSN: 2318-6968] Brazil, São Carlos 7 - 9 November 2018, pp. 115-122
summary	This work exposes the possible composition of a system composed of "crowd-working" of static, inert, flexible architecture elements, similar or identical entities, the "tesserae" and the integration with the link generated with Artificial Intelligence artifacts, a complex adaptive system, as a first experimental step to developments of Nanomaterials and systems that respond to the construction of the projective thought of the architectural envelope. The research responds to a general strategy of theoretical revision, with inductive and mixed methods. The exploration work examines the relative space within the idea of reason and the social function of architecture.
keywords	Self-organized; Decentralized; Nanorrobotic; Parametrism; Architectural Envelope
series	SIGRADI
email
last changed	2021/03/28 19:58

_id	sigradi2018_1616
id	sigradi2018_1616
authors	Rodrigues Alves, Manoel; Martins Abdalla, Alvaro; Tapia, Carlos
year	2018
title	Exploring Urban Interventions through Computational tools: genetic algorithm and urban connection patterns
source	SIGraDi 2018 [Proceedings of the 22nd Conference of the Iberoamerican Society of Digital Graphics - ISSN: 2318-6968] Brazil, São Carlos 7 - 9 November 2018, pp. 109-114
summary	This paper presents a particular approach to design processes in urban design, in a transdisciplinary environment. Exploring geotechnologies, information and communication technologies, artificial intelligence techniques and experimental softwares (fuzzy logic and generic algorithm), the workshop “Generation of Urban Connection Patterns”, developed by IAU-USP (Brazil) and ETSA-US (Spain), aimed: to investigate urban space connection patterns in areas of environmental and social vulnerability; to explore formal arrangements in urban design; to foster academic exchange and possibilities of collaborative workshops. The article also discusses the role of computational tools and the implementation of in-person and non-presential methods in the teaching/learning process.
keywords	Transdisciplinarity; Teaching and Learning; Genetic Algorithm; Urban Connection Patterns; Urban Design
series	SIGRADI
email
last changed	2021/03/28 19:59

_id	acadia18_146
id	acadia18_146
authors	Rossi, Gabriella; Nicholas, Paul
year	2018
title	Re/Learning the Wheel. Methods to Utilize Neural Networks as Design Tools for Doubly Curved Metal Surfaces
doi	https://doi.org/10.52842/conf.acadia.2018.146
source	ACADIA // 2018: Recalibration. On imprecisionand infidelity. [Proceedings of the 38th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA) ISBN 978-0-692-17729-7] Mexico City, Mexico 18-20 October, 2018, pp. 146-155
summary	This paper introduces concepts and computational methodologies for utilizing neural networks as design tools for architecture and demonstrates their application in the making of doubly curved metal surfaces using a contemporary version of the English Wheel. The research adopts an interdisciplinary approach to develop a novel method to model complex geometric features using computational models that originate from the field of computer vision. The paper contextualizes the approach with respect to the current state of the art of the usage of artificial neural networks both in architecture and beyond. It illustrates the cyber physical system that is at the core of this research, with a focus on the employed neural network–based computational method. Finally, the paper discusses the repercussions of these design tools on the contemporary design paradigm.
keywords	full paper, ai & machine learning, digital craft, robotic production, computation
series	ACADIA
type	paper
email
last changed	2022/06/07 07:56

_id	acadia18_108
id	acadia18_108
authors	Sanchez, Jose
year	2018
title	Platforms for Architecture: Imperatives and Opportunities of Designing Online Networks for Design
doi	https://doi.org/10.52842/conf.acadia.2018.108
source	ACADIA // 2018: Recalibration. On imprecisionand infidelity. [Proceedings of the 38th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA) ISBN 978-0-692-17729-7] Mexico City, Mexico 18-20 October, 2018, pp. 108-117
summary	The rise of platforms such as Facebook, YouTube, and Uber, initially celebrated as part of a disruptive new era of the internet, has slowly been reassessed as a problematic and unregulated form of twenty-first-century info-capitalism that contributes to inequality, mistrust, and user polarization. The internet has become a place for content creation, not only consumption, and the content freely created by the network of users has defined a self-organizing system of ad-hoc audiences following echo chambers organized through artificial intelligence, which amplifies previously identified trends. While a large portion of the content created by users seems to be aimed at personal forms of entertainment, a few remarkable projects, such as Wikipedia, have allowed hundreds of users to contribute to a collective goal. While we can observe that the platform model has appeared in diverse disciplines, allowing the creation of content from news articles to music, we have not seen the emergence of a robust design platform intended to proliferate and advance the discipline of architecture. This paper makes the case that video game technology and its audiences have reached a state of technical capability that could allow for architectural platforms to emerge, one in which players could learn, create, and share architectural designs. Such a platform comes with a series of ethical imperatives, questions of value proposition, and liabilities, as well as a high potential to communicate and proliferate architectural knowledge and know-how. Common’hood, currently under development, will be used as a case study to engage the development of an ethical architectural platform that develops a proposition towards authorship, ownership, and collective engagement.
keywords	full paper, platforms, capitalism, network, video game, combinatorics, information theory, entropy, co-ops, platform cooperativism, privacy, encryption
series	ACADIA
type	paper
email
last changed	2022/06/07 07:56

_id	ecaade2018_399
id	ecaade2018_399
authors	Cutellic, Pierre
year	2018
title	UCHRON - An Event-Based Generative Design Software Implementing Fast Discriminative Cognitive Responses from Visual ERP BCI
doi	https://doi.org/10.52842/conf.ecaade.2018.2.131
source	Kepczynska-Walczak, A, Bialkowski, S (eds.), Computing for a better tomorrow - Proceedings of the 36th eCAADe Conference - Volume 2, Lodz University of Technology, Lodz, Poland, 19-21 September 2018, pp. 131-138
summary	This research aims at investigating BCI technologies in the broad scope of CAAD applications exploiting early visual cognition in computational design. More precisely, this paper will describe the investigation of key BCI and ML components for the implementation and development of a software supporting this research : Uchron. It will be organised as follows. Firstly, it will introduce the pursued interest and contribution that visual-ERP EEG based BCI application for Generative Design may provide through a synthetic review of precedents and BCI technology. Secondly, selected BCI components will be described and a methodology will be presented to provide an appropriate framework for a CAAD software approach. This section main focus is on the processing component of the BCI. It distinguishes two key aspects of discrimination and generation in its design and proposes a new model based on GAN for modulated adversarial design. Emphasis will be made on the explicit use of inference loops integrating fast human cognitive responses and its individual capitalisation through time in order to reflect towards the generation of design and architectural features.
keywords	Human Computer Interaction; Neurodesign; Generative Design; Design Computing and Cognition; Machine Learning
series	eCAADe
email
last changed	2022/06/07 07:56

_id	caadria2018_052
id	caadria2018_052
authors	Fung, Enrica and Crolla, Kristof
year	2018
title	Choreographed Architecture - Body-Spatial Exploration
doi	https://doi.org/10.52842/conf.caadria.2018.1.101
source	T. Fukuda, W. Huang, P. Janssen, K. Crolla, S. Alhadidi (eds.), Learning, Adapting and Prototyping - Proceedings of the 23rd CAADRIA Conference - Volume 1, Tsinghua University, Beijing, China, 17-19 May 2018, pp. 101-110
summary	This paper presents a design-methodological case study that looks into the practical expansion of conventional conceptual architectural design media by incorporating contemporary technology of motion capture. It discusses challenges of integrating dance movement as a real-time input parameter for architectural design that aims at translating body motion into space. The paper consists of four parts, beginning with a historic background overview of scientists, physiologists, artists, choreographers, and architects who have attempted capturing body motion and turning the motion into space. The second part of the paper discusses the iterative development of the 'Dance Machine' as a methodological tool for the integration of motion capture into conceptual architectural design. Thirdly, the paper discusses tested design applications of the 'Dance Machine' by looking at two sited applications. Finally, the overall methodology is critically assessed and discussed in the light of continuous development of creative applications of motion capturing technology. The paper concludes by highlighting the architectural potential found in specific qualities of dance and by advocating for a broader palette of tools, techniques, and input methods for the conceptual design of architecture.
keywords	Choreographed architecture; Motion capture; Conceptual design media; Space design; Human body
series	CAADRIA
email
last changed	2022/06/07 07:50

_id	acadia18_156
id	acadia18_156
authors	Huang, Weixin; Zheng, Hao
year	2018
title	Architectural Drawings Recognition and Generation through Machine Learning
doi	https://doi.org/10.52842/conf.acadia.2018.156
source	ACADIA // 2018: Recalibration. On imprecisionand infidelity. [Proceedings of the 38th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA) ISBN 978-0-692-17729-7] Mexico City, Mexico 18-20 October, 2018, pp. 156-165
summary	With the development of information technology, the ideas of programming and mass calculation were introduced into the design field, resulting in the growth of computer- aided design. With the idea of designing by data, we began to manipulate data directly, and interpret data through design works. Machine Learning as a decision making tool has been widely used in many fields. It can be used to analyze large amounts of data and predict future changes. Generative Adversarial Network (GAN) is a model framework in machine learning. It’s specially designed to learn and generate output data with similar or identical characteristics. Pix2pixHD is a modified version of GAN that learns image data in pairs and generates new images based on the input. The author applied pix2pixHD in recognizing and generating architectural drawings, marking rooms with different colors and then generating apartment plans through two convolutional neural networks. Next, in order to understand how these networks work, the author analyzed their framework, and provided an explanation of the three working principles of the networks, convolution layer, residual network layer and deconvolution layer. Lastly, in order to visualize the networks in architectural drawings, the author derived data from different layer and different training epochs, and visualized the findings as gray scale images. It was found that the features of the architectural plan drawings have been gradually learned and stored as parameters in the networks. As the networks get deeper and the training epoch increases, the features in the graph become more concise and clearer. This phenomenon may be inspiring in understanding the designing behavior of humans.
keywords	full paper, design study, generative design, ai + machine learning, ai & machine learning
series	ACADIA
type	paper
email
last changed	2022/06/07 07:49

_id	acadia18_232
id	acadia18_232
authors	Kilian, Axel
year	2018
title	The Flexing Room Architectural Robot. An Actuated Active-Bending Robotic Structure using Human Feedback
doi	https://doi.org/10.52842/conf.acadia.2018.232
source	ACADIA // 2018: Recalibration. On imprecisionand infidelity. [Proceedings of the 38th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA) ISBN 978-0-692-17729-7] Mexico City, Mexico 18-20 October, 2018, pp. 232-241
summary	Advances in autonomous control of object-scale robots, both anthropomorphic and vehicular, are posing new human–machine interface challenges. In architecture, very few examples of autonomous inhabitable robotic architecture exist. A number of factors likely contribute to this condition, among them the scale and cost of architectural adaptive systems, but on a more fundamental conceptual level also the questions of how architectural robots would communicate with their human inhabitants. The Flexing Room installation is a room-sized actuated active-bending skeleton structure. It uses rudimentary social feedback by counting people to inform its behavior in the form of actuated poses of the room enclosure. An operational full-scale prototype was constructed and tested. To operate it no geometric-based simulation was used; the only communication between computer and structure was in sending values for the air pressure settings and in gathering sensor feedback. The structure’s physical state was resolved through the embodied computation of its interconnected parts, and the people-counting sensor feedback influences its next action. Future work will explore the development of learning processes to improve the human–machine coexistence in space.
keywords	full paper, fabrication & robotics, non-production robotics, materials/adaptive systems, flexible structures
series	ACADIA
type	paper
email
last changed	2022/06/07 07:52

_id	ecaade2018_315
id	ecaade2018_315
authors	Koehler, Daniel, Abo Saleh, Sheghaf, Li, Hua, Ye, Chuwei, Zhou, Yaonaijia and Navasaityte, Rasa
year	2018
title	Mereologies - Combinatorial Design and the Description of Urban Form.
doi	https://doi.org/10.52842/conf.ecaade.2018.2.085
source	Kepczynska-Walczak, A, Bialkowski, S (eds.), Computing for a better tomorrow - Proceedings of the 36th eCAADe Conference - Volume 2, Lodz University of Technology, Lodz, Poland, 19-21 September 2018, pp. 85-94
summary	This paper discusses the ability to apply machine learning to the combinatorial design-assembly at the scale of a building to urban form. Connecting the historical lines of discrete automata in computer science and formal studies in architecture this research contributes to the field of additive material assemblies, aggregative architecture and their possible upscaling to urban design. The following case studies are a preparation to apply deep-learning on the computational descriptions of urban form. Departing from the game Go as a testbed for the development of deep-learning applications, an equivalent platform can be designed for architectural assembly. By this, the form of a building is defined via the overlap between separate building parts. Building on part-relations, this research uses mereology as a term for a set of recursive assembly strategies, integrated into the design aspects of the building parts. The models developed by research by design are formally described and tested under a digital simulation environment. The shown case study shows the process of how to transform geometrical elements to architectural parts based merely on their compositional aspects either in horizontal or three-dimensional arrangements.
keywords	Urban Form; Discrete Automata ; Combinatorics; Part-Relations; Mereology; Aggregative Architecture
series	eCAADe
email
last changed	2022/06/07 07:51

For more results click below: