Pub Date : 2022-10-20DOI: 10.47330/dcio.2022.pfcu2358
Fadhil Fadhil
The course will introduce Modo software as a tool for component-based design. The design approach will focus on the fusion of discrete components with a distinct fabrication process. The course will introduce the process of starting with a primitive geometry that evolves into a multi-object component. The workshop will also address the upward and downward relationship between different parts of the component. We will introduce multiple techniques that explore multi-resolution surface modelling and the difference between 2D textures and 3D textures. Lastly, we will have a brief introduction of Modo real-time rendering and different ways of lighting strategies and texture mapping.
{"title":"Components Fusion in Modo Foundry","authors":"Fadhil Fadhil","doi":"10.47330/dcio.2022.pfcu2358","DOIUrl":"https://doi.org/10.47330/dcio.2022.pfcu2358","url":null,"abstract":"The course will introduce Modo software as a tool for component-based design. The design approach will focus on the fusion of discrete components with a distinct fabrication process. The course will introduce the process of starting with a primitive geometry that evolves into a multi-object component. The workshop will also address the upward and downward relationship between different parts of the component. We will introduce multiple techniques that explore multi-resolution surface modelling and the difference between 2D textures and 3D textures. Lastly, we will have a brief introduction of Modo real-time rendering and different ways of lighting strategies and texture mapping.","PeriodicalId":129906,"journal":{"name":"Design Computation Input/Output 2022","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128580322","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-20DOI: 10.47330/dcio.2022.ggnl1577
Victor Sardenberg, M. Becker
The existing methods for solution space navigation require numerical values to score solutions. The authors introduce a method of quantitative aesthetic evaluation utilizing Computer Vision (CV) as a criterion to navigate solution spaces. Therefore, aesthetics can complement structural, environmental, and other quantitative criteria. The work stands in the extended history of quantifying the visual aesthetic experience. Some precedents are: Birkhoff [1933] and Max Bense [1965] built an approach with experiments to empirically support a measure, whereas Birkin [2010], Ostwald, and Vaughan [2016] devised the first computational methods working on vector drawings. Our research automates and accelerates aesthetic quantification by utilizing CV to extract computable datasets from images. We are especially keen on architectural images as a shorthand to assign an aesthetic value to design, aiming to navigate the solution space in architecture. This work devises a method for rearranging parts in architectural images focusing on formal aspects, in opposition to semantic segmentation where objects unrelated to architectural design (cars, persons, sky…) are quantified to score images [Verma and Jana and Ramamritham 2018]. It uses Maximally Stable Extremal Regions (MSER) [Matas 2004] to recognize architectural parts because it is superior to similar methods such as SimpleBlobDetector in this task. Our method disassembles the parts in a diagram of scaled parts (Fig. 2) to analyze them in isolation, and a diagram of connectivity graph (Fig. 3), to evaluate relationships. These diagrams are examined to compare photos of buildings, cars, and trees to assess the applicability of such a method to a range of objects. Parts and connections are thus quantified, and these values are inputted in a refined version of Birkhoff’s formula to calculate an aesthetic score for each image for navigating the solution space. Finally, it tests the method to draw comparisons between the discrete and continuous paradigms (Fig. 1) in the contemporary discourse of architecture, comparing Zaha Hadid Architects` Heydar Aliyev Centre and Gilles Retsin´s Diamonds House to argue that there is a difference between the aesthetic effects of continuous and discrete designs, besides their distinction in tectonic logic. The method proved to be an efficient procedure for comparatively quantifying the aesthetic judgment of architectural images, enabling designers to incorporate aesthetics as a complementary criterion for solution space navigation in computational design. The method of computational aesthetic measure for solution space navigation and its calibrations via crowdsourced evaluation of images is further detailed in a paper by the authors being published at the 2022 eCAADe conference.
{"title":"Aesthetic Measure of Architectural Photography utilizing Computer Vision: Parts-from-Wholes","authors":"Victor Sardenberg, M. Becker","doi":"10.47330/dcio.2022.ggnl1577","DOIUrl":"https://doi.org/10.47330/dcio.2022.ggnl1577","url":null,"abstract":"The existing methods for solution space navigation require numerical values to score solutions. The authors introduce a method of quantitative aesthetic evaluation utilizing Computer Vision (CV) as a criterion to navigate solution spaces. Therefore, aesthetics can complement structural, environmental, and other quantitative criteria. The work stands in the extended history of quantifying the visual aesthetic experience. Some precedents are: Birkhoff [1933] and Max Bense [1965] built an approach with experiments to empirically support a measure, whereas Birkin [2010], Ostwald, and Vaughan [2016] devised the first computational methods working on vector drawings. Our research automates and accelerates aesthetic quantification by utilizing CV to extract computable datasets from images. We are especially keen on architectural images as a shorthand to assign an aesthetic value to design, aiming to navigate the solution space in architecture. This work devises a method for rearranging parts in architectural images focusing on formal aspects, in opposition to semantic segmentation where objects unrelated to architectural design (cars, persons, sky…) are quantified to score images [Verma and Jana and Ramamritham 2018]. It uses Maximally Stable Extremal Regions (MSER) [Matas 2004] to recognize architectural parts because it is superior to similar methods such as SimpleBlobDetector in this task. Our method disassembles the parts in a diagram of scaled parts (Fig. 2) to analyze them in isolation, and a diagram of connectivity graph (Fig. 3), to evaluate relationships. These diagrams are examined to compare photos of buildings, cars, and trees to assess the applicability of such a method to a range of objects. Parts and connections are thus quantified, and these values are inputted in a refined version of Birkhoff’s formula to calculate an aesthetic score for each image for navigating the solution space. Finally, it tests the method to draw comparisons between the discrete and continuous paradigms (Fig. 1) in the contemporary discourse of architecture, comparing Zaha Hadid Architects` Heydar Aliyev Centre and Gilles Retsin´s Diamonds House to argue that there is a difference between the aesthetic effects of continuous and discrete designs, besides their distinction in tectonic logic. The method proved to be an efficient procedure for comparatively quantifying the aesthetic judgment of architectural images, enabling designers to incorporate aesthetics as a complementary criterion for solution space navigation in computational design. The method of computational aesthetic measure for solution space navigation and its calibrations via crowdsourced evaluation of images is further detailed in a paper by the authors being published at the 2022 eCAADe conference.","PeriodicalId":129906,"journal":{"name":"Design Computation Input/Output 2022","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114777882","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-20DOI: 10.47330/dcio.2022.gpqp2161
M. Oskina, Z. Rusák, Peter Boom
More and more modern transport modalities are equipped with complex human-machine interfaces (HMI). HMI aim to narrow the information gap between the complex automation system and their human operator to ensure fast, effective interaction and decision making. We see HMI in the traffic controllers' rooms, the ADAS-equipped vehicles, the public transport drivers' rooms, and many other modern transport modes. Designers create HMIs to effectively draw the operator’s attention to the most necessary and critical information and to facilitate accurate and fast decision making. Whether these systems adequately support human operators and achieve the intention of their designer is difficult to test objectively. [Hamilton and Grabowki 2013] showed that visual, manual and cognitive distractions of ADAS-equipped vehicles tend to distract drivers, who in turn behave less safe on the roads. There is, however, no comprehensive overview about the typical cognitive challenges operators facing in different domains of HMI applications and how these challenges can be objectively assessed. We conducted a series of interviews on difficulties of operators’ Human-Machine interface experience with human factors experts working with in railway and ADAS systems and investigated Endsley's situation awareness theory in dynamic systems [Endsley 1995]. Our interviewees reported several typical issues from their HMI studies, including missing events on the HMI displays, information overload of operators, lack of contextual and situational awareness and, as a resulting mismatch in expected and performed operator actions. We aim to develop and objective approach based on mobile eye tracking technology that can be used to characterize operator situation awareness, decision making and task performance and validate HMI designs in specific mobility and industry applications. The first step of our method is HAZOP analysis of the Human-Machine events and operator tasks, which results in a set of use cases for the eye-tracking experiments. In the experiments, we use wearable eye-tracking glasses combined with AI based computer vision algorithms. Wearable eyetracking enables us to conduct studies in real world scenarios, while AI based computer vision helps use to automatically identify relevant events and streamline the eye tracking data analysis workflow. With the use of glasses, we collect hotspot analysis, sequence of eye movement analysis, time to capture alarms and other parameters. Finally, we use an AI (and open AI) component in the glasses to mark the event of interest and track when the eye interacts with an area or an event of interest. We process gained data to conclude the events engagement, mistakes in responses, and missed out information and explain the root causes. In the past period, we conducted a pilot study to validate the quality of data collected with the openeye eye-tracking equipment (https://kexxu.com/ ). In the next step, we will use validate our method in a fu
越来越多的现代运输方式配备了复杂的人机界面(HMI)。人机界面旨在缩小复杂的自动化系统与其人工操作员之间的信息差距，以确保快速、有效的交互和决策。我们在交通控制员的房间、配备adas系统的车辆、公共交通司机的房间以及许多其他现代交通方式中都看到了HMI。设计人员创建hmi是为了有效地将操作员的注意力吸引到最必要和最关键的信息上，从而促进准确、快速的决策。这些系统是否足以支持人类操作员并实现其设计者的意图很难客观地测试。[Hamilton and Grabowki 2013]表明，配备adas的车辆的视觉、手动和认知干扰往往会分散驾驶员的注意力，从而导致他们在道路上的行为不安全。然而，对于操作员在不同HMI应用领域面临的典型认知挑战以及如何客观评估这些挑战，目前还没有全面的概述。我们与在铁路和ADAS系统中工作的人因专家就操作员人机界面体验的困难进行了一系列访谈，并调查了Endsley在动态系统中的情况感知理论[Endsley 1995]。我们的受访者报告了他们在HMI研究中遇到的几个典型问题，包括HMI显示上缺失的事件、操作员的信息过载、缺乏上下文和态势感知，以及由此导致的预期操作和实际操作的不匹配。我们的目标是开发一种基于移动眼动追踪技术的客观方法，该技术可用于表征操作员的态势感知、决策和任务性能，并验证特定移动和工业应用中的HMI设计。我们方法的第一步是对人机事件和操作员任务进行HAZOP分析，从而产生一组用于眼动追踪实验的用例。在实验中，我们使用可穿戴式眼球追踪眼镜结合基于AI的计算机视觉算法。可穿戴式眼动追踪使我们能够在现实场景中进行研究，而基于AI的计算机视觉帮助我们自动识别相关事件，简化眼动追踪数据分析工作流程。通过使用眼镜，我们收集热点分析、眼动序列分析、捕捉报警时间等参数。最后，我们在眼镜中使用AI(和开放AI)组件来标记感兴趣的事件，并跟踪眼睛何时与感兴趣的区域或事件交互。我们对获得的数据进行处理，总结事件参与、反应错误、遗漏信息，并解释根本原因。在过去的一段时间里，我们进行了一项试点研究，以验证使用裸眼眼动追踪设备(https://kexxu.com/)收集的数据的质量。在下一步中，我们将在全尺寸实验中验证我们的方法。我们相信，我们的见解将有助于为人机交互的舒适性、安全性和有效性等人为因素研究的当前研究方法带来重大改进。我们还打算将我们的方法应用于培训和提高操作员的技能。”
{"title":"Eye on HMI - Assessment of Human-Machine Interface with wearable eye-tracking glasses","authors":"M. Oskina, Z. Rusák, Peter Boom","doi":"10.47330/dcio.2022.gpqp2161","DOIUrl":"https://doi.org/10.47330/dcio.2022.gpqp2161","url":null,"abstract":"More and more modern transport modalities are equipped with complex human-machine interfaces (HMI). HMI aim to narrow the information gap between the complex automation system and their human operator to ensure fast, effective interaction and decision making. We see HMI in the traffic controllers' rooms, the ADAS-equipped vehicles, the public transport drivers' rooms, and many other modern transport modes. Designers create HMIs to effectively draw the operator’s attention to the most necessary and critical information and to facilitate accurate and fast decision making. Whether these systems adequately support human operators and achieve the intention of their designer is difficult to test objectively. [Hamilton and Grabowki 2013] showed that visual, manual and cognitive distractions of ADAS-equipped vehicles tend to distract drivers, who in turn behave less safe on the roads. There is, however, no comprehensive overview about the typical cognitive challenges operators facing in different domains of HMI applications and how these challenges can be objectively assessed. We conducted a series of interviews on difficulties of operators’ Human-Machine interface experience with human factors experts working with in railway and ADAS systems and investigated Endsley's situation awareness theory in dynamic systems [Endsley 1995]. Our interviewees reported several typical issues from their HMI studies, including missing events on the HMI displays, information overload of operators, lack of contextual and situational awareness and, as a resulting mismatch in expected and performed operator actions. We aim to develop and objective approach based on mobile eye tracking technology that can be used to characterize operator situation awareness, decision making and task performance and validate HMI designs in specific mobility and industry applications. The first step of our method is HAZOP analysis of the Human-Machine events and operator tasks, which results in a set of use cases for the eye-tracking experiments. In the experiments, we use wearable eye-tracking glasses combined with AI based computer vision algorithms. Wearable eyetracking enables us to conduct studies in real world scenarios, while AI based computer vision helps use to automatically identify relevant events and streamline the eye tracking data analysis workflow. With the use of glasses, we collect hotspot analysis, sequence of eye movement analysis, time to capture alarms and other parameters. Finally, we use an AI (and open AI) component in the glasses to mark the event of interest and track when the eye interacts with an area or an event of interest. We process gained data to conclude the events engagement, mistakes in responses, and missed out information and explain the root causes. In the past period, we conducted a pilot study to validate the quality of data collected with the openeye eye-tracking equipment (https://kexxu.com/ ). In the next step, we will use validate our method in a fu","PeriodicalId":129906,"journal":{"name":"Design Computation Input/Output 2022","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124222795","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-20DOI: 10.47330/dcio.2022.kkrr9918
M. Colletti
The hyphenated term ‘post-digital’ was coined over twenty years ago by sonic artist Kim Cascone as a means of thinking about the aesthetics of the time beyond the digital revolution in electronic music, and then adopted in the arts, media, and architecture. In architecture, the term has been adopted to promote an anti-digital ‘downgrading’ aesthetic agenda: as post(minus)digital. However, if understood as post(plus)digital – in short: postdigital – the expression appears to be indicative of an evolution of the digital beyond its first implementation as numerical or virtual alternative to traditional analogue design and making techniques. What postdigital traits and characteristics can be identified? How can they be implemented in architectural design? Why does it matter?
{"title":"Postdigital Architecture. What Now?!","authors":"M. Colletti","doi":"10.47330/dcio.2022.kkrr9918","DOIUrl":"https://doi.org/10.47330/dcio.2022.kkrr9918","url":null,"abstract":"The hyphenated term ‘post-digital’ was coined over twenty years ago by sonic artist Kim Cascone as a means of thinking about the aesthetics of the time beyond the digital revolution in electronic music, and then adopted in the arts, media, and architecture. In architecture, the term has been adopted to promote an anti-digital ‘downgrading’ aesthetic agenda: as post(minus)digital. However, if understood as post(plus)digital – in short: postdigital – the expression appears to be indicative of an evolution of the digital beyond its first implementation as numerical or virtual alternative to traditional analogue design and making techniques. What postdigital traits and characteristics can be identified? How can they be implemented in architectural design? Why does it matter?","PeriodicalId":129906,"journal":{"name":"Design Computation Input/Output 2022","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117015203","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-20DOI: 10.47330/dcio.2022.ngwc1201
Dustin White
The lecture outlines the past five years of a research-based design practice with an interest how technology, craft, and materials come together in ways that explore the boundaries between design, architecture, and other disciplines. Specifically, the pedagogy of material based computational strategies supporting the integration of form, material, and structure by incorporating physical form-finding strategies with digital analysis and fabrication processes. In this approach material often comes before shape, with material explorations as the premise for making and fabricating, and design decisions that emerge from the results of the material experiments and testing. The work produced by my students and myself seeks to challenge digital technology and fabrication to further the relationship of material to machine and material to design. With the intent to develop and employ novel software techniques that aid in the translation from the virtual world to the physical medias we engage through craft and technology to hybridize design and making. The work presented varies in scale, technique, method, intent, and fabrication processes but is fascinated with thinking though material based computational design strategies.
{"title":"Material Based Computational Design Strategies","authors":"Dustin White","doi":"10.47330/dcio.2022.ngwc1201","DOIUrl":"https://doi.org/10.47330/dcio.2022.ngwc1201","url":null,"abstract":"The lecture outlines the past five years of a research-based design practice with an interest how technology, craft, and materials come together in ways that explore the boundaries between design, architecture, and other disciplines. Specifically, the pedagogy of material based computational strategies supporting the integration of form, material, and structure by incorporating physical form-finding strategies with digital analysis and fabrication processes. In this approach material often comes before shape, with material explorations as the premise for making and fabricating, and design decisions that emerge from the results of the material experiments and testing. The work produced by my students and myself seeks to challenge digital technology and fabrication to further the relationship of material to machine and material to design. With the intent to develop and employ novel software techniques that aid in the translation from the virtual world to the physical medias we engage through craft and technology to hybridize design and making. The work presented varies in scale, technique, method, intent, and fabrication processes but is fascinated with thinking though material based computational design strategies.","PeriodicalId":129906,"journal":{"name":"Design Computation Input/Output 2022","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132745074","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-20DOI: 10.47330/dcio.2022.nzjq3625
Gökhan Ongun, Mustafa Chehabeddine, E. Ramos, Ji Soo Han, Jonathan Dreyfus
{"title":"KPF Karst Mountains Inspired Facades Optimisation","authors":"Gökhan Ongun, Mustafa Chehabeddine, E. Ramos, Ji Soo Han, Jonathan Dreyfus","doi":"10.47330/dcio.2022.nzjq3625","DOIUrl":"https://doi.org/10.47330/dcio.2022.nzjq3625","url":null,"abstract":"","PeriodicalId":129906,"journal":{"name":"Design Computation Input/Output 2022","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125944838","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-20DOI: 10.47330/dcio.2022.mujm9015
Andreas Luka, Yong Guo
{"title":"Approach to Digital Twins of Woody Vegetation (Trees and Shrubs)","authors":"Andreas Luka, Yong Guo","doi":"10.47330/dcio.2022.mujm9015","DOIUrl":"https://doi.org/10.47330/dcio.2022.mujm9015","url":null,"abstract":"","PeriodicalId":129906,"journal":{"name":"Design Computation Input/Output 2022","volume":"42 23","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120820921","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-20DOI: 10.47330/dcio.2022.caxl3310
Omar A.I. Azeem, L. Iannucci
The current industrial practice used at the preliminary design stage of complex structures involves the use of multifidelity submodelling simulations to predict failure behaviour around geometric and structural design features of interest, such as bolts, fillets, and ply drops. A simplified global model without the design features is first run and the resulting displacement fields are transferred to multiple local models containing the design features of interest. The creation of these high-fidelity local feature models is highly expert dependent, and their subsequent simulation is highly time-consuming. These issues compound as these design features are typically repetitive in complex structures. This leads to long design and development cycles. Application of machine learning to this framework has the potential to capture a structural designer’s modelling knowledge and quickly suggest improved design feature parameters, thereby addressing the current challenges. In this work, we provide a proof of concept for a machine learning assisted preliminary design workflow, see Figure 1, whereby feature-specific surrogate models may be trained offline and used for faster and simpler design iterations. The key challenge is to maximise the prediction accuracy of failure metrics whilst managing the high dimensions required to represent design feature simulation parameters in a minimum training dataset size. These challenges are addressed using: a modified Latin Hypercube Sampling scheme adjusted to improve design of experiment in composite materials; a bi-linear work-equivalent homogenisation scheme to reduce the number of nodal degrees of freedom; a non-local volume-averaged stress-based approach to reduce the number of target features; and linear superposition of stacked bi-directional LSTM neural network models. This methodology is demonstrated in a case study of predicting the stresses of open hole composite laminates in an aerospace C-spar structure. Results highlight the high accuracy (>90%) and time saving benefit (>15x) of this new approach. This methodology may be used to faster correct and iterate the preliminary design of any large or complex structure where there are repetitive localised design features that may contribute to failure, such as in Formula 1 or wind turbines. Combined with exascale computing this methodology may also be applied for predictive virtual testing of digital twins.
{"title":"A machine learning assisted preliminary design methodology for repetitive design features in complex structures","authors":"Omar A.I. Azeem, L. Iannucci","doi":"10.47330/dcio.2022.caxl3310","DOIUrl":"https://doi.org/10.47330/dcio.2022.caxl3310","url":null,"abstract":"The current industrial practice used at the preliminary design stage of complex structures involves the use of multifidelity submodelling simulations to predict failure behaviour around geometric and structural design features of interest, such as bolts, fillets, and ply drops. A simplified global model without the design features is first run and the resulting displacement fields are transferred to multiple local models containing the design features of interest. The creation of these high-fidelity local feature models is highly expert dependent, and their subsequent simulation is highly time-consuming. These issues compound as these design features are typically repetitive in complex structures. This leads to long design and development cycles. Application of machine learning to this framework has the potential to capture a structural designer’s modelling knowledge and quickly suggest improved design feature parameters, thereby addressing the current challenges. In this work, we provide a proof of concept for a machine learning assisted preliminary design workflow, see Figure 1, whereby feature-specific surrogate models may be trained offline and used for faster and simpler design iterations. The key challenge is to maximise the prediction accuracy of failure metrics whilst managing the high dimensions required to represent design feature simulation parameters in a minimum training dataset size. These challenges are addressed using: a modified Latin Hypercube Sampling scheme adjusted to improve design of experiment in composite materials; a bi-linear work-equivalent homogenisation scheme to reduce the number of nodal degrees of freedom; a non-local volume-averaged stress-based approach to reduce the number of target features; and linear superposition of stacked bi-directional LSTM neural network models. This methodology is demonstrated in a case study of predicting the stresses of open hole composite laminates in an aerospace C-spar structure. Results highlight the high accuracy (>90%) and time saving benefit (>15x) of this new approach. This methodology may be used to faster correct and iterate the preliminary design of any large or complex structure where there are repetitive localised design features that may contribute to failure, such as in Formula 1 or wind turbines. Combined with exascale computing this methodology may also be applied for predictive virtual testing of digital twins.","PeriodicalId":129906,"journal":{"name":"Design Computation Input/Output 2022","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124927023","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-20DOI: 10.47330/dcio.2022.mmlw2640
E. Vermisso
This paper offers insights about the otherwise limited NLM-driven methodologies, supporting an examination of design creativity following the ‘process’ approach. [Abraham 2018] Recent application of AI models which rely on natural language processing (semantic references) is increasingly popular because of their directness and ease-of-use. Neural Language Models (NLMs) like VQGAN+CLIP, DALL-E, MidJourney) offer promising results, [Rodrigues, et al. 2021] seemingly bypassing the need for expensive datasets and technical expertise. Naturally, such models are limited because they cannot capture the multimodal complexity of architectural thinking and human cognition in general [Penrose 1989]. Alternative approaches propose the combination of NLMs with other artificial neural networks (ANNs) i.e. StyleGAN; CycleGAN which are custom-trained on domain-specific data. [Bolojan, Vermisso and Yousif 2022] Architects seek to expand their agency within such AI-assisted processes by controling the input encoding, so they can subsequently convert the generated outcomes to 3D models fairly directly. Still, AI models of computer vision like NLMs and GANs offer 2-dimensional output, which requires extensive decoding into 3-dimensional format. While this may seem severely constraining, it presents a silver lining when it comes to furthering design creativity. Designers are asked to scrutinize their methods from a cognitive standpoint, because these methodologies not only encourage, but demand thorough interrogation of the design intentionality, the design decision making factors and qualification criteria. Text-to-image correlation, on which NLMs rely, and their 2-dimensional output, ensure that certain important considerations are not circumvented. Instead of obtaining a 3D model, multiple possible -fragmented- versions of it are separately implied. Often, ‘fake’ images generated by the ANNs promote contradictory inferences of space, which require further examination. The hidden opportunity within the limited format of AI models echo Neil Spiller’s comments about the advantage of drawing over animation techniques twenty years ago: “Enigma is a creative tool that allows designers to see bifurcated outcomes in their sketches and drawings; it plays on the inability of drawings to faithfully record the distinct placement and extent of architectural elements”. [Spiller 2001] Comparing animations to static drawings, Spiller praised the drawing’s ability to hold “…an imagined past and an imagined future”. ‘Reading’ these results involves the (human) disentanglement of high and low-level features and consciously allocating their corresponding qualities for curation. The process of evaluating ‘parts-to-whole’ visual relationships is noteworthy because it depends on shifting our attention away from certain features, and an unconscious binding of visual elements. [Dehaene 2014] The philosopher Alain wrote that “The art of paying attention, the great art,…supposes the art of
{"title":"Fragmented Layers of Design Thinking: Limitations and Opportunities of Neural Language Model-assisted processes for Design Creativity","authors":"E. Vermisso","doi":"10.47330/dcio.2022.mmlw2640","DOIUrl":"https://doi.org/10.47330/dcio.2022.mmlw2640","url":null,"abstract":"This paper offers insights about the otherwise limited NLM-driven methodologies, supporting an examination of design creativity following the ‘process’ approach. [Abraham 2018] Recent application of AI models which rely on natural language processing (semantic references) is increasingly popular because of their directness and ease-of-use. Neural Language Models (NLMs) like VQGAN+CLIP, DALL-E, MidJourney) offer promising results, [Rodrigues, et al. 2021] seemingly bypassing the need for expensive datasets and technical expertise. Naturally, such models are limited because they cannot capture the multimodal complexity of architectural thinking and human cognition in general [Penrose 1989]. Alternative approaches propose the combination of NLMs with other artificial neural networks (ANNs) i.e. StyleGAN; CycleGAN which are custom-trained on domain-specific data. [Bolojan, Vermisso and Yousif 2022] Architects seek to expand their agency within such AI-assisted processes by controling the input encoding, so they can subsequently convert the generated outcomes to 3D models fairly directly. Still, AI models of computer vision like NLMs and GANs offer 2-dimensional output, which requires extensive decoding into 3-dimensional format. While this may seem severely constraining, it presents a silver lining when it comes to furthering design creativity. Designers are asked to scrutinize their methods from a cognitive standpoint, because these methodologies not only encourage, but demand thorough interrogation of the design intentionality, the design decision making factors and qualification criteria. Text-to-image correlation, on which NLMs rely, and their 2-dimensional output, ensure that certain important considerations are not circumvented. Instead of obtaining a 3D model, multiple possible -fragmented- versions of it are separately implied. Often, ‘fake’ images generated by the ANNs promote contradictory inferences of space, which require further examination. The hidden opportunity within the limited format of AI models echo Neil Spiller’s comments about the advantage of drawing over animation techniques twenty years ago: “Enigma is a creative tool that allows designers to see bifurcated outcomes in their sketches and drawings; it plays on the inability of drawings to faithfully record the distinct placement and extent of architectural elements”. [Spiller 2001] Comparing animations to static drawings, Spiller praised the drawing’s ability to hold “…an imagined past and an imagined future”. ‘Reading’ these results involves the (human) disentanglement of high and low-level features and consciously allocating their corresponding qualities for curation. The process of evaluating ‘parts-to-whole’ visual relationships is noteworthy because it depends on shifting our attention away from certain features, and an unconscious binding of visual elements. [Dehaene 2014] The philosopher Alain wrote that “The art of paying attention, the great art,…supposes the art of ","PeriodicalId":129906,"journal":{"name":"Design Computation Input/Output 2022","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123160990","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-20DOI: 10.47330/dcio.2022.fbeo7122
P. Baquero, Daoming Liu, Yota Adilenido
Deployable structures have many applications in architecture, from kinetic pavilions to temporary structures, to retractable rooftops. There are various advantages to building deployable elements in a factory, and then deploying them on site. It is simpler to join stripes in flat arrangements than to put them together in three dimensions. This study focuses on a novel approach for creating and simulating systems of elastic stripes and how they can be utilized to build 3d-surfaces. The goal is to find the correct 2D stripes geometry that when deployed corresponds to a given 3D designed surface. Deployment simulation is essential to the design phase, and it is believes that by designing, simulating, and re-using data from already tested physical models, kinetic design methodological framework would naturally transition from a Design-Fabrication-Simulation workflow into a Design-Simulation-Fabrication one. (Raviv et al. 2014). In order to find the stripes deployment and its proximity to the final 3d surface, three experiments are examined ere: Starting from a simple case, in order to get a negative Gaussian curvature (Figure 1, Top) and observe the transformation and distortion of the flat faces, a hexagonal flat model has been vertically extruded and anchored in two points. Then, investigating further deploying techniques for negative curvature surfaces, a 2D linear set of equal stripes is deployed evenly by adding a locker stripe on its ends (Figure 2, Bottom). From the prototype and the simulation deployment the distortion produced a negative Gaussian curvature. A more complex example of curved stripes, using a locker stripe at the start and connecting between them, a group of four curved deployable stripes were distorted and joined to create four arches while maintaining their opening state (Figure 3, Right). The same distortion was seen in both the arches simulation and the prototype.
可展开结构在建筑中有很多应用，从动态展馆到临时结构，再到可伸缩屋顶。在工厂中构建可部署元素，然后在现场部署它们有很多优点。将条纹以平面的方式连接起来比将它们以三维的方式组合起来要简单得多。本研究的重点是创建和模拟弹性条纹系统的新方法，以及如何利用它们来构建3d表面。目标是找到正确的2D条纹几何形状，当部署时对应于给定的3D设计表面。部署仿真对设计阶段至关重要，并且相信通过设计、仿真和重用来自已测试物理模型的数据，动态设计方法框架将自然地从设计-制造-仿真工作流过渡到设计-仿真-制造工作流。(Raviv et al. 2014)。为了找到条纹的分布及其与最终三维表面的接近程度，我们进行了三个实验:从一个简单的例子开始，为了得到一个负高斯曲率(图1，顶部)，并观察平面的变换和变形，将一个六边形平面模型垂直挤压并锚定在两点上。然后，进一步研究负曲率表面的部署技术，通过在其末端添加储物柜条纹，均匀地部署一组2D线性等条纹(图2，底部)。从样机和仿真部署来看，畸变产生负高斯曲率。一个更复杂的弯曲条纹的例子，在开始时使用一个储物柜条纹，并在它们之间连接，一组四个弯曲的可展开条纹被扭曲并连接在一起，形成四个拱门，同时保持它们的打开状态(图3，右)。在拱的模拟和原型中都可以看到相同的变形。
{"title":"Simulation and Fabrication of Elastic Deployable Stripe Structures","authors":"P. Baquero, Daoming Liu, Yota Adilenido","doi":"10.47330/dcio.2022.fbeo7122","DOIUrl":"https://doi.org/10.47330/dcio.2022.fbeo7122","url":null,"abstract":"Deployable structures have many applications in architecture, from kinetic pavilions to temporary structures, to retractable rooftops. There are various advantages to building deployable elements in a factory, and then deploying them on site. It is simpler to join stripes in flat arrangements than to put them together in three dimensions. This study focuses on a novel approach for creating and simulating systems of elastic stripes and how they can be utilized to build 3d-surfaces. The goal is to find the correct 2D stripes geometry that when deployed corresponds to a given 3D designed surface. Deployment simulation is essential to the design phase, and it is believes that by designing, simulating, and re-using data from already tested physical models, kinetic design methodological framework would naturally transition from a Design-Fabrication-Simulation workflow into a Design-Simulation-Fabrication one. (Raviv et al. 2014). In order to find the stripes deployment and its proximity to the final 3d surface, three experiments are examined ere: Starting from a simple case, in order to get a negative Gaussian curvature (Figure 1, Top) and observe the transformation and distortion of the flat faces, a hexagonal flat model has been vertically extruded and anchored in two points. Then, investigating further deploying techniques for negative curvature surfaces, a 2D linear set of equal stripes is deployed evenly by adding a locker stripe on its ends (Figure 2, Bottom). From the prototype and the simulation deployment the distortion produced a negative Gaussian curvature. A more complex example of curved stripes, using a locker stripe at the start and connecting between them, a group of four curved deployable stripes were distorted and joined to create four arches while maintaining their opening state (Figure 3, Right). The same distortion was seen in both the arches simulation and the prototype.","PeriodicalId":129906,"journal":{"name":"Design Computation Input/Output 2022","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122011972","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：