簡易檢索 / 詳目顯示

研究生: 莊政霖
Chuang, Cheng-Lin
論文名稱: 以機器學習工具輔助建築師與業主於設計初期之溝通
Facilitating Architect-Client Communication in the Pre-Design Phase through Machine Learning Based Tools
指導教授: 簡聖芬
Chien, Sheng-Fen
學位類別: 碩士
Master
系所名稱: 規劃與設計學院 - 建築學系
Department of Architecture
論文出版年: 2022
畢業學年度: 110
語文別: 中文
論文頁數: 84
中文關鍵詞: 初期設計建築計劃設計溝通機器學習類神經網路人工智慧MobileNetPix2Pix
外文關鍵詞: Pre-design, Architectural Programming, Design Communication, Machine Learning, Neural Network, Artificial Intelligence, MobileNet, Pix2Pix
相關次數: 點閱:130下載:51
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報
  • 一個建築專案從初期設計階段開始,就會開始研究和分析空間需求、建築基地機會和限制、成本及預算。這個階段也是建築師和業主進行探索建築計畫進而在想法及方向上達成共識的重要階段。由於業主通常不具設計專業知識以致難以表達其需求及偏好。
    從文獻回顧歸納建築師與業主溝通時使用的詞彙及圖像是不直觀的,造成溝通上有所限制以至於有誤解。大部分的輔助溝通工具都試圖將建築圖說轉為視覺化的建築空間,讓業主直觀的了解設計,而且皆強調業主直接參與設計,但都屬於設計中後期階段的溝通輔助。類神經網路機器學習是以資料驅動的運算機制,可以依資料進行預測,有機會幫助建築師對業主表達合理的判斷和協助設計思考。
    本研究以建築師與業主溝通時最常使用的三種媒材:口語、照片及草圖為基礎建構溝通輔助工具,分別為輔助分析的環境資料提示工具、提供建議的風格判斷工具及繪製草圖的繪圖工具。環境提示工具的核心智慧為以業主的觀點說明基地的相關文本資料訓練所得的單純貝氏分類模型。風格判斷工具及繪圖工具的核心智慧建構則是上網蒐集圖片後,根據風格及材料相關規則由建築專業者進行分類,最後分別使用MobileNet模型及Pix2Pix模型進行訓練而成。
    開發完成的機器學習輔助溝通工具進行實際使用的測試,並將溝通雙方的逐字稿進行三種文本分析:議題,回應及媒材,觀察是否達成增進溝通的結果。實驗結果符合實驗期待,有增進建築師及業主的溝通,其中風格判斷工具明顯有幫助建築師找尋合適的資料與業主進行溝通。但在實驗過程發現建築師在進行溝通時本身已有擅長的溝通媒材,在忽略他們的使用習慣下進行機器學習工具的使用,讓機器學習工具作為協助溝通的必要性效果甚微。
    以機器學習作為輔助溝通具有相當的潛力,尤其是將業主給予的照片篩選部分,先行篩選合適的風格幫助建築師省去不少搜尋的時間並且促進業主在議題的理解上有不錯的效果。另外在實驗中找出建築師在設計初期階段溝通實真正需要幫助的資料。在基地環境溝通方面,可以蒐集土地法規的資料;風格溝通方面,可持續對相關的建築風格進行照片的蒐集;繪圖溝通方面,繪製量體及基地配置圖仍有需求。

    The pre-design phase is an important stage for architects and clients to explore design requirements and finalize architectural programs. Communications between architects and clients are often slow and with many misunderstandings because clients cannot articulate their needs and preferences clearly, precisely, or comprehensively. The thesis employs machine learning technologies to develop tools that facilitate the communications. An environmental cueing tool for assisting verbal discussions of cite analysis; a style classification tool for assisting communications through photos; and a drawing tool for discussions using sketches. Experiments are conducted to examine the effectiveness of the tools. The results show that the style classification tool is the most effective and preferable among the three tools; and that photo is the best medium for communication.

    摘要 i ABSTRACT ii 謝誌 v 目錄 vi 圖目錄 vii 表目錄 viii 第1章 緒論 1 1.1 研究背景與動機 1 1.2 研究課題與目標 2 1.3 研究方法 2 1.4 論文架構 3 第2章 文獻回顧4 2.1 建築師在設計初期的溝通 4 2.2 溝通的媒材 5 2.3 溝通工具的相關研究 7 2.4 機器學習技術 9 2.5 小結 10 第3章 系統實作 12 3.1 系統需求分析 12 3.2 系統規劃 13 3.3 系統開發 14 第4章 系統實測評估 26 4.1 實驗假設 26 4.2 實驗設計 27 4.3 資料分析 28 4.4 實驗結果 30 4.5 小結 34 第5章 結論 35 5.1 研究成果 35 5.2 研究貢獻 35 5.3 後續研究 36 參考文獻 37 附錄 42 附錄A. 溝通輔助工具開發程式 43 附錄B. 溝通輔助工具操作示範 44 附錄C. 實驗採用之設計專案資料 48 附錄D. 實驗記錄:建築師的議題 54 附錄E. 實驗記錄:口語資料 71 附錄F. 實驗記錄:資料分析 72

    Abdelmohsen, S. M. A. (2011). An ethnographically informed analysis of design intent communication in BIM-enabled architectural practice [Doctoral dissertation, Georgia Institute of Technology]. Georgia Tech Library. http://hdl.handle.net/1853/41181
    Alkass, S., & Jrade, A. (2002). A web-based virtual reality model for preliminary estimates of hi-rise building projects. Sixth Design and Decision Support Systems in Architecture and Urban Planning - Part One: Architecture Proceedings, 27–34. http://papers.cumincad.org/cgi-bin/works/paper/ddssar0203
    Ang, G., Wyatt, D., & Hermans, M. (2001). A systematic approach to define client expectations of total building performance during the pre-design stage. CIB World Building Congress. https://www.irbnet.de/daten/iconda/CIB2813.pdf
    As, I., Pal, S., & Basu, P. (2018). Artificial intelligence in architecture: Generating conceptual design via deep learning. International Journal of Architectural Computing, 16(4), 306–327. https://doi.org/10.1177/1478077118800982
    Barrett, P., & Stanley, C. A. (1999). Better Construction Briefing. Wiley-Blackwell.
    Belém, C., Santos, L., & Leitão, A. (2019). On the impact of machine learning: architecture without architects? 18th International Conference CAAD Futures 2019, 148–167. http://papers.cumincad.org/cgi-bin/works/paper/cf2019_020
    Bragança, L., Vieira, S. M., & Andrade, J. B. (2014). Early stage design decisions: The way to achieve sustainable buildings at lower costs. The Scientific World Journal, 2014, 363364. https://doi.org/10.1155/2014/365364
    Broadbent, G. (1988). Design in Architecture: Architecture and Human Sciences (2nd ed.). David Fulton Publishers.
    Brown, G., & Gifford, R. (2001). Architects predict lay evaluations of large contemporary buildings: Whose conceptual properties? Journal of Environmental Psychology, 21(1), 93–99. https://doi.org/10.1006/jevp.2000.0176
    Cudzik, J., & Radziszewski, K. (2018). Artificial intelligence aided architectural design. Computing for a Better Tomorrow - Proceedings of the 36th ECAADe Conference - Volume 1, 77–84. https://doi.org/https://doi.org/10.52842/conf.ecaade.2018.1.077
    Emmitt, S., & Gorse, C. A. (2003a). Communication in construction. In Construction Communication (pp. 12–19). Blackwell.
    Emmitt, S., & Gorse, C. A. (2003b). Selecting appropriate communication media. In Construction Communication (pp. 117–131). Blackwell.
    Fawcett, W., Ellingham, I., & Platt, S. (2008). Reconciling the architectural preferences of architects and the public: The ordered preference model. Environment and Behavior, 40(5), 599–618. https://doi.org/10.1177/0013916507304695
    Greusel, D. (2008). Communicating with clients. In J. A. Demkin (Ed.), The Architect’s Handbook of Professional Practice (14th ed., pp. 222–232). Wiley & Sons, Inc.
    Hesse, C. (2017, February 19). Image-to-Image Demo. https://affinelayer.com/pixsrv/
    Howard, A. G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., Andreetto, M., & Adam, H. (2017). MobileNets: efficient convolutional neural networks for mobile vision applications. arXiv. https://doi.org/10.48550/ARXIV.1704.04861
    Huang, C.-H. J., & Krawczyk, R. J. (2007). Web based BIM for modular house development: query approach in consumer participatory design. The Third International Conference of the Arab Society for Computer Aided Architectural Design (ASCAAD 2007), 559–570. http://papers.cumincad.org/cgi-bin/works/paper/ascaad2007_044
    Isola, P., Zhu, J.-Y., Zhou, T., & Efros, A. A. (2017). Image-to-image translation with conditional adversarial networks. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 5967–5976. https://doi.org/10.1109/CVPR.2017.632
    Kalay, Y. (2004). Habitual methods of representation. In Architecture’s New Media: Principles, Theories, and Methods of Computer-Aided Design (pp. 87–117). MIT Press.
    Lertlakkhanakul, J., Choi, J. W., & Kim, M. Y. (2008). Building data model and simulation platform for spatial interaction management in smart home. Automation in Construction, 17(8), 948–957. https://doi.org/10.1016/j.autcon.2008.03.004
    Li, Y. H., & Jain, A. K. (1998). Classification of text documents. The Computer Journal, 41(8), 537–546. https://doi.org/10.1093/comjnl/41.8.537
    Mohamed, A., & Celik, T. (2002). Knowledge based-system for alternative design, cost estimating and scheduling. Knowledge-Based Systems, 15(3), 177–188. https://doi.org/10.1016/S0950-7051(01)00155-1
    Norouzi, N., Shabak, M., Embi, M. R. Bin, & Khan, T. H. (2015). The architect, the client and effective communication in architectural design practice. Procedia - Social and Behavioral Sciences, 172, 635–642. https://doi.org/10.1016/j.sbspro.2015.01.413
    Perkins, B. (2008). Design Phase. In J. A. Demkin (Ed.), The Architect’s Handbook of Professional Practice (14th ed., pp. 520–529). Wiley & Sons, Inc.
    Royal Institute of British Architects [RIBA]. (2020). The RIBA Plan of Work 2020 Overview. RIBA. https://riba-prd-assets.azureedge.net/-/media/GatherContent/Business-Benchmarking/Additional-Documents/2020RIBAPlanofWorkoverviewpdf.pdf
    Shahrin, F., & Johansen, E. (2013). Challenges in engaging the client during the capture, translation, transformation and delivery (CTTD) of client requirements (CR) within the BIM environment. In O. J. Klakegg, K. H. Kjølle, C. G. Mehaug, N. O. E. Olsson, A. T. Shiferaw, & R. Woods (Eds.), Proceedings from 7th Nordic Conference on Construction Economics and Organization 2013 (pp. 469–478). Construction Researchers on Economics and Organisation in the Nordic region (CREON) & Akademika Publishing. https://www.diva-portal.org/smash/get/diva2:1007649/FULLTEXT01.pdf#page=485
    Shen, W., Zhang, X., Shen, G. Q., & Fernando, T. (2013). The user pre-occupancy evaluation method in designer-client communication in early design stage: A case study. Automation in Construction, 32, 112-124. https://doi.org/10.1016/j.autcon.2013.01.014
    Shinde, P. P., & Shah, S. (2018). A review of machine learning and deep learning applications. 2018 Fourth International Conference on Computing Communication Control and Automation (ICCUBEA), 1–6. https://doi.org/10.1109/ICCUBEA.2018.8697857
    Shouman, B., Othman, A. A. E., & Marzouk, M. (2022). Enhancing users involvement in architectural design using mobile augmented reality. Engineering, Construction and Architectural Management, 29(6), 2514–2534. https://doi.org/10.1108/ECAM-02-2021-0124
    Sirikasem, P., & Degelman, L. O. (1990). The use of video-computer presentation techniques to aid in communication between architect and client. ACADIA ’90 Proceedings: From Research to Practice, 205–216. https://doi.org/10.52842/conf.acadia.1990.205
    Smith, K. S. (2012). Introduction and the sketch. In Architects’ Sketches: Dialogue and Design (pp. 1–25). Routledge. https://doi.org/10.4324/9780080878775
    Steinfeld, K. (2017). Dreams may come. Proceedings of the 37th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA), 590–599. https://doi.org/10.52842/conf.acadia.2017.590
    Taleb, H., Ismail, S., Wahab, M. H., & Rani, W. N. M. W. M. (2017). Communication management between architects and clients. AIP Conference Proceedings 1891, 020136. https://doi.org/10.1063/1.5005469
    Tessema, Y. A. (2008). BIM for improved building design communication between architects and clients in the schematic design phase [Master thesis, Texas Tech University]. Texas Tech University Libraries. http://hdl.handle.net/2346/11380
    Kailash, A(2020)。生成對抗網絡項目實戰(倪琛 譯,一版)。人民郵電出版社。(原著出版於2018年)
    何晗(2020)。NLP工程師養成術-自然語言處理入門(一版)。博碩文化。
    邱茂林、顏蘇禎(1998)。建築設計過程中設計表現形式與視覺溝通現象之研究。設計學報,3(2),87–110。https://www.jodesign.org.tw/index.php/JODesign/article/viewFile/738/387
    施威銘研究室(2020)。tf.keras 技術者們必讀!深度學習攻略手冊。旗標科技股份有限公司。
    飯塚豊(2018)。初步的基地勘查及聽取業主的需求。載於新手建築師の教科書(桑田德 譯,頁120–127)。原點出版。(原著出版於1997年)
    謝尚賢、郭榮欽、陳奐廷、蔡沅澔(2016)。認識BIM與塑模演練案例介紹。載於透過案例演練學習BIM:基礎篇(增訂一版,頁12–53)。國立臺灣大學出版中心。
    羅仁君(2011)。應用詞頻以改良多元貝式定理於文件分類之研究(碩士論文,中國文化大學)。中國文化大學機構典藏。http://ir.lib.pccu.edu.tw/handle/987654321/23850

    下載圖示 校內:立即公開
    校外:立即公開
    QR CODE