Two Generative Design Methods of Hospital Operating Department Layouts Based on Healthcare Systematic Layout Planning and Generative Adversarial Network

doi:10.1007/s12204-021-2265-9

Abstract

Abstract: With the increasing demands of health care, the design of hospital buildings has become increasingly demanding and complicated. However, the traditional layout design method for hospital is labor intensive, time consuming and prone to errors. With the development of artificial intelligence (AI), the intelligent design method has become possible and is considered to be suitable for the layout design of hospital buildings. Two intelligent design processes based on healthcare systematic layout planning (HSLP) and generative adversarial network (GAN) are proposed in this paper, which aim to solve the generation problem of the plane functional layout of the operating departments (ODs) of general hospitals. The first design method that is more like a mathematical model with traditional optimization algorithm concerns the following two steps: developing the HSLP model based on the conventional systematic layout planning (SLP) theory, identifying the relationship and flows amongst various departments/units, and arriving at the preliminary plane layout design; establishing mathematical model to optimize the building layout by using the genetic algorithm (GA) to obtain the optimized scheme. The specific process of the second intelligent design based on more than 100 sets of collected OD drawings includes: labelling the corresponding functional layouts of each OD plan; building image-to-image translation with conditional adversarial network (pix2pix) for training OD plane layouts, which is one of the most representative GAN models. Finally, the functions and features of the results generated by the two methods are analyzed and compared from an architectural and algorithmic perspective. Comparison of the two design methods shows that the HSLP and GAN models can autonomously generate new OD plane functional layouts. The HSLP layouts have clear functional area adjacencies and optimization goals, but the layouts are relatively rigid and not specific enough. The GAN outputs are the most innovative layouts with strong applicability, but the dataset has strict constraints. The goal of this paper is to help release the heavy load of architects in the early design stage and present the effectiveness of these intelligent design methods in the field of medical architecture.

Key words: healthcare systematic layout planning (HSLP)| generative adversarial network (GAN)| genetic algorithm
(GA)| plane layout design| hospital

摘要： With the increasing demands of health care, the design of hospital buildings has become increasingly demanding and complicated. However, the traditional layout design method for hospital is labor intensive, time consuming and prone to errors. With the development of artificial intelligence (AI), the intelligent design method has become possible and is considered to be suitable for the layout design of hospital buildings. Two intelligent design processes based on healthcare systematic layout planning (HSLP) and generative adversarial network (GAN) are proposed in this paper, which aim to solve the generation problem of the plane functional layout of the operating departments (ODs) of general hospitals. The first design method that is more like a mathematical model with traditional optimization algorithm concerns the following two steps: developing the HSLP model based on the conventional systematic layout planning (SLP) theory, identifying the relationship and flows amongst various departments/units, and arriving at the preliminary plane layout design; establishing mathematical model to optimize the building layout by using the genetic algorithm (GA) to obtain the optimized scheme. The specific process of the second intelligent design based on more than 100 sets of collected OD drawings includes: labelling the corresponding functional layouts of each OD plan; building image-to-image translation with conditional adversarial network (pix2pix) for training OD plane layouts, which is one of the most representative GAN models. Finally, the functions and features of the results generated by the two methods are analyzed and compared from an architectural and algorithmic perspective. Comparison of the two design methods shows that the HSLP and GAN models can autonomously generate new OD plane functional layouts. The HSLP layouts have clear functional area adjacencies and optimization goals, but the layouts are relatively rigid and not specific enough. The GAN outputs are the most innovative layouts with strong applicability, but the dataset has strict constraints. The goal of this paper is to help release the heavy load of architects in the early design stage and present the effectiveness of these intelligent design methods in the field of medical architecture.

关键词: healthcare systematic layout planning (HSLP)| generative adversarial network (GAN)| genetic algorithm
(GA)| plane layout design| hospital

CLC Number:

TU 246.1

ZHAO Chaowang (赵朝望), YANG Jian (杨健), XIONG Wuyue (熊吴越), LI Jiatong (李佳潼). Two Generative Design Methods of Hospital Operating Department Layouts Based on Healthcare Systematic Layout Planning and Generative Adversarial Network[J]. J Shanghai Jiaotong Univ Sci, 2021, 26(1): 103-115.

References 28

[1]	JAMES P, NOAKES T. Hospital architecture [M]. 3ed. Harlow, UK: Addison-Wesley Longman Ltd., 1994.
[2]	GOODFELLOW I J, POUGET-ABADIE J, MIRZAM, et al. Generative adversarial nets [C]//Proceedingsof the 27th International Conference on Neural InformationProcessing Systems. Cambridge, MA, USA:MIT Press, 2014: 2672-2680.
[3]	HUANG W X, ZHENG H. Architectural drawingsrecognition and generation through machine learning[C]//Proceedings of the 38th Annual Conference of theAssociation for Computer Aided Design in Architecture.Mexico City, Mexico: [s.n.], 2018: 156-165.
[4]	KRUPKA D C, SANDBERG W S. Operating roomdesign and its impact on operating room economics[J]. Current Opinion in Anaesthesiology, 2006, 19(2):185-191.
[5]	KHOSHKENAR A, TAAFFE K, MUHS M, et al.Simulation-based design and traffic flow improvementsin the operating room [C]//Proceedings of the 2017Winter Simulation Conference. Las Vegas, NV, USA:IEEE, 2017: 2975-2983.
[6]	BAYRAMZADEH S, JOSEPH A, ALLISON D, et al.Using an integrative mock-up simulation approach forevidence-based evaluation of operating room designprototypes [J]. Applied Ergonomics, 2018, 70: 288-299.
[7]	CHAILLOU S. AI+ architecture: Towards a new approach[R]. Cambridge, MA, USA: Graduate School ofDesign, Harvard University, 2019.
[8]	HARADA M, WITKIN A, BARAFF D. Interactivephysically-based manipulation of discrete/continuousmodels [C]//Proceedings of the 22nd Annual Conferenceon Computer graphics and Interactive Techniques.New York, USA: ACM, 1995: 199-208.
[9]	SAVOV A, TESSMANN O. Introduction to playablevoxel-shape grammars [C]//Proceedings of the 37thAnnual Conference of the Association for ComputerAided Design in Architecture. Boston, USA: [s.n.],2017: 534-543.
[10]	MARSON F, MUSSE S R. Automatic real-time generationof floor plans based on squarified treemaps algorithm[J]. International Journal of Computer GamesTechnology, 2010, 2010: 624817.
[11]	WANG X Y, YANG Y, ZHANG K. Customization andgeneration of floor plans based on graph transformations[J]. Automation in Construction, 2018, 94: 405-416.
[12]	AZIZI V, USMAN M, PATEL S, et al. Floorplanembedding with latent semantics and human behaviorannotations[C]//The 11th Annual Symposiumon Simulation for Architecture and Urban Design(SimAUD2020). [s.l.]: SCS, 2020: 343-350.
[13]	SIMON J. Evolving floorplans [EB/OL]. [2020-05-11].https://www.joelsimon.net/evo floorplans.html.
[14]	ZHENG H, REN Y. Architectural layout designthrough simulated annealing algorithm[C]//Proceedings of the 25th International Conferenceon Computer-Aided Architectural DesignResearch in Asia (CAADRIA 2020). Hong Kong,China: CAADRIA, 2020: Paper 24.
[15]	ISOLA P, ZHU J Y, ZHOU T H, et al. Image-toimagetranslation with conditional adversarial networks[C]//2017 IEEE Conference on Computer Visionand Pattern Recognition (CVPR). Honolulu, HI,USA: IEEE, 2017: 5967-5976.
[16]	CAMPO D, MANNINGER S, SANCHE M, et al.The Church of AI: An examination of architecturein a posthuman design ecology [C]//Proceedings ofthe 24th International Conference on Computer-AidedArchitectural Design Research in Asia: Intelligentand Informed, CAADRIA 2019. Hong Kong, China:CAADRIA, 2019: 767-772.
[17]	CHO D, KIM J, SHIN E, et al. Recognizing architecturalobjects in floor-plan drawings using deep-learningstyle-transfer algorithms [C]//Proceedings of the 25thInternational Conference on Computer-Aided ArchitecturalDesign Research in Asia (CAADRIA 2020).Hong Kong, China: CAADRIA, 2020: 717-725.
[18]	ZHENG H, AN K, WEI J X, et al. Apartmentfloor plans generation via generative adversarial networks[C]//Proceedings of the 25th International Conferenceon Computer-Aided Architectural Design Researchin Asia (CAADRIA 2020). Hong Kong, China:CAADRIA, 2020: Paper 15.
[19]	NEWTON D. Deep generative learning for the generationand analysis of architectural plans with smalldatasets [C]//Proceedings of the 37th eCAADe and23rd SIGraDi Conference. Porto, Portugal: [s.n.],2019: 21-28.
[20]	National Health Commission of the People’s Republicof China. Code for design of general hospital: GB51039-2014 [S]. Beijing, China: China Planning Press,2014 (in Chinese).
[21]	WANG H D, LIU X P. A preliminary study on thecomparison of architectural design standards of emergencydepartments of general hospitals in China withAnglo-American standards [J]. Chinese and OverseasArchitecture, 2019 (1): 77-80 (in Chinese).
[22]	Facility Guidelines Institute. Guidelines for design andconstruction of hospitals and outpatient facilities 2014[S]. Chicago, USA: American Hospital Association,2014.
[23]	Veterans Health Administration Washington. Spaceplanning criteria (PG-18-9). Chapter 102: Intensivecare nursing units [S]. Washington, DC, USA: NationalInstitute of Building Sciences, 2016.
[24]	SHARMA D, GUPTA N, CHATTOPADHYAY C, etal. DANIEL: A deep architecture for automatic analysisand retrieval of building floor plans [C]//2017 14thIAPR International Conference on Document Analysisand Recognition (ICDAR). Kyoto, Japan: IEEE,2017: 420-425.
[25]	HASDA R K. Contribution to the optimizationof unequal area rectangular facility layoutproblem [EB/OL]. [2020-05-11]. https://hal.archivesouvertes.fr/tel-01716010.
[26]	DINO I G. An evolutionary approach for 3D architecturalspace layout design exploration [J]. Automationin Construction, 2016, 69: 131-150.
[27]	HAUPT R L, HAUPT S E. Practical genetic algorithms[M]. 2 ed. Hoboken, NJ, USA: John Wiley &Sons, 2004.
[28]	LIN Y Z, LIN Y C. Applying an immune ant colonysystem algorithm to solve an integrated flexible bayfacility layout problem with input/output points design[J]. Lecture Notes in Management Science, 2015,7: 57-62.