An overlooked solution for climate-friendly buildings: Better floor plans

Posted on

Image: AI generated by Anthropocene magazine

When researchers auto-generated small apartment floor plans to make them efficient and comfortable, they found they could often reduce carbon emissions more than by changing the building exterior

By Sarah DeWeerdt
October 8, 2024 in Anthropocene magazine

A new algorithm-based method of designing floor plans for apartment buildings could reduce greenhouse gas emissions by ensuring more efficient use of space, according to a new study.

Buildings account for 40% of greenhouse gas emissions, but until now most policies to encourage more sustainable construction have been focused on heating and cooling systems or building materials such as insulation and windows.

“These findings question the way we currently incentivize new construction or give tax breaks to ‘supposedly’ low-carbon buildings,” says study team member Ramon Weber, an architect at the University of California in Berkeley, who conducted the work as a graduate student at the Massachusetts Institute of Technology. That is, pack a McMansion full of insulation and stick a set of solar panels on the roof, and by current standards you’ve got yourself an environmental winner.

It’s well known that energy use in buildings is roughly proportional to building size, but there has been little attention to making apartments as small as possible while making sure they still function well for living.

Making buildings more space-efficient would reduce greenhouse gas emissions associated with building operation to a greater degree than making changes to the building exterior in (up to) 72%

“As an architect I find it very encouraging that we could prove that space matters,” Weber says. “For too long sustainability has been a question that we only try to solve with ‘technical’ means. I hope that my research contributes to emphasizing that spatially efficient designs are crucial for creating low carbon buildings.”

Making buildings more space-efficient would reduce greenhouse gas emissions associated with building operation to a greater degree than making changes to the building exterior in 72% of a sample of buildings in Zurich, Switzerland, 61% in New York, New York, and 33% in Singapore, Weber and his collaborators report in a paper published in the journal Nature Communications.

To demonstrate this, Weber and his collaborators turned to a newly developed tool called a “hypergraph,” which encodes floor plans as a collection of nodes and edges representing rooms and the connections between them.

newly developed tool called a “hypergraph,” which encodes floor plans as a collection of nodes and edges

“I was actually surprised how robust and scalable the method turned out,” says Weber. “I had surveyed different approaches and it became clear that we could not just use existing methods from computer science research to represent a floor plan. The hypergraph framework that I developed is specifically made for architectural design and building floor plans, and as such the first of its kind.”

The hypergraph can help researchers evaluate existing floor plans as well as automatically generate new floor plans to subdivide a given building footprint into apartments or to subdivide a given apartment unit into rooms.

looking at a how building floor plans relate to a building’s structure

A separate computer analysis ensures that the floor plans can accommodate all of the necessary furniture for a given room’s function (for example, a bed, dresser, and cabinet in a bedroom) while minimizing excess space.

The researchers also modeled the floor plans in three dimensions to analyze daylight and other aspects of comfort for people living in the apartments. The automatically generated floor plans increased apartments’ access to daylight by up to 24% in Zurich, the researchers report, suggesting that this strategy could make dwellings more pleasant to live in as well as more environmentally sound. (They found smaller improvements in access to daylight for buildings in New York and Singapore.)

automatically generated floor plans increased apartments’ access to daylight by up to 24%

In the future, Weber aims to expand his analysis from apartments to whole neighborhoods. “On a building scale we have to figure out what design decisions have the biggest carbon impact,” he adds. “For instance looking at a how building floor plans relate to a building’s structure and with it the carbon emissions from the materials used(the embodied carbon).”

Source: Weber R.E. et al. “A hypergraph model shows the carbon reduction potential of effective space use in housing.” Nature Communications 2024.

Abstract

Humans spend over 90% of their time in buildings, which account for 40% of anthropogenic greenhouse gas emissions and are a leading driver of climate change. Incentivizing more sustainable construction, building codes are used to enforce indoor comfort standards and minimum energy efficiency requirements. However, they currently only reward measures such as equipment or envelope upgrades and disregard the actual spatial configuration and usage. Using a new hypergraph model that encodes building floorplan organization and facilitates automatic geometry creation, we demonstrate that space efficiency outperforms envelope upgrades in terms of operational carbon emissions in 72%, 61% and 33% of surveyed buildings in Zurich, New York, and Singapore. Using automatically generated floorplans in a case study in Zurich further increased access to daylight by up to 24%, revealing that auto-generated floorplans have the potential to improve the quality of residential spaces in terms of environmental performance and access to daylight.

Introduction

Current estimates predict that the global built area may grow by 250 billion square meters by 2050 to house a growing population1,2. Such estimates are necessarily extrapolations from current building practices. While many decades of building science research and practice have enabled design teams across the world to precisely predict carbon reduction savings that can be attained through any number of upgrades for building operation3,4,5,6 and materials7,8, surprisingly little attention has been paid to space evaluation methods. The ubiquitous energy use intensity (EUI) metric, defined as energy use per conditioned floor area, has become the de facto benchmark for high-performance buildings, leading to sometimes absurd situations where over-sized single-family homes with rooftop photovoltaics are hailed as beacons of sustainability despite their significant material and space use per occupant.

Given that energy use roughly scales with building size, reducing the floor area per apartment unit while maintaining good indoor environmental conditions, offers a complementary path toward a net zero building stock. Traditional architectural design workflows are unsuitable for this type of exploration since they rely on a human manually drawing interior walls while considering a plethora of architectural, safety, and Americans with Disabilities Act (ADA) requirements9. In residential construction, the position of these interior partitions obviously impacts access to daylight, thermal comfort, and views of the outside. Many design decisions are intuitively made by architects based on prior experience or reference projects10,11 but without assessing their impact on building performance12 due to the time, effort, and technical sophistication required to conduct this type of analysis. However, coupling methods of design with quantitative simulation feedback in an early design stage has the potential to significantly improve design outcomes13.

While the construction industry has long shied away from quantitatively evaluating space use, the urgency of the climate crisis, along with a shortage of architects to meet the global housing demand, has led to some, mostly developer-driven and proprietary attempts to automatically generate floor plans14. Most current implementations are linked to financial cost models, evaluating multiple ways to divide a building footprint into a desired number of apartment units15,16,17. Current approaches for within-unit room divisions are an active area of computer graphics research but are not presently used in the architecture field due to various limitations, including only being able to represent rectangular18 or orthogonal boundary conditions19, or solely responding to either topological or spatial or boundary constraints20,21,22,23,24. On a technical level, machine learning (ML) based models create neural networks that relate the geometric graph structures from room walls to an adjacency graph (vector25,26 or pixel-based27) or use reinforcement learning to subdivide a space28. This results in a linear, one-sided generation process, where a room adjacency graph is converted into a visually real and geometrically valid floor plan. Inherently these statistical processes do not allow for exact specifications of room sizes, boundary conditions, or further geometric manipulation of parts of the final output, as are needed in architectural design. Furthermore, implicit geometric relationships are difficult to train, and floor plan training data is sparse, scarce, and unvetted; thus, such approaches can neither guarantee architectural quality nor environmental performance14.

In this work, we present the hypergraph, a generalizable shape generator and descriptor for floor plans. The hypergraph represents key characteristics of the shape divisions of any given floor plan layout, enabling both the mapping and benchmarking of suitable, high-performing floor plans, as well as their automatic generation. A hypergraph is created from existing building floor plans and can be applied to new conditions. This allows for translating cultural conventions and practices into new designs, and a fully transparent source attribution. We introduce a spatial analysis workflow to minimize “excess space” while retaining the same spatial functionality of a given floor plan. The concept of excess space is based on the notion that a room with a certain program, for example, a bedroom, has minimum functional requirements in terms of furniture (bed, dresser, cabinet) and space around that furniture that supports its proper use. Areas beyond those functional requirements are then defined as excess. Furthermore, an automatic integration of environmental analysis methods, assessing energy use and daylight, allows us to benchmark high-performance designs and maximize occupant comfort conditions.

space sufficiency can become a highly impactful carbon mitigation strategy, informing future building energy policy, and should guide the standards and building codes of cities in the future

Hypergraphs can be used to generate and evaluate floor plans

To describe a whole building, we can apply the hypergraphs to an apartment boundary, generating detailed floor plans for each apartment unit. A fitting procedure is shown in detail in Fig. 2, in which an apartment boundary polygon (Fig. 2a) is subdivided by a library of different hypergraphs (Fig. 2b) to create different internal apartment configurations (Fig. 2c). We then use the apartment boundary polygon and its orientation toward the building circulation to filter floor plans with similar orientations and façade to adiabatic wall ratios. The hypergraph method removes the need for manual drawing of floor plans and preparation of geometry for different environmental simulation procedures. It allows the complex structure of a floor plan to be described as a graph, a quantifiable and searchable data structure that encodes key parameters of a design. To filter geometrically valid but spatially inadequate outputs, a series of heuristics filter and rank feasible results (see “Methods”, Apartment validity heuristic). With this, we can generate architecturally feasible floor plans where rooms have an aspect ratio and size that makes them usable for their specified use, have access to a façade, and are configured in a way that allows access within the apartment and to the building’s circulation. For assessing the spatial validity of a floor plan, we propose an automatic version of the spatial scoring system developed by the City of Berlin’s public housing provider30. Using automatic placement of furniture blocks, we can assess if rooms are large enough to result in livable spaces and compare the overall area to reference floor plans with the same occupancy (Fig. 2d) (see “Methods,” Furniture placement).

Fig. 2: Steps for fitting an apartment using the hypergraph method.
figure 2

An apartment boundary is extracted from a building (a) and combined with a library of hypergraphs (b). The applied hypergraphs generate different internal subdivisions for the apartment boundary (c). A spatial evaluation using placement of furniture, accessibility, and room geometry is performed to filter feasible solutions (d). Energy and daylight analysis are performed to further evaluate the resulting floor plan (e), and a chosen plan is inserted into the building (f).

Full size image

To estimate daylight and energy performance, the selected floor plans are automatically converted into a simple 3D model, with walls and windows, that creates a building energy model of the apartment. Focusing on the spatial layout, the sizing of windows was kept constant (at a window-to-wall ratio of 0.6) and placed automatically on walls of rooms requiring daylight. This 3D model can be used to create a building energy model (BEM), a heat-transfer and mass-flow simulation that is industry standard for energy use predictions31. However, current BEM practices do not break up apartments (or floors) into detailed sub-spaces, creating a lack of data and standards on detailed room models and how they relate to occupancy. Because of this, we create a simplified apartment BEM that does not differentiate energy use by room program or occupancy and uses the same temperature setpoint across all zones. While this use is justified in the context of the United States, where apartments typically have unified setpoints by a single thermostat, this does not hold in Zurich or Singapore. However, in a sample apartment simulation in different cities, we found only a modest impact of differentiating BEM zones (see Supplementary Note 1.10). Furthermore, we calculate daylight access in the apartments by assessing the spatial daylight autonomy (sDA), a metric for interior spaces that, through a yearly illuminance simulation with physics-based raytracing and local weather data32, predicts the percentage of hours per year when a minimum light level of 300 lux can be achieved with daylight (Fig. 2e). While whole building energy models typically do not have the geometric resolution of single rooms, the models generated with the hypergraph method will allow more detailed energy performance analysis that can capture effects of airflow and natural ventilation for more accurate predictions. Detailed room geometry further allows for more accurate daylight predictions than simple shoebox models or whole building massings, as the internal configuration of a floor plan will determine how light is obstructed inside an apartment.

References

  1. IEA, The Critical Role of Buildings, Perspectives for the Clean Energy Transition, International Energy Agency, Paris, https://www.iea.org/reports/the-critical-role-of-buildings (2019).

  2. IEA, Net Zero by 2050 – A Roadmap for the Global Energy Sector, International Energy Agency, https://www.iea.org/reports/net-zero-by-2050 (2021).

  3. Reyna, J. L. & Chester, M. V. Energy efficiency to reduce residential electricity and natural gas use under climate change. Nat. Commun. 8, 14916 (2017).

  4. Zhong, X. et al. Global greenhouse gas emissions from residential and commercial building materials and mitigation strategies to 2060, Nat. Commun. 12, 6126 (2021).

  5. Ang, Y. Q., Berzolla, Z. M., Letellier-Duchesne, S. & Reinhart, C. F. Carbon reduction technology pathways for existing buildings in eight cities, Nat. Commun. 14, 1689 (2023).

  6. Baniassadi, A., Heusinger, J., Gonzalez, P. I., Weber, S. & Samuelson, H. W. Co-benefits of energy efficiency in residential buildings. Energy 238, 121768 (2022).

    Article Google Scholar

  7. Simonen, K., Rodriguez, B. X. & De Wolf, C. Benchmarking the Embodied Carbon of Buildings. Technol. Des 1, 208–218 (2017).

    Google Scholar

  8. Röck, M. et al. Embodied GHG emissions of buildings – The hidden challenge for effective climate change mitigation. Appl. Energy 258, 114107 (2020).

    Article Google Scholar

  9. Civil Rights Division, Disability Rights Section, 2010 ADA Standards for Accessible Design. https://www.ada.gov/assets/pdfs/2010-design-standards.pdf (2010).

  10. T. Jocher and S. Loch, Raumplilot Grundlagen. kraemerverlag, (2012).

  11. O. Heckmann, F. Schneider, and E. Zapel, Floor Plan Manual Housing, Fifth, Rev. Birkhäuser, https://doi.org/10.1515/9783035611496 (2018).

  12. Çavuşoğlu Ö. H. & Çağdaş, G. Why do we need building information modeling (BIM) in conceptual design phase? In Computer-Aided Architectural Design. Future Trajectories, G. Çağdaş, M. Özkar, L. F. Gül, and E. Gürer, Eds., Singapore: Springer Singapore, pp. 121–135 (2017).

  13. Burnell, E., Stern, M., Flooks, A. & Yang, M. C. Integrating design and optimization tools: A designer centered study, In Volume 7: 29th International Conference on Design Theory and Methodology, Cleveland, Ohio, USA: American Society of Mechanical Engineers, (2017).

  14. Weber, R. E., Mueller, C. & Reinhart, C. Automated floorplan generation in architectural design: A review of methods and applications. Autom. Constr. 140, 104385 (2022).

    Article Google Scholar

  15. Sidewalk Labs – Part of Google, Delve. New York, 2024. https://www.sidewalklabs.com/products/delve (2024).

  16. Archistar. Sydney, https://www.archistar.ai/ (2024).

  17. Test Fit. Dallas https://www.testfit.io/ (2024).

  18. Ślusarczyk, G., Strug, B., Paszyńska, A., Grabska, E. & Palacz, W. Semantic-driven graph transformations in floor plan design. Comput. Aided Des. 158, 103480 (2023).

    Article MathSciNet Google Scholar

  19. Bisht, S. et al. Transforming an adjacency graph into dimensioned floorplan layouts. Comput. Graph. Forum 41, 5–22 (2022).

    Article Google Scholar

  20. Para, W., Guerrero, P., Kelly, T., Guibas, L. & Wonka, P. Generative Layout Modeling using Constraint Graphs. In 2021 IEEE/CVF International Conference on Computer Vision (ICCV), Montreal, QC, Canada: IEEE, 6670–6680 (2021).

  21. Wu, W., Fan, L., Liu, L. Wonka, P. MIQP-based layout design for building interiors, Comput. Graph. Forum 37, 511–521 (2018).

  22. Nauata, N., Chang, K.-H., Cheng, C.-Y., Mori, G. & Furukawa, Y. House-GAN: Relational Generative Adversarial Networks for Graph-Constrained House Layout Generation, presented at the Computer Vision – ECCV 162–177. https://doi.org/10.1007/978-3-030-58452-8_10 (2020).

  23. Hu, R. et al., Graph2Plan: Learning floorplan generation from layout graphs. ACM Trans. Graph. 39, 118 (2020).

  24. Sun, J. et al. WallPlan: synthesizing floorplans by learning to generate wall graphs. ACM Trans. Graph. 41, 1–14 (2022).

    CAS Google Scholar

  25. Shabani, M. A., Hosseini, S., & Furukawa, Y. Housediffusion: Vector floorplan generation via a diffusion model with discrete and continuous denoising. In 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, BC, Canada: IEEE, 5466–5475 (2023).

  26. Tang, H. et al., Graph transformer GANs for graph-constrained house generation. In 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, BC, Canada: IEEE, 2173–2182 (2023).

  27. Carrera, L. et al. The impact of architecturally qualified data in deep learning methods for the automatic generation of social housing layouts. Autom. Constr. 158, 105238 (2024).

    Article Google Scholar

  28. Kakooee R. & Dillenburger, B. Reimagining space layout design through deep reinforcement learning, J. Comput. Des. Eng. 11, 43–55 (2024).

  29. De Berg, M., Cheong, O., Van Kreveld, M. & Overmars, M. Computational Geometry: Algorithms and Applications. Berlin, Heidelberg: (Springer Berlin Heidelberg, 2008).

  30. Howoge, Grundrissbewertungssystem. https://www.howoge.de/wohnungsbau/expertise-bau/grundrissbewertungssystem.html (2023).

  31. Polly, B., Kruis, N. & Roberts, D. Assessing and Improving the Accuracy of Energy Analysis for Residential Buildings. National Renewable Energy Lab. (NREL), Golden, CO (United States, 2011).

  32. Illuminating Engineering Society of North America. ANSI Standard: IES Spatial Daylight Autonomy (sDA) and Annual Sunlight Exposure (ASE). https://webstore.ansi.org/preview-pages/IESNA/preview_IES+LM-83-12.pdf (2024).

  33. Goldstein, B., Gounaridis, D. & Newell, J. P. The carbon footprint of household energy use in the United States. Proc. Natl. Acad. Sci. USA 117, 19122–19130 (2020).

    Article ADS CAS PubMed PubMed Central Google Scholar

  34. Asensio, O. I. & Delmas, M. A. The effectiveness of US energy efficiency building labels. Nat. Energy 2, 17033 (2017).

    Article ADS Google Scholar

  35. ASHRAE 90.1-2022 (I-P). https://www.techstreet.com/ashrae/standards/ashrae-90-1-2022-i-p?product_id=2522082 (2023).

  36. USGBC, LEED Certification, U.S. Green Building Council. https://www.usgbc.org/ (2023).

  37. Hu, S. et al. A systematic review of building energy sufficiency towards energy and climate targets. Renew. Sustain. Energy Rev. 181, 113316 (2023).

    Article Google Scholar

  38. Grynbaum, M. & Mac, R. The Times Sues OpenAI and Microsoft Over A.I. Use of Copyrighted Work. New York Times, New York, 28, (2023).

  39. Epstein, Z. et al. Art and the science of generative AI. Science 380, 1110–1111 (2023).

    Article ADS CAS PubMed Google Scholar

  40. Schneider, T. & Till, J. Flexible housing: opportunities and limits. Arq Archit. Res. Q. 9, 157–166 (2005).

    Google Scholar

  41. Tarkhan, N., Mokhtar, S., Weber, R. E. & Reinhart, C. Natural ventilation in a warming climate: An evaluation of computational simulation methods and metrics.In 2022 Annual Modeling and Simulation Conference (ANNSIM), San Diego, CA, USA: IEEE, 694–705 (2022).

  42. Poleg, D. The Next Crisis Will Start With Empty Office Buildings. The Atlantic. https://www.theatlantic.com/ideas/archive/2023/06/commercial-real-estate-crisis-empty-offices/674310/ (2023).

  43. Rowden, Jacob U. S. Office Outlook, Cyclical challenges persist but tentative green shoots emerging in second quarter. JLL, https://www.us.jll.com/content/dam/jll-com/documents/pdf/research/americas/us/jll-us-office-outlook-q2-2023.pdf (2023).

  44. Loh, T. Hadden, Terplan, E. & Rowlands, D. Myths about converting offices into housing—and what can really revitalize downtowns. (The Brookings Institution, Washington, 2023).

    Google Scholar

  45. Stamm-Teske, W., Fisher, K. & Haag, T. Raumpilot: Wohnen. Krämer, (2010).

  46. Zapel, E. Birkhäuser. Floor plan manual housing, (2017).

  47. Badger, E. & Buchanan, L. Here’s How to Solve a 25-Story Rubik’s Cube, New York Times, New York, NY, https://www.nytimes.com/interactive/2023/03/11/upshot/office-conversions.html (2023).

  48. Architectural Database. Birkhauser, https://bdt.degruyter.com/ (2023).

  49. Wee Kim, W. Census of population, 2020: statistical release 2: households, geographic distribution, transport, and difficulty in basic activities. Singapore: Department of Statistics, Ministry of Trade and Industry, Republic of Singapore, (2021).

  50. Gaumer, E. The 2021 New York City Housing and Vacancy Survey: Selected Initial Findings, NY: Department of Housing Preservation & Development, New York, https://www.nyc.gov/assets/hpd/downloads/pdfs/services/2021-nychvs-selected-initial-findings.pdf (2022).

  51. Statistik Stadt Zürich, Präsidialdepartement, Haushalte in Neubau-Wohnungen nach Bauperiode, Zimmerzahl, Eigentumsart, Haushalttyp und Stadtquartier, seit 2006. https://data.stadt-zuerich.ch/dataset/bau_neubau_whg_hh_whnfl_bp_zizahl_eart_hhtyp_quart_od6991 (2023).

  52. Department of Statistics Singapore, Singapore Department of Statistics | Population Trends 2021, https://www.singstat.gov.sg/-/media/files/publications/population/population2021.pdf (2021).

  53. Guo, M. et al. Data-Efficient Graph Grammar Learning for Molecular Generation. International Conference on Learning Representations (ICLR). https://doi.org/10.48550/arXiv.2203.08031 (2022).

  54. Krenn, M., Häse, F., Nigam, A., Friederich, P. & Aspuru-Guzik, A. Self-Referencing Embedded Strings (SELFIES): A 100% robust molecular string representation. Mach. Learn. Sci. Technol. 1, 045024 (2020).

    Article Google Scholar

  55. Zhao, A. et al. RoboGrammar: graph grammar for terrain-optimized robot design. ACM Trans. Graph. 39, 1–16 (2020).

    Article Google Scholar

  56. Whalen, E. & Mueller, C. Toward reusable surrogate models: Graph-based transfer learning on trusses. J. Mech. Des. 144, 021704 (2022).

    Article Google Scholar

  57. Merrell, P. Example-based procedural modeling using graph Grammars. ACM Trans. Graph.42, 1–16 (2023).

    Article Google Scholar

  58. Fiser, M. et al. Learning geometric graph grammars. In Proceedings of the 32nd Spring Conference on Computer Graphics, Slomenice Slovakia: ACM, 7–15 (2016).

  59. Robert McNeel & Associates, Rhino3D®and Grasshopper3D®. https://www.rhino3d.com(2022).

  60. C. Ruegg, J. Van Gael, J. Kalias, and J. Larsson, Math.Net open source library. https://www.mathdotnet.com/ (2023).

  61. A. Johnson, Clipper2. https://github.com/AngusJohnson/Clipper2 (2023).

  62. A. Hagberg, D. Schult, and P. Swart, NetworkX. https://github.com/networkx/networkx(2023).

  63. Kamada, T. & Kawai, S. An algorithm for drawing general undirected graphs. Inf. Process. Lett. 31, 7–15 (1989).

    Article MathSciNet Google Scholar

  64. Crawley, D. B., Pedersen, C. O., Lawrie, L. K., Winkelmann, F. C. & Juve, G. Energy plus: Energy simulation program’. ASHRAE J 42, 49–56 (2000).

    Google Scholar

  65. Solemma, Climate Studio. Solemma LLC., https://www.solemma.com/ (2023).

  66. Deitke, M. et al. ProcTHOR: Large-Scale Embodied AI Using Procedural Generation. Advances in Neural Information Processing Systems, 35, 5982–5994. https://proceedings.neurips.cc/paper_files/paper/2022/file/27c546ab1e4f1d7d638e6a8dfbad9a07-Paper-Conference.pdf (2022).

  67. Fisher, M., Ritchie, D., Savva, M., Funkhouser, T. & Hanrahan, P. Example-based synthesis of 3D object arrangements. ACM Trans. Graph. 31, 1–11 (2012).

    Article Google Scholar

  68. Yu, L.-F. et al. Make it home: automatic optimization of furniture arrangement. ACM Trans. Graph. 30, 1–12 (2011).

  69. Jacoby, S., Arancibia, A. & Alonso, L. Space standards and housing design: typological experimentation in England and Chile. J. Archit. 27, 94–126 (2022).

    Article Google Scholar

  70. Design for London and Mayor of London, London Housing Design Guide, London Development Agency, London, http://www.designforlondon.gov.uk/uploads/media/Interim_London_Housing_Design_Guide.pdf (2010).

Pledge Your Vote Now
Change language