Optimization design of age-friendly homestay spaces integrating Space syntax and SegNet

Abstract

Objective

This study addresses critical gaps in age-friendly homestay design, an area that currently relies heavily on experience-based layouts and evaluation methods that remain largely decoupled from the spatial generation process. The objective is to develop quantifiable, interpretable, and automated spatial optimization methods for age-friendly homestay designs. “Age-friendly homestay spaces” refers to interior accommodation spaces designed or renovated for older adult guests in short-term lodging scenarios, rather than urban-scale planning or institutional elderly-care facilities. The proposed method does not directly measure long-term elderly user experience. Instead, it optimizes spatial proxy indicators related to accessibility, fall-risk exposure, visual recognizability, and wayfinding complexity, which are used to support age-friendly layout decision-making.

Methods

An optimization approach for age-friendly homestay spaces is proposed, integrating SegNet and space syntax. The lightweight SegNet model is able to detect elements in architectural floor plans. Space syntax modeling, along with Visibility Graph Analysis, provides a multidimensional space for age-friendliness. The method also integrates Non-dominated Sorting Genetic Algorithm II and Simulated Annealing for algorithmic optimization.

Results

Experimental results demonstrate that the proposed model maintains high computational efficiency. The mean Intersection over Union reaches 0.89. Precision and recall both exceed 0.90. In spatial optimization, the hybrid solving algorithm outperforms the single genetic algorithm. The hypervolume indicator increases to 0.72. The inverted generational distance decreases to 0.22.

Conclusion

The proposed framework effectively facilitates the generation of age-friendly spatial configurations that balance accessibility efficiency, caregiving safety, and cognitive clarity. This study provides scientific decision support for precise renovation of homestay spaces. It also offers practical value for improving the quality of health-oriented tourism accommodation.

Keywords

Space syntax / Age-friendly design / SegNet / Semantic segmentation / Homestay space

References

[1] Aleksandrov M, Amoozandeh K, Winter S, Zlatanova S, Tomko M. A visibility-based multitree of a space subdivision for indoor localization. Transactions in GIS, 2024, 28(7): 2397-2411

[2] Bae SS, Cave K, Görg C, Rosen P, Szafir DA, Bearfield CX. Bridging network science and vision science: Mapping perceptual mechanisms to network visualization tasks. IEEE Transactions on Visualization and Computer Graphics, 2025, 31(10): 7137-7152

[3] Cai W, Daneshvar Z, Miller K, Dash M, Jiang S. Exploring campus configuration and patient arrival behaviors at large hospitals: An integrated post-occupancy evaluation (POE) protocol. HERD-HEALTH ENV RES, 2024, 17(4): 45-64

[4] Carta S, Shekhawat K, Ambrosini L, Turchi T, Vidal M. Non-trivial self-organised floor plans: An optimisation strategy. Nexus Network Journal, 2025, 27(1): 161-186

[5] De Nardin A, Zottin S, Toma A, Piciarelli C, Foresti GL. A multibranch and multiattention framework for floor plan segmentation. Computer-Aided Civil and Infrastructure Engineering, 2025, 40(23): 3674-3687

[6] El Samaty HS, Feidi JZ, Refaat AM. The impact of glazed barriers on the visual and functional performance of transition spaces in college buildings using space syntax. Ain Shams Eng J, 2023, 14(9): 1-17

[7] Fattahi Tabasi S, Rafizadeh HR, Andaji Garmaroudi A, Banihashemi S. Optimizing urban layouts through computational generative design: Density distribution and shape optimization. Architectural Engineering and Design Management, 2024, 20(5): 1164-1184

[8] Feng X, Zhang H, Wei S. Automation in architectural design: Optimization and decision-making in interference design. Architectural Science Review, 2024, 67(5): 415-436

[9] Martens J, Blankenbach J. Vox2bim+-a fast and robust approach for automated indoor point cloud segmentation and building model generation. PFG-Journal of Photogrammetry, Remote Sensing and Geoinformation Science, 2023, 91(4): 273-294

[10] Mazurkiewicz W, Sitek M. Spatial and behavioural cognition of wayfinding system in hospital buildings in terms of oncology patients’ comfort. Architecture Civil Engineering Environment, 2024, 17(4): 67-82

[11] O. Güler and H. Demirkan (2024) Evaluating the relationship between visual privacy and work-process interactions in open-plan offices: a space syntax approach,” Archnet-IJAR: International Journal of Architectural Research, 17(2), 243–266. https://doi.org/10.1108/ARCH-03-2024-0123

[12] Omrani Azizabad S, Mahdavinejad M, Hadighi M. Three-dimensional embodied visibility graph analysis: Investigating and analyzing values along an outdoor path. URBAN ANALYTICS AND CITY SCIENCE, 2025, 52(7): 1669-1684

[13] Peponis J. Space syntax and design. URBAN ANALYTICS AND CITY SCIENCE, 2024, 51(5): 1073-1078

[14] Shirini K, Aghdasi HS, Saeedvand S. Multi-objective aircraft landing problem: A multi-population solution based on non-dominated sorting genetic algorithm-II. Journal of Supercomputing, 2024, 80(17): 25283-25314

[15] Sun Q, Fang C, Liu S, Sun Y, Shang Y, He Y. PolyGraph: A graph-based method for floorplan reconstruction from 3D scans. IEEE Transactions on Visualization and Computer Graphics, 2025, 31(10): 7350-7362

[16] Tao R, Tao L, Su B, Javanmardi E. Neighbourhood-search-enhanced non-dominated sorting genetic algorithm-III for multi-objective assembly line balancing problem considering operator skill levels and carbon footprint. Engineering Optimization, 2025, 57(10): 2984-3006

[17] Wang X, Yang M, Li W, Yuan L, Ren Q, Xie Q, Liu R. Natural movement: Measuring optimal pathway configuration of museum layout and pedestrian accessibility through visibility graph analysis and intelligent agent simulation. Frontiers of Architectural Research, 2025, 14(5): 1228-1248

[18] Wang Z, Sun N. Offset-guided attention network for room-level aware floor plan segmentation. IEEE Access, 2023, 11: 63667-63677

[19] Wei L, Lai C. Wall segmentation in house plans: Fusion of deep learning and traditional methods. Visual Computer, 2024, 40(9): 6015-6031

[20] Xie H, Ma X, Mei Q, Chui YH. A semi-supervised approach for building wall layout segmentation based on transformers and limited data. Computer-Aided Civil and Infrastructure Engineering, 2025, 40(10): 1295-1313

[21] Yu E, Choi J. Development of building information modeling-based automation assessment process for universal design of public buildings. Journal of Computational Design and Engineering, 2023, 10(2): 641-654

[22] Yuan H, Zhou Y. The impact of crowds on visibility in emergency department: Integrating agent-based simulation and space syntax analysis. Frontiers of Architectural Research, 2025, 14(5): 1398-1414

[23] Z Xu, C Yang, S Alheejawi N. Jha S Mehadi, M Mandal. (2025) Automatic floor plan analysis using boundary attention-based deep network.” IJDAR, 28(1), 19–30. https://doi.org/10.1007/s10032-024-00487-6

RIGHTS & PERMISSIONS

The Author(s)