In contemporary architectural practices, case studies are used for learning from precedents and improving design efficiency and quality. However, the complexities of public building layouts pose considerable challenges, such as difficulty in interpreting designs and quantifying associated constraints. Existing automated retrieval methods often focus on simpler residential layouts and overlook the complexity of the constraints associated with public buildings. This paper proposes a graph-based layout retrieval approach that leverages similarities between the function and topology of building layouts and user requirements. The proposed approach is applied to a self-compiled library dataset and validated through professional architect evaluations. Results show that the proposed method aligns with evaluations of experienced architects, while providing more consistent recommendations than inexperienced architects in terms of design quality and behavioral suitability. This study advances machine understanding of complex layouts, enhancing automation and rationality in early-phase design and improving the integration of design information with construction materials.
@article{guo2026retrieval,title={Function-aware graph-based layout retrieval for academic libraries},author={Guo, Song and Kee, Tris and Zhuang, Weimin},year={2026},journal={Automation in Construction},doi={https://doi.org/10.1016/j.autcon.2026.106881},url={https://www.sciencedirect.com/science/article/abs/pii/S0926580526001226},dataset={https://data.mendeley.com/datasets/xk9n6jnvp7/1},}
Visual AI
Influence of objective and perceived exposures to urban nature on people’s happiness
Exposure to nature influences urban dwellers’ well-being and happiness, thereby impacting urban sustainability. However, urban dwellers are exposed to nature in different ways: indirect exposure through window views, incidental exposure when walking along streets, and intentional exposure when visiting parks. Moreover, objective exposure does not necessarily align with how people perceive their exposure to nature. This study examines how three types of objective nature exposure—indirect, incidental, and intentional—provided by greenery and water bodies, along with perceived exposure, impact happiness in Tokyo, Japan. To measure the objective exposure, we use 3D photorealistic city information models, street view imagery, road network datasets, and remote sensing imagery. To measure happiness and perceived nature exposure, we use data from a national survey, focusing on the results from 10,798 residents in 801 neighborhoods in Tokyo. Results showed that perceived nature exposure has higher explanatory power for happiness than objective exposure. Views of greenery from windows (indirect exposure) and proximity to parks (intentional exposure) influenced perceived nature exposure and happiness the most. The quantitative evidence suggests that urban planning align with human behavior, e.g., by prioritizing the improvement of greenery views from windows and park accessibility in Tokyo to promote urban sustainability.
@article{li2025urbanhappiness,title={Influence of objective and perceived exposures to urban nature on people's happiness},author={, Maosu Li and Guo(Co-first), Song and Duarte, Fábio and Kumar, Ashutosh and Kobori, Norimasa and Xue, Fan and Zhuang, Weimin and Yeh, Anthony G.O. and Ratti, Carlo},year={2026},journal={npj Urban Sustainability},doi={https://doi.org/10.1038/s42949-025-00306-9},url={https://www.nature.com/articles/s42949-025-00306-9},}
2025
Visual AI
Urban visual uniqueness: A landmark-free framework to quantify city’s identity and distinctiveness from everyday scenes
The visual appearance of a city is shaped by a complex interplay of factors, including cultural backgrounds, geographical features, historical developments, and policy decisions. But measuring cities’ visual uniqueness remains a challenge. Previous studies often focused on iconic landmarks, neglecting everyday scenes that people are likely to encounter. By examining how and to what extent different visual patterns build up unique characteristics of cities, we propose a data-driven framework to measure visual uniqueness in terms of identity and distinctiveness. We performed bottom-up visual clustering on Google Street View (GSV) images in the six most visited Japanese cities. We found that 8 representative visual clusters explain each city’s visual identity and relative distinctiveness. This research demonstrates how artificial intelligence applied to visual data can reveal subtle differences in urban environments. In the era of growing globalization, with frequent tourism and intercity visits, the cultivation of a city’s unique visual characteristics can help avoid the homogenization of urban landscapes, and stimulate the development of urban tourism by shaping an imageable city.
@article{GUO2025102351,title={Urban visual uniqueness: A landmark-free framework to quantify city's identity and distinctiveness from everyday scenes},journal={Computers, Environment and Urban Systems},volume={122},pages={102351},year={2025},issn={0198-9715},doi={https://doi.org/10.1016/j.compenvurbsys.2025.102351},url={https://www.sciencedirect.com/science/article/pii/S0198971525001048},author={Guo, Song and Jang, Kee Moon and Duarte, Fábio and Kang, Yuhao and Ratti, Carlo},keywords={Urban uniqueness, Place identity, Place distinctiveness, Street view images, Landmark-free, Visual artificial intelligence},}
Visual AI;Sensors
Greenery Scanner: urban greenery health sensing using low-cost multispectral camera in drive-by condition
@article{iacussi2025greeneryscanner,title={Greenery Scanner: urban greenery health sensing using low-cost multispectral camera in drive-by condition},author={Iacussi, Leonardo and Guo(Co-first), Song and Mora, Simone and Speridi\~{a}o, Andrei and Oyinlola, Oluwatobi and Papitto, Giancarlo and Fugiglando, Umberto and Chiariotti, Paolo and Cigada, Alfredo and Ratti, Carlo},year={2025},journal={IEEE Sensors Journal},note={To be submitted},}
2024
Architectural Studies
Enhancing spatial configurations in informal learning spaces: An early-phase design decision-making framework
Traditional design decision-making is usually a process of adapting and reusing the paradigm at the architect’s disposal based on intuition and experience, which helps to provide acceptable solutions to known design problems quickly but makes it challenging to generate new spatial types beyond the existing paradigm. The emergence of informal learning behaviors has caused changes in public learning spaces, and traditional design decision-making processes have struggled to provide better solutions. Based on 620 questionnaires, this study combines statistical and space syntax analysis to obtain the relationship between user preferences and space. It uses multi-agent simulation to interpret preferences into spatial forms to assist design in obtaining a more suitable spatial layout pattern for informal learning scenarios. Proposing a design decision-making method to solve complex behavior-driven problems effectively, this research will help to integrate research and design effectively, improve the efficiency and accuracy of design decision-making.
@inproceedings{10732175,author={Guo, Song and Zhuang, Weimin and Yan, Xiaoyue},booktitle={2024 Annual Modeling and Simulation Conference (ANNSIM)},title={Enhancing spatial configurations in informal learning spaces: An early-phase design decision-making framework},year={2024},volume={},number={},pages={1-13},doi={10.23919/ANNSIM61499.2024.10732175},}
Analysis, configuration, and optimization: The prospect of Graph theory applied in architectural programming
Air infiltration rate is one of the fundamental parameters in the design and analysis of building environment, which is the fundamental mode of air exchange between indoor and outdoor in non-mechanically ventilated residences when doors and windows are closed. Most Chinese residential buildings have no mechanical ventilation equipment installed. Previous researches on Chinese housing have primarily focused on urban residences. However, the number of both urban households and rural households accounts for almost half of Chinese households. Considering the significant urban-rural differences in Chinese housing, there is an urgent need for researching the air infiltration rate in rural area. This study employed a multi-zone network airflow model approach to simulate the air infiltration rate of 111 rural residential samples in 13 Chinese provinces, obtaining the hour-by-hour results for a whole typical weather year of each sample. Thus, the distributions of air infiltration rate for each rural region were finally summarized. Additionally, we performed field measurements in 23 samples to validate the simulation results. According to the results, the mean value of air infiltration rate is larger than that of urban residences, and its distribution in each climatic zone generally follows a log-normal distribution. The results are expected to be applied for indoor air quality studies, building energy consumption analysis, and other frontiers related to architectural environmental design and analysis.
@article{WANG2024111284,title={Air infiltration rate distribution across Chinese five climate zones: A modelling study for rural residences},journal={Building and Environment},volume={252},pages={111284},year={2024},issn={0360-1323},doi={https://doi.org/10.1016/j.buildenv.2024.111284},url={https://www.sciencedirect.com/science/article/pii/S0360132324001264},author={Wang, Ye and Shi, Shanshan and Zhou, Zhengxu and Guo, Song and Zhao, Bin},keywords={Air infiltration rate, Rural, Indoor air, Simulation, CONTAM},}
2023
Digital Heritage
Historical Architecture Map of Nei Mongol(内蒙古古建筑地图)
Lin
Yuan, Sarina
Zhao, Qianyi
Li, Xueyue
Bai, Yichi
Zhang, and Song
Guo
— Book Chapter, Tsinghua University Press,
Apr 2023
2021
Digital Heritage
A study on the spatial scenes of Jingqingzhai at Beihai Park during the Qianlong Period
Based on the comparative analysis of texts, images and material objects of jingqingzhai (Jingxinzhai) at Beihai Park, this paper deduces jingqingzhai’s space duringthe Qianlong period. On this basis, it attempts to clarify the concept of "scene" in landscape research, and tentatively represents and analyzes some spatial scenes of jingqingzhai as depicted in Qianlong’s poems, in hopes of upgrading the understandings of the "scene" as an experience and design unit.
@article{郭凇2021乾隆时期北海镜清斋的空间场景研究,title={A study on the spatial scenes of Jingqingzhai at Beihai Park during the Qianlong Period},author={Guo, Song and Li, Luke and Qian, Bo},journal={Architectural Journal(建筑学报)},number={11},pages={10},year={2021},}
Greenery Scanner: urban greenery health sensing using low-cost multispectral camera in drive-by conditionIEEE Sensors Journal , 2025To be submitted
Historical Architecture Map of Nei Mongol(内蒙古古建筑地图)—