|
|
ENQUIRE PROJECT DETAILS BY GENERAL PUBLIC |
| Project Details |
| Funding Scheme : | General Research Fund | ||||||||||||||||||||||||||||||||||||||||
| Project Number : | 17203920 | ||||||||||||||||||||||||||||||||||||||||
| Project Title(English) : | Enhancing Underground Development Users' Health through an Automated Risk Assessment System for Facilities Management | ||||||||||||||||||||||||||||||||||||||||
| Project Title(Chinese) : | 地下設施自動風險評估系統: 提升地下建築用戶的健康水平 | ||||||||||||||||||||||||||||||||||||||||
| Principal Investigator(English) : | Prof Chan, Yee Shan Isabelle | ||||||||||||||||||||||||||||||||||||||||
| Principal Investigator(Chinese) : | |||||||||||||||||||||||||||||||||||||||||
| Department : | Department of Real Estate and Construction | ||||||||||||||||||||||||||||||||||||||||
| Institution : | The University of Hong Kong | ||||||||||||||||||||||||||||||||||||||||
| E-mail Address : | iyschan@hku.hk | ||||||||||||||||||||||||||||||||||||||||
| Tel : | 28597984 | ||||||||||||||||||||||||||||||||||||||||
| Co - Investigator(s) : |
|
||||||||||||||||||||||||||||||||||||||||
| Panel : | Engineering | ||||||||||||||||||||||||||||||||||||||||
| Subject Area : | Civil Engineering, Surveying, Building & Construction | ||||||||||||||||||||||||||||||||||||||||
| Exercise Year : | 2020 / 21 | ||||||||||||||||||||||||||||||||||||||||
| Fund Approved : | 684,520 | ||||||||||||||||||||||||||||||||||||||||
| Project Status : | Completed | ||||||||||||||||||||||||||||||||||||||||
| Completion Date : | 31-12-2024 | ||||||||||||||||||||||||||||||||||||||||
| Project Objectives : |
|
||||||||||||||||||||||||||||||||||||||||
| Abstract as per original application (English/Chinese): |
香港地處陡峭，城市發展長期受限。受世界各地的成功地下發展項目啟發，香港政府於年前開始了“地下空間發展”計劃，預計將會有越來越多市民需要於地下空間工作或停留更長時間。數十年來，社會對於地下空間發展的接受程度一直很低。而且，地下發展的成本可以是地面發展的1. 3至6倍。如果在沒有充分考慮地下環境可引致的健康風險的情況下發展地下項目，經濟負擔將會更高。例如，在英國，與建築物有關的健康問題每年使英國國家醫療服務體系損失14億英鎊。 事實上, 先前文獻已指出了地下用戶面臨著各種健康風險, 例如，由於缺乏自然光照射導致的睡眠障礙（生理健康風險）、對地下環境缺乏感知控製而產生的焦慮（心理健康風險）、與地面聯繫不足而造成的孤立感（社交健康風險），以及由於聽覺舒適性不佳而導致注意力分散（認知健康風險）。設施管理具有在建築環境中協調物理環境及其對人之影響的功能，所以長期以來被認為是提升建築用戶健康的有效途徑。然而，由於缺乏對多維地下設施管理（包括空間管理、屋宇裝備和配套設施）和地下空間用戶健康風險的認知、低效率的傳統信息提取和交換過程及過於分散的項目團隊等因素，使以減低地下空間用戶的健康風險為前提的設施管理變得困難重重。 基於申請人正在進行的兩項試點項目中獲得的顯性和隱形知識，本提案旨在通過開發以BIM（建築信息模型）為基礎的地下設施自動風險評估系統，來支持以減低地下空間用戶的健康風險為前提的設施管理決策, 從而提升地下空間用戶的健康水平。目標包括：（1） 開發地下設施管理與健康風險數據庫， （2） 開發地下設施自動風險評估系統以提升用戶的健康水平，及（3）使用實際案例驗證地下設施自動風險評估系統。 |
||||||||||||||||||||||||||||||||||||||||
| Realisation of objectives: | Obj 1 - To develop underground FM-H databases Based on the empirical studies, the project has resulted in two major underground FM-H databases, with one focusing on underground greenery and occupant health and the other focusing on underground connectivity and occupant health. The FM-H databases were developed based on on-site empirical investigations. Twelve underground sites and twelve corresponding aboveground sites in Hong Kong were selected for a large-scale questionnaire survey, resulting in over 650 survey samples containing rich health data measured with multi-item scales (including both physical and mental health). In addition, on-site data collections were conducted on these sites to objectively capture underground facilities (e.g., greenery, physical, visual and implicit connectivity facilities) data. Statistical methods were conducted to examine the relationships between the identified significant underground FM elements and occupant health. Examples include: (1) conducting a polynomial regression analysis to explore the relationships between underground greenery and health, providing empirical evidences for the inverted U-shape relationships between underground greenery coverage ratio and health, (2) conducting a multivariate linear regression analysis to explore the relationships between visual connectivity (proxied by the ratio of outdoor spaces in a panoramic view), implicit connectivity (proxied by the ratio of glass facade in a panoramic view), physical connectivity (proxied by the ratio of vertical traffic in a panoramic view) and underground occupants’ health. Implications & foundation for next step: The databases, followed by statistical analyses, provided empirical evidence that the relationship between underground greenery and occupants’ health adheres to an inverted U shape. When underground greenery coverage ratio achieves an optimized point, the optimal health outcome is achieved, while the health performance declines when the green element is greater or below the optimization point. Meanwhile, visual, implicit, and physical connectivity are associated with underground occupants’ health through a complex mechanism. The health variable is operationalized as the difference between the health score of underground occupants and the average health score of their aboveground counterparts revealed from the database. The rationale behind is to enhance the health score in underground environment so that it’s equivalent to or even better than that of their aboveground counterparts. Obj 2 - To develop an automated FM-H assessment system for underground developments Based on the results of Objective 1, two automated FM-H assessment platforms were developed. The first one was developed to optimize underground greenery for occupant health, and the second one was developed to enhance underground connectivity facilities for human health. The two automated platforms were developed in the Building Information Modelling (BIM) environment (the Revit software), with the former one developed based on the 2D CNN object recognition algorithm for greenery object recognition, while the latter based on the Few-Shot CLIPSeg image segmentation algorithm for connectivity element extraction. The plugins can provide a real-time health score based on the automatic identification of the relevant facilities in an underground space, which can then inform health-oriented facility design to yield better health outcomes. Obj 3 - To validate the automated FM-H assessment systems using real case A real-world underground case was used to validate the developed automated platforms. The systems were applied to different zones of an underground project and provided a health score accordingly, allowing designers or FM managers to adjust facility design for better health outcomes. The empirical study unveiled that both insufficient and excessive greenery negatively impact health performance. The optimization point of greenery coverage ratio in underground space is found as 4.85%. The results suggested that merely introducing high quantities of greenery does not guarantee improved well‑being and may even lead to aesthetic fatigue. As such, the automated platforms developed in this research aim to assist underground space designers and FM managers in making appropriate, context‑sensitive decisions to promote health-centric underground greenery design. Based on the automated health-oriented underground greenery assessment system, the greenery-related health scores of different zones in the underground BIM were calculated; providing insights for optimizing the green ratio across various areas. Specifically, a comparative analysis was conducted between the scenarios with and without the automated FM-H assessment platform. The results indicated that, in the absence of the automated system, the Revit API was unable to recognize all greenery‑related elements, leading to omissions in the assessment process. In the contrary, when the automated system was adopted, all greenery elements were automatically identified and semantically enriched as greenery objects, enabling an accurate evaluation of greenery–health implications within the underground BIM environment. The automated platform for underground connectivity and health was employed to automatically assess the connectivity-related health implications across different zones until the connectivity level—encompassing physical, visual, and implicit connectivity facility elements—reached an acceptable threshold. In the validation case study, three groups of test images were processed using CLIPSeg to segment physical, visual, and implicit connectivity elements. The results confirmed that the model accurately identified relevant objects and generated appropriate segmentation masks. Additionally, the proportional coverage of each connectivity category relative to the entire image was computed based on the model’s output. An eight-floor shopping mall was selected as the case study due to its integration of above-ground and underground spaces. To evaluate system performance, five rooms on the underground floor—including the atrium, public eating area, room FC-302, room FC-303, and room FC-304—were analyzed. The automated system first generated sampling points at predefined intervals. At each point, five directional images—simulating the human field of view—were captured and processed by CLIPSeg to segment the three connectivity types and compute their area ratios. The average physical, visual, and implicit connectivity ratios at each sampled point were calculated, representing the occupant’s perceptual experience of connectivity. These values were then used to derive health indicators for each space (e.g., atrium: −0.482; room FC-302: −0.477; public eating area: −0.518; room FC-304: −0.531; room FC-303: −0.530) through connectivity-health equations. By adjusting the three connectivity indicators, corresponding health scores were subsequently computed, providing valuable insights for optimizing health-oriented underground connectivity facility design. | ||||||||||||||||||||||||||||||||||||||||
| Summary of objectives addressed: |
|
||||||||||||||||||||||||||||||||||||||||
| Research Outcome | |||||||||||||||||||||||||||||||||||||||||
| Major findings and research outcome: | The comprehensive literature reviews and empirical data collected from large-scale questionnaire surveys and field measurement studies provided robust evidence on the critical role of greenery and aboveground-underground connectivity facilities in underground environments for human health (Chan & Chen, 2023; 2024; Chen & Chan, 2024; Chan et al., 2024; 2025; Chan, Twum-Ampofo et al., 2025; Chan, Chen, & Ma, 2025). Building upon these solid databases, the research team developed an automated Facility Management–Health (FM-H) assessment system tailored for underground development projects. The empirically derived FM-H relationships are summarized as follows. First, the relationship between underground greenery and occupant health follows an inverted U-shaped curve. Based on empirical findings, optimal health benefits are achieved when greenery occupies approximately 4.85% of the visual field of underground occupants. Deviations beyond this optimal point—either more or less greenery—lead to diminished health performance (Chan, Chen, Zhang, & Zhang, 2025a). Second, the relationship between visual, implicit, and physical connectivity and occupant health is complex. Based on the empirical data, the underground connectivity facility related health value was modeled with the following equation: Underground mental health = 0.337 × visual_connectivity – 0.269 × implicit_connectivity + 0.490 × visual#implicit – 0.245 × visual#physical + 0.022 × implicit#physical – 0.531. A health value equal to or greater than zero indicates that the health levels of underground occupants are comparable to or exceed those of their aboveground counterparts. This serves as a key criterion for optimization in subsequent automated platform development (Chan, Chen, Zhang, & Zhang, 2025b). [OBJ 1] The development of the automated evaluation platform was based on these empirically supported relationships. The greenery-health assessment plugin automatically detects greenery objects within underground BIM models, using the inverted U-shaped relationship to compute health scores. Similarly, the connectivity-health assessment plugin extracts connectivity indicators—including ratios of physical, visual and implicit connectivity—via automated multi-view image capture and calculates health scores based on the linear connectivity-health relationship (Chan, Chen, Zhang, & Zhang, 2025a, 2025b). [OBJ 2] Finally, validation with real underground case demonstrated the effectiveness of these systems. The greenery-health platform successfully identified green coverage ratios across various areas within the BIM model, indicating that adjusting greenery coverage toward the optimal 4.85% can enhance occupant health. Additionally, the CLIPSeg segmentation algorithm achieved 78.63% accuracy in identifying greenery and connectivity elements, enabling reliable health estimation through precise extraction of connectivity indicators. (Chan, Chen, Zhang, & Zhang, 2025a, 2025b). [OBJ 3] | ||||||||||||||||||||||||||||||||||||||||
| Potential for further development of the research and the proposed course of action: |
The FM-H relationship established was derived from a limited set of underground sites within Hong Kong, a high-density city with a unique, international cultural background. To enhance the robustness and broader applicability of these relationships, future research is recommended to expand investigations to include additional underground sites across different regions. Additionally, extending the database to incorporate underground environments in other cities and regions outside Hong Kong would facilitate the development of cross-cultural automated systems. These systems could incorporate region-specific formulas tailored to local conditions and human characteristics, supporting human-centric underground space design on a global scale. Establishing collaborations with international research teams for transnational site investigations and underground FM-H databases developments will be essential. Such partnerships are expected to improve the generalizability of the FM-H relationships and expand the applicability of the automated assessment platform across diverse geographic and cultural contexts. | ||||||||||||||||||||||||||||||||||||||||
| Layman's Summary of Completion Report: | The development of underground spaces involves significantly higher costs compared to above-ground construction. This financial burden can be further increased if user health considerations are not adequately integrated into the built environment. Typically, health evaluations are conducted only after occupancy, and health-related knowledge is generally not emphasized in traditional architecture and facility management (FM) professional training. This situation limits opportunities for implementing health-centric design and FM strategies for underground spaces. To address this gap, this study examined the relationships between facility management and health (FM–H) through on-site surveys and field measurements. The databases developed from these investigations were used to develop automated systems for underground FM–H assessment. Using real-world validation case from underground building information model (BIM), the study demonstrated how these tools can be applied to optimize underground greenery and improve connectivity facilities. The automated platforms are designed to support evidence-based decision-making during early design phases, helping to prevent suboptimal underground space designs that overlook health considerations. They can also be employed during post-occupancy phases to assist FM decision-making in underground projects. | ||||||||||||||||||||||||||||||||||||||||
| Research Output | |||||||||||||||||||||||||||||||||||||||||
| Peer-reviewed journal publication(s) arising directly from this research project : (* denotes the corresponding author) |
|
||||||||||||||||||||||||||||||||||||||||
| Recognized international conference(s) in which paper(s) related to this research project was/were delivered : |
|
||||||||||||||||||||||||||||||||||||||||
| Other impact (e.g. award of patents or prizes, collaboration with other research institutions, technology transfer, etc.): |
|||||||||||||||||||||||||||||||||||||||||
| SCREEN ID: SCRRM00542 |