CAREER:HCC: Using Virtual Reality Gaming to Develop a Predictive Simulation of Human-Building Interactions: Behavioral and Emotional Modeling for Public Space Design

许多人都经历过在一个大的、复杂的、不熟悉的建筑物中试图找到去医生办公室、登机口或会议室的路的挫折。这些航行困难造成的经济和人员损失往往被低估。研究人员发现，糟糕的建筑设计造成了巨大的经济损失，包括数百万次错过预约和错过航班，延迟交付，新员工和临时员工的表现下降，以及为游客提供方向的时间。压力建筑的直接人力成本也可能很大，特别是对于那些在设计中经常被忽视的人，如老年人和年轻人，以及有认知障碍的人。空间焦虑和令人沮丧的低效率是非常真实的现象，对那些必须在设计不友好的建筑物中工作和导航的人的福祉有影响。为了使建筑环境更容易导航，现在可以使用虚拟仿真工具来分析建筑物的布局，检查“问题点”和令人困惑的设计特征，并评估改变特定环境特征的人为影响。从长远来看，在建筑物建造之前对设计文件进行这种分析检查可以节省大量的财务费用和人类痛苦。该项目的目标是使用来自实际人类导航经验的数据来帮助改进此类计算设计评估工具。更好的设计审查将有助于更舒适和愉快的公共空间，并通过提高医院和机场等设施的运营效率产生经济效益。本研究提出了一种基于证据的设计（EBD）的变革方法，这一领域传统上依赖于后勤复杂和昂贵的方法，如邮政，与建筑物使用者一起进行占用研究和参与式设计会议，为设计决策提供信息。该研究将利用虚拟现实（VR）的独特功能，在各种设施中创建引人入胜的寻路评估任务，并收集有关参与者如何在这些虚拟建筑中导航并对关键环境特征做出反应的数据。然后，研究人员将通过机器学习在广泛的结果数据集中确定可预测的趋势。这种方法将使我们能够用基于证据的认知代理模型（EBCAM）取代现有人群模拟模型中使用的合理化寻路算法（与实际人类行为相比非常不准确），以产生更逼真的模拟人类对建筑特征的反应。作为最后一步，研究人员将通过比较，并在必要时微调，其预测与从两个真实世界建筑物中的人类参与者收集的数据来验证EBCAM模拟。该项目还通过考虑空间体验中更广泛的因素，如不确定性，情绪反应和空间记忆，将人类建筑交互分析扩展到寻路性能结果之外。在寻路过程中将进行的关于心理因素的连续数据收集的程度，在以前的研究中是前所未有的，将允许在空间导航领域测试新的理论框架，以及增强基于证据的模拟工具。以模拟工具的形式应用这些实证研究结果大大增加了设计工作流程中的可访问性，可能使更广泛的设施能够获得循证审查的好处。公众可以从这种基于证据的设计分析的扩展中受益，到目前为止，这种扩展主要限于富裕的城市环境。该研究将通过纳入老年人等被忽视的群体，开发反映其特定需求的模型，进一步增强设计的包容性。该工具将使设计师能够更加自信和创造性地工作，因为他们知道基于证据的评估有助于确认或反驳创新想法的价值。该奖项反映了NSF的法定使命，并被认为值得通过使用基金会的知识价值和更广泛的影响审查标准进行评估来支持。

Many people have experienced the frustration of trying to find their way to a doctor’s office, departure gate, or meeting room in a large, complex, and unfamiliar building. The economic and human toll of these navigational difficulties is often underestimated. Researchers have found tremendous economic losses resulting from poor architectural design, measured in millions of missed appointments and missed flights, delayed deliveries, reduced performance of new and temporary staff members, and time spent providing directions to visitors. The immediate human costs of stressful buildings can also be substantial, particularly for those whose needs are frequently overlooked in design, such as the elderly and the young, and people with cognitive impairments, among others. Spatial anxiety and frustrating inefficiency are very real phenomena with implications for the wellbeing of those who must work in and navigate through buildings whose designs are not user-friendly. To make the built environment easier to navigate, it is now possible to use virtual simulation tools that analyze a building’s layout, check for “problem spots” and confusing design features, and evaluate the human effects of altering specific environmental features. Performing such analytical checks on design documents before the building is constructed can save a great deal of financial expense and human grief in the long run. The goal of this project is to use data from actual human navigational experiences to help improve such computational design-evaluation tools. Better design review will contribute to more comfortable and enjoyable public spaces, and produce economic benefits through improved operational efficiency in facilities such as hospitals and airports.This research proposes a transformative approach to evidence-based design (EBD), a field that has traditionally relied on logistically complex and expensive methods such as post-occupancy studies and participatory design sessions with building users to inform design decisions. The research will leverage the unique capabilities of virtual reality (VR) to create engaging wayfinding evaluation tasks in various facilities, and collect data about how participants navigate through these virtual buildings and react to key environmental features. The researchers will then identify predictable trends in the extensive resulting dataset via machine learning. This approach will allow us to replace the rationalized pathfinding algorithms used in existing crowd-simulation models (which are wildly inaccurate in relation to actual human behavior) with an Evidence-Based Cognitive Agents Model (EBCAM) to produce more realistic simulated human responses to architectural features. As a final step, the researchers will validate the EBCAM simulation by comparing, and, if necessary, fine-tuning, its predictions against data collected from human participants in two real-world buildings. This project also extends human–building interaction analysis beyond wayfinding performance outcomes, by considering broader factors in spatial experience such as uncertainty, emotional response, and spatial memory. The extent of continuous data collection that will be performed regarding psychological factors during wayfinding, unprecedented in previous studies, will allow for testing new theoretical frameworks in the spatial navigation field, as well as enhancing the evidence-based simulation tool. The application of these empirical findings in the form of a simulation tool vastly increases their accessibility in design workflows, potentially making it possible for a much broader range of facilities to receive the benefits of evidence-based review. The general public can benefit from such expansion in evidence-based design analysis, which until now has mostly been limited to wealthy, urban contexts. The research will further enhance the inclusiveness of design by incorporating overlooked groups such as older adults, developing a model that reflects their specific needs. The tool will empower designers to work more confidently and creatively, knowing that evidence-based evaluations can help to confirm or refute the value of an innovative idea.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.