Plaza Life Revisited: Machine Learning for Public Space Research

Plaza Life Revisited: Machine Learning for Public Space Research – 广场生活回顾：公共空间研究的机器学习

SWA (Anya Domlesky, XL Lab at SWA)

Plaza Life Revisited is a research project by XL Lab, SWA’s innovation lab undertaking practice-based research. The project reconsiders the writer William H. Whyte’s Street Life Project and seminal study The Social Life of Small Urban Spaces (1980). It seeks to understand how types of new public spaces have shifted some 40 years after he published his book and companion film, what has changed in how people use public realm spaces, and what makes well used spaces. The project looked at 10 plazas in Manhattan by 10 different designers, constructed or renovated in the last 15 years. The sites range from the type of bonus plazas Whyte was observing, to infrastructural leftovers, alleys, transit plazas, private campus spaces, and tactical urbanist interventions. The team used new analytical tools such as a machine learning algorithm (a type of artificial intelligence) that employed object detection and tracking on video footage of peak daytime use. These resulted in heat maps describing dwell time, frequent and infrequent usage, and preliminary pedestrian counts. The team also used some of the same techniques Whyte did—behavioral observations, site measurements, and hand tabulation. The goal was to identify common behavioral patterns, collective activity, programming, physical elements, and understand context across the sites in order to inform future public realm design. Findings and methods were published in a booklet called Field Guide to Life in Urban Plazas.

For the “Eyes of the City” of UABB, the research team is experimenting with an extension of the New York study using a different data input. Infrared video footage registers human and animal body heat instead of light, which allows for both evening site usage to be accurately captured and analyzed, as well as automatic individual anonymization. Two local SWA-designed sites are engaged: Shekou Coastal Promenade in Shenzhen, and Xiqu Centre in Hong Kong.

该项目是XL实验室（SWA旗下从事实践研究的创新实验室）的一项研究项目。项目回顾了作家威廉·怀特的“街道生活”和开创性研究《小城市空间的社会生活》（1980年）。本项目试图了解在怀特出版书籍和放映同名电影之后的大约40年里，新型公共空间的类型是如何变化的，人们使用公共领域空间的方式发生了什么变化，以及是什么促成了空间的充分利用。该项目选取了曼哈顿过去15年内由10个不同设计师建造或翻新的10个广场。这些广场分布的范围，从怀特所观察的奖金广场，到基础设施遗留部分、小巷、交通广场、私立校园空间，到战术性城市干预空间。项目团队使用新的分析工具，比如机器学习算法（人工智能的一种），对白天高峰时段的视频片段进行目标检测和跟踪，并由此生成关于停留时间、使用频率以及基本行人数量的热图。项目团队还使用了一些与怀特相同的方法——行为观察、场所测量和手工制表，以识别常见的行为模式、集体活动、编程、物理元素，并理解场所之间所形成的语境，以便为将来的公共领域设计提供信息。研究结果和方法已发表在《都市广场生活图鉴》的小册子上。

在深港城市\建筑双城双年展的“城市之眼”板块，研究团队正在尝试使用不同的数据输入对纽约的研究进行扩展。红外视频片段记录了人和动物的体温，而非光线，因此可准确地捕捉和分析晚间场所的使用情况，并自动对个人信息作匿名化处理。研究纳入了SWA设计的两个本地场所，即深圳蛇口海滨长廊和香港戏曲中心。

Credits:

Anya Domlesky, Emily Schlickman, Tom Balsley, Chella Strong, Jen Saura, Hallie Morrison, Bill Tatham, Julie Eakin, Paul Wehby, Xiaoyin Kuang 邝晓茵.