A Real-Time Twin in Venice explores how interactive objects in public space can anchor activities and impact spatial dynamics. It features a simple robot that moves autonomously throughout public spaces in Venice, providing information about the Biennale. In the Arsenale, a virtual interface displays a digital reconstruction and analysis of the space of the robot, allowing visitors to control the robot and test AI-predicted crowd behaviour. The real robot adjusts its position hourly based on visitor preferences, improving AI accuracy over time. This installation highlights extracting and visualising urban data in real-time, using AI to solve design problems like optimal object placement, and constructing an AI agent by incrementally learning behavioural patterns.
A Real-Time Twin in Venice
Zhang Li, Urban Ergonomics Lab, School of Architecture, Tsinghua University, Xie Qixu, Deng Huishu, Wang Ziheng, Pang Lingbo, Chen Yuhong, Mei Xiaohan, Ye Yang
