Synthetic and Tangible Agents for an Activity-based Urban Planning Tool

AI is on the rise. Powerful cloud platforms and networked software components can perform increasingly complex data evaluations and simulations. Recent research and development projects1 show how great the potential of artificial intelligence is for urban planning. However, despite the impressive, technical possibilities, it currently remains unclear how planning stakeholders and the affected population can be meaningfully involved in the intelligent processes of the "black box". The authors are of the opinion that sustainable urban development planning not only requires acceptance of the spatial planning result, as has been the case up to now, but also requires acceptance of the increasingly digitally supported planning process. For this reason, it must also be possible for laypersons to understand the digital analysis and evaluation processes and to comprehend their relevance and spatial interactions. Consequently, simulations must not only run in the computers of the respective planning or engineering offices, but require a simple, haptic analog translation that can also be used in participation processes as already shown in the CityScope projects2. For this project, the big revitalization project of Deutzer Hafen in Cologne to a future district with more than 9.500 daily users is used as a case study in building a decision support system for urban planning. It is composed of three parts: an agent-based model, a tangible user interface and a synthetic population. The project enables users to get in touch with an agent-based model (ABM) without any knowledge in coding or even interacting with computers. It connects physical objects to digital information. Based on the theories of Castiglione et.al.3 , Gehl4 , Shannon5 and Jacobs6 this project shows how to use an artificial and analog simulation model to measure the urban vitality of the public spaces in the district, based on the activity and travelling patterns of the population. This is done by testing different scenarios in which we change interactive parameters of the model: the use of the buildings and the demographics of the population. We can then determine which scenarios benefit the most life in the public spaces of the district, by finding areas of interest or problematic ones.