Article Preview
TopIntroduction
While AI presents the most recent wave in computational approaches in urban planning and design, comprehensive scientific treatment and impact evaluation is still needed. The attempts undertaken to explore the potentials of Machine Learning and Autonomous Reasoning in support of urban planning and management tasks, are yet disparate and fragmented. While research institutions such as the MIT Media Lab´s City Science Group investigate the fundamental technical, societal, and ethical potentials of AI in the urban context, companies such as Sidewalk Labs or ESRI push AI-based products and services to the market, targeting especially authorities and municipalities. They focus on the automated analysis, management, and optimization of large urban data sets, as generated in digitally networked urban systems such as construction, transportation management, or energy infrastructures.
For planners and decision-makers in urban development and design and citizens affected by it, AI remains a black box technology that raises fundamental questions in respect to legal responsibility, authorship, and transparency of decision-making and underlying algorithmic structure. On that basis, this paper analyses, discusses, and evaluates the advantages and shortcomings of AI approaches from technological, socio-ethical, and planning perspectives. Consequently, we identify urban participation processes as a high-potential application field for AI and discuss methodological frameworks adapted from information technology studies and Design Thinking to embrace AI usage in the urban context. We summarize crucial developments and outline potential benefits from state-of-the-art AI approaches such as Natural Language Processing (NLP) for the analysis of large quantities of feedback from citizens, the automated analysis and evaluation of large numbers design propositions (Discriminative AI), as well as the algorithmic generation of design propositions themselves (Generative AI).