The session will provide a window into the opportunities of AI & Planning using a real-world application developed by the City of Des Moines and Iowa State University. The research team explored ‘supervised’, ‘self-supervised’ and the latest ‘zero shot learning’ paradigms to train an AI model to assess property conditions. This assessment is an important indicator in the typology used by the city to determine targeted planning interventions. Oftentimes, such efforts are limited by resources and other factors. The session will illustrate how AI can enable efficiency, overcome resource challenges and most importantly, make data consistent and free from bias in such assessments.
Planning is built on a bedrock of equity and inclusion, putting people at the core of its functions. The profession strives to eliminate bias and disparate impact while working relentlessly to serve everyone. An AI model using anonymized data has the potential to eliminate value judgements that may arise from an interpreter’s geographic or contextual bias. It can augment efforts to target limited city resources efficiently for the benefit of all. This not only harnesses the potential of rapidly evolving technology but also results in sound policy making, keeping people at the forefront of planning decisions.
