Abstract
Urban healthcare infrastructure in high-density environments faces intense operational pressures, characterized by spatial constraints, complex multidisciplinary workflows, and stringent clinical protocols. This study investigates the integration of artificial intelligence (AI)-driven generative design methodologies to optimize spatial floor plans in healthcare facilities, focusing on patient flow velocity and clinical staff operational efficiency. Utilizing a multi-objective evolutionary algorithm (MOEA) coupled with discrete-event simulation and agent-based spatial modeling, we generated and evaluated layout permutations across three representative hospital department typologies in New York City. The computational framework simultaneously optimized spatial adjacency graphs, visual connectivity, walking travel distances, and infection control zoning constraints. Comparative performance assessments against existing baseline floor plans demonstrate that AI-generated layouts achieved an average reduction of 22.4% in cumulative staff transit distances and an 18.6% decrease in non-value-added patient transit times during peak operational hours. Furthermore, sightline visibility from central nursing substations to critical patient beds improved by 31.2%. The findings underscore the efficacy of algorithmic space planning in transforming legacy architectural workflows into evidence-based, performative design paradigms capable of navigating the dense spatial footprints typical of urban medical centers.
