Use of GIS-supported semi-autonomous planning bot in site selection studies

The site selection process for a planned hospital in Karaburun, İzmir, was conducted using a semi-autonomous planning bot supported by Geographic Information Systems (GIS). A site selection model based on environmental sustainability and aligned with natural processes was developed using Ian McHarg's ecological planning principles outlined in Design with Nature. Four different GPT (Generative Pre-trained Transformer) models were utilized for information provision and analysis. The first model provided a comprehensive knowledge base on McHarg's planning strategies. The second model compiled demographic, economic, and environmental data specific to İzmir. The third model offered spatial data for Karaburun, including topography, land use, and transportation networks. These three models served as data sources only, without participating in the decision-making process. The fourth model synthesized all inputs, identified site selection criteria, related them to planning strategies, and calculated criteria weights using the Analytical Hierarchy Process (AHP). The resulting weights were applied in GIS-based overlay analyses, producing thematic maps classifying areas as "suitable," "moderately suitable," or "unsuitable." This study demonstrates that integrating AI-based systems into ecological planning can enhance the accuracy and efficiency of decision-making. It also highlights the adaptability of McHarg's principles to modern technologies and offers an innovative approach to sustainable site selection.