A Multi-Stage Method for Spatial Demands Prediction in Healthcare Buildings

Abstract

As urbanization accelerates and population structures change, healthcare buildings are becoming increasingly crowded. Predicting functional area demand is crucial to adapting to this trend and providing high-quality services. This paper introduces an innovative multi-stage method to forecast unbuilt projects using operational data from hospital information systems and building design models to automatically calculate spatial demands. The study’s main findings demonstrate that our method successfully outputs regional demand data, supporting hospital design validation and operational decision-making. Through processing and analyzing log data, this research identified the dynamic characteristics of user activities in hospital buildings and converted them into a time series data format. This method has iterative self-validation and self-optimization features and can maintain flexibility in different scenarios and frequently changing design drawings. This method will provide technical support for a wide range of hospital building stakeholders and has the potential to be applied to more types of buildings. These findings will contribute to various fields, including medical facility planning, design, and public health.

Keywords

• spatial demand
• time series
• process mining
• BIM
• healthcare building