Mobility modelling in a process constrained environment: Modelling the movements of nurses in a neonatal intensive care unit

Abstract

Understanding the movement of people constrained by process is of practical importance. It may enable process improvements and more accurate provision of space in buildings (such as hospitals, laboratories and airports) and thus contribute to making safer and more efficient built environments. We present an empirical study of the movement of nurses working at a neonatal intensive care unit (NICU) within a UK hospital. The aim of this study is to model the mobility of individuals within a process constrained built environment. Our objective is to create a model that recreates room occupancy distributions – this implies that we require a room transition model that predicts a person’s next destination as well as a dwell time model that predicts how long a person will stay in a room. This class of situation is of theoretical and practical significance because nurses’ movements are driven by sequences of purposeful activity that are spatially, logically and temporally constrained i.e. process constrained. We used Ekahau Wi-Fi location tracking tags to collect room transitions of 10 day-shift nurses within a NICU for a period of 28 days. We use this dataset to evaluate four proposed models of room transition: (1) random model; (2) an occupancy and distance model; (3) an attractiveness model; (4) a Markov model.We evaluate the models’ goodness-of-fit by comparing our empirical dataset with model predictions.