LAUSR

This article explores the ability to predict contextual variations in the size of a local population using a model developed by fusing local sensed data sources. The purpose is to demonstrate that pervasive computing and data science can improve knowledge and forecasts about people–place interactions, as an alternative to urban simulations that rely upon static administrative statistics and generalized models of behavior. To demonstrate, a simple question is asked: can we forecast the size of a population in an urban open public space and how it will vary due to dynamic environmental and social conditions? To answer the question, a prediction model is developed from a year of daily WiFi device counts and sensed weather conditions.