Characterization and Analysis of Human Vertical Mobility in Complex Urban Environments: Interplay With Horizontal Mobility and Implications for High-Density Urban Design

Abstract

The study of human mobility deals with the movement of human beings (as a group or as an individual) across space and time. Our movements in space affect and are influenced by the built environment, economy, transportation, and the natural environment. These factors are nowhere more intensely interlinked than in cities. Cities are increasingly growing dense and vertical to adapt to the rapid urbanization of our planet. However, a complete understanding of the influence of the vertical dimension on the dynamics of the cities and the lives of its inhabitants needs to be improved. Urban environments are increasingly studied as networks across the three physical dimensions. However, the processes that occur on them, specifically mobility, are still studied as planar (i.e., two-dimensional) processes considering only horizontal mobility. The latter has been extensively studied using various data sources, e.g., census data, travel surveys, call detail records, and GPS. It can therefore be said that there is a general lack of data and methodology to track and derive vertical movements from large-scale human experiments. It is critical for urban planners to understand the patterns of mobility across three physical dimensions and their impact on the built environment. For instance, such knowledge would greatly help improve traffic management, optimize the allocation of facilities and transportation networks, and study several spreading processes (e.g., epidemic diseases). This research analyzes the patterns of human mobility, encompassing the deeply intertwined horizontal and vertical movements. By studying mobility, valuable insights can be gained, and novel ways can be offered to create, design, and enhance urban built environments with the ultimate goal of prioritizing residents' well-being and quality of life. First, we introduce a city-scale experiment conducted in Singapore, tracking a large population of 50,000 students carrying a wearable device with many sensors. This experiment provides the foundation for accurately deriving vertical displacement based on understanding barometric pressure behavior and developing a machine learning framework. The second study offers novel contributions to the mobility literature by revealing the spatiotemporal patterns of vertical mobility for the first time, discovering key aggregate statistics of vertical mobility, and showing the intricate interplay and correlation between horizontal and vertical mobility. The third and final study aims to identify and understand the factors that influence user movement in an integrated development, where residential, commercial, and public facilities are distributed in a vertical, high-density built environment. With Kampung Admiralty development as a case study, we show that network-theoretic methods offer a powerful way of looking at data-driven designs at the building scale by associating the network process (user movement) with the network structure (built environment layout).