Matsunaga, Takahiro (The University of Tokyo), Hato, Eiji (The University of Tokyo)

Three-Dimensional Pedestrian Route Choice Model with Quantization of Observational Uncertainty

Scheduled for presentation during the Regular Session "Modeling, Simulation, and Control of Pedestrians and Cyclists I" (WeAT8), Wednesday, September 25, 2024, 12:10−12:30, Salon 16

2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC), September 24- 27, 2024, Edmonton, Canada

This information is tentative and subject to change. Compiled on December 26, 2024

Abstract

Travel behavior in urban spaces is influenced by choices within three-dimensional network structures, highlighting the significance of 3D route choice as a key area of research for future urban design, where transportation infrastructure seamlessly integrates with architecture. Continuous observation of travelers within buildings remains challenging, with the accumulation of large datasets from BLE, WiFi, and AI cameras, despite privacy concerns. The challenge is compounded by significant observation errors from data by fixed sensors such as BLE and WiFi observations, which measure ID and signal strength, rendering traditional GPS-based behavioral models less effective. This study introduces a new 3D route choice model that accounts for these observation uncertainties and addresses issues related to the management of vast data, the accuracy of observations, and estimation distortions caused by externalities, employing quantization of observed data and observation manifolds. By applying this methodology to route choice behaviors at Tokyo's iconic Shibuya station, we have achieved notable improvements in model accuracy. We also simulated 3D flow based on data assimilation with the estimated model and camera data, which can capture the real number of passing pedestrians.