Mathematical Modelling of Pedestrian Route Choices in Urban Areas Using Revealed Preference GPS Data



Mathematical Modelling of Pedestrian Route Choices in Urban Areas Using Revealed Preference GPS Data

Nominated for The Neil Mansfield Award

Authors

Eka Hintaran, Atkins

Description

The purpose of this research is to estimate a pedestrian route choice model from revealed preference GPS data.

Abstract

Knowledge about pedestrian route choice behaviour is modest, while this is very relevant for planning and designing pedestrian-oriented cities. This research studies the relationship between pedestrian route choice behaviour and different environmental street characteristics. Knowledge about this relationship could help us predict pedestrian’s preferences in urban areas.

The purpose of this research is to determine which street characteristics have an influence on the route choice process of pedestrians in urban areas. The objective is to develop a mathematical pedestrian route choice model in order to determine these characteristics.

This model has been developed using revealed preference GPS data collected in Zürich. At the time of this research, this was the first pedestrian route choice model estimated from GPS data on a scale of a city. In order to use the GPS data for modelling purposes, the data required extensive post-processing: filtering, smoothing, cleaning and map-matching of the GPS data. The observed routes were used to calculate the alternative non-chosen routes, using a new algorithm that searches for shortest alternative routes while reducing the remaining network. The last step is to calculate the route attributes for all routes.
The main finding was that maximum rise has the largest influence on route choice behaviour of pedestrians in Zürich. Road types for pedestrian and cyclists only were preferred by pedestrians.

The conclusion is that maximum rise and road type have the most influence on the route choice process of pedestrians in urban areas. However, in order to scientifically answer the research question, this study should be carried out in other types of cities too. The applied modelling techniques of using GPS data for route choice modelling could be used by local governments to determine pedestrian’s preferences in urban areas. This knowledge could be used to design and plan pedestrian facilities and pedestrian-friendly areas.

Publisher

Association for European Transport