Comparing ideal and real cycling routes to support public active mobility policies

• Grant number: 25/21593-7
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: December 01, 2025
• End date: November 30, 2026
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Higor Amario de Souza
• Grantee: Gabriel Campanelli Iamato
• Host Institution: Escola Politécnica (EP). Universidade de São Paulo (USP). São Paulo , SP, Brazil

Abstract

This research proposal aims to compare ideal bicycle travel routes, as suggested by recommendation algorithms, with the real routes taken by cyclists in the city of São Paulo, using data from a recent survey conducted with cyclists and data from the Strava application. The growing relevance of the bicycle as a means of transportation brings several environmental, social, and economic benefits to the city and its citizens. São Paulo has high cycling potential, with a growing cycling infrastructure over the last decade, though it is still smaller than what is planned for 2028. Thus, understanding the travel behavior of cyclists can contribute to the implementation of new cycling infrastructure. Furthermore, understanding the similarities and differences between the routes suggested by routing algorithms and the actual behavior of cyclists allows for an evaluation of whether the suggested ideal routes can assist in planning cycling infrastructure in locations where there is no real cyclist travel data. Information from actual routes can also be used to improve route suggestions. The objectives of this research project include developing algorithms to suggest more suitable routes, identifying factors that influence choices, analyzing areas of greatest interest, and proposing locations for the installation of new bike lanes. The methodology involves a literature review, collection and processing of data from Strava and routing service APIs, comparison of real and ideal routes, and the use of pathfinding algorithms such as Dijkstra and A*, as well as machine learning-based techniques. The analysis will seek differences in distance, time, road type, infrastructure, presence of points of interest, and environmental conditions. The project also includes a planned research internship abroad (BEPE) at an internationally recognized university for cycling mobility. It is expected to contribute to urban planning for active mobility. (AU)