The city of Lund has a high share of trips done by bicycle, and ambitious goals of becoming climate-neutral by 2030. Still the municipality, like many other cities, has difficulty in reaching the goals of increasing cycling and needs to find the real potential. To understand the untapped cycling potentials of a city, and generally to promote it, it is important to first assess how bike is used and why. This can be done by observing the bike flows and understanding the reasons for the use or misuse of bikes as a transport mode. Particularly, for the scope of identifying untapped potential, the preferred time scale should be smaller than the traditional count based on annual average figures. Also, opposite to motorized traffic, bicycle volume can rapidly change inside a city. All this caused the need to search for more effective methods to estimate bicycle volume like crowdsourcing data that can be provided by mobile apps which are usually collected for other purposes. The present study introduces a method of bicycle volume estimation based on GPS data from the TravelVu application, an app-based tool to collect travel survey data. 

A spatial analysis has been chosen to investigate the correlation between the data from the Lund municipality counting and the GPS data from TravelVu app. The analysis has been performed by comparing the average daily bike volume with the aggregated data in 3 months for the GPS track. Counting data refers to 2018 and 2021 and the TravelVu data refers to 3 months starting from August 2018 and 2021, where 2018 data are collected through crowdsourcing and 2021 are random sampling. All the weekend data have been removed since the manual counting was always carried out during weekdays (and should be represented by those). At first, all the data from the Lund municipality report were digitalized in a GIS map and then clustered per intersection to have a more robust regression when compared to the TravelVu data and to reduce the possible spatial errors.