Workshop synthesis: Data Fusion and Integration Techniques

This paper summarizes the discussions and findings from the workshop on Data Fusion and Integration Techniques during the 13th International Conference on Transport Survey Methods. The workshop addressed a broad range of approaches for combining traditional survey data with emerging data sources, with a particular focus on Mobile Network Operator (MNO) data derived from mobile phone signaling events. Methodological contributions included probabilistic record linkage, fuzzy matching, and Dynamic Time Warping, as well as models for transportation mode prediction and the use of app-based travel diaries. Spatial features and their role in integrating diverse data streams were also discussed. A recurring theme was the potential of these integration techniques to advance mobility analysis, improve the understanding of travel behavior, and support urban and transport planning. The following recommendations were made during the workshop: Future research should focus on standardized and transparent data fusion protocols tailored to diverse applications, while balancing traditional and novel data sources for analytical depth and reliability. Emphasizing open science practices will enhance reproducibility and accessibility. Hybrid modelling that integrates statistical and machine learning methods offers new potential, supported by international collaboration through shared data infrastructures and governance.