Complete Trip dataset links multimodal journeys for population-level analysis

Existing human mobility datasets typically capture only isolated travel behaviors, missing the linked multimodal journeys that reflect real-world movement. The Complete Trip dataset, developed by researchers from multiple institutions, fills this gap by reconstructing complete travel episodes from passively collected smartphone location-based services (LBS) data. Its first release covers six counties in Utah for all of 2020, representing journeys across car, bus, rail, and active transportation (walking/biking). The dataset employs a four-stage workflow: trip identification (detecting movement segments), mode imputation (classifying transport type), route reconstruction (mapping to digital transportation networks), and trip linking (connecting sequential segments belonging to the same travel episode). This preserves journey-level relationships and enables network-based route representations.

By including statistically calibrated expansion weights, Complete Trip supports population-level inference, making it valuable for urban science, public health, disaster resilience, and transportation planning. The dataset addresses a critical gap by providing a holistic view of how individuals combine multiple modes across a single journey—something prior datasets rarely achieve. Researchers can now study modal shifts, travel behavior under disruptions, or accessibility patterns with unprecedented granularity. The paper includes 16 pages, 9 figures, and 5 tables detailing the methodology and validation. This open resource (arXiv:2607.15436) could become a benchmark for multimodal mobility research, especially as LBS data becomes more prevalent in smart city initiatives.