UbiComp 2025 Mobile Sensing Dataset (UbiLife Study)

This dataset contains multi-modal mobile sensing and self-reported data collected during a demonstration [1] at the UbiComp/ISWC 2025 conference in Espoo, Finland, as part of the UbiLife study. A total of 82 participants contributed data over three weeks (October 9–29, 2025), with a focus on the main conference days (October 13–16). Data was collected using a smartphone-based sensing application built on the CARP Mobile Sensing (CAMS) framework [2]. The dataset includes more than one million records across 15 heterogeneous data streams, covering: Passive sensing (step count, activity recognition, location) Device context (battery state, screen events) Environmental data (weather, air quality) System signals (heartbeat / keep-alive) Active self-reports (stress, sleep, wellbeing, personality, event ratings) The dataset enables the study of human behavior, mobility, and well-being in a real-world academic conference setting, and supports research in mobile sensing, digital phenotyping, and ubiquitous computing. Processed data is provided in Parquet format with a unified schema. Documentation on GitHub provides a full data dictionary and details, and Python scripts on the preprocessing steps. This dataset is released alongside a paper (to be added when the paper is available). GitHub repo: https://github.com/dtu-digital-health/UbiLife/tree/main References Jakob E. Bardram and Nigel Sjölin Grech. 2026. The CARP Mobile Sensing Framework – Demonstration. In Companion of the 2025 ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp Companion '25). Association for Computing Machinery, New York, NY, USA, 434–438. https://doi.org/10.1145/3714394.3754442 Bardram, Jakob E. "The CARP mobile sensing framework--A cross-platform, reactive, programming framework and runtime environment for digital phenotyping." arXiv preprint arXiv:2006.11904 (2020).