User-Side IoT Sensing Data with 20.6 Million Bluetooth Beacon Records

20,612,286 Bluetooth beacon data points with 15 attributes including sound, temperature, humidity, and air quality were collected for a workshop on People-Centric Sensing. The dataset, associated with a 2019 workshop paper, includes beacon records, device descriptions, and user check-in timeseries from three Android tablets. It was authored by researchers from The University of Texas at Austin.

Use Cases

• Modeling human presence and interaction patterns based on the check-in timeseries data.
• Analyzing correlations between environmental variables like sound, temperature, and air quality (eCO2, TVOC) from the beacon sensor data.
• Studying Bluetooth signal strength (RSSI) for device proximity estimation or localization.
• Mapping Bluetooth device addresses to physical device identities for device profiling.

Strengths

• Contains 20,612,286 primary data points, providing a substantial base for analysis.
• Each beacon record includes 15 attributes covering multiple sensor modalities (acoustic, environmental, radio).
• Includes sample tables (1,000 data points) for easier initial exploration and validation.
• Provides companion tables for device identity mapping and user interaction timeseries, enabling multi-table analysis.

Limitations

• Column-level documentation is absent; field semantics must be inferred from attribute names or the source paper after download.
• Last update date is unknown; freshness unverified.
• The specific temporal and geographic coverage of the data collection is not detailed in the provided metadata.

Provenance

Source

Associated with the paper 'Dataset: User side acquisition of People-Centric Sensing in the Internet-of-Things' presented at the 2nd Workshop on Data Acquisition To Analysis (DATA'19).

Collection Method

Likely collected via user-side sensing devices and Android tablets, as described in the associated workshop paper.