A dataset for room level indoor localization using a smart home in a box

Area of Science:

• Biomedical Engineering
• Computer Science
• Ubiquitous Computing

Background:

• Accurate indoor localization is crucial for health monitoring and rehabilitation.
• Existing systems often require complex infrastructure or are costly.
• The Smart Home In a Box (SHIB) system offers a low-cost, deployable solution.

Purpose of the Study:

• To present an annotated dataset for evaluating room-level indoor localization methods.
• To provide data for training and testing localization algorithms using a home-based system.

Main Methods:

• Utilized a wrist-worn wearable device measuring tri-axial accelerometer data.
• Collected data via Bluetooth Low Energy (BLE) packets transmitted to four room-based gateways.
• Recorded Received Signal Strength Indicator (RSSI) values to infer location.
• Gathered calibration data from ten participants and conducted a free-living experiment.

Main Results:

• The dataset includes calibration sequences for ground-truth room data.
• A free-living experiment provides real-world data for testing localization algorithms.
• The SHIB system demonstrates potential for unobtrusive, home-based monitoring.

Conclusions:

• The presented dataset facilitates research in low-cost indoor localization systems.
• This data can advance the development of rehabilitation monitoring technologies.
• The SHIB system's design supports ease of deployment and data collection in domestic settings.