Semi-Automated Processing of Harmonized Accelerometer and GPS Data in R: AGPSR

Area of Science:

• Human behavior research
• Sensor data analysis
• Mobility and health studies

Background:

• Increasing use of sensors in mobility, health, and place research generates vast amounts of data.
• Existing software pipelines for cleaning and harmonizing global positioning system (GPS) and accelerometer data are limited.
• Proprietary software, personnel time, and human error contribute to the high cost and inefficiency of current data management approaches.

Purpose of the Study:

• To address the bottleneck in managing sensor data influx.
• To develop an open-source software pipeline for cleaning and harmonizing GPS and accelerometer data.
• To provide simultaneous information on location and activity status from integrated sensor data.

Main Methods:

• Development of the open-source AGPSR pipeline in the R software environment.
• Combining global positioning system (GPS) and accelerometer data.
• Creation of a unified dataset integrating time, space, and movement information.

Main Results:

• The AGPSR pipeline successfully integrates GPS and accelerometer data.
• The pipeline yields simultaneous information on location and activity status.
• Demonstrations and validity data confirm the pipeline's effectiveness.

Conclusions:

• The AGPSR pipeline offers a cost-effective and efficient solution for sensor data management.
• This open-source tool facilitates objective quantification of human behavior patterns.
• AGPSR (Ver 1.0) is available for researchers studying simultaneous GPS and accelerometer wear.