Software Correction of Speed Measurement Determined by Phone GNSS Modules in Applications for Runners

Area of Science:

• Sport Science
• Geomatics Engineering
• Signal Processing

Background:

• Consumer wearables like smartphones and sports watches use Global Navigation Satellite System (GNSS) for tracking activities.
• GNSS speed and distance measurements often suffer from inaccuracies due to signal fluctuations and device limitations.
• Existing methods for error correction may be complex or costly, limiting their application in low-cost devices.

Purpose of the Study:

• To develop and evaluate a software-based correction method for GNSS speed and distance measurements.
• To assess the effectiveness of digital low-pass filters in reducing measurement errors.
• To improve the accuracy of GNSS data from affordable consumer devices for sports applications.

Main Methods:

• Implemented digital low-pass filters to process GNSS speed and distance data.
• Utilized real-world data from popular running applications on smartphones and smartwatches.
• Simulated various running scenarios, including constant speed and interval running.
• Compared results against a high-accuracy GNSS receiver as a reference.

Main Results:

• Reduced traveled distance measurement error by up to 70% compared to uncorrected data.
• Decreased speed measurement error in interval running by as much as 80%.
• Demonstrated that the low-cost software solution significantly enhances GNSS data quality.

Conclusions:

• Software correction using digital low-pass filters effectively improves GNSS speed and distance accuracy in consumer wearables.
• The proposed method offers a cost-effective way to achieve high-quality distance and speed estimation.
• This approach makes precise GNSS data accessible for a wider range of sports and fitness applications.