Assessment of speed of human locomotion using a differential satellite global positioning system


Medicine & Science in Sports & Exercise:

SCHUTZ, Y., and R. HERREN. Assessment of speed of human locomotion using a differential satellite global positioning system. Med. Sci. Sports Exerc., Vol. 32, No. 3, pp. 642–646, 2000.

Purpose: The objective was to explore whether a satellite-based navigation system, global positioning system used in differential mode (DGPS), could accurately assess the speed of running in humans.

Methods: A subject was equipped with a portable GPS receptor coupled to a receiver for differential corrections, while running outdoors on a straight asphalt road at 27 different speeds. Actual speed (reference method) was assessed by chronometry.

Results: The accuracy of speed prediction had a standard deviation (SD) of 0.08 km·h−1 for walking, 0.11 km·h−1 for running, yielding a coefficient of variation (SD/mean) of 1.38% and 0.82%, respectively. There was a highly significant linear relationship between actual and DGPS speed assessment (r2 = 0.999) with little bias in the prediction equation, because the slope of the regression line was close to unity (0.997).

Conclusion: the DGPS technique appears to be a valid and inconspicuous tool for “on line” monitoring of the speed of displacement of individuals located on any field on earth, for prolonged periods of time and unlimited distance, but only in specific environmental conditions (“open sky”). Furthermore, the accuracy of speed assessment using the differential GPS mode was improved by a factor of 10 as compared to non-differential GPS.

Determination of speed of displacement of individuals in free-living conditions is extremely important for numerous studies on energy metabolism of walking and running as well as for assessment of total daily energy expenditure of individuals (10). Until now, this determination could hardly be done and with poor accuracy. The simplest and also most common device used in free-living conditions is the pedometer, which roughly estimates the covered distance (and therefore the speed if associated with chronometry) from the number of gait cycles done (8). With this device, errors of measurements can be made, because the ratio between stride rate and speed is clearly variable between individuals. Despite this limitation, this technique can be useful for obtaining an overall physical activity index. Furthermore, it offers, in contrast to activity questionnaire, an objective tool for epidemiological study of physical activity (11).

Another approach developed more recently is the accelerometry technique. Body accelerations are recorded and integrated to estimate daily energy expenditure (3,7). In our laboratory (1), we parameterized body accelerations measured by four sensors and then calculated by sophisticated mathematical tools (neural network) the speed of walking and running (6) as well as the incline of roads with reasonably good accuracy. However, the main disadvantage of this method is that a calibration is necessary for each individual to obtain an accurate speed assessment. The accuracy of the latter method is also very sensitive to uncontrolled modifications in the style of walking or running, which will occur for instance with fatigue or with a change in the nature of the road or terrain.

So, until now, technology failed to offer to physiologists an accurate and reliable tool to determine outdoor speed of displacement of individuals, in particular with “on line” calculations.

In view of a preliminary study performed in our laboratory (9), the space-based system of global positioning (GPS) developed in United States provides accurate, continuous, all weather, three-dimensional position, velocity, and time and therefore could meet this criterion.

In the present study, the possibility of determining speed of displacement of individuals was further explored by using a differential global positioning system (DGPS) because the latter is known to be much more accurate than standard GPS measurements.

Author Information

Institute of Physiology, Faculty of Medicine, University of Lausanne, CH-L005 Lausanne, SWITZERLAND

Submitted for publication December 1998.

Accepted for publication February 1999.

Address for correspondence: Y. Schutz, Institute of Physiology, Faculty of Medicine, University of Lausanne, Rue du Bugnon 7, CH-L005 Lausanne, Switzerland. E-mail:

©2000The American College of Sports Medicine