An Open-Source Wearable Sensor System for Measuring the Duty Factor of Runners

A runner’s duty factor (DF) is defined as the ratio of ground contact time (GCT) to stride time. Fast runners tend to have short GCTs as well as a small DF. In the current method of DF measurement, the runner needs to run on a treadmill and use a high-speed motion capture camera for video recording to examine manually when the runner’s foot touches and leaves the ground. This method is labor costly, slow, and inefficient. To ease the DF measurement, we proposed a novel method by designing a special wearable sensor system, the Tag, can collect the acceleration of runners and compute DFs automatically. The Tag can be installed on the head, waist, or ankle to obtain the acceleration of runners for DF calculation. However, different runners will generate significantly varying characteristics of acceleration as their body shapes and running habits may not be similar. Therefore, a machine-learning algorithm was introduced to overcome this issue. The proposed system was evaluated on 27 runners with different running professions, genders, heights, and weights. Results indicate that by using acceleration data measured from the runner’s head and training data based on the runner’s profession category, the proposed design can accurately measure the DF, with a mean absolute error (MAE) of 5%. To facilitate the development of this domain, this study features the first open-source wearable sensor design for this application. New sensing components and data processing algorithms may be introduced to enhance the performance and open additional possibilities to apply this technology in this area.