2.1.2. Field Study

The goal of the field study was to evaluate the algorithm performance regarding total distance in a real-world scenario. We recorded twelve subjects who performed a self-paced 3.2-km run by completing eight rounds on a 400-m tartan track. We used this setup to be able to obtain a reference distance accurately.

The equipment (IMUs, shoes) and settings were the same as described in the lab study (Section 2.1.1) to enable a direct comparison of the results. The subjects participating in the field study were not part of the lab study. Additionally, we recorded GPS data using a smartphone (Galaxy S8, Samsung Inc., Seoul, South Korea) and the fitness application Strava (Strava, Strava Inc., San Francisco, CA, USA), which is, with 136 million uploaded runs in 2017, one of the most popular fitness apps worldwide [23]. It also has the capability to export the GPS track in the GPX-format [24] allowing for a computation of the distance of the running track. The accuracy of the implemented algorithms can be compared to the GPS data for the total distance of the run. We used the great circle distance to compute the total distance of the GPS measurements [25]. Our computed total distance from the exported GPX file matched the distance that Strava provided via its services. Thus, we could compare the accuracy of the different algorithms to state-of-the-art running platform distance measurements.
