Some ultra-diffuse galaxies (UDGs) are known to host very large globular cluster (GC) systems, clearly exceeding the expectations for galaxies with such low stellar mass and density and thereby challenging our understanding of star cluster formation and its connection to galaxy evolution. As bright point-sources, GCs in UDGs have been used as tracers of stellar populations and kinematic tracers of the total dynamical mass. For the well-discussed UDG NGC1052-DF2 this kinematic approach has found a total mass that might be consistent with the stellar mass only, thereby revealing a lack of dark matter (van Dokkum et al. 2018, Nature, 555, 629). However, this result strongly depends on the assumed distance to the galaxy as several studies have discussed (Trujillo et al. 2019 MNRAS, 819, 20; Danieli et al. 2019, ApJ, 895, 4), and at the distance suggested by van Dokkum et al. 2018, NGC1052-DF2 would have an extremely luminous GC system. This talk will present a novel way of constraining the distance to GC-rich UDGs using the internal velocity dispersions of GCs. Analoguous to the Faber-Jackson relation for elliptical galaxies, GCs follow a well-defined relation between their internal velocity dispersions and absolute magnitudes, offering an independent way of measuring distances. We will present the underlying concept of this GC velocity dispersion distance relation with data of nearby galaxies. Then, we will present initial results of our ongoing project using VLT/FLAMES high resolution spectra of NGC1052-DF2 GCs to derive their mass-to-light ratios and precision velocity dispersions to provide an independent distance measurement to this well-debated system.