Ian Roederer led this study of the kinematics (using Gaia proper motions) of some 30+  fields stars in the stellar halo which were selected because they were highly  r-process enhanced. Using Gaia kinematics we computed their orbits and their actions (which are quantities like angular momentum that are conserved). By looking for clustering in action space we identified 8 possible clusters with at least 3 stars. Interestingly the stars in each cluster have a much smaller spread in iron abundance relative to the spread of the whole sample implying that they could actually be signaling the mass of the satellite they came from. Using the standard mass-metallaicity relation this implies that these 8 possible satellites were of similar mass to known satellites. The paper is here: http://adsabs.harvard.edu/abs/2018AJ….156..179R