Electric scooter (e-scooter) sharing has recently known a wide success in many cities all around the world. Nevertheless, it has also generated issues due to risky and improper behaviour of its users. Wild parking, namely parking without caring about the rules of the road, has in particular become a major issue and has induced an increasing number of cities to impose bans and fines to e-scooter sharing. To tackle wild parking, we introduced the figure of the beautificator, an agent hired by a sharing company with the specific task to reposition e-scooters for guaranteeing urban decorum. In this paper, we propose to increase the effectiveness of the beautificators by integrating Unmanned Aerial Vehicles (UAVs) in their activities: remotely controlled UAVs equipped with cameras are deployed to fly across the sharing service area for identifying e-scooters that require beautification with priority. Thanks to the UAVs, the beautificators do not have to operate blindly, touring locations of parked e-scooters without knowing their parking condition, but can readily learn which e-scooters require their immediate attention. We formulate the problem of optimally scheduling the joint actions of beautificators and UAVs, taking into account beautification constraints, battery limits of UAVs and the possibility of swapping exhausted UAV batteries. For tackling this problem, we propose a mixed integer programming model and a heuristic for accelerating the convergence to the optimum of a state-of-the-art optimization solver, reporting results of computational tests over realistic instances.