As automation is being used increasingly in the industry, this internship explores ways of solving task scheduling problems using a mobile robot in charge of deburring the holes of a part to machine. As an infinite amount of robot configurations can theoretically reach a given hole, we model this problem as a Generalised Travelling Salesman Problem (GTSP). We therefore explore several ways of solving this GTSP, whether it is from a large initial set of base poses or from a large set of arm configurations where the base poses are chosen beforehand. We also implement several variants of these two different approaches and compare all the described methods in order to evaluate their strengths and weaknesses from an industrial point of view.