In this study, the main focus is the scheduling problem arising in two-machine flexible robotic cells consisting of CNC machines in which sets of multiple part-types are produced. In such manufacturing cells, it is possible to process lots that contain different types of parts. The completion time of the production depends on the robot moves as well as the part assignments and processing times of the parts. The considered problem is to find the robot move sequence, the part sequence and the allocation of processing times of the parts on each machine, that jointly minimize the completion time of a particular set of production. A simulated annealing based algorithm is proposed in order to solve the problem of determining the best schedule in a two-machine cell. Experimental results show that this approach works well and can be extended to further cases.