In this paper, the cyclic scheduling problems in 2-machine robotic cells have been studied. We investigated the timed Petri network graph for modeling the part sequencing and the optimal robot moves sequence in robotic manufacturing cells. The robotic manufacturing cell considered in this study has two identical machines and one single gripper robot. Also, we have assumed that the manufacturing cell is capable of producing identical and different parts. The main objective of this study is to minimize the cycle time. To solve this problem, we have proposed two meta-heuristic algorithms called particle swarm optimization (PSO) and simulated annealing (SA) and compared the obtained results with the exact solutions by LINGO. Also the complexity of the proposed model has been analyzed.