The Simultaneous Localization and Mapping (SLAM) by an autonomous robot is an intensive computational problem and is considered to be a time consuming process. A major limitation of the pose tracking is the real-time constraint. The pose estimation should be done at an acceptable latency to get accurate position information. In this paper, FastSLAM 2.0 approach is proposed, where the computational process is divided into two parallel tasks, the pose tracking and the map optimization. The presented work depends on a distributed architecture where the tracking and mapping tasks concurrently operate as a service in the Cloud. Therefore, the robot onboard system is freed from all the heavy computations. The experiments are performed on public dataset comparable to state-of-the-art techniques. The results show that the computational cost of the tracking process in the Cloud is reduced by 83.6% as compared to its execution on a single robot platform.