Simultaneous localization and mapping (SLAM) constitutes the core challenge in autonomous navigation while avoiding obstacles for mobile robots. Traditional robots require close-range operators during mapping, and remote control remains difficult in a complex environment. Furthermore, the height of sensors limits the detection of small obstacles in 2D mapping. This paper presents a robot system that solves these problems. The system operates SLAM remotely, navigates narrow paths, and estimates the location of small obstacles beyond the detection range of 2D lidar. The experiments utilized the Robot Operating System and an open source GMapping software package. Lidar, a camera, and an inertial measurement sensor unit enabled the remote monitoring of the robot in real-time via Rivz, Rqt, and V-rep. Experiment results demonstrate the advantageous operability and reliability of the system.