Wheeled robots are widely used in many industrial fields. The wheeled robot needs to have implemented an autonomous navigation system to improve work efficiency. In this research, a map-based indoor navigation system is implemented on wheeled robot with Robotics Operating System (ROS) platform using Hector Mapping algorithm. The algorithm Multisensor Data Fusion using Extended Kalman Filter (EKF) which fuses Wheel Odometry data with IMU sensor data for localization, Field Dynamic A-Star algorithm for path planning, and ON-OFF controller for trajectory tracking. Field Dynamic A-Star algorithm is chosen because it solves general path planning algorithm's main issue that limits robot's orientation movement for every 45 o (suboptimal and subnatural path). The robot has ODROID-XU4 as controller to perform map-based indoor navigation, Arduino Mega 2560 to drive motors, RPLIDAR A2 LASER rangefinder for mapping, and VEX Integrated Encoder with Sparkfun Razor 9DoF IMU for localization. The navigation system is successfully implemented on wheeled robot with ROS platform. Robot has successfully mapped indoor environment with 0.174 meter error rate, and has successfully done localization with average error rate of 0.05m on x coordinate, 0.028m on y coordinate, and 1.506 o on orientation angle. Path planner is proved capable of generating path that is not limited every 45 o orientation. Path planner yields 62.5% success rate in generating traversable path and the robot yields 75% success rate in following the path. Robot yields average error rate of 0.046m in moving towards target's x coordinate, 0.072m in moving towards target's y coordinate, and 5.163 o in turning towards target's orientation angle.