A mobile robot can move autonomously in its environment, meaning it can travel from one place to another without needing to be fixed in one location. These robots are designed to perform tasks in various environments, such as factories, warehouses, hospitals, and homes, and can be remotely controlled or programmed to operate autonomously. One of the applications used is for rescue in rural areas; these robots would have to be designed to operate in complex and rugged terrains and overcome obstacles such as rocks, logs, and branches. Additionally, they would also need to operate in environments with low visibility, such as areas with a lot of smoke, dust, or fog. On the other hand, a surface recognition and reconstruction system is a technology used to capture the shape and texture of three-dimensional objects and create digital models. The system uses 3D scanning techniques to capture data from the object, process it, and generate a digital model in real time. This project aims to integrate a surface recognition and reconstruction system into a terrestrial mobile robot to support rescue operations in rural areas. Additionally, a reinforcement learning algorithm, explicitly Q-learning, is incorporated into the mobile robot to teach it to make correct decisions in an unknown environment. Finally, functional tests of the mobile prototype assembly and total integration were conducted, resulting in favorable outcomes in the surface reconstruction where the robot had moved.