In this paper, we present the results of the study of autonomous mobile robot movement characteristics in a laboratory environment using small-scale ground vehicles. We consider methods of localization, obstacles detection and path planning for the research robot travelling along a coast. In our approach, we provide techniques for simulation of outdoor conditions in the laboratory using special indoor track with various types of road surfaces (sand, stones, grass). The results of the experiment on autonomous movement in off-road conditions are considered and travelling paths of the robot during a computer simulation and movement in a real environment are compared. Position error is estimated for different types of road surfaces and further solutions are proposed.