The paper describes mapping, localization and navigation on uneven terrain of an autonomously driven vehicle with variable geometry. The task of the work was to research and implement algorithms for automatic mapping, localization and navigation on uneven terrain for an autonomously driven vehicle. In doing so, it was necessary to study the existing algorithms for automatic mapping, localization and navigation based on the available literature and select the appropriate ones, adapt the selected approach to the problem of navigation on uneven terrain and implement the proposed solution on an autonomously driven vehicle. A solution was proposed in the form of a lidar filter node that has unfiltered values of lidar reading distances and rotation angles from the inertial sensor as an input. Based on this data, the filter publishes compensated values on a specific topic within ROS. Finally, the proposed solution was tested experimentally and compared with the existing solution with the help of the Vicon system.