Abstract Unmanned ground vehicles need accurate sensors to detect obstacles and map their surroundings. Laser-based distance sensors offers precise results, but 3D off-the-shelf sensors may be too expensive. This paper presents a 3D sensing system using a 2D laser sensor with a rotation system. Point cloud density analyses are presented in order to achieve the optimal rotation speed depending on the vehicle speed, distance to obstacles, etc. The proposed system is able to generate real-time point clouds, detect obstacles and produce maps, with high accuracy and a reasonable price (less than 5, 000 USD).