Terrain analysis as a design tool for autonomous vehicles in difficult terrain

The paper addresses the problem of choosing a conceptual design for autonomous vehicles for difficult terrain in an outdoor environment. The aim is to find measures that can be used to estimate the ability of a vehicle to cope with the terrain topography. Then we can compare different conceptual designs in an quantitative way and even though it will never be possible to quantify every aspect of the vehicles performance we can use these measures together with other selection criteria when trying to select the best conceptual design. As a basis of the analysis we will use a DEM (Digital Elevation Model) of some terrain area a specific area that the robot will be working in or a characteristic representation of some type of difficult terrain. Such models have been obtained by high resolution photogrammetric techniques like laser scanning or by traditional ground measuring. From the terrain model we can calculate scale dependent measures like slope, curvature and envelope of the ground and scale independent measures like fractal dimension. The vehicle is conceptually described by parameters like friction coefficient μ to he ground and characteristic size of vehicle and support area (wheels or feet). The parameters of the vehicle is compared to the measures calculated from the terrain model to give an estimate of the vehicles ability to cope with the terrain.