DEM-Based Traversability Map Generation for 2.5D Autonomous Multirobot Navigation

Autonomous mobile robots operating in outdoor environments must have an understanding of the surrounding terrain geometry to ensure efficient and safe navigation. This article presents a DEM-based intelligent traversability mapping framework to transform open-source geospatial data into slope-aware cost maps for multirobot autonomous navigation within the ROS2 framework. The proposed cv_gdal algorithm automatically processes GeoTIFF elevation data using adaptive slope thresholding based on each robot’s physical capabilities, generating ROS-compatible cell occupancy maps. Six regions of Spain were used to evaluate terrain representation accuracy and navigation performance in kilometer-scale DEMS. This framework enables autonomous perception-to-planning pipelines and supports the deployment of multirobot systems for search and rescue (SAR) tasks. By bridging geospatial analytics with robotic perception and adaptive decision-making, this work contributes to the development of intelligent, self-configuring robotic systems capable of operating safely in complex outdoor environments.