A full 3D map of the environment, with all visible surfaces included in the map. Dense 3D maps are used for operations within the autonomy system when the robot requires full detail for more complex tasks such as autonomous trailer loading and unloading. A similar type of 3D map, but with less detail, is the sparse 3D map.
How do robots create a dense 3D map?
Dense 3D maps are generated using integrated sensor data such as stereo cameras, depth sensors, LiDAR, or a combination, processed using advanced algorithms such as SLAM (Simultaneous Localization and Mapping), stereo vision, and volumetric fusion techniques. These maps enable robots to perceive their environment with fine granularity, supporting precise navigation, obstacle avoidance, and real-time motion planning.
Why are dense 3D maps important?
By maintaining a live, accurate dense 3D map, autonomous robots can operate safely in complex, unstructured environments, dynamically recognizing and adapting to changes around them. Dense 3D mapping is key for applications ranging from aerial drones to warehouse AMRs, facilitating effective exploration, path planning, and interaction with the environment.