LAMP: Large-scale Autonomous Mapping and Positioning

About this technology

Overview

The software, Large-scale Autonomous Mapping and Positioning (LAMP) is a complete package for Simultaneous Localization and Mapping (SLAM) that is adaptable to a large range of robotic systems for exploration of unknown indoor and underground environments. LAMP will process sensor inputs, such as lidar scans, IMU, and odometry estimates, and produce an accurate global location, global map and global pose-graph. This capability is also extended to multi-robots systems, through a centralized fusion of pose-graphs, with low data-rate communications of sparse pose-graph information.The lidar-odometry front-end can be leveraged as part of LAMP, or separately, to produce an accurate odometry estimate, by processing lidar scans, along with IMU and other, less accurate, odometry estimates. LAMP is designed to be a modular architecture and can have modules adapted to include different information into the global pose-graph, to leverage what sensing information is available for maximum localization accuracy. LAMP runs in the ROS ecosystem and expects sensor inputs as ROS messages, and outputs information as ROS messages.