MSM - Multisensor Selfreferencing 3D-Mapping System

Download MSM summary

The MSM project at HSKA, funded by the Baden-Württemberg Ministry of Science, Research and Art (MWK) in the frame of the research programme "Innovative Projects", has predominantly focused on indoor navigation and mapping (without GNSS). Indoor scenes pose a challenge for navigation system as GNSS access is virtually impossible. The MSM project builds on the already developed "Navka" navigation and sensor fusion algorithms by implementing the system in ROS (Robot Operating System). This involves the development of ROS interfaces and wrappers that enable the system to seamlessly integrate with other open source ROS based mapping, navigation and SLAM packages.

The mapping component of the system is based on the Velodyne VLP16 3D laser scanner, a stereo camera rig and wheel odometry. Even if sensors like laser scanners and cameras are usually used only for mapping, the MSM project uses data from such exteroceptive sensors to estimate the platform's state and motion trajectory, which can be of a robot, a UAV or even a hand held mapping/navigation platform. The MSM mapping and navigation system substitutes such sensors to compensate for lack of GNSS in indoor scenarios. Further more the system detects loop closures so as to constrain drift and hence enhance the navigation state estimation. The final out of the system is an accurately registered 3D point cloud map with the respective platform trajectory estimation.

A top view of part of a foyer scene with the sparse point clouds from the VLP16 laser scanner registered. The integration of a navigation component significantly improves the accuracy of such sparse 3D point cloud registration and optimized trajectory computation.


Indoor 3D Mapping with VolksBot, 3D laser scanner and NAVKArine-MSM

Indoor 3D Mapping Based on the VolksBot mobile robot platform, VLP16 3D laser scanner and
NAVKArine-MSM, as a further develpment of the G1MC multi-sensor navigation module and NAVKA-software
In an addition small computer (NUC) was used as central input processing unit (GPIO) for the extended NAVKA-algorithms.

As further hardware component as laserscanner (Velodyne) was used.

This short animation shows a 3D point cloud map of the HSKA Building-B ground floor mapped with
the Multi sensor self-referenceing 3D mapping platform NAVKArine-MSM developed at HSKA under the MSM project "MSM - Multisensor Selfreferencing 3D-Mapping System", funded by the Baden-Württemberg Ministry of Science, Research and Art (MWK) in the frame of the "Innovative Projects" research program.

Development platform: ROS
Hardware: VolksBot mobile robot,
Velodyne VLP16 3D laser scanner
NAVKAarine-G1MC navigation module (GNSS receiver, accelerometer, gryroscope, barometer and magnetometer)
Packages and Algorithms used: NAVKA-navigation and sensor fusion algorithms, togehter with Lidar Odometry and Mapping + Generalized ICP