This node is still experimental. I have been using it to help understand the odometry process. I will also be looking at KISS ICP. There are differing requirements depending on the end use. Currently I am experimenting with start/stop higher resolution global mapping. The start/stop is moving to a position then stationary at that position while a higher resolution map is acquired, this is saved and then moved to the next position to repeat this process over. Then the saved sets of maps are merged into a global map that becomes the basis for localization.
Realtime odometry is also being worked on. I am still trying to see if the IMU from the L2 can be used.
The overall process has been:
1) create a L2lidarClass to use in apps
2) create a l2diagnostic app to understand how the L2 operates
3) create a l2lidar_node publisher for ros2 that can be used to source point cloud and IMU data from the L2
4) work on odometry using just the L2 published stream
5) use the L2 published stream to create a global map for localization
1,2 are fairly completed. Mostly minor fixes or enhancements as needed or requested.
3 fairly complete, needing assistance with robot state URDF handling
4 work in progress, have explored quite a few different approaches but still needs a lot more learning
5 Maps sets being generated. Establishing data flow for more automated processing. Experimenting with rosbag data collects as part of that process.
This node is still experimental. I have been using it to help understand the odometry process. I will also be looking at KISS ICP. There are differing requirements depending on the end use. Currently I am experimenting with start/stop higher resolution global mapping. The start/stop is moving to a position then stationary at that position while a higher resolution map is acquired, this is saved and then moved to the next position to repeat this process over. Then the saved sets of maps are merged into a global map that becomes the basis for localization.
Realtime odometry is also being worked on. I am still trying to see if the IMU from the L2 can be used.
The overall process has been:
1) create a L2lidarClass to use in apps
2) create a l2diagnostic app to understand how the L2 operates
3) create a l2lidar_node publisher for ros2 that can be used to source point cloud and IMU data from the L2
4) work on odometry using just the L2 published stream
5) use the L2 published stream to create a global map for localization
1,2 are fairly completed. Mostly minor fixes or enhancements as needed or requested.
3 fairly complete, needing assistance with robot state URDF handling
4 work in progress, have explored quite a few different approaches but still needs a lot more learning
5 Maps sets being generated. Establishing data flow for more automated processing. Experimenting with rosbag data collects as part of that process.