Integrated rmagine's Vulkan backend into MICP-L. Use it to accelerate correspondence finding e.g. on NVIDIA Jetson devices.

Integrated the first pre-release of Ray Monte Carlo Localization (RMCL). The RMCL node uses high-performance ray tracing in triangle meshes and scene graphs as the sensor model in MCL.

Key features:

• RTX-accelerated ray casting and likelihood updates
• Point-wise likelihood updates, i.e. one sensor update per single 1D range value, which naturally handles motion distortion
• Flexible configuration of motion update, sensor update, and resampling rates to better fine-tune the use of available computing resources

A minimal example to get started is available at https://github.com/amock/rmcl_examples.

Patch:

• Fixes for OptiX correspondences.
• New examples in rmcl_examples

Stability Update

• Dedicated subscriber and publisher queues for each sensor
• All sensors now run in separate threads
• Replaced all occurrences of point cloud 1 with point cloud 2
• Sensor time delay issue: Warning now takes longer to be triggered
• Added new examples demonstrating how to combine different sensor sources
  The new code has been successfully tested during extended runs on actual robot hardware (FRE2025).

After conducting real-world tests, we refactored the MICP-L node to better integrate it into the ROS 2 ecosystem and added several new features:

• Limited the possible inputs to only rmcl_msgs. Instead, we provide nodes and instructions to convert commonly used range sensor messages into rmcl_msgs.
• MICP-L can now be launched as a composable node.
• Separated correspondence search from optimization without losing much efficiency. This allowed us to add classic closest-point correspondences (CP), in addition to ray-casting correspondences (RC) (only available for embree backend).
• Improved time synchronization between combinations of sensors and odometry.
• Added many new examples and small demos for a quick start: https://github.com/amock/rmcl_examples

We had to do minor structural changes to the repository in order to better integrate new features into RMCL. This repository is now devided into

• "rmcl" which is a ROS-agnostic library that can be compiled and installed as regular CMake project,
• "rmcl_ros" which contains all the nodes,
• "rmcl_msgs" which are the message moved from to this repository. The original msgs repository is not required anymore.

Using the latest rmcl version might break your launch files as the nodes are now located in "rmcl_ros" package. However, it's rather simple to fix that. The new versions of RMCL are v2.1.0 for ROS 2 and v1.3.0 for ROS 1.

Additional patch-changes:

• using rmagine's internal functions for singular value decomposition and covariance math
• unified correspondence search API for EmbreeCorrectors

Restructuring for the ROS 1 noetic version. We have introduced a new folder structure and changed some namespaces:

• ROS-agnostic library parts were moved to "rmcl"
• ROS-dependent parts are moved to "rmcl_ros"
• Messages are now directly integrated into this repository

Patch:

• integrated rmagine 2.2.4 support
• removed unused ROS wrappers

New:

• implemented added closest point correspondences to all embree correctors

ROS2 release. Functionalities almost unchanged to noetic version. Still it is not backwards compatible. So we decided to increase the version to v2.0.0

• Main MICP-L executable has now full integration for O1Dn scan patterns

• added findRCC functions to all embree simulators
• added rmagine dependency to package.xml since rmagine can now compiled with a ROS workspace