Skip to content

lidarslam_ros2 v0.13.0

The supported Humble and Jazzy rosbag2-to-Autoware-compatible map workflow, the first release from the renamed repository rsasaki0109/lidar_slam.

v0.13.0 is not v1.0. One of its five blocking accuracy profiles, newer_college_math_hard, was not measured for this release; see Release verification.

Shortest supported paths

These are unchanged from v0.12.0:

  • lidarslam-map doctor, then lidarslam-map demo <new-output-dir> for the fixed public MID-360 demo;
  • lidarslam-map doctor /path/to/rosbag2, then lidarslam-map start /path/to/rosbag2 for your own bag;
  • pip install lidarslam, then lidarslam-map start /path/to/rosbag2 --engine core without ROS 2 (preview).

Changes since v0.12.0

  • GNSS anchors without ROS 2. With use_gnss (for example --lidarslam-param lidarslam_rtk_gnss.yaml), graph_slam_offline_runner and the lidarslam package anchor the map to the bag's RTK GNSS as the live node does. With use_gnss false every output is byte-identical.
  • Velodyne packets without ROS 2. The lidarslam package decodes velodyne_msgs/msg/VelodyneScan with velodyne_pointcloud's decoder (vendored, BSD-3-Clause), byte-identical to velodyne_transform_node's cloud. lidarslam-map run|start --engine core maps a VelodyneScan + Imu bag with the RKO-LIO profile; the classic NDT packet profile still needs ROS 2.
  • Quantized scan store (opt-in). submap_cloud_max_error_m keeps each submap's stored scan within that many metres in 0.28-0.44x the bytes; the default 0.0 stays lossless.
  • Refinement keeps neighbours together. A map refinement window that moves two consecutive submaps apart by more than 0.10 m or 1 deg against the pose graph is discarded (refine_window_max_adjacent_change_m / _deg).
  • Repository rename. The project is now rsasaki0109/lidar_slam. This release's images are ghcr.io/rsasaki0109/lidar_slam:v0.13.0-humble and ghcr.io/rsasaki0109/lidar_slam:v0.13.0-jazzy; v0.12.x and older images stay under ghcr.io/rsasaki0109/lidar_slam_ros2.

The full list is in CHANGELOG.md.

Release verification

All gates ran on the clean carrier commit 37216efff1b9a5411e1183d0634113c1afcd2ca0 (develop) with a Jazzy build of that commit (lidarslam-map --version: lidarslam_ros2 0.13.0). The tagged commit differs from it only in this section, the changelog heading, the README links, a recorded exception in RELEASING.md and the release-gate snapshot of Benchmarking.

Gate Command Result
Python product suites bash scripts/run_product_python_tests.sh graph_based_slam 1,772 passed / 46 skipped; lidarslam 1,246 passed, 3 failed for host reasons (see Notes).
Strict documentation python3 -m mkdocs build --strict PASS
First-map verification package python3 scripts/check_first_map_verification_package.py --json READY
v1 readiness contract python3 scripts/check_v1_readiness.py --json valid; NOT_READY 8/10 (distribution; independent adoption)
Release bundle reproducibility python3 scripts/check_release_bundle_reproducibility.py PASS: two byte-identical builds, 303 files, 17,661,896 bytes, SHA-256 9dedb3746385c14ee454193cc7f691ab9d90d8109eff13f7f5574a7d71d178cd
Installation check lidarslam-map doctor PASS (READY)
Fixed first-map demo lidarslam-map demo <new-dir> --viewer none PASS: map verifier PASS (8/8), first-map receipt PASS (7/7)
Default workflow (build and package tests) GitHub Actions ci on Humble and Jazzy PASS on 37216eff (run 44208689357)
Blocking release profiles bash scripts/run_release_readiness_checks.sh --skip-default-ci --benchmark-root <root> --fail-on-profiles exit 2: NO_DATA for newer_college_math_hard only (exception below)

Blocking release profiles at the carrier commit, each with clean provenance for that exact commit:

Profile Metric v0.13.0 v0.12.0 Pass Target Status
newer_college_math_hard APE RMSE vs ground truth not measured not measured 0.10 m 0.08 m NO_DATA
ntu_viral_tnp_01 APE RMSE vs ground truth 0.995 m 0.995 m 1.00 m 0.30 m PASS
mid360_gt_rtkslam_construction_seq2 APE RMSE vs checkpoints 0.062 m 0.062 m 0.30 m 0.15 m TARGET_MET
mid360_gt_rtkslam_construction_seq1 APE RMSE vs checkpoints 0.205 m 0.205 m 0.55 m 0.30 m TARGET_MET
leo_drive_applanix_velodyne_cross APE RMSE vs Applanix 0.142 m 0.142 m 1.50 m 0.50 m TARGET_MET

Exception: newer_college_math_hard was not measured

The official Maths-Hard recording, its ground-truth CSV and the Collection 3 calibration are still distributed only through the dataset's request form: on 2026-11-18 (02:02 UTC) the download page embedded only that Google Form and carried no direct download link. The form was not submitted: this release run was not allowed to submit web forms, and the repository must not scrape or rehost the non-commercial assets. Without the data the profile is NO_DATA. The maintainer chose to release v0.13.0 without this profile, as a third one-time exception recorded in RELEASING.md. The profile stays blocking for later releases.

There is no Maths-Hard accuracy number for v0.13.0, v0.12.0 or v0.11.0; the last one is v0.10.0's 0.070 m, which does not carry over (see the v0.11.0 notes).

Notes

  • Host. An Ubuntu 24.04 cloud VM (4 vCPU, 15 GiB) with ROS 2 Jazzy from apt. The workspace was outside the checkout, built --packages-up-to the default targets (Release, BUILD_TESTING=OFF) with RKO_LIO_FETCH_CONTENT_DEPS=OFF against the system robin-map, Sophus, TBB and nlohmann-json. The benchmark scripts source the workspace next to the checkout they run from, so the profiles ran from a clean git worktree of the carrier commit inside that workspace; every metrics.json names the workspace's offline_node, scanmatcher_node and graph_based_slam_node.
  • Python suites. Run in a full clone. The three lidarslam failures are host conditions, not product faults: test_run_source_onboarding_probe needs the ip command, which this VM lacks, and test_sensor_setup_wizard (also re-run by test_check_publication_slice_plan) saw an older ROS workspace installed inside that checkout. Both pass from the clean worktree.
  • NTU VIRAL. 0.995 m (raw odometry 0.985 m), 9,200 pose pairs, the same as v0.12.0 and v0.11.0, again within the 1.00 m threshold by 5 mm. A fresh preparation starts with the write- and network-free bash scripts/download_ntu_viral_tnp01.sh --dry-run; its plan needs 49.2 GB, more than this host had free. The official archive was read as a stream and checked against its published size and MD5 (8,736,253,605 bytes, 82588ea4f29e311447f3d716865a022b); only /os1_cloud_node1/points, /imu/imu and /leica/pose/relative were kept, converted to rosbag2 (zstd message compression, to fit the host's disk) and restamped by the download script's own step. The restamped cloud stream matched v0.11.0's (5,795 messages, same first and last stamps).
  • RTK-SLAM. Construction Hall 2 0.062 m (median 0.059 m) and Hall 1 0.205 m (median 0.095 m), 16/16 checkpoints each, identical to v0.12.0. Both sequences were fetched with scripts/download_rtk_slam_dataset.py at the pinned dataset revision (exact size and SHA-256) and run with scripts/run_rtk_slam_accuracy_suite.py. A first Hall 2 run read the 10.7 GB bag cold and stopped at the benchmark's 200 s stall guard before the first pose; with the bag in the page cache it completed.
  • Leo Drive. The Velodyne driver overlay was built at upstream revision 56fc178d, the revision of the July evidence. The bag matched the July input identity (metadata.yaml and the compressed DB3 SHA-256). 571 corrected pose pairs; map verifier PASS.