drive_03.pcd · 412 MB
For autonomous vehicle teams
When road-test data keeps ending in one-off scripts, LidarFlow gives you one browser path from rosbag or MCAP to a mapped point cloud you can review across the team.
Use cases
If you record LiDAR or camera data and still write glue scripts to turn it into 3D, this page is for you.
LidarFlow replaces the pile of bag-conversion commands, topic checks, mapping scripts, and handoffs that grows around a repeated rosbag or MCAP workflow — whether you reconstruct from LiDAR or a single camera.
/livox/lidar · LiDARok
/gnss/fix · NavSatFixok
/imu/data · Imuok
/tf_static · TFok
For robotics field validation
Use one workflow for upload, topic validation, run status, and artifact delivery instead of rebuilding the post-processing chain every time a new recording lands.
mapping
For site capture and reconstruction
If your team records LiDAR, or just a forward camera, and then spends days cleaning up the path from capture to a delivered point cloud, LidarFlow replaces that glue. With only a camera, the monocular metric-depth path still gets you metric 3D.
/head/lidar · LiDARok
/torso/imu · Imuok
/feet/contact · ContactStateok
/cam_front · Imageok
For humanoid robotics teams
Bipeds and manipulation platforms generate dense multi-sensor logs every test session. LidarFlow turns each run into a reviewable 3D map so locomotion, perception, and behavior engineers can debug against the same reconstruction.
SLAM · Trajectory
lowhigh
For drone localization
GNSS-degraded flights, indoor inspection, and BVLOS missions need a localization ground truth that does not rely on the onboard EKF. Drop the bag in, get a georeferenced trajectory and point cloud you can compare flight to flight.
Scan · 360°
For UGV and mobile robot fleets
Wheeled and tracked UGVs running outdoor patrols, last-mile delivery, or yard logistics use LidarFlow to turn every shift into a comparable map. No CUDA box, no per-platform SLAM tuning, just upload and review.
Recognize your workflow?
Stop maintaining the glue. Get a mapped point cloud per run.
Who this platform is for
- Teams already recording LiDAR and tired of stitching together post-processing scripts
- Operators who need visible QA around mapping and georeferencing runs
- Leads who want a repeatable workflow without standing up a full internal SLAM platform first
Why teams use it
The value is not only map creation. It is also the audit trail: what topics were selected, what settings were used, what worker processed the run, and which artifacts came back out.