Structured ingest
Reconstruct from LiDAR or a camera. GNSS and IMU are optional, but the UI makes their presence visible and useful when the data exists.
About LidarFlow
Upload. Reconstruct. Georeference. Download.
There is still a gap between recording robotics data and getting back usable 3D. Viewers help you inspect a bag; desktop tools export frames. LidarFlow fills the missing step: rosbag or MCAP in, metric 3D out — reconstructed from LiDAR (GLIM SLAM) or a single camera (monocular metric depth), georeferenced when GNSS is present.
What the product does
From upload to artifact delivery in one place
Upload. Reconstruct. Georeference. Download. All in your browser.
Production-minded mapping
Runs are tracked, artifacts are downloadable, and every selected setting is visible instead of living in an ad hoc notebook cell.
Clear attribution
LidarFlow builds on serious upstream work and makes that dependence explicit instead of hiding it behind generic platform copy.
Foundations
Built on top of strong mapping and georeferencing projects
GLIM
GLIM
GLIM powers the precision mapping path behind LidarFlow and the canonical map export path.
FlexCloud
FlexCloud
FlexCloud powers the GNSS-backed georeferencing stage that turns SLAM outputs into geographically meaningful deliverables.
Next step