AMR aisle navigation with dynamic obstacles
Fleet-relevant AMR runs with human and equipment interference in warehouse aisles, captured with intervention and replan metadata.
This is a custom capture program for autonomous mobile robot (AMR) navigation through warehouse aisles full of dynamic obstacles. We record LiDAR and RGB-D runs with the human and equipment interference that triggers stops, slowdowns, and replans, tagged with intervention and geospatial metadata. It targets the messy, shared-space navigation that simulation under-represents and that drives most real-world disengagements.
What we collect
Dynamic obstacle fields, stop/start interventions, and replan episodes with geospatial metadata, capturing the interference patterns, people, forklifts, spills, parked equipment, that real aisles contain.
Sensors and modalities
LiDAR and RGB-D matched to your fleet's sensor stack, time-aligned through our multi-sensor synchronization service so perception and motion correspond.
How capture works
A pilot confirms rig parity and intervention labeling, then capture scales across aisles, shifts, and traffic conditions following our robotics data collection workflow.
QA and metadata
Episodes carry obstacle type, intervention and replan tags, aisle IDs, and geospatial metadata. QA validates sensor sync, calibration, and label completeness against agreed criteria.
Who it is for
AMR and fleet teams working on warehouse automation where navigation must hold up in shared, dynamic spaces. The interference focus pairs with deliberate edge-case data collection for rare near-miss events.
Scenario FAQ
Yes, rig parity is defined during the discovery call so capture matches the LiDAR and camera configuration on your fleet.
People, forklifts, pallets, spills, and parked equipment, the dynamic interference that triggers stops and replans in real aisles.
Yes. Stop/start interventions and replan events are tagged with geospatial metadata for navigation-robustness evaluation.
Scope your capture program
Book a discovery call to align on your stack and data requirements.