Project background
A floor-stacked warehouse operated by a large contract manufacturer has 3,000+ storage locations. Goods and pallets move frequently, leaving few fixed references. The customer also did not want QR codes on the floor. The unmanned forklift must detect low and overhanging obstacles such as people and cartons in the aisles.Technology selection
In areas where load positions change frequently or where large open spaces provide few stable features, laser SLAM can be affected by changing contours and insufficient references. The project therefore used a Bluebotics visual-SLAM unmanned forklift integrated with MRDVS visual localization, pallet docking, and 3D visual obstacle avoidance.
Solution components
Actual localization accuracy and operating performance depend on site structure, environmental changes, installation, and calibration. Validate the system on site before deployment.
