On white goods production lines, compressor handling constitutes a major bottleneck in robotic automatic loading and unloading. Manual handling brings high labor intensity and frequent positioning errors, whereas conventional robotic arms without vision guidance fail to cope with pallet stacks featuring multiple compressor sizes and random orientations.
Recently, Transfer Technology rolled out a 3D vision intelligent loading and unloading system for a leading domestic manufacturer. The integrated single-camera and single-robot setup addresses the automatic depalletizing challenges at dual pallet stations for compressors.
01 Project Conditions
The customer’s production line features mixed production of multi-specification compressors with an integrated main and auxiliary tank structure. Each pallet layer holds a maximum of 40 units, with incoming stacks presented in single-orientation and dual-orientation layouts. The legacy automation solution suffers from three core operational bottlenecks:
1.Inadequate positioning accuracy. The curved and reflective workpiece surfaces, coupled with narrow stacking gaps, result in significant depth detection errors for conventional vision systems, frequently causing gripping collisions and workpiece abrasion.
2.Poor production flexibility. Frequent tooling reconfiguration is required for mixed-model production, and the original system fails to support compressors of varying diameters, limiting production compatibility.
3.Unsatisfactory cycle efficiency. The traditional system can only identify a single workpiece per scan. Long image acquisition and algorithm computation times restrict overall production line throughput.
Furthermore, fluctuating workshop lighting and oil and water stains on workpiece surfaces create harsh operating conditions, placing stringent demands on 3D imaging stability.
02 Solution
1.Station Layout
The system adopts a single-camera dual-pallet architecture. One Epic Eye Laser L V2S industrial camera paired with a single robot covers two independent pallet stations. With a working distance of 1200–3000 mm, the camera delivers full field-of-view coverage for entire pallet stacks.
2.Closed-Loop Workflow
■ The robot and PLC trigger scanning commands;
■ The 3D camera acquires full-layer point cloud data of all compressors in a single capture;
■ The vision algorithm performs batch computation to solve the pose and coordinate data of multiple workpieces;
■ The system outputs accurate picking points and pose data in response to robot instructions;
■ The robotic arm executes grabbing and places the workpieces at the designated unloading pallet station;
■ After the completion of single-layer picking, on-site operators replace empty pallets for cyclic continuous operation.
03 Solution Highlights
Flexible Dual-Pallet Operation The integrated single-camera and single-robot setup serves both left and right pallet stations. The vision system automatically identifies stocked pallets and guides robotic picking, effectively boosting overall equipment utilization.
Multi-Specification Adaptability The system supports both single-orientation and dual-orientation incoming stacking layouts, enabling stable adaptation to mixed-model production scenarios.
High Scalability and Low Maintenance Cost The system is fully commissioned and compatible with three compressor models on-site. Subsequent new product specifications can be independently configured by trained customer engineers, lowering long-term technical reliance and operational costs
Start Your
3D Vision Journey
Whether it's product selection, custom solutions, or technical support — our expert team is always ready to help.
Service Hotline
4000-191-161
Business Email
marketing@qianyi.ai
19110438869
18768118149
Office Locations
Beijing HQ
Floor 14, Building 2, Weilai Park South Zone, Changping District, Beijing