HomeOnline Exhibition Hall
EV Battery Banner
Home/Cases/EV Battery/AI + 3D Vision-Guided Depalletizing of Battery Cells

Large new-energy company

AI + 3D Vision-Guided Depalletizing of Battery Cells

Customer Demand

A leading new-energy company required an automated solution to pick and feed high-value battery cells from large, densely packed trays. The cells are fragile and tightly arranged, and any contact with the cell tabs during picking may cause short circuits and serious safety incidents. Incoming cells also show variable placement and appearance, so the system must deliver both precision and high throughput to meet downstream takt times for assembly, welding, and inspection. The customer needed an AI-powered, flexible solution to increase safety, reliability, and line autonomy.

Mech-Mind General-Purpose "Eye + Brain" for Robots

AI + 3D Vision-Guided Depalletizing of Battery Cells Project Site

Workflow:

Mech-Mind's “Eye + Brain” solution (including high-precision 3D cameras and AI software suites) equips robots to autonomously locate battery cells and empty trays, verify cell pose, plan collision-free pick paths, and perform stable, high-speed feeding. The system is engineered for seamless integration with robots and conveyors to maintain material flow and match line cycle requirements.

AI + 3D Vision-Guided Depalletizing of Battery Cells

Advantages:

  1. The Mech-Eye LSR XL-GL 3D camera (ETH) and Mech-Eye LSR L-GL 3D camera (EIH) together capture fine-detail, high-quality 3D point clouds of battery cells.

  2. Deliver ultra-high resolution and a wide scanning range, with superior resistance to reflections and ambient light.

  3. Handle reflective surfaces and complex structures effectively, reducing rework.

  4. Provide a working distance of over 3 m, enabling safer and more flexible robot operation without sacrificing accuracy.

  5. Precisely recognize white trays with grooves, enabling robust pick-and-place.

  6. Integrate built-in AI path-planning and collision-detection algorithms for safer, faster motions.

  7. Guide robots to pick two entire layers from a large tray, increasing throughput.

  8. Offer easy installation and versatile deployment—generalize across multiple cell specifications to support rapid production-line changeovers.

Results:

  • Stabilized cycle times and improved pick accuracy.

  • Increased throughput and minimized idle time.

  • Enabled fast deployment and shorter changeovers.

  • Achieved highly efficient, flexible, and intelligent production.  

Click here to watch the real-world case video.