Piece Picking using Yaskawa and Mech-Mind

This automation solution combines logistics and sorting automation by utilizing Mech-Mind 3D Vision and the Yaskawa GP8 robot to classify and transfer a diverse range of unstructured objects.

By leveraging a highly cost-effective 3D vision system to recognize both the position and geometry of items, the robot can easily sort products even if their locations change or multiple types of objects are mixed together. This is the ideal solution for environments with diverse workpieces or for companies looking to automate manual tasks with minimal changes to their existing workspace.

The project was implemented to automate order picking in fulfillment centers and eliminate the manual labor required for sorting individual packaged goods. Furthermore, it aims to achieve high-efficiency sorting for unstructured and mixed-SKU inventory.

*Note: This content is based on a demo; actual deployment may require additional equipment based on specific customer requirements.

Components

Robot	Robot: Yaskawa GP8 6-axis high-speed articulated robot. Payload: 8kg Reach: 727mm Weight: 32kg Repeatability: ±0.02mm Vision System: Mech-Mind 3D Vision (AI-driven recognition for mixed-item environments).

Workflow

STEP 1.	Recognition: The system identifies and classifies individual products within an input bin containing a mixture of various items.
STEP 2.	Transfer: The robot picks the recognized item using a vacuum gripper and transfers it to the discharge bin. * The vision system dynamically recalculates the picking coordinates regardless of any changes in the input bin's position.

※ The content provided by Marosol is protected under U.S. and Canadian copyright and intellectual property laws. Unauthorized reproduction, distribution, or use of this content is strictly prohibited.

Key Features

Rapid and Accurate Sorting Performance

Industrial-Grade Speed: Leverages high-performance industrial robots for swift operational throughput.

Efficient Cycle Times: Sorting an individual item takes approximately 3 to 4 seconds (variable based on performance tuning).

Parallel Processing: Executes vision recognition simultaneously with robot movement to drastically reduce overall cycle times.

Intuitive Robot Teaching

User-Friendly Programming: Utilizes a flowchart-based programming tool for easy setup and operation.

Flexible Logic: Simplifies teaching for complex or mixed-SKU environments where items are randomly oriented.

Smart Motion Planning: Automatically generates optimal paths that prevent collisions based on the specific geometry of the workpiece.

Unbeatable Vision Value & Performance

Breakthrough Pricing: A high-performance 3D vision solution at a highly competitive price point in the early 20 million KRW range.

Deep Learning Recognition: Exceptional ability to identify diverse objects using AI-based 3D vision.

Shape Agnostic: Capable of recognizing and picking items regardless of geometric complexity.

Compact and Simple Installation

Minimal Hardware Requirements: Eliminates the need for specialized feeders since input and discharge bins do not require fixed positioning.

Space Efficiency: The system can be integrated directly into existing manual workspaces with minimal layout changes.

Implementation Results

ROI Period	14.4 months
Key Metrics	Significant reduction in personnel required for manual sorting tasks. Drastic decrease in sorting errors (Human Error) during fulfillment.
Client Feedback	We previously delayed automating our sorting due to the sheer variety of product types and shapes. However, thanks to deep learning, we can now sort items without the need for manual registration, which has exponentially increased our automation efficiency. The decision was final once we saw the accurate recognition rates combined with such an affordable vision price.

※ The content provided by Marosol is protected under U.S. and Canadian copyright and intellectual property laws. Unauthorized reproduction, distribution, or use of this content is strictly prohibited.

Please log in to view

The content provided by Marosol is protected under U.S. and Canadian copyright and intellectual property laws. Unauthorized reproduction, distribution, or use of this content is strictly prohibited.

※ The content provided by Marosol is protected under U.S. and Canadian copyright and intellectual property laws. Unauthorized reproduction, distribution, or use of this content is strictly prohibited.

Components

Workflow

Key Features

Implementation Results

Piece Picking using Yaskawa and Mech-Mind

Recommended Use Cases