Stay In Your Lane: Sorting Boxes with 99.9% Accuracy

Combining machine vision and 3D laser scanning to identify, inspect, and sort variable-sized boxes at up to 60 boxes per minute.

Automated packaging lines need to identify, inspect, and route products quickly without slowing production. For this application, Concept Systems developed a machine vision sorting system that combines camera inspection and 3D laser scanning to identify variable-sized boxes, verify their dimensions, detect irregularities, and route each case to the correct destination.

The Problem: Multi-sized Boxes

Boxes of varying sizes and products traveled along the same automated packaging line and needed to be accurately routed to one of three stackers. The system needed to identify each box using its case code while also verifying its physical dimensions.

Sorting by label alone wasn’t enough. Boxes with open flaps or other dimensional irregularities also needed to be detected and rejected before continuing through the process. The challenge was creating a reliable inspection and sorting process that could perform all of these checks without slowing production.

The Solution: Machine Vision

Concept Systems developed an automated inspection and sorting solution using an Allen-Bradley ControlLogix PLC, two Basler Gigabit Ethernet machine vision cameras, and a Hermary SL-1880 3D laser scanner.

As each box travels down the line, the cameras identify its case code while the laser scanner creates a three-dimensional model to verify its dimensions. Combining these technologies allows the system to confirm both the identity and physical condition of each box before determining where it should be routed.

The 3D inspection also identifies open flaps and other dimensional irregularities that could interfere with downstream handling. Boxes that meet the required parameters are routed to the appropriate stacker. If one or more parameters fall outside the acceptable range, the system generates a reject code and separates the box from the line.

The Results

The completed machine vision sorting system provided a consistent, automated method for identifying, inspecting, and routing boxes while detecting irregularities before they continued downstream.

• 99.9% identification and sorting accuracy
• Throughput of up to 60 boxes per minute
• Automated verification of box dimensions and case codes
• Detection and rejection of open flaps and other dimensional irregularities
• 9-month return on investment

Project Details