Tetra Pak Code Inspection Machine
1.Self developed testing software, processing speed is fast,the fastest up to 36000 packets / hour
2.Replace labor, cost saving, high cost performance
3.Unique lighting design,not affected by th e paper package shading
I. Equipment Overview
1. Purpose
The system is designed for online inspection of inkjet printing and straw quality on paper-packaged products in the Linying Jiayin production line. Defective packages are automatically detected and rejected to ensure production quality.
2. Key Features
Non-contact inspection, fully online, with no risk of product damage.
High detection speed, capable of handling up to 24,000 packages per hour.
User-friendly HMI, supporting multi-level access rights for different operators.
Accurate rejection mechanism: An encoder mounted on the conveyor’s synchronous motor records defective package data, ensuring rejection accuracy regardless of line stops or speed fluctuations.
Industrial-grade cooling system, preventing downtime or system errors caused by extended high-temperature operation.
Customizable functions tailored to the customer’s production requirements.
Automatic defect archiving, storing daily defect images with adjustable retention time.
Continuous rejection and line-stop function, halting production after a preset number of defects.
24/7 remote access and control for troubleshooting and support.
360° inkjet print inspection, covering the entire print area.
Zero-delay inspection, eliminating false rejects caused by bag friction, jams, or manual placement.
II. Detection Capabilities
Presence of code (e.g., missing code, inverted package)
Blurred code (distorted, illegible, or unreadable characters)
Incorrect code
Missing characters
Code recognition (decoding and verification)
Position deviation (beyond defined tolerance)
Duplicate codes (multi-code detection)
Straw detection (missing straw)
III. Installation Location
The equipment is installed on the conveyor line, 0.5–1.0 m downstream of the tube applicator, where inkjet coding and straw application can be simultaneously checked.
IV. Detection Principle
A Gigabit Ethernet digital camera captures top-down images of each package. The images are digitally analyzed, and any nonconforming products are flagged for rejection.
When a package passes the positioning sensor, the sensor logs its arrival, and the control unit records both the package ID and encoder signal. The sensor then triggers the camera to capture an image of the inkjet code and straw, which is transferred to the image processor via the network. The image processor analyzes the data and simultaneously sends the results to:
the HMI, where they are displayed in real time, and
the control unit, which signals the rejector to remove the defective package.
Algorithm Principle
The system is built on deep learning technology, minimizing manual configuration. For model changes, only ~300 product images need to be collected. Our team develops and trains the data model, and once it encompasses all product types, the inspection process becomes virtually error-proof. Screenshots of selected software functions are shown below.
V. Equipment Configuration and Inspection Accuracy
(I) Equipment Specifications
A. Technical Parameters
Equipment Name: Inkjet Printing Inspection System
Model: MT-LL-PM-C1-01X
Maximum Production Speed: 24,000 bags/hour
Inspection Scope: Inkjet code verification and straw inspection
Rejection Method: Automated removal of defective bags identified by the inspection unit
Protection Level: IP65, ensuring reliable performance in industrial environments
B. Primary Applications
This system is widely applied in paper-packaged beverage production lines, where it inspects inkjet codes and straws on various package types and automatically rejects defective products.
C. Key Advantages
Significantly reduces the need for manual inspection, lowering labor and production costs.
Enhances overall production efficiency and ensures stable product quality.
Guarantees that inkjet inspection results comply with the manufacturer’s technical specifications.
D. Detection accuracy:
Serial Number
| Inspection Items | Testing Conditions and Accuracy | Rejection rate
| False kick rate
|
1 | Inkjet Printing Inspection | -- | -- | < = 0.03% |
1.1 | No inkjet printing (missing code, reversed package) | -- | >99.99% | |
1.2 | Missing inkjet printing Blurred inkjet printing | Overall code missing >50% | ≥99.99% | |
Overall code missing >30% | ≥99.96% | |||
1.3 | Incorrect inkjet printing | Code characters unreadable to the human eye | >99.93% | |
1.4 | Position offset | Incorrect production date, incorrect factory code, incorrect security code, deformation (elongation or shortening) | >99.99% | |
1.5 | Duplicate code | Exceeding the expansion zone | >99.99% | |
1.6 | Straw inspection | Multiple codes printed within the code area | >99.99% | |
2 | Inspection Items | No straw | >99.99% | <= 0.01% |
Note: The rejection rate for the "No Coding" test item in the table indicates that for every 10,000 uncoded packages tested by the equipment, one package might not be rejected. This is theoretical experimental data only, and the same applies to other test items. The above test items are tested in real time for each paper package, not just during shift changes to determine coding errors.
