Print Verification System for Wire and Cable with Anti-Twist Measures
Mr. Craig Girdwood, Mr. Max Fondyga
Taymer International Inc., Markham, Ontario, Canada, +1-905-479-2614
Abstract
The use of the Print Verification (PV) system in detecting print defects greatly improves quality assurance on wire and cable. Print defects on the jacket are continuously detected by a machine vision system at production line speed. Images of the defects are processed and saved. This technology improves the print quality of the cable produced, thereby preventing defective product from reaching the end customer
Introduction
Print defects are generally the most common single source of defects in power cable and fiber optic cable manufacturing facilities.
Incorrect messages, print smudging, faded print, and missing print are all print defects that should not be present on good quality product. These types of defects are not only cosmetic problems as they can also lead to misuse of the cable when it reaches the end-user. By being able to detect these defects, corrective actions can be taken at the production level and allow for reprinting or discarding of defective pieces.
The PV system uses machine vision technology to analyze, isolate, and identify print defects. The system runs inline and saves all defect image data, including length and defect size, in a database. The camera is generally positioned close to the printer to simplify print legend identification and print location on the cable. Images of the print legend are captured and processed through high speed image processing algorithms and character verification techniques. Any illegible or incorrect print legends are detected by the software, alerting the operator to take corrective action. The defective print images are stored together with length information, allow for further analysis or report generation.
The alert provides operators with instant feedback on current printing conditions resulting in not only corrective actions, but more importantly allowing the operator and process engineers to identify and reduce the root causes of print defects. This helps reduces future print defects before they happen; improving print quality, reducing rework cost, wasted consumables, and defective products.
The PV system can be easily integrated to existing production lines and works well for all types of printed products (twisted pair, jacketed, bare wires, and convoluted profiles). The system can analyze print legends from any type of printing method such as hot foil, inkjet, laser, and gravure printing.
Maintenance on the unit involves a quick wipe when the unit is dusty. The lifetime of the lights can be over 50,000 hours of usage and are the only consumable part in the whole system.
Current Print Inspection Techniques and Limitations
Inline Visual Inspection using Stroboscope
This type of inspection involves using a stroboscope inline. A stroboscope flashes light at the proper period, making the print legend appear to slow moving or stationary. This allows the operator to verify the existence or quality of the print. However, there is often insufficient time to thoroughly inspect length print messages. This method is also sample based, which means that lengthy sections of product are not inspected. Additionally, the flashing strobe lights can be a nuisance to the operator.
Offline Viewing through Magnifying Lens
Another sample-based technique to detect print defects is having operators inspect the print manually through a magnifying glass. Although the operator can see the individual print defects and categorize them accordingly, this method is limited to small portions of the total product length. It can also only be used up to a relatively slow speed, so is usually only used for customers who require it. As print legends can be lengthy, there is a high chance of human error during the inspection. The repeatability of this method is also limited as each operator makes a judgement call of what is considered ‘bad’ print.
Image Capture and Sampling
This process involves a camera acquiring an image based on an encoder signal and a set sampling rate. Captured images are then shown on a display for operators to assess if there are any print defects. This method has several of the same limitations as stroboscope and offline inspection. Fatigue, operator’s individual judgments, and slower processing speed all reduce the effectiveness of image capture and sampling. The possibility of missing print defects is also fairly high, depending on operator attentiveness and the sampling rate.