Unilux Integrates Machine Learning To Coil Edge Inspection System

SADDLE BROOK, N.J: Unilux, the world's leading designer and manufacturer of surface and edge inspection systems for the metals industry, has just announced the ability to provide Automatic Defect Detection using the industry's first database of coil edge defects.

High definition examples of typical trim defects, captured by the Edge Tech remote inspection system, provide a rare in-process view of coil edge defects typically only seen on finished coils.

We've gathered over 2 million images to identify the telltale signs of a bad edge just as it is beginning, said Eric Zaremski, Engineering Manager. Current and future Unilux customers will use this learning to identify bad edges in real time, so adjustments can be made to correct the issue, such as changing knives or correcting critical settings, before they become quality issues.

Edge Tech improves safety by giving operators a real-time view of edge quality from the safety of the pulpit. Knife changes are based upon actual knife wear, limiting operator interaction to only when necessary. Cost savings are realized by moving beyond traditional indicators of knife life such as the number of coils produced or length of strip trimmed.

Historically, coil edge quality is addressed after the coil is complete, where inefficient options for corrections are re-trim and down-grading. For mills providing solutions where edge quality is important, such as appliance and automotive grades, the extra step of re-trim reduces profitability. With Edge Tech, quality is monitored and corrected in real-time to reduce waste and increase yield.

Like This Article?
Unilux Integrates Machine Learning To Coil Edge Inspection System