Jam Detector for Steel Pickling Lines Using Machine Vision

High efficiency and availability in industrial processing lines are requirements to produce top-grade steel at a minimum cost. One of the most important aspects in achieving these goals is efficient automation, which ensures high performance and reduces the cost of production. This work proposes a new system to improve the automation of a steel processing line: a jam detector based on machine vision. The proposed system is designed to detect jams in a crucial step in steel production: pickling. The proposed machine-vision application acquires images from the pickling line and detects the jam based on the number of pieces ejected from the side trimmers. State-of-the-art methods are used for image processing, providing a fast and robust detector for the industrial line. Tests and the results obtained after more than one year of operation in a steel processing plant indicate that the proposed system meets production needs.