Carbon Block Tracking Package Based on Vision Technology

Abstract

Consistent and high anode quality is a key when smelters look for continuous economical and operational performance. Resulting from complex and long processes with various influent factors, anode quality deviations are difficult to anticipate in real-time, or to qualify when they start to destroy value on the pot. To address this problem, continuous and real-time anode quality monitoring solutions emerged recently as part of the digital revolution. To be efficient, such solutions must include a reliable anode tracking system. This paper presents a new vision-based anode tracking system. Using digital signature of the anode, like for a fingerprint, the solution went through a successful proof of concept at Aluminium Dunkerque carbon plant. Hundreds of anodes from cooling tunnel down to rodding shop, passed through several vision-based identification stations to evaluate the anode identification rate. This solution is part of a global Carbon Material Tracking and Process Optimization system.