Aluminum Surface Defects Classification with Deep Learning Algorithms for Classification Improvement and Automatic Coil Release
Processing pipelines were the fundamental components of the industrial revolution that our world experienced during the 20th century. Every sector in the manufacturing process, i.e. aluminum processing, was gradually upgraded with automated systems resulting in reduced production cost but most importantly in minimizing the human-factor induced error. Since the beginning of this century, we are experiencing an unprecedented rate of technological advance with a significant impact on every aspect of human endeavors. However, the pace at which the Industry is adapting to the change severely lags behind other fields. The recent technological advancements have led to new end-product specifications that require stricter and less error-prone automation rules during the manufacturing process. Thus, processing pipelines that rely on conventional automated systems will soon be deemed as obsolete, in front of the upcoming fourth industrial revolution (Industry 4.0). Sophia is our AI-powered solution for the classification of aluminum surface defects and the improvement of automatic coil release. Sophia is based on Deep Learning algorithms that operate on data derived from sensors and IoT devices that are already part of the processing pipeline, offering real-time intelligent analytics and automated decision making.
Accurate surface defect annotation with state-of-the-art computer vision algorithms on live camera feed from the production pipeline.
Visualize the quality of entire coils in the form of an interactive defect map that can be controlled by defect category or camera id filters. Extract quality reports and automate the decision making process.
Generate coil statistics, quality reports and automate the decision making process.