In January 2024, Japan’s Tokyo Steel Co., Ltd. and Singapore’s Volkswagen Steel successively signed contracts to introduce the laser-mesh scrap intelligent grading project, aiming to realize unmanned scrap grading and automate the scrap settlement process.
As the largest electric furnace steel mill in Japan, the former plays a decisive role in the Japanese steel industry. The amount of scrap steel used every day is huge, and the site is extremely complex. Multiple application scenarios such as covered dock loading, dumping unloading, semi-trailer unloading, and unfixed unloading frequently occur. Therefore, the company is very eager for intelligent scrap grading technology and equipment to realize unmanned scrap grading and automate the scrap settlement process. But which supplier was introduced? The company did not make a decision blindly, but after more than two years of investigation and communication, shopping around, and more than 20 plan modifications, it finally chose the laser-mesh scrap steel intelligent grading system.
The same situation also occurs in the latter. As an electric furnace steel mill that uses scrap steel as the main raw material, its demand for scrap steel is also huge, and most of it is light scrap steel that is difficult to determine. The company also looked forward to the intelligent scrap grading system. After comparing and inspecting multiple suppliers, it finally chose Leimu technology and products.
The signing of the contract between Tokyo Steel Co., Ltd. and Volkswagen Steel marks that the Radium Smart Scrap Intelligent Grading System has the largest performance among steel plants in the industry (excluding scrap supplier small yard grading projects), covering yard types and unloading methods. The database includes the most types of scrap steel, the largest library of tagged pictures, the most officially launched projects, and the largest number of intelligent scrap steel grading systems that can be inspected by steel mills. Previously, Leimu’s first overseas scrap steel intelligent grading system was successfully put into use at Chiyoda Steel Industry Co., Ltd. in Japan in February 2023, and will cover the scrap grading of the entire plant in June 2023. The grading accuracy of its scrap steel system is stable at over 93%, and the weight deduction accuracy is around 95%. AI can completely replace manual scrap grading around the clock and at all times, which can greatly reduce the labor intensity of employees and promote the safety and accuracy of scrap steel inspection work. , efficient operation. This fruitful cooperation has also made Radium Corporation and Chiyoda Steel Industry Co., Ltd. a long-term strategic partner.
The laser scrap steel intelligent grading system has not only been highly recognized by domestic and foreign steel companies, but also promoted the intelligent development of steel companies. During the cooperation, the two parties not only reached a consensus on the intelligent scrap grading system, but also conducted in-depth discussions on vehicle logistics informatization and scrap inventory management automation, bringing more efficient and accurate scrap management solutions to enterprises and further improving It has improved the production efficiency and competitiveness of enterprises and injected new impetus into the sustainable development of the steel industry.
The Leimu Scrap Intelligent Grading System has been developed since mid-2017. It has 6 years of research and development experience and has rich implementation experience in the intelligent scrap grading of electric furnace steel plants. This makes the Leimu Scrap Intelligent Grading System stand out among its peers. It stands out and has been recognized by many steel companies. After 6 years of sharpening the sword, the sword of the Leimu Scrap Steel Intelligent Grading System has become even sharper. It has been put into use in 21 steel mills around the world, with a total of 168 points deployed and a total of 283,210 vehicles graded. It can be implemented quickly and can be put into use in just 1 month. The recognition accuracy of a single piece of scrap steel is as high as 93%, and manual measurement is supported. The material type recognition rate is as high as 95%, and the vehicle-level accuracy is as high as 97%.
In order to adapt to the user’s grading standards, the Leimu scientific research team hired skilled workers with long-term scrap grading experience in steel plants to provide guidance to the company, regularly held scrap steel process seminars and team leader teaching meetings, and reviewed the annotations of each annotator. Systematic management of the situation. At present, the laser marking team has marked a total of 473,757 scrap steel images, marking a total of 9,208,313 pieces of scrap steel.
With the support of the above-mentioned huge database, the Leimu Scrap Intelligent Grading System uses deep learning methods, scrap gravity detection, cluster analysis algorithms, and big data fitting algorithms to conduct multi-dimensional and multi-level detection of scrap steel. The recognition capability is becoming more and more powerful, and now it can recognize more than 200+ material types, covering thickness recognition from 1mm to 25mm.