AI solutions for the steel industry

Application of smart steel in scrap steel recycling

Scrap steel, known as “urban minerals”, is an important pillar of the green and sustainable development of the steel industry. Increasing the ratio of scrap steel is an important path to achieving carbon peak and carbon neutrality in the steel industry.

1. Scrap unmanned centralized control


2. Scrap steel automatic inspection system

The scrap steel automatic inspection system and centralized control platform based on the industrial Internet platform and the new digital intelligence base of the 5G private network can realize the visual recycling of scrap steel and avoid unqualified impurities in the recycled scrap steel, effectively It solves the industry pain point of long-term reliance on manual rough judgment and lack of universal, unified, objective and accurate inspection standards in the offline trading of scrap steel.

The types and quantities of steel scrap received are miscellaneous, and the quality inspection of scrap steel received has become an important problem in the iron and steel smelting process. The scrap steel intelligent inspection and receiving system can realize the workflow and data management of the scrap steel inspection work of the target group of customers. It minimizes the inspection time, improves the accuracy and efficiency of the inspection, and promotes the fairness, accuracy, and efficiency of scrap steel inspection management.

The development of the intelligent inspection and receiving system has realized the remote monitoring of scrap steel inspection and the automatic capture of material type pictures, which significantly reduces the labor intensity of employees, reduces the mixed loading problems of scrap steel type doping and impurities, and improves the quality of the inspection. And unloading speed, so that scrap steel inspection is fair, accurate, and efficient.

The system can also be connected with the bulk material weighing system to obtain vehicle weighing information, and at the same time control the vehicle to prohibit the skin from being returned before the goods are graded. It can save the complete cargo inspection and approval process, and support retrospective query and analysis after the event.

1) The grading process is efficient, intelligent, and accurate, free from interference from human factors, and avoids emotional and empirical quality verification;

2) The grading process is safer and more efficient, and on-site inspectors do not need to climb ladders to observe the scrap steel in the car;

3) Adopt the whole process of visual management to avoid human intervention, and the unloading efficiency is more efficient and reasonable;

4) The judgment data is automatically pushed to the settlement system to avoid errors in the manual handwritten data transmission process;

5) The judgment data is automatically saved, avoiding mutual buck-passing between suppliers and inspectors.

In the process of using the scrap steel intelligent grading system, there is no need for manual intervention. The system’s grading process responds in seconds, and the grading results can be displayed after the completion of each truck’s unloading. The application of the scrap steel intelligent grading system greatly relieves the scrap steel business department. It realizes the digital management of the whole process of steel scrap from entering the factory, grading, inspection, unloading, settlement, etc.


3. Scrap Intelligent Algorithm

The intelligent steel scrap quality inspection system adopts artificial intelligence technology, takes deep learning technology as the core, and applies neural network algorithms. AI algorithm model.

(1) AI algorithm for steel scrap area extraction.


scrap steel recycling

The algorithm uses a convolutional neural network and uses ResNet-FPN to extract multi-scale features from scrap car images, with 100+ convolutional layers and tens of millions of parameters. The pixel-level recognition of the compartment position is realized through the head network. It can automatically identify the position of the vehicle; carry out pixel-level recognition and extraction of the compartment, judge the quality of the extracted image, and ensure that the picture of the compartment is clear; automatically identify whether the vehicle is loading and unloading, and identify the status of the material level in the vehicle.

(2) Crane sucker (grab) tracking AI algorithm.


tracking AI algorithm

Using a convolutional neural network, using CSPdarknet53 as the backbone feature extraction network, combined with SPP network to achieve multi-scale feature extraction, 70+ convolutional layers, 60 million parameters to extract grab features, real-time automatic tracking of scrap steel grabs, the model has a high accuracy The tracking speed ensures the real-time performance of the tracking grab, and has high positioning accuracy to ensure the accuracy of the tracking grab. It can accurately track the position of the grab and suction cup to realize automatic positioning.

(3) AI algorithm for steel scrap type identification.


AI algorithm for steel scrap identification

Using a convolutional neural network, using the spatial pyramid pooling module and the network of encoder-decoder structure to extract pixel-level multi-scale features of scrap steel images, it can accurately identify large, medium, and small scrap steel, while ensuring the recognition speed. Pixel-level segmentation of different types of steel scrap can be realized, and the proportion of different thickness types of steel scrap can be predicted. It can accurately identify the quantity, thickness, area, and other characteristic information of different types of steel scrap in the scrap mixture, and calculate the weight ratio of different types of the mix.

(4) Foreign object detection AI algorithm.


Foreign object detection, AI algorithm

A convolutional neural network is used to combine efficient multi-scale feature fusion and model scaling to extract features from scrap steel images and detect dangerous goods such as closed containers, hydraulic cylinders, explosives, etc. Accurate recognition, maintaining a very high recognition speed in complex backgrounds, making foreign object detection more efficient. It can accurately identify airtight containers such as airtight cylinders, cylinders, oil barrels, fire extinguishers, and overlong pieces, and automatically alarm, automatically identify debris, broken soil, debris, and other information, and automatically deduct miscellaneous tons.

(5) Tare calculation AI algorithm.

Through the output results of the scrap steel recognition algorithm model and the foreign object detection algorithm model, the self-learning feature is used to realize vehicle grade judgment and tamper prediction.


What is scrap used for in the steel industry?

As a recyclable resource, steel scrap is mainly used in steelmaking and foundry production. At present, the total amount of scrap steel produced in the world every year is 300-400 million tons, accounting for about 45-50% of the total steel output, of which 85-90% are used as raw materials for steelmaking, and 10-15% are used for casting, ironmaking, and recycled steel products. The consumption of scrap steel mainly lies in steelmaking, and the consumption of scrap steel in the foundry industry accounts for a small proportion and the consumption is relatively stable.

Scrap demand changes are mainly affected by steelmaking demand. From the process point of view, steelmaking is divided into two types: long process and short process. Scrap steel is an important source of ferrite raw materials in the two steelmaking processes. Steel scrap is the main source, and molten iron is the supplement.

Importance of steel scrap to steelmaking

Scrap iron and steel, as a green energy source, is a renewable resource that can be recycled many times and has become the main source of materials in iron and steel smelting. Therefore, whether the quality of scrap iron and steel is strictly controlled directly affects the variety and quality of products. , It may also be related to the control of time in the smelting process, which is the starting point of scrap steel ironmaking, and the quality of scrap steel will eventually affect the economic benefits and performance of enterprises.

1. Steel scrap is an energy-saving resource

For large-scale iron and steel production enterprises, from iron ore to coking, sintering, ironmaking, and steelmaking, energy consumption, and pollution emissions in the entire process flow are mainly concentrated in ironmaking and pre-iron processes, generally accounting for 60% of comprehensive energy consumption. Compared with iron ore, using steel scrap can save this part of the energy and greatly reduce the overall energy consumption of steel production.

2. Reduce pollution

Steel scrap is a low-carbon resource. Scrap steel can reduce the generation of wastewater, waste slag, and waste gas in pre-iron processes such as ironmaking, coking, and sintering, with 76% of wastewater, 72% of waste slag, and 86% of waste gas. If the tailings slag is produced in the iron ore beneficiation process and the dust produced in the coking and sintering process can be reduced by 97%. The use of scrap steel for steelmaking can greatly reduce the generation of “three wastes” and reduce carbon emissions.

According to environmental protection data, using scrap steel to make steel produces 86% fewer pollutants than using ore, 76% less water pollution, and 72% less waste residue. It can be seen that the use of scrap steel for steelmaking will obtain the greatest economic and social benefits. The quality of scrap steel has an important impact on the quality and output of steelmaking production. How to strengthen the quality control of scrap steel and how to make the judgment of scrap steel acceptance grade more intelligently will become an important research topic for iron and steel enterprises.

3. Steel scrap is a recycling resource

Steel – Forming – Use – Scrap, a cycle every 8-30 years, can be used repeatedly in an infinite cycle, and the natural loss is very low. Every 1 ton of steel scrap can reduce the consumption of 1.7 tons of concentrate powder while reducing the mining of 4.3 tons of raw ore and the discharge of 2.6 tons of iron and steel tailings. In addition, using more steel scrap and less iron ore is an important measure to alleviate this dilemma under the current high iron ore price. Increasing the supply capacity of scrap steel is an important way to alleviate the dependence on iron ore.

4. Steel scrap is an important steelmaking resource

Scrap iron and steel are different from other renewable resources such as scrap copper, scrap aluminium, waste plastic, waste paper, etc., as the number of cycles increases, the physical and chemical performance indicators will not be reduced, and the product quality will not be reduced. In a sense, steelmaking is the process of purifying molten steel. Compared with the original iron ore resource, steel scrap is a high-quality steelmaking resource, and it is also the only high-quality raw material resource that can replace the gradually depleted iron ore resource. Therefore, the industry should gradually encourage the increase of scrap steel ratio and actively recycle scrap steel, so as to take the lead in the face of high iron ore prices and the requirements of low-carbon economic development.

Specific application analysis of intelligent manufacturing technology in scrap steel transfer crane system


scrap steel transfer crane, unmanned crane


1 Application in crane positioning technology module

In the steel transfer crane system, the application of intelligent manufacturing technology can realize its intelligent positioning. The first is the positioning of the large vehicle. Because the distance between the working area of the 1# crane in this project is about 450 m, in the specific design, the DL1000 laser range finder developed by SICK is selected. The maximum range The range can reach 1500 m. In the specific installation, it is necessary to install a reflective sticker on a wall in the factory building and install the rangefinder on the crane to achieve good laser emission and reception effects. Then, with the help of the time-of-flight function, you can Accurately calculate the distance from the crane to the factory building. In this way, precise and intelligent positioning effects can be achieved.

The second is the positioning of the trolley. In this project, the span of the trolley is 24 m. In order to realize the accurate positioning of the electromagnet on the trolley’s stroke, the DL100 laser ranging sensor developed by SICK is specially used in the specific design. In the specific installation, it is necessary to install the reflective sticker on the end beam of the trolley and install the laser rangefinder on the main beam, and its installation position should correspond to the track of the trolley, so as to achieve good laser emission and reception effects. With the help of the time-of-flight function, the distance from the crane to the factory building can be accurately calculated. In this way, precise and intelligent positioning effects can be achieved.


2 Application in the module of crane communication technology

Applying intelligent manufacturing technology to the communication module of the scrap steel transfer crane can also achieve a good intelligent control effect. In this project, combined with the actual needs of the site, data communication can be divided into two types: wired and wireless. Among them, wired communication is mainly to complete data exchange with the factory’s current system; wireless communication is to achieve wireless coverage for the crane. In order to achieve this goal, it is necessary to set the corresponding wireless device between the ground control station and the driving control station to achieve a good wireless communication effect. The Crane position identification system, absolute position encoder, and frequency converter all need to be connected with PLC by means of PROFINET. For each component, its own geological number should be set. The address needs to be set in the station, configured through the master station PLC, and then stored in the hardware configuration in the PLC. On the ground operation station, W788-1PRO and wireless access point settings need to be performed; on the crane, W748-1PRO client settings need to be performed. After the directional antenna is connected, the maximum communication distance between the ground operation station and the crane can reach 1 km.

3 Application in the material identification technology module of the crane

Applying intelligent manufacturing technology to the scrap steel transfer crane can also achieve intelligent material identification.

①Intelligent silo scanning

in the specific design, it is necessary to install the laser scanner on the double-sided shears, so as to realize the accurate identification of the position and height of scrap steel in the entire hopper. In the on-site reservoir area, a three-dimensional benchmark model map should be established. By comparing the data obtained by each scan with this model, the area where the outline of the reservoir area changes can be obtained, and then through corresponding calculations, the sky can be calculated. The coordinate position where the car needs to be hoisted or put down next time is determined scientifically.

②Intelligent vehicle scanning

In order to achieve the automatic loading and transportation effect of the crane, it is necessary to use a laser to scan the vehicle and the body, so as to realize the identification of its location and loading and unloading position. During this process, the camera can be used to observe the scrap steel pit and the vehicles on the ground. If manual intervention is required, the management personnel can directly conduct corresponding commands and operations in the central control room. If there is scrap steel falling on the ground, after the management personnel issue corresponding instructions, the crane will clean up the scrap steel on the ground according to the corresponding route and put it into the scrap steel pit.

4 Application in the technical module of crane dispatching management

In the scheduling management of cranes, intelligent manufacturing technology can also give full play to its application advantages.

①In the application of the equipment operation management system, in this process, it is necessary to write the appropriate program into the PLC, and according to the display of the host computer, apply various logics to judge the specific operating conditions of various equipment, so as to To realize the intelligent monitoring of equipment operation. At the same time, it is also necessary to set up an independent remote monitoring system in the system, to achieve a comprehensive understanding of the operation of each device by analyzing the background data, and to inform the user in time before the device failure occurs, so as to minimize the occurrence of failures probability, save the time for abnormal processing, and realize a good guarantee for the operation efficiency of the crane.

②Application in the crane safety information management system. In this process, the software and hardware in the engineer station need to be configured according to the relevant requirements in GBT 28264-2012 of “Safety Monitoring and Management System for Hoisting Machinery”. Then, the specific monitoring status of the bridge metallurgical crane is used as a reference to collect the corresponding parameter information. At the same time, the monitoring camera should be set at the position of the spreader in the crane and other important positions in the workshop.

Notify of
Inline Feedbacks
View all comments

With machine vision as the core, DBM can provide intelligent equipment R&D, manufacturing, and transformation services such as unmanned factories in the metallurgical industry and other industries, machine vision, and robot application design.

Product category

Customer testimonials

Tell us your needs and a technician will contact you immediately.

Friendship Link

We are committed to becoming a “leader in visual applications”, advocating the core values of “integrity, cooperation, innovation, expertise, and excellence”, taking the road of “technology shaping the brand, innovation leading the industry”, and attaching importance to the investment in scientific research and technology. The research and development of new products and the construction of the team make unremitting efforts for the goal of scientific and technological development.

We have been supplied to world’s top steel manufacturer Arcelormittal, TATA Steel, EZZ steel etc. We do OEM for Concast and Danieli for a long time

Would love your thoughts, please comment.x
Metallurgical Intelligence
Change the way of manufacturing

Save labor, improve working environment, improve efficiency, reduce error rate and save cost!

Металлургический интеллект
Изменить способ производства

Сэкономьте труд, улучшите рабочую среду, повысьте эффективность, уменьшите количество ошибок и снизьте затраты!