Jingli Technology Solves Quality Inspection Problems in High-Temperature Metallurgical Working Conditions with AI Vision, Seeks Media Coverage
In the extreme production environment of high temperature, dust, and heavy loads in the metallurgical industry, traditional manual inspection and sorting methods have long faced multiple pain points, including high safety hazards, low accuracy, and difficulty in matching the production line rhythm. "Jingli Technology" takes AI visual recognition, 3D detection, and industrial robot control as core technologies to create intelligent solutions covering three scenarios: surface defect detection of high - temperature workpieces, intelligent sorting and palletizing of special steel, and detection and grinding of titanium alloy bars. It has formed a technical product matrix that can be scaled up for implementation.
1. Industry Pain Points and Market Opportunities
As a typical process - type heavy industry, the metallurgical manufacturing industry generally has extreme working conditions such as high temperature, dust, and heavy loads in its production environment. In the context of the current industry's transformation towards intelligence and unmanned operation, key links such as quality inspection, sorting, and grinding still rely heavily on manual operations, which has become the core bottleneck restricting the industry's quality improvement and efficiency increase.
Currently, the industry generally faces three major pain points:
Prominent safety hazards. Manual close - range operations are required for the surface inspection of high - temperature workpieces, resulting in a high risk of accidents such as scalding and mechanical injuries. When manually handling large - weight special steel, high - intensity operations are prone to cause fatigue - related work injuries, putting great pressure on the enterprise's safety management.
Inadequate detection and processing accuracy. Manual visual inspection is easily affected by visual fatigue, resulting in a high rate of missed detection of minor defects such as cracks, pits, and scale. The processing links such as grinding rely on experience for judgment, resulting in uneven processing effects, which directly affects the product yield and consistency.
Difficulty in matching the production line rhythm. The speed of manual detection is limited. The sorting and palletizing operations are labor - intensive and error - prone, making it difficult to meet the needs of large - scale continuous production. When changing product specifications, the traditional equipment debugging is cumbersome and cannot meet the requirements of flexible production.
Meanwhile, the state continues to promote the intelligent transformation of the manufacturing industry and the upgrading of safety production standards. Metallurgical enterprises are facing multiple pressures such as difficulty in recruiting workers, high labor costs, and rising safety compliance costs. The demand for intelligent solutions that can replace manual labor, adapt to extreme working conditions, and ensure safe production shows a rigid growth trend. The market urgently needs an automated technology system integrating detection, sorting, and grinding.
2. AI Visual Solutions and Technology Matrix
"Jingli Technology" has independently developed three sets of intelligent solutions integrating software and hardware for three typical scenarios in the metallurgical industry.
The first solution is an online detection system for surface defects of workpieces under high - temperature working conditions. Aiming at the high - temperature and dusty environment of the metallurgical production line, this system integrates machine vision technology, high - temperature - resistant imaging technology, and AI defect recognition algorithms to construct a non - contact online detection solution.
At the hardware level, the system selects high - temperature - resistant industrial area - scan cameras, telecentric lenses, and high - temperature protective shells, which can withstand working conditions above 800°C. At the software level, it is equipped with a self - developed visual inspection platform and a lightweight CNN defect recognition model. The workflow is as follows: After the high - temperature workpiece is transported to the detection position, the visual system automatically triggers imaging. After image pre - processing to remove high - temperature noise, the AI model accurately identifies defects such as cracks, pits, and scale. The real - time judgment results are transmitted to the production line control system, and defective products are automatically sorted. The detection data is simultaneously uploaded to the MES system for archiving.
This solution has a detection speed of 50 frames per second, fully matching the high - speed rhythm of the production line, and the recognition rate of 0.1mm - level micro - defects reaches 99.5%.
The second solution is an intelligent sorting and palletizing system for special steel. Aiming at the characteristics of diverse specifications and large weight of special steel, this system integrates machine vision recognition, industrial robot arm control, and AI specification matching algorithms to create an unmanned and flexible sorting and palletizing solution. The hardware includes heavy - duty industrial robot arms, industrial vision cameras, laser distance sensors, and production line conveyor rollers. The software is equipped with a visual positioning platform, a self - developed steel specification recognition model, and an intelligent palletizing path planning system. The workflow is as follows: After the steel is transported to the recognition position, the visual system collects contour and barcode information. The AI model determines the specification and model. The robot arm grabs the steel according to the planned path and accurately palletizes it according to the preset process. The data is uploaded to the production management system in real - time.
This solution supports one - key switching of multiple - specification steel, with an identification accuracy rate of 99.8%. The operation speed of the robot arm is more than three times that of manual operation, and it can operate continuously for 7×24 hours.
The third solution is a defect detection and automated grinding system for titanium alloy bars. Aiming at the characteristics of high detection difficulty and high grinding requirements of titanium alloy bars, this system integrates 3D visual detection technology, industrial robot control, and AI trajectory planning algorithms to construct an integrated automated solution of "detection - positioning - grinding - tracing". The hardware includes 3D cameras, industrial robot arms, force - controlled grinding heads, and PLC control systems. The software is equipped with a self - developed visual recognition platform, a grinding trajectory planning system, and a production monitoring and management system. The workflow is as follows: After the information is entered by scanning the code during feeding, the RGV transports the bar to the detection station. The 3D camera scans and identifies the manually marked defects. The 2D - 3D mapping extracts the spatial data of the defects. The AI plans the optimal grinding path. The robot performs constant - force grinding. After secondary detection after flipping, the finished product is unloaded. The whole - process data is uploaded to the MES system for archiving.
In this solution, the defect positioning deviation is ≤0.05mm, the recognition rate reaches 99.8%. The force - controlled grinding head achieves precise force control of 5 - 300N, and the grinding depth is controlled within 0.05 - 0.1mm. It supports titanium alloy bars with a diameter of 85 - 480mm and a weight of 100 - 3000kg, and the overall equipment utilization rate is increased by 60%.
3. Business Model and Market Space
In terms of business model, "Jingli Technology" adopts the model of "solution output + hardware integrated sales". Based on the self - developed core algorithms and software platforms, the team integrates them into hardware such as industrial cameras, robot arms, and grinding equipment, and delivers them to metallurgical manufacturing enterprises in the form of an integrated system. Customers can choose single - scenario deployment or multi - scenario integration solutions according to their own needs.
The target users are focused on manufacturing enterprises in the fields of steel smelting, special alloy processing, and titanium material production. As a basic industry of the national economy, China's crude steel production accounts for more than 50% of the world's total. The production capacity of high - end materials such as titanium alloy and special steel continues to expand. Under the multiple factors of the improvement of safety production standards, the intensification of the problem of difficulty in recruiting workers, and the promotion of intelligent transformation policies, the demand of metallurgical enterprises for automated detection, sorting, and grinding equipment is changing from "optional" to "standard", with a broad market space.
In the long - term strategy, the team plans to expand horizontally to more metal processing scenarios, such as casting detection, weld flaw detection, and surface treatment automation, based on the proven AI vision and robot control technology platforms, to create a reusable standardized product matrix, and move from "single - point breakthrough" to "full - process intelligent empowerment".
4. Team Background and Current Progress
The core team of "Jingli Technology" is led by technical leaders with more than 15 years of experience in the fields of automation, machine vision, and industrial robots. The team members have in - depth accumulations in the fields of computer vision, embedded systems, industrial control, and AI algorithms, forming a full - chain ability from algorithm research and development to hardware integration, from on - site deployment to after - sales service.
In terms of current progress, the prototype research and development and laboratory verification of the core technologies of the three solutions have been completed. The high - temperature workpiece surface defect detection system has passed the high - temperature working condition simulation test, and the imaging scheme operates stably in an environment above 800°C; the special steel sorting and palletizing system has completed the recognition and grabbing verification of multi - specification steel; the titanium alloy bar detection and grinding system has realized the joint debugging test of 3D visual positioning and force - controlled grinding.
The current focus of work is shifting from product research and development to on - site verification and customer delivery. The team is in contact with several metallurgical manufacturing enterprises for pilot cooperation and plans to deploy the system to real production lines for stress testing to verify the stability and effectiveness of the solution in a continuous production environment.
The overall operation of the project is steadily promoted with self - raised funds from the founding team. This report aims to showcase the technical achievements and solutions, and hopes to establish connections with more metallurgical manufacturing enterprises and industrial partners to jointly promote the intelligent upgrading of the metallurgical production process.