New Energy Battery Electrode Coating Inspection: 3D AI AOI Replaces Manual Labor, Reducing Labor Costs by −35%

DaoAI's 3D AI AOI equipment, leveraging its proprietary 3D camera and 3D morphology reconstruction, accurately identifies micron-level morphological defects and 2D optical blind spots during new energy battery electrode coating, reducing labor costs in this segment by −35%. This solution not only significantly improves the accuracy and efficiency of electrode quality inspection but also offers enterprises a clear path to optimizing return on investment (ROI) within the current industrial AI trend, effectively alleviating human resource pressures for new energy battery manufacturers.

In the critical new energy battery manufacturing process of electrode coating, achieving comprehensive online quality inspection, especially the identification of micro-morphological defects and defects within 2D optical blind spots, is crucial for improving battery performance and safety. However, traditional manual visual inspection is inefficient, costly, and inconsistent. DaoAI's 3D AI AOI equipment, with its proprietary 3D camera and advanced 3D morphology reconstruction/point cloud technology, can accurately capture thickness variations, edge defects, surface foreign objects, and micron-level pores in the coating layer, many of which are undetectable by 2D vision. Through 2D-3D fusion inspection, DaoAI has successfully reduced labor costs in the electrode coating segment by −35%, providing a reliable automated quality inspection solution for leading new energy battery manufacturers.

Pain Points: Why This Hurdle Is Difficult to Overcome

The quality of new energy battery electrode coating directly determines the battery's energy density, cycle life, and safety. However, inspection in this segment faces multiple challenges: First, **detection accuracy and coverage**. The coating layer thickness typically ranges from tens to hundreds of microns, and any minor thickness variations, coating scratches, edge overflow, surface pores, or particulate matter can lead to reduced battery performance or even safety hazards. Traditional 2D AOI struggles to acquire depth information and is ineffective against 3D morphological defects hidden beneath the surface or lacking clear contrast, leading to high false-negative rates. Second, **labor costs and personnel management**. Electrode coating lines often operate at high speeds, requiring a large number of skilled workers for long hours of high-intensity manual inspection. This not only results in high labor costs but also makes fatigue-induced missed detections common, with poor stability and consistency for human eyes to identify micron-level defects. According to statistics, a leading manufacturer's annual labor cost for the electrode coating section alone amounted to tens of millions of CNY, compounded by difficulties in recruitment and high employee turnover. Third, **production rhythm and data traceability**. Manual inspection speeds cannot match high-speed production line rhythms, and inspection results are difficult to standardize and digitize, hindering defect traceability and process optimization. These factors collectively constitute a significant challenge for quality control in electrode coating.

The root cause of these pain points lies in the complexity of the coating process and the optical properties of the materials. After coating, electrode materials (such as positive and negative electrode slurries) may exhibit certain reflections or textures on their surface, making it difficult for traditional 2D vision systems to stably extract features. Furthermore, many critical defects are anomalies in 3D morphology, such as electrochemical performance differences caused by uneven coating thickness, or the impact of internal structures of tiny pores on ion transport. These defects may not be obvious or even completely obscured in a 2D planar projection, creating 2D optical blind spots. Traditional rule-based AOI systems or detection methods relying on grayscale and color differences prove inadequate when faced with these challenges, failing to provide reliable inspection results. In the current industrial AI landscape, where ROI is emphasized, enterprises urgently need an intelligent solution that can effectively replace manual labor, reduce employment costs, and simultaneously improve detection accuracy and efficiency.

Technical Principles

DaoAI's 3D AI AOI equipment demonstrates outstanding performance in online inspection of electrode coating, with its core lying in its self-developed high-precision 3D camera and advanced 3D morphology reconstruction algorithms. Our 3D camera employs multi-frequency structured light projection combined with the Phase Measuring Profilometry (PMP) principle, projecting encoded gratings onto the electrode surface. By analyzing the deformation of the reflected gratings, it precisely calculates the 3D coordinates of each point. This method achieves micron-level Z-axis (depth) measurement accuracy, far exceeding the limitations of traditional 2D vision, which only acquires X-Y plane information. Through rapid scanning, the system can generate high-density point cloud data of the electrode surface in real-time, fully presenting the 3D morphology of the coating layer.

Based on this high-precision 3D point cloud data, DaoAI's AI AOI software system utilizes deep learning algorithms for defect identification. Unlike traditional rule-based AOI which relies on preset thresholds and feature extraction, our AI model learns from a vast amount of 3D morphological data of electrode sheets to automatically identify and classify various types of defects, including coating thickness anomalies, edge defects, pores, scratches, and foreign objects. Particularly for 2D optical blind spot defects like pores, the AI model can make accurate judgments by combining depth information and surrounding morphological features, effectively avoiding missed detections common in traditional 2D vision. Furthermore, 2D-3D fusion inspection enhances robustness; 2D images capture surface features like color and texture, while 3D data provides precise morphological information. This complementarity allows DaoAI's 3D AI AOI equipment to achieve comprehensive and accurate detection of electrode defects, reducing the false-negative rate to <0.4%, significantly outperforming manual inspection and traditional 2D AOI.

Typical Application Scenarios

• **Coating Thickness Uniformity Detection:** In the electrode coating process, the uniformity of coating thickness directly affects the battery's internal resistance and capacity. DaoAI's 3D AI AOI equipment, through 3D point cloud data, can generate real-time coating thickness distribution maps, precisely identifying localized areas that are too thick or too thin, with micron-level accuracy. Traditional methods struggle to achieve high-precision online full-width thickness measurement, whereas our solution provides continuous, high-resolution thickness data.
• **Edge Defects and Overflow Detection:** The quality of electrode coating edges is crucial, as burrs, overflow, or incomplete coating can lead to short circuits or capacity loss. 3D morphology reconstruction technology can clearly display the three-dimensional structure of the electrode edges, accurately identifying various edge defects, including tiny edge damage or irregular shapes. Traditional 2D vision is susceptible to lighting and contrast variations in complex edge structures, leading to higher false-positive rates.
• **Surface Pores and Pits Detection:** Pores and pits in the electrode coating layer can affect the compaction density of the slurry and the infiltration of the electrolyte, thereby impacting battery performance. These defects are often not obvious in 2D images, or may even be obscured by reflections. DaoAI's 3D AI AOI equipment can identify micron-level pores and pits through precise depth information, effectively compensating for 2D optical blind spots and improving defect detection rates.
• **Surface Foreign Objects and Scratches Detection:** Tiny foreign objects or scratches may be introduced during the coating process, affecting the battery's electrochemical performance. 3D morphology detection can more accurately identify these protruding or sunken morphological anomalies. Combined with color and texture information from 2D images, it enables precise classification and localization of various foreign objects and scratches, reducing false positives caused by lighting changes in traditional 2D inspection.
• **Substrate Wrinkle and Deformation Detection:** The electrode substrate (such as copper foil, aluminum foil) may exhibit wrinkles or slight deformation before or during coating, affecting coating quality. The 3D camera can perceive the overall flatness of the substrate, timely detecting and warning of potential substrate deformation issues, avoiding waste in subsequent processes.

Case Study

A leading domestic new energy battery manufacturer's electrode coating production lines have consistently faced severe labor cost pressures and quality control challenges. This manufacturer operates multiple high-speed coating lines, each requiring at least 6 skilled workers for three-shift manual visual inspection to identify coating defects. Due to the high speed of the electrode sheets and the tiny, diverse nature of defects, manual inspection suffered from high false-negative and false-positive rates, and prolonged worker fatigue led to inconsistent product quality. The enormous annual labor costs and losses from rework and scrap due to quality issues significantly impacted the company's profitability. After learning about the advantages of DaoAI's 3D AI AOI equipment, the manufacturer decided to conduct a pilot deployment on one of its representative coating lines.

The deployment process was very smooth. The DaoAI team quickly integrated and debugged the 3D AI AOI equipment according to the client's production line characteristics. Leveraging the powerful APDT positive/few-shot learning capabilities of the DaoAI AI AOI software system, initial defect model training was completed within 5 minutes using just 15 good electrode samples, significantly shortening the deployment cycle. After a one-month trial run and data validation, the results were remarkable: Before deployment, the average false-negative rate for manual inspection on this line was approximately 1.2%, with a false-positive rate as high as 8%; after deployment, DaoAI's 3D AI AOI equipment **consistently maintained a false-negative rate of <0.4%** and **reduced the false-positive rate by −70%**, greatly improving detection accuracy. More importantly, the equipment successfully replaced all manual inspection positions on that production line, requiring only a small number of maintenance personnel for equipment monitoring. Through this deployment, the manufacturer achieved a significant **reduction in labor costs by −35% for a single production line**, with an estimated return on investment period within 18 months. The client reported that DaoAI's 3D AI AOI equipment not only solved their recruitment challenges but also fundamentally improved product quality stability and traceability, providing a solid technical guarantee for their subsequent production line expansion.

DaoAI's 3D AI AOI equipment is not merely a tool for enhancing detection accuracy, but a crucial investment for enterprises to optimize their human resource structure and achieve sustainable development.

DaoAI Solutions and Products

DaoAI's core solution for new energy battery electrode coating is centered around the DaoAI 3D AI AOI equipment, supported by a powerful software system. This equipment integrates our self-developed high-speed, high-precision 3D camera, capable of real-time acquisition of 3D point cloud data from the electrode surface, with pre-processing performed by a built-in edge computing unit. At its heart is the DaoAI AI AOI software system, which employs advanced vision foundation models and deep learning algorithms, featuring APDT positive/few-shot learning capabilities. It requires only 1–20 good samples to complete model training in 5 minutes, enabling 0-code automatic programming and significantly lowering deployment and maintenance thresholds. Addressing the complexity of electrode coating, DaoAI's AI AOI software system particularly excels at handling micron-level morphological defects, 2D optical blind spot defects, and various hidden defects. Its semantic false-positive filtering function effectively reduces misjudgments caused by environmental interference or material properties. The equipment supports 100% local private deployment, with all inspection data processed and stored within the client's factory, ensuring data security and compliance. Furthermore, DaoAI World as a unified foundation ensures the solution's semantic understanding and cross-scenario generalization capabilities, allowing the system to continuously learn and optimize from production line feedback, constantly improving detection performance.

Through DaoAI's 3D AI AOI equipment, clients can achieve automated and intelligent quality control throughout the electrode coating process. This solution not only significantly improves the defect detection rate to over 99.6%, reducing the false-negative rate to <0.4%, but also effectively lowers operating costs by replacing a large number of manual inspection positions. Data shows that in practical applications, this solution can save clients an average of −35% in labor costs, while seamlessly integrating the production line rhythm with inspection speed, avoiding bottlenecks caused by manual inspection. DaoAI provides not just advanced hardware, but a complete, continuously learning and optimizing AI vision solution, helping clients maintain a leading edge in fierce market competition and achieve high-quality, low-cost intelligent manufacturing.