Should factory invest in an AOI or an AI fabric inspection machine?

For the manufacturing industry, quality and yield are always the most essential factors. AOI vision inspection, which was formerly only utilized in the electronics industry, is now widely employed in the textile and garment industries. However, not all users, manufacturers, or software vendors can achieve what they want, so let's look at the requirements for visual inspection in the textile and garment business.

What is AOI?

AOI (Auto Optical Inspection) is a system for non-contact image recognition inspection. It is a combined optical, mechanical, electrical control, and software system. A high-resolution camera captures the image, which is subsequently analyzed and manually sorted into a faulty feature. Simply put, AOI systems employ manually discovered defects as a baseline to automatically find other comparable defects.

How diverse are AOI applications?

AOI is used in a wide range of applications, including battery products, PCBs, precision components, food packaging, medical packaging, biomedical science, fingerprint matching, electronics, machinery, molds, and the automotive industry. The introduction of AOI can significantly reduce inspection processing time and hence production costs.

The market's reliance on technology has led to increasing demand for automated optical inspection equipment. It is estimated that the global AOI market will be worth US$2.45 billion by 2028, up from US$1.34 billion in 2022. More factories are adopting automation. and the AOI market is expected to experience a compound annual growth rate of over 12% by 2024.

How easy is visual recognition in the textile and garment industry?

AOI is frequently used in the semiconductor and electronics industries to inspect the appearance of circuit board (PCB) assembly lines. With the market saturated as a result of the industry's popularity of AOI technology, many manufacturers have begun to focus on the development of visual recognition in textiles and the garment sector.

The textile business is notoriously labor-intensive, and almost all textile and garment products worldwide are still inspected manually by eye, which not only has a low detection rate but is also slow, lower birth rate makes recruitment difficult and costly. The average fabric inspector in the textile and garment business can find up to 200 fabric defects per hour, with a detection rate of 70%. As a result, more and more factories are investing in AOI machines.

However, the drawback of AOI fabric inspection machines is their low precision, which necessitates double-checking by skilled employees despite their high detection rate.

Many businesses, particularly AOI companies, are enthusiastic in applying technology to the textile and apparel industry since AOI fabric inspection sounds simple. Utilizing visual recognition technology to its full potential necessitates a thorough understanding of fabric and how it behaves when processed through a machine. Without these must-have factors, AOI fabric inspection machines are inefficient.

What are the challenges to implementing visual recognition technologies in fabric applications?

Although AOI automation can replace manual labor, AOI image processing still has potential for development in terms of product complexity and inspection speed. PCBs, for example, have strict yield criteria and argue that they would rather make a mistake than let it go. As a result, customers have established severe AOI parameters, resulting in a high percentage of misjudgment, forcing users to repeat the process again, increasing expenses even further.

Many AOI manufacturers attempted to transfer the technology to the textile sector, but most were unsuccessful. The fabric with perforations allows light to pass through. Optical inspection is limited by light, angle, resolution, and other elements, and even if the parameters are adjusted to the best settings, there are numerous fabric faults that must be comprehended by skilled employees. Second, because the fabric is such a sensitive commodity, the primary downside of the AOI fabric inspection machine is that the camera has a particularly high mistake rate in grey or dark inconspicuous areas, and an improper working surface setup can significantly degrade the fabric inspection findings.

The application of visual recognition in the textile garment industry necessitates not only the skilled worker’s experience but also the actual time spent at the user's site by the AOI machine manufacturer (approximately 1-6 months depending on the number of fabric types and the type of defects of the user) to further modify the recognition model. This requires long-term collaboration between machine suppliers and users.

Despite the fact that many software vendors or manufacturers have invested in the field of fabric vision recognition and are expected to lead the way, without the necessary fabric flaw experience, time, and manpower costs investment, suppliers will become increasingly frustrated, and users will give up using AOI fabric inspection machines halfway through the process.

Overall, both the customer and the machine vendor must devote a significant amount of time. Without the front-line work’s expertise, user tolerance and attitude toward AOI fabric inspection machines, and manufacturers’ comprehension of fabric defects, not only will users quit up halfway, but machine suppliers will also become increasingly frustrated.

Why should textile and garment markers use an AOI fabric inspection machine?

• Detect and eliminate problems solving

Using an AOI fabric inspection equipment allows for the immediate discovery of fabric flaws, allowing the production to immediately remedy the problem and reducing defective recalls and waste. Priority defect discovery improves customer satisfaction and increases the likelihood of repeat orders.

• Reduce human error

Factory workers had to spend lengthy hours finding problems with their eyes before AOI fabric inspection machines replaced manual labor. Human error, weariness, and a lack of light all hampered manual examination, allowing substandard products to slip through the cracks unnoticed. The AOI machine can help to mitigate this danger.

• Reduce customer complaints and improve return on investment

If adequate fabric defect data is entered into the system, the AOI fabric inspection machine may quickly detect a wide range of fabric problems. Not only is the inspection procedure simplified, but human error is reduced as well. Larger orders can be accepted by the factory, enhancing profit potential and return on investment.

• Reduce production costs

Manual inspection can take up to 20-30 minutes and has an average speed of 20-30cm/s. The inspection speed of the AOI fabric inspection machine may reach 40-60cm/s without halting the machine. Training, trainee quality, and retention concerns, for example, can all lead to rising expenses.

With AOI, do we still need AI?

Although an AOI fabric inspection machine can replace manual inspection, without a certain level of understanding and patience, using AOI may be just as expensive, if not more expensive, than manual testing.

The idea of the AOI fabric inspection machine is to utilize pre-defined parameters as a benchmark, therefore manual judgment of samples and definitions is critical, and much depends on the instructor's knowledge. However, this does not imply that the AOI fabric inspection machine is all-powerful - the AOI fabric inspection machine is a machine that listens and relies on 'people' to summarize defect definitions. In other words, if a new fabric defect develops or is vaguely described, the system will be unable to recognize it.

The fundamental distinction between AOI and AI fabric inspection machines is that AI can assume and generalize experience on its own. In contrast to the stringent principles of AOI, AI learns in-depth and aggregates the results based on human judgment. This is not to say that the information provided to the machine does not need to be filtered. Bad and faulty fabric defect information, or improper classification by humans at the start of the process, can also reduce the accuracy of the AI model.

AOI and AI are complementary, and a solid AI foundation stems from defect pre-feeding and model training. Whether it is an AOI or an AI machine, the pre-purchase stage is a long and necessary journey in which both the user and the machine provider must collaborate.

How to choose a fabric inspection machine provider?

Many people in the textile or garment sector have the mistaken assumption that they can get started right away with an AOI or AI fabric inspection machine. In truth, AOI and AI are not a cure-all. They require a substantial initial investment in both work and time, and the more fabric faults collected and the more precise the classification, the smarter the machine will be.

A professional fabric inspection machine provider not only understands the machine's construction but also has the capability to deal with fabric flaws to reduce the time it takes the user to run the machine. OSHIMA EagleAi is not only capable of overcoming knitted fabric inspection by keeping tension under 5%, but we also collaborate with our long-term partner, a world-leading textile manufacturer, importing thousands of fabric defect images to help the next user reduce the time required to collect fabric defects, with the ultimate goal of allowing them to import into their existing production lines painlessly.

With over a half-century of experience in garment manufacturing, OSHIMA has a long history of working with the industry's top-tier customers to understand their problems and deliver solutions. We are also one of the few garment machinery suppliers in the world to offer machines that are used in more than 85% of plants, reducing the contact window and facilitating communication. Our strength is that we not only manufacture machines, but also offer a whole solution that includes factory planning, manufacturing, machine sales, assembly, and after-sales service. Let us know your fabric inspection requirements, and we’ll assist you choose the most suitable machine for you.