Step On The "Hot Wheels" Of AI Technology. Xiao Cloth Intelligent Quality Inspection Robot Will Appear In Intertextile Autumn Winter Noodles Accessories Exhibition.

"The development of AI AI can not be separated from the combination of industry. Landing in the industry is the direction that many companies are looking for. At present, the better combination is intelligent surveillance, finance, face recognition and other fields, but AI has not been well applied in traditional industries, especially in the textile industry. Our company is committed to promoting the application of artificial intelligence in the textile industry, boosting the textile industry's information and intelligent industrial upgrading. Recently, Shanghai Kai Quan Intelligent Technology Co., Ltd. CEO, technology director Liu Zhen briefed reporters on their own "AI solution".


For textile enterprises, quality is the life of products. How to increase the quality of products is a must answer for every enterprise. The product named "AIQC Xiao Bu intelligent quality inspection robot", developed by Kai Quan intelligent, can help textile enterprises complete the "big problem" of fabric product quality testing.


There are many difficulties.


The textile industry, as China's traditional dominant industry, has been playing an important role in export products. In recent years, China's textile exports are often faced with technical barriers and green barriers, partly due to the fact that domestic and foreign testing organizations are not uniform in testing standards for textile products, some are tested according to local standards, some are judged by laboratory testing standards, others are tested and compared according to the standards provided by suppliers. It is difficult to form a unified testing standard.


In addition, in the process of fabric quality inspection, "immediately" is also very important, early detection of problems, early remedial measures can be taken. It takes time to regenerate the fabric. In particular, some special specifications of fabrics, such as warp and weft yarns or warp and weft density, are different from those commonly used in the market.


"In the past, the common spot checks were 5% to 10% of the total volume, but if the fabric is of high value, or when customers need it, 100% tests should be carried out after the sampling is qualified." Liu Zhen told reporters that the contents of fabric inspection mainly include two aspects: appearance and quality. It includes color, measurement, weight, warp and weft, fabric structure, color fastness, shrinkage and so on. Appearance inspection is an important part of fabric inspection. "The error rate of manual detection is very high. For skilled workers, labor intensity is high and efficiency is low. Often there are cases of missing inspection, wrong inspection, or even no inspection. If a foreign trade enterprise is required to conduct a full inspection in a professional inspection institution, the enterprise must bear a high inspection fee. Liu Zhen said.


Technology breakthroughs need to be hardened.


Based on the reality of the above textile industry, AIQC Xiao Bu intelligent quality inspection robot arises at the historic moment. According to Liu Zhen, this robot uses the intelligent defect recognition algorithm developed by Kai Quan intelligent self-developed, combines the high-precision multi camera defect capture system, and takes the textile fabric defect big data, inspection process and use scenarios as support, and develops the intelligent robot for textile appearance quality inspection, aiming at the needs of fabric appearance inspection and quality evaluation.


"In the process of design and development, we focus on three main advantages: first, the unified inspection standard. This robot has applied artificial intelligence algorithm. Through the study of defective big data, we can identify more than twenty major categories of textile defects, and the detection accuracy can reach more than 95%. The second is a high-speed inspection and simplification operation process with 40~80 M / min, which can save testing time and efficiency, at the same time, the function integration is high, the occupying area is small, the operation mode is humanized, simplifying the work of workers inspection, making cloth inspection work easy to start. Third is to test the information to realize the sharing of things, real-time feedback in the detection process, and also to the upstream and downstream enterprises to share information. With the increasing demand for quality in the textile industry at present, the detection equipment of AI artificial intelligence has played a very favorable role in promoting the quality of the whole industry. Liu Zhen said.


A powerful AIQC textile intelligent quality inspection software equipped with smart fabric inspection robot has many functions, such as fabric appearance inspection, defect classification, color difference analysis and inspection report. It also adopts IOT industrial Internet of things management and ERP enterprise management system. It can effectively improve the management of fabric information data, accurately record the results of each meter fabric, and make the fabric appearance inspection information, visualization and standardization. In addition, textile inspection as a practical scenario, through artificial intelligence deep learning, instead of the textile inspection industry's original low specialization, low informatization, low standardization mode. Upgrading the industry with intelligent technology, enabling AI to serve enterprises, and making China create changes in the world.


"The original intention of making this product is to bring intelligent technology into the textile industry, subvert the inspection methods of textile fabrics, standardize the inspection standards of textile fabrics, and promote the upgrading of traditional industries." Liu Zhen said, "like the hardware part of the camera is also our own research and development, the purpose of this is to get rid of the control of core equipment abroad, effectively control the cost of equipment. Compared with other industries, the profit of the whole textile industry is not high. We just want to develop a low-cost, easy to use product that can really serve the textile industry, and create the AI technology for the upgrading of the textile industry. It is reported that Xiao Bo intelligent quality inspection robot will be held on September 25~27, China international textile fabrics and accessories (Qiu Dong) Expo 5.2 Pavilion C22 booth for the first time.


Think deeply and realize the distance.


2019 can be said to be the world's "5G first year", and China's 5G development has also entered the sprint stage. Under the "dual core drive" of 5G+AI, the development of industrial Internet will enter the "fast lane". For the vast majority of traditional textile enterprises, if we want to survive and develop in a competitive international market, we must conform to the trend.


Liu Zhen said frankly: "at present, we will strive to continue to deepen the application of the textile industry, seize the tide of 5G, and further promote the application of AI in the textile industry. Now it is only a part of fabric finished product inspection. It will extend to the upstream of the production end and avoid the problem of fabric defects from the source of the production process. On the other hand, relying on the existing advantages, by setting up the RFID 'ID card' tag with the quality inspection information and using all kinds of interconnection brought by 5G technology, we can provide the downstream enterprises with the hardware foundation of intelligent warehousing and logistics.


Hardware input is very hard, and software construction is not "soft". This is Liu Zhen's "ingenuity" and is also the "ingenuity" of every "made in China" entrepreneur. We can not yet imagine what revolutionary applications the textile industry will bring, but these applications will certainly emerge in the practice of 5G+AI. Many times, the application of technology is not based on one-sided prediction and planning. What enterprises need to do is to make technology go ahead and prepare for the underlying technology to become more mature and flexible, so as to accelerate the breakthroughs and landing, step on the "Hot Wheels" of AI technology, solve many major challenges faced by human society, and pave the way to the future world.