Over the years, the cloth/fabric inspection machine has become a necessary special equipment for the inspection of fabrics before the production of the garment takes place as it replaces manual labour, automatically inspects and classifies the fabric, and counts the defects of the inspected fabric before storing the entire detailing for future access. It is true that fabric inspection or fabric defect identification technology has been there for years now but only a few companies have gone beyond traditional technology to foray into AI-driven technology for inspection of fabrics.
Once such name is Bullmer which is one of the first companies to launch a next-gen AI-driven intelligent technology – AI 01 – for the inspection of fabric. As a world’s leading brand of automatic cutting equipment, Bullmer – a Germany-based wholly owned subsidiary of China’s Jack Stock – has set up an AI intelligent fabric inspection machine team in China’s Shenzhen to bring the latest and most cutting edge Al intelligent technology to the apparel and textile industry to overcome the high defective rate of clothing and textile factories due to fabric defects, high compensation rate and low manual recognition rate.
The role of such technology is even more important because manual identification information in fabric inspection process cannot be effectively transmitted, and the industry cannot form realistic pain points such as unified standards, hence can’t reform it.
Here is what Bullmer’s intelligent fabric inspection technology is all about…
The parent company of Bullmer – Jack Stock – made some commendable developments in a difficult year like 2020 and one of those developments was the acquisition of a Chinese technology company named LINTSENSE to open up new areas of intelligent cloth inspection. The technology has a number of trademarks, patents and software copyrights in its name and Bullmer, with its years of technical expertise and experience, aims to better it.
As shown in Image 1, the machine is equipped with high-precision power/non-powered rollers with an automatic edge alignment/centering system as well as discharging station. From this point of view, multi-process wrinkle removal happens which is followed by an automatic deviation correction; constant tension winding; and high-precision meter/yard recording.
Once the above mentioned processes take place in the machine, an automatic defect marking process identifies defect and location on the fabric by analysis of an in-built data of over 1000 types of fabric tests, because of which the defect identification is said to be accurate. Further to this, a visual defect map is automatically generated for the user to understand what sort of actions can be taken. “Our machine uses American standard four-point scoring system throughout this fabric defect identification process,” says the company.
As stated above, the machine AI 01 can store the data of defects in inspected fabric, form self-learning ability through ML programming after accumulating data and finally improve the fabric inspection process – all through cloud data service. Bullmer, with this new technology, focuses on advancing the cloth inspection process to help apparel and textile industry save enormous cost.
Simply put, below are the three core factors that make this AI-driven fabric defect identification machine a significant technology:
-
Continuous self-evolution in fabric inspection with patented AI-based Deep Model Self-Learning technology! In the complex fabric environment, the machine can learn more and more intelligent, automatically.
-
According to the operation of trillions level characteristics, millisecond response can be achieved using this technology. It integrates a variety of image processing technologies and the world’s most advanced line scan uses an industrial camera to ensure the stability and anti-interference of imaging and detection.
-
It is the first ecosystem for textile industry that includes fabric defects, types and industry information using big data platforms.
The AI intelligent fabric inspection machine team has visited 100+ customer sites, communicated in-depth with 500+ traditional textile enterprise quality inspections and senior textile practitioners, collected 10 million+ metres and 1000+ fabric test data, and continuously optimised and upgraded the data. China’s well known clothing brands Semir, Anta and Li Ning etc., have implemented this technology at their respective suppliers’ end in China.
Benefits the users can get out of this Bullmer Technology AI 01:
‘Datamation’ of fabric management: Defect maps can be directly formed using the machine, and the data can be directly connected to customer company management systems such as ERP, MES, etc. to facilitate judgment, management and problem handling.
Fabric management standardisation: The machine reduces the high defective rate of enterprises which generally arises due to fabric problems and reduction rate is around 75 per cent. This is achieved due to the standardisation of the defect identification parameters.
Fabric inspection automation: The machine supports 24 hours of uninterrupted work and one fabric inspection machine can save four fabric inspection workers. One person can operate multiple machines, and the single day fabric inspection can cross 18,000 metres in 10 hours of working. The speed of the equipment is 30 metres per minute and the ongoing R&D of the company will see it reach 50 minutes/minute in coming months.
Simplified service and maintenance: The equipment can be diagnosed and maintained remotely, and the core components such as the visual light source and the line scan camera can last up to 10 years. What’s noteworthy is the simplified training process and even a low skilled worker can learn to operate the machine in just under 3 hours.
Intelligent fabric inspection: Through Al deep learning calculation, and periodic improvement of the database, the fabric defect recognition rate can reach 90 per cent, and the highest recognition accuracy can reach 0.01 mm. The company says that the defects are divided in more than 20 categories according to the characteristics of the fabric block and top 5 defects that are identified as common by machine algorithm contribute to around 90 per cent of the total defects in fabrics. However, until now, the machine is suitable to detect defects in plain fabrics both knitted and woven, but the ongoing R&D will make this machine capable enough to detect defects in printed and plaid fabrics.
Flexible fabric inspection: Users can independently set the types of defects to be inspected according to the needs of the fabric.
Source : https://in.apparelresources.com/