adv-06

Artificial Intelligence in the Textile Industry Concept and Technology Review

adv-07

Maeen Md. Khairul Akter

Managing Editor, Textile Focus

image001We often get pop-up advertisements in our social media apps about products and services that we have just been thinking of! This is how artificial intelligence (AI) has revolutionized today’s information world. It is so intense nowadays that human minds are being read. This technology is changing the concepts and creating scopes of innovative developments to solve technical and managerial problems. This article attempts to shed light on the concepts of AI application in the textile industry and the review of available technologies.

image003Artificial Intelligence (AI) Explained

Intelligence, the capability to acquire knowledge and apply them as skills, has been the sole supremacy of human beings. Intelligence can be characterized by the capacity for logic, understanding, self-awareness, learning, emotional knowledge, reasoning, planning, creativity, critical thinking, and problem-solving. If this ability can be produced artificially, it may be termed artificial intelligence or AI in short. This is what happened, and today, AI is one of the most influential breakthrough inventions in the scientific world. So, AI is the simulation of human intelligence in machines in a way that it can mimic one or some of the human intelligence actions. Specifically, a machine that can learn and solve problems at its own capacity is considered a machine with AI technology.

Concept of AI in the Textile Industry

The application of AI technologies in the textile industry is becoming visible as solving complex problems quickly and accurately is becoming key to competitiveness. The textile industry is a labor-intensive industry where many operations, including production and quality control, are predominantly handled by human hands. With AI innovation, attempts are made to develop technologies to improve efficiency and effectiveness in such operations. The concept of AI technology is based on basic actions like detection, identification, inspection, grading, machine vision, prediction, etc. All these actions are primarily done by humans, often considered tedious and often with ambiguous outcomes. AI can augment these basic actions’ efficiency and accuracy with technologies categorized as expert systems, artificial neural networks, fuzzy logic, algorithms, and natural language.  The concepts used in developing AI solutions in the textile industry can be explained in table-1.

Table-1: AI Logics behind the basic actions in potential application areas in textile industry

  Basic Actions Potential Application Areas AI Logic
Detection Detection of any anomaly or irregularity in yarn, fabric, or garments during the production process A high-speed camera is used to snap millions of images of the target object continuously and processing them in a computer to detect whether there is a deviation in the pattern of the images. Any deviation may suggest the detection of some anomaly or irregularity in the object.
Identification/ Recognition Identify whether the anomaly or irregularity matches any prior events or preset parameters to categorize the event; fault identification, or pattern recognition. Preset images of yarn, fabric, or garment faults/ patterns are stored in a computer. The detected deviation by the AI camera is continuously matched with such images to precisely identify its kind of fault. The frequency of the fault occurred is also calculated simultaneously.
Machine Vision Replace or augment human examinations; areas where human eye examinations are not effective—for example, color matching, pattern identification, design extraction etc. AI camera is used as a replacement of the human eye, which scans elements according to desired programmed outcomes with a computer.
Prediction Predict the outcome of subsequent processing stages like in yarn quality prediction from fiber quality or color prediction from dye recipe. Use of big data and neural network technology to predict probable outcomes before the actual production process.

 AI Technology Review in the Textile Industry

The use of AI in textile technology is not widespread yet. The complexity, the operating skill requirement and ROI are the probable reasons behind the limited diffusion of such technologies. However, rigorous research and development are ongoing in different labs and technology companies to develop AI technologies for the textile industry. Reliable technology forecast suggests that in the next 5-10 years, many textile manufacturing processes will adopt AI-based solutions to enhance efficiency. Table-2 demonstrates the current and potential future AI technologies and their reviews in the textile industry. These technologies are fundamentally based on the basic actions described in table-1.

 

Table-2: Review of AI technologies for the textile industry

The Technology Description Advantage
WiseEye

 

Developed by Institute of Textile and Clothing, Hong Kong Polytechnic University

Source: https://rb.gy/nfczoh

 

§  This device detects textile defects during weaving.

§  Installing machine vision into looms, it can continuously scan the weaved cloth and identify 40 common fabric defects with an accuracy resolution of 0.1mm/pixel, minimizing the chances of producing sub-standard fabric by up to 90 percent.

§  A high-powered LED light bar equipped with a high-resolution charge-coupled device camera is mounted on a rail, which is driven by an electric motor to monitor the fabric as the machine produces it. The captured images are preprocessed and fed into an AI system which has been preloaded with data from thousands of yards of fabric. The system compares the new weave with its models in real-time and generates and displays analytical statistics and alerts when needed.

Currently, quality control in textile production is done by humans inspecting random lengths of fabric by eye and matching with a supplied fault sheet. Fatigue and simple human errors make this process unreliable and inconsistent. WiseEye is proved to be 90% more efficient in fault detection compared to the conventional process.
Cognex ViDi

 

Developed by Cognex Corp., an American manufacturer of machine vision systems, founded in Boston in 1981

Source: cognex.com

 

§  Technology tailored for fabric pattern recognition in the textiles industry.

§  Cognex ViDi platform can automatically inspect aspects of fabric patterns such as weaving, knitting, braiding, finishing, and printing.

§  The platform requires no development period for integrating it into a manufacturing system, and it can be trained using predefined images of what a good fabric sample looks like.

The system outperforms the best human fabric pattern analyzers. It requires no software development. Extremely simple installation process where the software algorithm trains itself on a set of known good samples/patterns to create its reference model.
Datacolor

 

Developed by Datacolor, founded in Lucerne, Switzerland in 1970

Source: datacolor.com

§  AI Tolerancing for Fabric Color Matching, Color Measurement, Formulation, Lab dye and Dispensing.

§  It has an AI Pass/Fail (P/F) feature to improve the accuracy and efficiency of instrumental tolerance.

§  Datacolor Spectrophotometers provide a platform for increased efficiency and color measurement confidence while delivering precision and accuracy.

§  Datacolor offers dispensing systems concentrated colorants, liquid pigments, pastes, mediums and thickeners. Prepare the most accurate solutions in the least amount of time for dyeing industries.

 

High accuracy and efficiency of Datacolor compared to manual or conventional color management and dyeing recipe formulation systems in the textile industry.

Among the three technologies, Datacolor is already in operation in different textile industries in Bangladesh. The other two technologies are not yet been introduced but carry huge potentiality to deliver benefits. Some other technologies developed (or in developing stage) target the textile industry in addition to the three AI-based technologies described in this article. For example, an AI-based system for virtual modeling of yarn from fiber properties has been developed at Cornell University, USA. This technology will allow predicting probable yarn quality parameters accurately from yarn property data. Similar research has also been done in the Textile Department, Amirkabir University, Iran, which involved grading yarn appearances from fiber data. In Fraunhofer Institute, Germany simulation of mechanical textile properties has been done using their I solutions named TexMath software. Another process of fabric wrinkle measurement has been developed by Zhejiang University, China, where image processing method is used. Also, 3-D modelling of articles based on technical data can come in handy for high tech technical textile industries. There are other AI-based solutions that can benefit the textile business and decision making. These technologies are based on big data and artificial neural networking technique that tries to establish or identify critical co-relationships between different supply chain players including production, quality, retailing, marketing, costing, consumer feedback etc. Most of the Ai based technologies are only being adopted by tech-savvy textile industries predominantly in the developed countries who produce very high end products or have a very sophisticated operation process. However, solutions can also be designed and adopted in the mass manufacturing textile industries, especially in Bangladesh and other South Asian countries, to improve their production efficiency, design, and quality control beyond the level achievable by only human intervention.

Conclusion

The application of AI in textile technology is not still widespread, even in developed countries. However, technologies are being developed especially in Germany, the USA, and China, to cater to the complex problems prevailing in their textile supply chain. Yet, potential AI technologies have great scopes to benefit textile industries in developing countries where production and quality control problems persist. The key here is to get our industries ready to adopt such technologies. If we want to design the industry of the future, we have to look into things seriously and work collaboratively to develop our skill base and technology management ability.

failed

We send out weekly newsletter.It's free!

Our intelligent team curate fresh news & updates to entertain our valued audience

You are almost done. Please check your inbox or spam folder to confirm your subscription.