ARTIFICIAL

INTELLIGENCE &
INTELLIGENT
SYSTEMS IN
GARMENT
MANUFACTURING
BY – DEBASRUTI DAS

NOVEMBER 2020
NEED OF AI IN GARMENT MANUFACTURING
The apparel industry in India is an important contributor to the
economy with regard to export earnings and employment
generation. Although India is among the top exporters to the world,
there is a growing realization that the full potential of the industry
in India is not being realized. Also, the apparel industry is facing
stiff competition from countries like Cambodia, Bangladesh, Sri
Lanka, Turkey, Indonesia, Vietnam& China.
To stay relevant and competitive, Indian manufacturers will need
to make a much-needed leap. Thus, it is important to reconsider
the options available to companies, of which technology appears
to be the best solution. The wave of new and emerging
technologies presents new means of cost-saving and revenue
generation every day.
Artificial Intelligence is one of the technologies that is going to be
a game changer in the coming years because of its potential to
EXPORT IN... 15% enable change at a rapidly accelerating pace. There are so many
potential use cases for AI in manufacturing today, making this area
GDP 4%
one of the most invested-in by the global venture capital
community. Technologies that emanate from AI, called cognitive
INDUST... 14%
technologies, include machine learning; computer vision; natural
language processing; speech recognition; robotics; optimization;
rules-based systems; and planning & scheduling.
 Applications of Artificial Intelligence in Garment Industry

Prediction of fabric properties .

Process control and online
monitoring

Classification and grading. SCM and retailing

Identification and analysis of faults. FORECASTING

SOME INTELLIGENT SYSTEMS USED IN MANUFACTURING

01 Automated ERP entries

Even though ERP has progressed a lot in last one decade but the biggest challenge i.e. manual data entry is sti ll not resolved
fully. Arti fi cial intelligence as an integrated part of the ERP system will affect the very essence of daily operati ons. AI soluti ons
will most likely take over routi ne tasks like entries by merchandiser for BOM (Bill of material). This will give a big boost to
companies looking to increase their effi ciency.

02 CAD
CAD systems are used in garment manufacturing for creati ng designs, patt ern-making, and grading operati ons. Even though CAD
soft ware helps in achieving high producti vity and improved quality compared to manual operati on, the CAD soft ware cannot be used
to automati cally generate clothing patt erns or designs for a specifi c garment style. In additi on, in many garment industries, the
traditi onal method of garment patt ern generati on is sti ll done by experienced designers and does not include the use of CAD,
although there is the scope of using AI in patt ern generati on. Several researches have been done to implement the AI that can help
to develop basic clothing patt erns automati cally.

03
.
Fabric inspection
Generally, fabric inspecti on is performed by skilled workers using lighted tables or equipment. This process is rather slow
and many ti mes can allow faults to pass to the garment. Furthermore, the effi ciency of the fabric inspectors will be
reduced quickly with fati gue. However, the use of AI can perform this task at a faster rate, with much higher accuracy, and
without fati gue.

04 Production Planning
AI-based producti on planning system can help determine the most appropriate producti on line-based on style type,
SMV, previous history data like bott lenecks, delays etc

05 Cutting
Cut order planning is one of the most complex processes involved in garment making. AI can make the planning effi cient and
faster. For example, Jaza Soft ware has developed an AI integrated Cut planning soft ware-  OptaCut which can esti mate fabric
required for ordering almost accurately and makes the cut planning process automated.
 APPLICATIONS OF AI SYSTEMS IN APPAREL MANUFACTURING:

The garment manufacturing process is becoming more automated to cater the

increasing demand of consumers, reduce the number of faults, and keep the
production cost low. AI is increasingly used to predict the performance of a
sewn seam, designing of clothing, in PPC, in various sewing operations, and in
quality control. AI can be applied for intelligent manufacturing of clothing to
predict the clothing properties after a particular process. So it can be used for
suitable garment designing by fabric engineering and monitoring the garment
manufacturing processes.
Performance of sewn seam:
In sewn garments seams and stitches are used to join two
or more pieces of fabric together. The ease of seam
formation and the performance of the seam are the
important parameters, which are judged by the term
known as “sewability.” Fabric low-stress mechanical
properties such as tensile, shear, bending, and surface
can affect the sewability. AI system can be used to find the
sewability of different fabrics during garment production.
Fabric mechanical properties affect the performance
during spreading, cutting, and sewing.

A good quality seam is essential for a good quality

garment. The performance of a sewn seam depends on
the type of fabrics and sewing thread combination; seam
and stitch type; and sewing conditions, which includes
needle size, stitch density, and the sewing machine
condition. The performance properties of the seam are
evaluated by seam puckering, seam slippage, and yarn
severance, which can be predicted by AI.
Computer-aided design systems:
One of the important steps in garment manufacturing is pattern making, where paper patterns
are made by the designers and subsequently digitized to a computer. Several two dimensional
(2D) patterns are prepared for a garment, which are the basic blocks of a 3D garment. Various
CAD software are used in the garment industry for patternmaking, digitizing, grading, and
marker planning. The CAD software helps in achieving high productivity and improved quality.
The designers involved in clothing designing create numerous designs by using the CAD
software.

However, the CAD software cannot be used to automatically generate clothing patterns or

designs for a specific garment style. In addition, in many garment industries the traditional
method of garment pattern generation is still done by experienced designers and does not
include the use of CAD, although there is the scope of using AI in pattern generation.

Several researches have been done to implement the AI that can help to develop basic
clothing patterns automatically. For example, Inui had developed an AI integrated CAD system
(combination apparel CAD and GA) that can be used to search apparel designs that the
system users prefer. The search process involves the man– machine interaction cycles, where
the user assesses the examples produced by the systems.

CAD systems are used in garment manufacturing for creating designs, patternmaking, and
grading operations. Several attempts have been made by researchers to integrate AI with
CAD systems to generate designs automatically. Experienced designers are needed for
appropriate pattern design of different clothing styles. However, the AI system can be used to
provide the expert knowledge of experienced designers.
Production planning and control:

PPC coordinates between various departments of production so that delivery

dates are met and customers’orders are delivered on time. Various research
activities focused on the problems related to PPC and avoid the bottlenecking.
Majority of the studies were based on the problems in PPC relating to sewing
floor such as fixing the machine layout, line balancing in sewing, and
managing operators in the sewing floor. AI can be used to solve or optimize
the problem of machine layout, operation assignment, and sewing line
balancing. This can help in achieving the objectives of PPC.

AI based decision support system was used in decision making to determine

the most appropriate manufacturing plant for a particular customer order. A
GA-based real-time segmentation for rescheduling was developed by Wong to
deal with the PPC-related problems in sewing floor during: (1) marker making,
(2) fabric spreading, (3) cutting, and (4) bundling.

In another study a GA-based system was designed by Guo for production

scheduling for each production order to appropriate assembly lines. This
system was designed by considering various factors for production delay,
production uncertainties (such as processing time, orders, and lead arrival
times), and other bottle necking.
 Final garment inspection:

The application of AI in final garment inspection includes:

automatic classification of general faults in shirt collars (for
mono colored materials) using machine vision; application
of AATCC (American Association of Textile Chemists and
Colorists) wrinkle rating for evaluation of wrinkle by using a
laser sensor; detection and classification of stitching
defects using wavelet transform and BP NN; seam pucker
evaluation by using self-organizing mapping; and designing
of a smart hanger for garment inspection. In manufacturing
the seamless garments, AI can be used to detect faults
online. The image of the final garment can be captured and
compared with the standard and any variation from the
standard is reported as a fault that can be mended at that
time or a marking is done where the fault occurs.

Different fabric defects inspected (arrow indicates defects) by artificial

intelligence: (a) gout, (b) warp float, (c) draw back, (d) hole, (e) dropped
stitches, and (f) press-off
USAGE OF ROBOTICS IN APPAREL INDUSTRY:
Robots are used in many methods and research projects to guide textiles
through the sewing process.

1. Integrated three-dimensional sewing system:

Integrated 3D-sewing system refers to a system where the cut elements are
sewn in a room in a 3D manner. The system was created to sew a car seat
head cushion cover.

2. Three-dimensional sewing with robot for performers:

A special advancement of the last decade has been the one-sided sewing
process, in which the textile is fixed and the sewing head is operated over the
textile. These sewing methods are mainly attractive for composite applications,
especially because of the complex geometric stitching of the material. Since
the movement is driven by the sewing head, very large elements can be sewn
in one-sided sewing techniques and the bottom of the workpiece has a
seamlessly moved gripper or low thread system. The use of robots to guide the
sewing head has been established. Important sewing methods can all be
managed by a robot or sewing portal.

Robot-based sewing systems have already been developed for small-scale production only. Therefore, on the one hand the method of
producing garments and on the other hand the method of producing lots of products (such as headrests, airbags, etc.) have been invented.
There are strict restrictions on its use in various fields of textiles. Further problems for production automation are sewing techniques (e.g.,
double locked stitching, double chain stitching), limited design possibilities, initial and subsequent process steps or handling times. Although
there are several disadvantages before the use of automation and robotics, the days are not far away when the garment industry will be fully
controlled by robots.
 COMPANIES DEVELOPING INTELLIGENT SYSTEMS
 Cognex – Fabric Pattern Inspection
Cognex Corp., founded in Boston in 1981 and with over a 1000 employees today is an American manufacturer of machine vision systems,
software, and sensors. The company offers its purportedly machine vision-based Cognex ViDi platform tailored for fabric pattern recognition in
the textiles industry.

Cognex claims that the Cognex ViDi platform can automatically inspect aspects of fabric patterns such as weaving, knitting, braiding, finishing,
and printing. The company also suggests its 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.
Based on the description provided by Cognex, the product seems to work as follows:

 Textile manufacturers might save on costs and time taken for inspecting the quality of the final fabric end-product by replacing visual inspection with
use the Cognex Vidi platform. Typically the manufacturer might install the camera-based inspection system in their factories and input a few hundred
images of “good” final samples, and “bad” samples (see image above).

 The platform learns the weaving pattern, yarn properties, colors and tolerable imperfections from these images and after a training period of a couple
of weeks and might potentially be able to detect defects (like wrong knitting patterns) in the textile end-product, saving humans from the manual task
of assessing hundreds of yards of material manually.

Below are a few snapshots from Cognex’s brochure illustrating its features, and what kinds of textile defects can potentially be detected by the
machine vision system:
According to Cognex, several challenges are inherent in
inspecting fabric patterns, namely their complexity, variability
and the sheer numbers of fabric types. Reto Wyss,
Computer Science PhD and the CTO the Director of
Software at Cognex was CTO at ViDi for 5 years before the
first was acquired by Cognex.
 Datacolor – AI Tolerancing for Fabric Color Matching

Datacolor, founded in Lucerne, Switzerland in 1970 with over 380 employees offers color management instruments and software.

To ensure that the original design colors match the colors in a finished textile product businesses usually assign a “color tolerance” – a limit to how big
the difference in color between a sample and the requirements of the customer can be, before the sample is considered acceptable. These tolerance
values are generally agreed upon internally by the manufacturer or between supplier and customer to determine if the sample passes or fails
inspection.

While traditional color tolerancing was done based on numeric descriptions of color through ”instrumental tolerancing systems”, that method generally
had a lot of false positives compared with visual inspections, causing delays in the approval process because of the need for careful human
intervention.

Datacolor claims it has developed an artificial intelligence Pass/Fail (P/F) feature to help improve the accuracy and efficiency of instrumental tolerance.

Datacolor suggests that its AI feature can take into account historical data of visual inspection results from human operators while creating the
tolerances that in turn result in instrumental inspections matching more closely the samples of visual inspections.

Datacolor’s AI P/F procedure purportedly works as follows:

 The textile expert first visually reviews all the individual batches that had been manufactured

 The operators enter the color measurements and tolerances for all the batches in the Datacolor software to help train the AI P/F system

 The AI P/F system can then be tested for new batches to automatically set AI tolerances, training the system to determine which samples pass and fail

The snapshot below shows how textile manufacturers might use the platform to set tolerances for a number of manufactured batches for one customer.
The green circle around the center of the graph represent the batches with “ideal” color values, thus passing the test and the yellow circle represent the
acceptable tolerance limits:
In the real-world, this application might benefit both textile
manufacturers and their customers to improve the speed and
accuracy of the inspection processes for color matching.

For example:

 A tablecloth business might buy textile raw materials from a

fabric manufacturer and place an order for a particular type
of fabric in a certain specific color – the color is usually
specified in digital terms as a specific number.

 The fabric is the manufactured in several batches with each

batch varying slightly in terms of actual color obtained due to
variables in the dying process like the amount of color added
or the differences in the color bonding (to the fabric) levels in
each batch.

 Samples from each batch might be quality checked using

Datacolor’s platform to set a tolerance limit. Any samples
lying outside the tolerance limit are rejected (see red dots in
the figure above)

 The manufacturer’s human inspection officers are first tasked with entering visual tolerance readings for each batch
(a few hundreds of samples) into the AI Pass/Fail feature to help identify what samples can be considered good and
what samples need to be rejected.

 The AI P/F feature is then fed with images of the fabric from other batches where it can potentially assign the
tolerance limits automatically by “learning” from the human inspectors – potentially saving significant time and
human effort for the manufacturers by automating tedious color matching tasks.
SOME RESEARCHES IN THE
FIELD OF GARMENT
MANUFACTURING
 Analysis of the modeling methodologies for predicting the sewing
thread consumption

• Purpose – Aims to provide a rapid and accurate method to predict the amount of sewing thread required to
make up a garment.
• Design/methodology/approach – Three modeling methodologies are analyzed in this : theoretical model,
linear regression model and artificial neural network model. The predictive power of each model is evaluated
by comparing the estimated thread consumption with the actual values measured after the unstitching of the
garment with regression coefficient R2 and the root mean square error.
• Findings – Both the regression analysis and neural network can predict the quantity of yarn required to sew a
garment. The obtained results reveal that the neural network gives the best accurate prediction.
• Research limitations/implications – This study is interesting for industrial application, where samples are
taken for different fabrics and garments, thus a large body of data is available.
• Practical implications – This has practical implications in the clothing and other textile-making-up industry.
Unused stocks can be reduced and stock rupture avoided.
• Originality/value – The results can be used by industry to predict the amount of yarn required to sew a
garment, and hence enable a reliable estimation of the garment cost and raw material required.
 
 Two‐stage approach for nesting in two‐dimensional cutting
problems using neural network and simulated annealing

Nesting of two-dimensional patterns on a given raw sheet has applications in a number of industries. It is a
common problem often faced by designers in the shipbuilding, garment making, blanking die design, glass and
wood industries. It presents a new two-stage layout approach for nesting two-dimensional patterns by using the
self-organization assisted layout and simulated annealing. The nesting approach consists of two stages: initial
layout stage and layout improvement stage. This heuristic algorithm generates a 'good' initial layout by using the
self-organization assisted layout (SOAL) algorithm and then improves the layout by using the simulated
annealing (SA) algorithm. Some examples are treated for showing the effectiveness of this approach in nesting
the two-dimensional irregular patterns with and without holes.
 

Two‐stage approach for nesting in two‐dimensional cutting

problems using neural network and simulated annealing

Neural networks are used to predict the performance of fabrics in clothing manufacturing. The
predictions are based on fabric mechanical properties measured on the KES-FB system. The influence
of the number of input and hidden nodes on the convergence speed and the prediction accuracy are
investigated. Tests indicate that these artificial neural networks are effective for predicting potential
problems in clothing manufacturing.
 Expert-based customized pattern-making automation:
Basic patterns

• Purpose –This paper aims to present a flattening method for developing 2D basic patterns from 3D designed
garments. The method incorporates the techniques of professional pattern development for the purpose of
pattern-making automation. The aims of the flattening method are to improve the dressing suitability and to
produce pleasing figures by reversing design procedures .
• Design/methodology/approach – A flattening method is presented in this paper for developing 3D
undevelopable NURBS surfaces in 2D. The automatic operation embeds the expertise of pattern makers by
reducing total area differences between the designed garments in 3D styles and the two-dimensional patterns.
Basic pattern-making invokes the boundary constraints which apply mesh alignments techniques.
• Findings –The global area difference between the original 3D designs and the 2D-developed pattern is
controlled within 5 percent in order to reach the final outcomes of basic patterns, whose shapes are similar to
the drawing patterns currently utilized in the industry.
• Research limitations/implications –This study currently handles simple designs, such as basal designs,
and can only flatten garments in symmetric styles. The direct flattening method is developed by this study. In
addition, this study is supplemented by expert-based knowledge, and it establishes basic boundary conditions
for various garment patterns to increase the feasibility of flattening automation .
• Originality/value –This study introduces the fundamental theories and methodologies used in the automatic
making of basic patterns from 3D garment designs. It proposes a flattening method with pattern expertise
embedded by real-time approximations of the global area of the 3D undevelopable designs to the 2D patterns. 
 FUTURE APPLICATIONS

• On the production side, AI might be applied to detecting

visual defects in shirts or collars, or it may be applied
to automatically detecting and measuring wrinkles on
fabric.
• Measuring wrinkles in fabric material is vital as it
influences and decides the visual aesthetics of a garment.
AATCC (American Association of Textile Chemists and
Colourists) methods are commonly used in measuring
fabric wrinkle performance but the process is tedious –
and trained experts still disagree frequently about the
results. Machine vision-based wrinkle measurement might
help textile manufacturers cut costs and time required for
this process.
• Another theme we might see in the future for AI in textiles includes using machine learning to identify previously hidden patterns from raw data. We might also see textile players
taking a cue from finance or healthcare industry players and adopt “transferrable: data science and data-mining techniques as explained below:

• In the textile industry, the manufacturing of products (such as t-shirts, tablecloths, etc) generates a large amount of data regarding the raw materials used, machine settings for
production, and quality parameters of the product.  Machine learning can potentially enable business to find patterns and correlations between fiber properties, process
parameters, and yarn properties or among yarn properties, machine settings, and fabric performance. This might help businesses of the future discovers relationships that were
previously undiscovered thus aiding in improving efficiency and maintenance. This is very similar to AI applications we see elsewhere in manufacturing, where data about the
production process can be collected to improve outcomes.

• Fabrics play an important role in design and prototyping in many industries, such as the design of upholstery in cars or T-shirt designs. Artificial intelligence might help design
engineers in the textiles industry ‘3D-model’ yarn fibers in their designs and prototypes.

• Traditional methods of modeling fibers are very tedious and simpler procedural models are too regular and not realistic enough in appearance. Researchers from Cornell have
developed an AI algorithm which can model the yarn and fiber properties automatically and realistically without much human intervention.

• In the traditional process, human 3D modeling experts would have to create a yarn from individual ‘virtual fibers’ making it time-consuming and tedious. The research paper from
Cornell claims to have developed a method where images of single strands of yarn are scanned using a CT (Computed Tomography) scanner. An AI algorithm is used to convert
data from the micro CT scan into a 3D fiber model as shown in the figure below.
 REFERENCE LIST

https://www.allerin.com/blog/artificial-intelligence-in-the-
apparel-industry

https://emerj.com/ai-sector-overviews/artificial-intelligence-
in-the-textile-industry-current-and-future-applications/

https://medium.com/@stitchdiary/artificial-intelligence-in-
the-apparel-industry-db0bc3ddbb60
https://www.researchgate.net/publication/258196424_Appl
ications_of_artificial_intelligence_in_the_apparel_industry
_A_review
Automation in Garment Manufacturing by by Rajkishore
Nayak and Rajiv Padhye
Garment Manufacturing Technology by Rajkishore Nayak
Rajiv Padhye
Apparel Manufacturing Technology by T. Karthik, P.
Ganesan, D. Gopalakrishnan
http://stitchdiary.com/artificial-intelligence-apparel-industry/
Thank You