JTATMVolume 1, Issue 1, September 2000

Volume 1, Issue 1, September 2000

RAPID PROTOTYPING IN THE TEXTILE & APPAREL INDUSTRY:A PILOT PROJECT

Cynthia L. IstookNorth Carolina State University

ABSTRACT

Technological advances during the past decade have opened many new doors for the Textile and Apparelindustries, especially in the area of rapid prototyping and related activities. We need to have a betterunderstanding of how some of this technology may benefit the textiles industry, how these developingsystems can be most successfully implemented, and how we can use these systems to the greatest advantageto serve our customers, today and in the future. A project was developed to explore integration of 3-D bodyscanning, CAD/CAM, and digital printing technologies to support prototpying and mass customizationactivities. Patterns altered to fit each specific customer were cut from their digitally printed designs, sewninto the desired garments, and tested for fit. This project allowed us to test the developing rapidprototyping and mass customization paradigms and look for the bugs that are inherent in new technologies.While the process was ultimately successful, several areas were uncovered where the implementation ofthese processes might be problematic for industry.

1 INTRODUCTION

During the past decade, the textile and apparelcomplex has been scrambling to adjust to arapidly changing business environment. Withincreasing imports and rising labor rates, theindustry has seen a drastic change in itsappearance. Businesses have closed andemployment within the industry has decreased,as production of textile products moved to othercountries. Industry leaders have been forced toevaluate this business shift and the ultimateeffect on the consumer to determine ways inwhich the industry might maintain, if not regain,market share. These evaluations have led to thedevelopment of quick response, rapidprototyping, and mass customization strategies.

Quick response (QR) strategies were developedduring the 80s, with the expected benefits to theindustry of reduced overhead, handling costs,product inventories, financial investments, andmarkdowns [1,2,3]. Theoretically, the adoptionof these strategies would allow businesses torespond to consumer demand more quickly thanmass production strategies, because product

decisions and commitments would be madecloser to product entry in the marketplace—weeks rather than months before introduction. Inorder to implement QR strategies mosteffectively, a significant amount of trust andcommunication between members of the supplychain was imperative. The Demand ActivatedManufacturing Architecture (DAMA) projectwas created in 1993 to help the IntegratedTextile Complex (ITC) manage supply chainpartnerships that would support quick responsestrategies (DAMA, 2000) [4].

Unfortunately, the industry as a whole has haddifficulty implementing these strategies, for anumber of reasons. First, the development timerequired for a new textile product was especiallylengthy. New print designs took weeks ormonths to produce the first strike off for samplefabrics. The design and development processwas not only lengthy, it was also extremelycostly—costing almost $30,000 to set up eachdesign, whether the design was actually put intoproduction or not [5]. Second, businesses in theITC were not normally powerful enough to beable to trust that their industry partners would

JTATMVolume 1, Issue 1, September 2000

2

“come through” for them if their order quantitieswere small and placed closer to delivery times.Finally, technologies supporting the ITC did notallow integration of activities or encouragecommunication between channel members. The7 years that the DAMA project was in effectwere spent developing and enablingcommunication streams and educating the entireITC about potential benefits of quick responsestrategies.

A 1997 DAMA survey of mid-sized apparelmanufacturers uncovered a number of issues thatwere believed to impact a firm’s ability toparticipate in quick response strategies. Amongthe issues were two that have significance to thispaper. The first was the need for technologiesthat would help reduce the number or order ofthe product development processes and the timeinvolved [6]. Manufacturers felt that theprocesses were so arduous and prolonged that itwas extremely difficult to meet their customers’needs in a timely manner. Because of expenseand long lead times, they were often forced tomake product development decisions based onsome other company’s perceptions of theircustomer’s wants and needs, rather than theirown. This often contributed to less than uniqueofferings and a significant amount of unsoldstock. Technological advancements wereperceived to be a strong way to affect importantchanges. The second important issue was theneed for better communication and integrationbetween product development systems. The factthat there were many successful CAD/CAMsystems used by many members of the ITC madethe product development process even moredifficult to successfully complete. It had becomeessential that companies be able to communicateexplicitly, regardless of the specific hardware orsystem that each company used.

Information technology and automation are avital part of rapid prototyping because theyconstitute the connection between theconsumer’s wants and needs and the ability of amanufacturer to create the products accordingly,and in a timely manner [7]. Taking advantage oftechnological developments in combination withproximity to the target market is the differentialadvantage that domestically manufacturedproducts will have over foreign produced goods.However, while technology may enable theseefforts, the processes involved are far fromautomatic. A significant amount of “behind thescenes” effort is still required in order to providethe color selection and fit of each garment that

might be requested for sampling or individualcustomers.

2 DEVELOPING TECHNOLOGIES

When evaluating consumers’ dissatisfactionswith apparel that has been mass-produced and isreadily available in the marketplace, there areseveral recurring themes. The first is anoverwhelming criticism about how garments“fit”. While fit is a subjective variable, at best, itis nevertheless a problem that almost everyconsumer has had at some time. In the apparelindustry, accurate measurements are veryimportant as the first step to determine correctsizing and to create garments customized to aspecific target market. Historically, tailors andfashion designers used measuring tapes to obtainthe physical measurements of the bodies theycreated for. This method has been timeconsuming, invasive, and often inaccurate, basedon who took the measurements and how theytook them. Until just recently, only tailors andcouture houses actually still used real bodymeasurements to create or alter the clothing theyproduced. Mass production strategies of the past50 years encouraged the move from garmentsmade to fit to garments made to size.Unfortunately, the sizing systems that havedeveloped through the years are neitherstandardized nor related to the average human’sbody measurements While we know thatgarment sizing can be impacted by productioninaccuracies, we must face the fact that manygarments were sized or proportioned incorrectlyfor the target consumer, from conception [8,9].This specific problem has encouraged thedevelopment of 3-D body scanning technologiesthat will enable rapid and accurate extraction ofindividual consumer’s measurements. Thisability will enable redevelopment of currentsizing systems, as well as production of made-to-measure garments.

Another recurring criticism about garments thatare currently available is that they are “ugly”.This can be translated to mean that the clothingis not as long or short, wide or skinny, tight orloose, etc. as the consumer would like it to be.While this may sound like a fit issue, it is reallymore of a design manipulation issue and is onethat designers/apparel manufacturers labor witheach season. The need to respond more quicklyto customer’s wants has encouraged thecontinuing development of Apparel CADsoftware and systems.

3

A final complaint by many consumers is thatthey don’t like the color, print, or fabrication ofthe garments available in the marketplace. Thelikelihood of finding a garment in the right color,and the right fiber, and the right fabrication isrelatively slim considering the current massproduction processes. Development of digitalprinting technology has occurred to respond tothe desires of the market place and enable shortproduction runs. The efficient integration ofthese technologies will allow manufacturers tocustomize their product offerings to better meetthe needs of their consumers.

2.1 Body Scanning

The development of 3 dimensional body-scanning technologies may have significantpotential for use in the apparel industry, for anumber of reasons. First, this technology has thepotential of obtaining an unlimited number oflinear and non-linear measurements of humanbodies (in addition to other objects) in a matterof seconds. Because an image of the body iscaptured during the scanning process, thelocation and description of the measurements canbe altered as needed in mere seconds, as well.Second, the measurements obtained using thistechnology have the potential of being moreprecise and reproducible than measurementsobtained through the physical measurementprocess. Third, with the availability of aninfinite number of linear and non-linearmeasurements the possibility exists for garmentsto be created to mold to the 3 dimensional shapesof unique human bodies. Finally, the scanningtechnology allows measurements to be obtainedin a digital format that could integrateautomatically into apparel CAD systems withoutthe human intervention that takes additional timeand can introduce error.

Researchers have developed 3-dimensional bodyscanners that can capture the outside surface ofthe human body by using optical techniques, incombination with light sensitive devices, withoutphysical contact with the body. Body scanningsystems consist of one or more light sources, oneor more vision devices, software, a computersystem, and monitor screen in order to visualizethe data capture process. The two basic types ofbody scanning systems are laser and light.

One method uses a laser scanning triangulationprocess to acquire the 3D image [10,11,12,13].Cyberware in Monterey, California built onesuch system. This system uses 8 laser diodes, an

arrangement of mirrors, and cameras within fourscanning heads to record the laser images thatare distorted by the body’s shape. The data fromthe four heads are combined to create a completeimage of the scanned body. The entire processtakes approximately 17 seconds and creates anobject with around 400,000 x, y, z coordinates.

In 1998, the Civilian American and EuropeanSurface Anthropometry Resource program(CAESAR) at Wright-Patterson Air Force Baseinitiated the largest scale anthropometric surveyperformed in over 30 years. It is the firstinternational survey of its kind to utilize body-scanning technology. The Cyberware WB4whole body scanner was used in this study[14,15]. The collected data will be used bymultiple industries, including the military,automotive, and apparel).

AnmeClCaMaphigthecaclocoareneCoprintthegadathaobme

JTATMVolume 1, Issue 1, September 2000

other technology uses white light phaseasurement (PMP) [12,16,17]. Textile

othing Technology Corporation ([TC]2) inry, North Carolina developed the Bodyeasurement System (BMS) using the PMPproach because of the short acquisition time,h accuracy, and relatively low cost. One ofir systems uses the data collected from six

meras (48 images in all) to create a 3D pointud of the body scanned. The initial raw data

nsists of approximately 250,000 points, which many more than the extraction software

eds to obtain accurate body measurements.nsequently, the raw scan data must be further

ocessed to filter stray points, segment the bodyo individual parts (trunk, arms, legs), smooth low level “noise” from the scan, fill small

ps, and compress the data into a less denseta set. The extraction software allows moren 124 pre-defined measurements to betained, as well as the creation of newasurement definitions by the user.

CCyybbeerrwwaarree

JTATMVolume 1, Issue 1, September 2000

4

Wicks and Wilson has developed scanningtechnology that allows the creation of a 3-Dcolor picture of the object being scanned [12,18].The premise behind this technology is that moreaccurate measurements might be obtained if theresearcher could “see” actual points on the body.This is related to the fact that all of the scanningsystems have problems identifying keymeasurement points (i.e. acromium, knee, wristbones, etc.) on a human body when faced with250,000 or more points in a point cloud.

Many of these systems (such as, Cyberware,Telmat, Tecmath, and (TC)2) claim the ability tointegrate their data into some of the importantapparel CAD systems (i.e. Gerber, Lectra,Assyst, PAD). Ideally, these systems should beable to work together much like Microsoft’sExcel and Word software programs. At this pointin time, however, no integration occursautomatically between commercially availablescanning and apparel CAD systems.

2.2 Apparel CAD

Adoption of CAD/CAM technology over thepast few decades has increased the speed andaccuracy of developing new products, reducingthe manpower required to complete thedevelopment process. Unfortunately, thistechnology has also encouraged manufacturers tosimplify the design of garments, allowing a moreefficient use of materials and making massproduction much easier [19]. These systemsinitially only made an effort to adapt traditionalmanual methods instead of encouraginginnovation in design or fit adaptations. Currentdevelopments in the area of informationtechnology help build on the traditionalCAD/CAM functions and offer a new way oflooking at and using the systems for design andproduct development [20].

2.2.1 Assyst Bullmer

Assyst Bullmer has an array of products thathave been developed to support rapid productcreation. Cad.assyst is an apparel patterndevelopment and modification software.Strengths of this software include ease of use andincreased productivity. All of the mostimportant and routine tasks require minimumsteps and mouse clicks. Many pattern pieces canbe worked on simultaneously, withoutrestrictions. A new feature in this module ispattern dependency that allows any changesmade on a source pattern to also occur on thepatterns created from it. Macros can be easilycreated for repetitive tasks, such as the additionof seam allowances, the creation of facings, andthe placement of buttons and buttonholes. Pieceand grading information can be imported from

TTCC22 WWhhoollee BBooddyy SSccaannnneerr

WWiicckkss && WWiillssoonn

TTeeccmmaatthh

JTATMVolume 1, Issue 1, September 2000

5

Gerber, Lectra, PAD, and anything else that canexport using the AAMA .dxf file format.

Smart Pattern is a new product scheduled to bereleased in October, which is a modification toolrather than a construction tool. The basis of thissoftware is to automate work done repeatedly bycreating macros that are organized by type ofactivity. Compound macros can also be createdto work together. The system will come with500 pre-defined modules that can be used toeasily create a background macro. Aftertraining, the user will be able to create their ownmodules for activities not initially covered.

Leonardo, a system due to be released in 2001, isa very new and unique way of managing made-to-measure or custom products. The system ismade up of 3 parts that currently appear to bevery involved and time consuming. Basically,the software is being developed to recreate aspecific pattern (or solve a problem) based on themeasurements that are defined each time.Currently, patterns being developed initially inthis system require more time than in thecad.assyst system. The benefit to be derived willbe a significant reduction in time and effort forrepeated garments made from differentcombinations of body measurements. Thissoftware is a very unique departure from thecomplicated alteration processes employed bymost of the competition’s systems in their masscustomization or Made-to-Measure strategies.2.2.2 Gerber Technology

Gerber Technology’s PDS 2000 allows the userto move multiple lines at once, in the samedirection, or in reverse, as desired. The systemcan also remember where pieces were on thescreen when they were saved so that they canreturn to that position when recalled. Multiplepieces can be selected at one time using themarquis function. Pieces can also come into thework area in a full-scale view. This allowspatterns to be created or refined in less time,which in turn increases profitability.

APDS 3D is the software from Asahi that allowsa virtual try-on of garments created in PDS 2000.This system comes with a variety of dressforms—men’s, ladies’, children’s, Juniors, andwith legs—all of which can be altered by theuser, to some degree. An ease table is built intothe system, which allows fabric drape to bedemonstrated based on Kawabata values.Essentially, this software allows the user to drape2D patterns on-screen and make pattern

revisions, check new designs and graded patternson a virtual drape model, and input fabric designto create a virtual sample.

Artworks Studio is a design, merchandising,style development, and illustration system thatallows the user to easily create the artworkneeded to design, manufacture, and illustrate avariety of products. This system provides thepower to enhance and communicate ideas fasterand more efficiently than ever before. Since it isa modular system, the user can customize apackage that would best meet their needs.

Web PDM is a 3-tier system that is databaseindependent. Customers can choose the databasethey want (Oracle, Sequel, Informax, etc.) basedon what they may already use. It also allowsinterface with SAP, JBA, and other ERPsystems. Database entry is easier with thissystem since the user is allowed to enter theinformation directly into the datatbase or in theform where the information will be applied.Datacolor and Gretag Macbeth systems can beused to submit color calibration information,providing more precise product specificationsenabling more positive results in the productmanufacturing process.

Web PDM has style files to provide informationspecific to one specific style. It also has RangePlans to coordinate styles within a season. Thereis an Accumark interface that is linked toautomatically update information as Accumarkpattern files change. The retailer interface isthrough the Made to Measure (MTM) softwarein the classic system. Internet access is allowedwithout software support on the system usingpassword controls. This allows read only accessfor customers or other constituents for all distantlocations.Web PDM has a 3D costing module based onsize, fit, and color. Wizards help set up theforms to meet the needs of the users, originally.Labor forms within the system, interface with

JTATMVolume 1, Issue 1, September 2000

6

GSD. The Standard Costing Application (SCA)works within the forms created using the Wizardinterface.

Made to Measure (MTM) allows options to bedeveloped or determined for each garment in thesystem. These decisions are set-up in the modeleditor. Alteration rules are developed tocoordinate actions to be done to each garment forwhich access is allowed through MTM. A batchprocess runs all MTM actions, generates markerorders, submits or process the marker order,makes the marker, and creates either a plot file ora cut file. These actions occur at themanufacture level through directions in MTM,possibly at the retail level.

2.2.3 Investronica

INVESTRONICA SISTEMAS has introduced a3D system from an alliance with Toyobo calledDressing Sim FDK. This system acts as a virtualtry-on of garments created through the PatternGeneration System (PGS). The user will importa body form, import a pattern created in PGS,and import pattern characteristics. After liningeach garment piece up to the body, it can beevaluated from each side of the body. Onceproperly aligned, the garment will be virtuallyseamed together. Fabric variations can beapplied to simulate fabric drape and relaxed toconform to the body. This system is still indevelopmental stages and requires some time tooperate correctly in order to get the most correctvirtual representation of the intended garment.Knowledge of flat pattern engineering, as well asdraping, would be important for successfuloperation.

Another new product is the 3D-Design system,which will be released at the Milan show inMarch 2001. This system will come with 50 to60 body shapes to cover a wide range of ages,sizes, and shapes of people. The user will alsohave the ability to change the measurements ofthe key dimensions, as desired.INVESTRONICA handles the issue of ease intheir Body Garment 3D Design system byallowing the user to define the specific amountof ease desired in specific locations. Once theease layer has been defined, the designer will beable to draw design lines on the 3 dimensionalform. These lines will then be used to identifypattern shapes that will be flattened to 2D forexport into their PGS system. The PGS systemwill continue to be used as the platform for the

addition of seam allowances and otherproduction features.

The INVESTRONICA Made to Measure (MTM)system works much the same as other systems onthe market. All the pattern pieces have to beprepared for alteration and alteration tablescreated to guide the alteration process. Thesetables can then be applied to all garments in thedatabase, if prepared adequately andappropriately in advance. This module enhancesthe process of receiving orders for garments fromthe point of sale. Web MTM is a tool developedfor retailers to use to communicate customizationof manufactured garment styles.

InvesPM is the specification package that linkswith the PGS system. This system organizes allthe product information generated during thedesign cycle. Users define garment attributes,such as the season style, category, brand, line,base size, size offerings, etc. User variations areallowed in the Bill of Materials and informationabout specific materials can be added “on thefly” to the materials database from with the Billof Materials page.

The WebPM is a product management andcoordination system that brings together all ofthe information associated to a given product,coordinating the various departments andcompanies that participate in the developmentand production of a specific product. It can belinked to many management or enterpriseresource-planning systems for import and export.A catalog of garment categories comes up forevery authorized user that accesses the web site.This allows access to general data, measurementspecs, piece lists, and marker descriptions,.

2.2.4 Lectra

LECTRA Systemes had a number of newproducts to debut at IMB 2000. Catalogue wasintroduced to make the sales/ordering process ofcustomized goods more easily controlled andcommunicated. The retailer who offers amanufacturer’s products for sale would use thesystem. A database containing all of theavailable styles and fabrications can be managedby Catalogue, allowing the consumer to definefabrication for each garment component. FitNetwill take the product defined in Catalogue andcreate a customer order for it. Customerinformation is entered in or pulled from acustomer database, critical measurements may beentered in (from manual measurements) or

JTATMVolume 1, Issue 1, September 2000

7

imported in from Tecmath 3-D body scanningsoftware, and an order created and sent to themanufacturer over the Internet. Once thisinformation is received at the manufacturer’ssite, a marker can automatically be made for thatspecific consumer.

Another system targeted to the Retailer was 3-DVisual Merchant (3DVM). This package can beplaced on a stand-alone system or be managedusing a server, allowing access to manymerchandisers within a store. It will allowmerchandisers to design their floor layouts,complete with virtual garments in this season’scolors, before they ever pull the merchandise outof the box. There is a catalogue of 150 genericfixture, lighting, color, and apparel choicesincluded with the system, but the user can alsoadd, create, or define their own. This productwill not only allow retailers to more easilycommunicate their merchandising wishes to alltheir branch stores, it will also enable retailers tokeep their storefronts fresh and new for theconsumer in a more creative and/or pleasingway.

A new version of Graphic Spec wasdemonstrated at IMB 2000. This softwarestrengthens the creation of technical drawingsthat support specification packages created forproducts going to production. The files createdin this package can be imported into Gallery,LECTRA’s new specification software package.Design details, including silhouette, fabrication,and color choices are included, as well as sizespecifications, sewing instructions (orconstruction details), bill of materials, bill oflabor, costing, pattern thumbnails, markerlayouts, attachments (cut files, video files, faxes,etc.) and work in progress. The pattern andmarker files are linked to Modaris and Diaminoso that any future changes in either file willautomatically occur in the Gallery file, as well.

The Gallery software allows many users to beworking in the file at one time, as long as theyeach are working in a different section of thepackage. The work in progress section keepstrack of every change that occurs and identifiesthe person who made the change. Technicalreports can be custom created by the user, if the50 included with the package are not enough.

Yet another new Lectra product to debut at IMBwas Virtual Garment. Using special glasses, theuser is able to see how a garment might fit anddrape on the wearer. It also allows modification

of the fabric design and color, so that a morepleasing garment might be developed.

Modaris and Diamino, LECTRA’s pattern designand marker making software, also had some newcapabilities. Modaris now has Macro capability,which allows the system to memorize certaintasks that are done over and over again. Themacros let the user be involved with certain partsof each process (such as determining the amountof seam allowance to add), without having torepeatedly select an assortment of tools.Diamino has expanded marker-makingcapabilities, with a new “shake” utility thattweaks an automatically created marker toimprove efficiency.

2.2.5 PAD System Technologies

PAD SYSTEM TECHNOLOGIES introducedtheir PAD System 4.0 software that should beavailable in the fall. Users will see a new“cloning” tool that allows any changes made to amaster piece to also occur in the pieces clonedfrom that master. For example, any blouse,dress, or jacket that was cloned from one masterblock will automatically receive the changesmade to the master block.

Within the pattern design system, each piece canbe colored according to its use, as determined bythe user. All pieces to be made from the fashionfabric could be blue, from the lining might beyellow, and from interlining might be gray. Andsince each of these might be cloned from theoriginal, any changes to the neckline edge of themain “blue” piece, will also occur at thecorresponding point on the “yellow” lining pieceand the “gray” interlining piece. Frames havebeen updated to also affect changes on clonedpieces.

Another new functionality includes the ability towork on pieces, while they are in grade view.The user can now see the effects of any changesin all sizes, at the same time. Technical drawingsare viewable at every stage of patterndevelopment and grading, as well. In addition,plan, piece, and grade windows are all available.

3D Sample software allows a 2 dimensional setof patterns to be placed on a 3D form that is thecorrect size for a virtual try-on of the virtuallysewn garment. The user can also create atechnical drawing of the garment, export it toAdobe Illustrator and/or a pattern file. If the

JTATMVolume 1, Issue 1, September 2000

8

garment is too small to fit correctly, body partswill show through.

PAD SYSTEMS Manager allows the user tocreate specification packages for productscreated with PAD SYSTEMS Pattern. Technicaldrawings can be created in Adobe Illustrator, inaddition to other similar drawing packages andimported into Manager.

2.2.6 CAD Alterations of Developed Patterns

Almost every CAD system used in apparelpattern making has some method that enablespattern alterations based on individualmeasurements. Although each has created aninterface somewhat different from all of theothers, most (Gerber Technologies, LectraSystems, Investronica, & Assyst) have threepreparational activities in common that willallow “automatic” alterations to occur. Twoother systems (PAD and Scanvec) haveattempted to enable automatic alterations in aslightly different manner; however, the basicunderlying theory is the same. Thesepreparational activities are laborious in thebeginning, but ultimately allow the automaticalteration of existing garment patterns. Thisset of activities requires a strong knowledge ofgarment design, grading, and garmentconstruction, as well as an understanding of howcomputer programs "think”.

The fact that this process continues to becumbersome and time consuming at best, willimpact the creativity involved in garmentsdesigned for mass customization. Whiledevelopers have been working to integrate themeasurements extracted from 3-dimensionalbody scanners [16], the process is still far fromautomatic and often provides less than desirableresults.

2.3 Digital Printing

Initial developments in ink-jet technology weredirected at improving the decision-makingprocess and the cost of sampling involved in thetextile mass production operation [21]. Sincethat time, key players in the textile and apparelindustry have begun to think outside the box andimagine ways in which this technology might beused to meet consumer needs, not justmanufacturing needs [7]. Continuedimprovements in the areas of ink-drop formation,pigmented or dye-based ink formulations, colorcontrol, pre- and post-treatment methods, and

fabric handling systems will be important if masscustomization strategies are to be implementedand allowed to succeed at the highest level.Ink-jet technologies are of two basic types: drop-

on-demand and continuous [21,22]. The drop-on-demand method produces ink droplets onlywhen and where required. The drops can beproduced using thermal “bubbles” or themechanical shocks created with piezoelectrictechnology. Encad printers use thermaltechnology and Mimaki printers use piezoelectrictechnology. Both of these technologies aresignificantly slower, yet more affordable, thancontinuous ink jet. Challenges for drop-on-demand technology continue to be related toincreasing the production speed of the image,increasing the number of colors that can beproduced, and solving problems with print headdeterioration.

The continuous ink-jet method involves the useof a highly pressured, continuous stream of ink,which has been forced through a microscopicnozzle. Tiny drops are created from this streamby high frequency stimulation and directed to thedesired location by a variable electric charge[21,22]. This technology allows relatively high-speed production of digitally printed fabric.Challenges still to be overcome include reducing

EEnnccaadd----TThheerrmmaall

SSttoorrkk AAmmbbeerr----PPiieezzoo

JTATMVolume 1, Issue 1, September 2000

9

the high cost of the technology, management ofrecirculated inks, and the need for extremelypurified and expensive dyes. Stork’s Amethystprinter is an example of this technology.

At this point in time, most of the companies thatsupply ink-jet printers are only willing to warrantthe quality of the image created using theirprinters if you use their pre-treated fabrics, theirinks, and their post-treatment (usually steaming)equipment and methods. This requirementplaces extreme limitations on the productdevelopment process, as a whole, and also makesit cost prohibitive, in many cases. The futuresuccess of digital printing technology will bebased on the cooperative efforts of hardwaredevelopers, software developers, pigment anddye-based ink developers, and textilemanufacturers and finishers.

3 PILOT STUDY

During a summer research program held at theuniversity over two summers, seven separatecustomization studies were done. Each studylasted for a period of 4 days, 4 hours a day, andincluded five high school aged subjects (17 to 18years of age). Most of the subjects hadexpressed an interest in apparel management,which was why they were included in our study.None of the subjects had any apparel design orproduction background or experience. Two tothree subjects had an artistic background andwere fairly proficient using graphical designprograms on the computer.

The goal for each week was to enable eachsubject to create and physically produce agarment that would fit their specific body andthat was made from their choice of fabric/design,within limitations. The number of garmentdesigns they could choose between, the fabric

selection, and the computer software availablefor use limited the subjects.

3.1 Preparation

Mass customization is based on the idea ofcustomizing previously designed products. Bydefinition, this places limitations on the designprocess and the ultimate consumer. RapidPrototyping, on the other hand, is the ability toconceive a design idea and see a physicalcreation of that idea in a very short period oftime. This physical prototype needs to becreated out of the materials that would be usedwhen retailing the product. Ideally, this abilityto actually see a physical representation of aproduct idea in a very short period of time,would reduce the overall development time andallow better decisions to be made more quickly.To more accurately imitate the masscustomization paradigm and to make rapidprototyping a realistic possibility with untrained“designers”, a small “line” was designed andtested out with sample garments.

3.1.1 Apparel Design

3.1.1.1 Product Development

The participants selected to be part of this projectwere males and females, 17 to 18 years of age.Because of this, our initial product developmentactivities centered on determining apparel stylesthat would appeal to this market. We usedtraditional information gathering techniques todetermine the kinds of garments that might bethe most desirable. The development processwas impacted by two limitations. The first wasthe fact that our subjects had little or no apparelproduction experience and they were to actuallyproduce the garments. The second limitationwas the fact that there were only four days inwhich to complete the project.

Six total garments were designed using Gerber’sPattern Design System (PDS); three for females(Garments A, B, & C) and three for males(Garments D, E, & F). Garment A was an aboveknee-length fitted shift with waistline fittingdarts, bust darts, and a square neckline. GarmentB was a straight sleeveless shift, above knee-length, with a v-neckline. Garment C was aplain pair of drawstring shorts. Garment D was abasic tank type shirt. Garment E was a pair ofdrawstring shorts and Garment F was a pair ofdrawstring pants. Sample garments were made

Sttoorrkk AAmmeetthhyysstt----CCoonnttiinnuuoouuss

JTATMVolume 1, Issue 1, September 2000

10

and displayed for the subjects to see and try on,if desired.

3.1.1.2 Grading

As mentioned previously, most of the apparelCAD programs enable previously createdgarments to be altered, based on the sizinggrades that have been applied to the garments.Grade rule tables were created in Gerber’sAccumark system. Based on our expected targetcustomers, we chose to grade Garments A-C inMissy sizes 2-14. Garments E & F were gradedin Men’s sizes 28 through 42 and Garment Dwas graded as Men’s S, M, L, XL.

3.1.1.3 Alteration Rules

An alteration rule table was created within theGerber Accumark system for each of thegarments in our line. This table was createdaccording to how the structural lines of thegarment should move based on the differencebetween actual body measurements of thesubject and the physical measurement uponwhich the sizing grade was based at keylocations. Key measurement locations for thefitted shift included the bust, waist, hip, back

waist length, and waist height. Alteration ruleswere created much like grade rules in that thegrading movement was significant to theorientation of the pattern piece on the computerscreen.

3.1.1.4 Prototype Test

Once each of the garments in the line weregraded and prepared for alteration, the accuracyof the alteration decisions was tested out. To dothis, we three dimensionally scanned the body ofa test subject, extracted the key measurementsneeded for each garment, and physically insertedthose measurements within the Size Code tablesin the Gerber Accumark System. A customizedmarker was ordered and made for each garmentusing the Grade Rule, Alteration, and Size Codetables developed. These garments were then fittested on our test subject.

3.1.2 Textile Design3.1.2.1 Motif library

A textile designer worked to develop a library ofmotifs that could be used by our subject“consumers” to create their own designs. Motifswere selected that might appeal to a wide varietyof people. They were organized by subject andwere created in a variety of colorways.

3.1.2.2 Textile Design Layout

The designer also created a number of designlayouts using the motifs and colorwayspreviously created. Examples of border prints,overall prints, and complementary prints werecreated so that the subject “consumers” could beled or “helped” when making their textile designdecisions for their own garments.

JTATMVolume 1, Issue 1, September 2000

11

3.1.2.3 Samples printed

Samples of most of the designs were printedusing the Encad TX1500 Digital Printer andreactive inks. The samples were steamed,washed, and dried so that our “consumers” couldsee the end result of the process. This was alsoimportant, because the appearance of the designon the computer screen had a significantlydifferent color appearance than the final digitallyprinted fabric.

3.1.3 Fabric Pretreatment

The feasibility of using digitally printed fabricsin mass customization efforts will be based, inpart, on the availability of affordable, pre-treatedfabrics. Because of this we were encouraged todevelop and test our own pretreatment methodfor the selected fabrics, rather than purchasingthe pretreated fabrics from the printer or inkvendors.

3.1.3.1 Treatment selection

A treatment formula was developed based onscreen-printing experience and trial and error. Amixture of urea, sodium bicarbonate, and sodiumalginate was created to pad the fabrics selectedfor use. This formula was consideredappropriate because it was similar to the onesuggested by Encad and was being prepared forcotton fabrics and reactive inks. Our goal was topretreat the fabric so that image quality would behigh (little diffusion or bleeding of the inks), theimage would be colorfast, and there would be noneed for a stabilizing, paper backing for thefabric.

3.1.3.2 Fabric Padding

A100% cotton twill and a 100% cotton knit werechosen for use in this project. Approximately 35yards of the twill and 10 yards of the knit werepadded using the Xorella autoclave/steamer thepilot lab at the university. The speed andpressure of the machine were set to allow about70% wet pickup. The fabric was draped andallowed to dry naturally. We did not run thefabrics through the tenter frame because of therelatively small lengths being padded and theexpense related to that effort.

3.1.3.3 Printing Preparation

Once the fabrics had been padded and dried, theywere cut down to 59” widths so that they wouldwork in the Encad printer. Ten yard lengths ofthe fabrics were then rolled to prepare forprinting.

3.2 Subjects

There were five subjects in each of the sevenone-week research sessions. The first, third,fifth, and last week each had one male; all of theothers were female. All of the subjects were inthe senior (final) year of high school and wereinvolved in the summer research program as partof their search for a college and a major area ofstudy.

3.3 Procedures

On the first day of each session the studentswere: a) introduced to the goals of the projectand the idea of mass customization, b) safetytrained in the apparel design lab, c) physicallymeasured, d) three dimensionally body scannedusing the (TC)2 BMS system, and e) introducedto the apparel design and textile design decisionsthat would need to be made the next day. Thesecond day of the project was spent selecting thegarment design, making minor adjustments to

get the look the student desired (changingnecklines and lengths, adding slits and pockets,etc.), developing the size code tables based ontheir scanned measurements, and creating thedesired textile design. A CAD specialist inapparel design or textile design assisted thestudents in each of these steps, although theywere allowed and encouraged to do their ownwork, if they felt capable or inclined.

The students spent the third day of the sessionprinting their fabric designs and steaming,washing, drying, and preparing the fabric forcutting.

and preparing a presentation to be made to theother research groups.

3.4 Outcome

At the end of each session, every student hadproduced a garment. Regardless of anythingelse, the fact that the students survived theintensive product development process andlearned some rudimentary sewing skills qualifiedthe project as a success. In our view, however,the sessions had varying levels of success.

Session one was plagued with a variety oflogistics problems that were difficult toovercome. The digital printer developed inkdelivery problems (air in the lines and cloggedheads) that required a significant amount of timeto correct and yards of fabric re-prints beforeeach student’s fabric was successfully printed.The Xorella steamer had to be engineered tohandle the larger quantity of digitally printedfabric, which was also time consuming. Thesetwo situations caused a reduction in the timeinitially planned for garment cutting andproduction. It wasn’t until the garments werebeing sewn together that we discovered an error

JTATMVolume 1, Issue 1, September 2000

12

On the fourth day, each student’s customizedgarment marker was cut from his or her fabric,using the Gerber Cutting Edge Cutter. The restof the day was spent constructing their garments

in the Alteration Rules that were developedbefore the sessions began. This error had notbeen apparent during the testing phase becauseour test model did not have a significant fitproblem. At that point it was too late to fix theproblem. We were able to adapt the design ofthe garments that had a problem, however, sothat each would be wearable.

Session two had a group of students with a widervariety of fit problems. Our initial Missy GradeRule had to be adjusted to accommodate more ofthe larger sizes, and was revised to include up toa size 22. We had none of the printing orsteaming problems that we had experienced theweek before and found ourselves with enoughtime on our hands to have the students pad out

JTATMVolume 1, Issue 1, September 2000

13

some more fabric to be used in the next session.(This was necessary because of the excess fabricused during the printer problems in the firstsession.) Students successfully completed theirgarments. Some of the were disappointed,however, in the final length of their garment.This was not a problem that occurred because oferrors in the Alteration Rules, but rather becausethe students were unaware of how the garment fitthe fit model who was used for measurementcomparisons.

Sessions three and six were the most successfulsessions. We had repaired the problems with theAlteration Rules and the digital printer that we’dhad the previous sessions. We did have aproblem, however, with the quality of thedigitally printed image and the colorfastness ofthat image. We ultimately determined that thisproblem occurred because we had used fabricpadded some time before, which may not havebeen stable. The outcome was that blacksbecame brown and rich burgundies became pink.While the garments were still beautiful by mostobservers’ standards, they weren’t what thestudents had in mind when they developed theirtextile designs.

4 CONCLUSIONS

We learned two significant things from this pilotproject. The first was that there is a possibilityfor successful use of mass customizationstrategies that may benefit the textile and apparelindustry. The second, and possibly mostimportant thing we learned was that there is agreat deal of work yet to be done to enable thatto happen. There are some significant issueswith current technology that must be addressedbefore these technologies will actually beattractive to industry.

Each of these technologies, alone, is amazingand powerful and provides benefits to theindustry, in some way. For mass customizationstrategies to be profitable and successful,however, these technologies must all worktogether, seamlessly. Critical developmentsmust occur in the areas of: a) integration between3D body scanned measurements and apparelCAD systems; b) software that will allowautomatic alteration of patterns according toscanned measurements; c) software that willallow the integration between apparel CADsystems (vector programs) and textile CADsystems (raster programs) so that the textiledesign can be more creatively applied to the

apparel design; d) inks that might reduce the pre-and post-treatment requirements of the fabrics;and e) hardware that will allow faster printing,higher quality printing, and better control of thefabric. Ultimately, success will depend on theability of many people to work towards acommon goal, in complementary andcooperative ways.

ACKNOWLEDGEMENTS

The author would like to thank the NationalTextile Center for their support of this research,as well as graduate students Karla Simmons,Lashawnda McKinnon, and Lisa Parrillo-Chapman.

REFERENCES [1] Bianchi, A. “Quick-response apparel.”

Inc. Magazine, 15(11), 35.

[2] Dickerson, K. Textiles and Apparel in theInternational Economy, 186, MacMillan,1991.

[3] Glock, R & Kunz, G. ApparelManufacturing Sewn Product Analysis, 2nd

Edition, 22-23, Prentice-Hall, 1995.

[4] DAMA. 1996 DAMA Annual Report.DAMA-G-29-96, May, 1996.

[5] Tortora, P. “Textile printing and design.”Understanding Textiles, 426-453, NY:Macmillan Publishing ,1992.

[6] Wimple, C. & Vosti, E. “CurrentInformation Technology Needs of Small toMedium Sized Apparel Manufacturers andContractors.” DAMA Enterprise Modelingand Simulation, UCRL-ID-129187,DAMA-G-2-98, November, 1997.Available online athttp://tc2.sandia.gov/needs.html.

[7] Little, T. “It’s a CAD, CAD World.”WSA, 13-15, Winter 1997.

[8] Anderson, L.J., Brannon, E.L, Ulrich, A.B.,Jenkins, A.B., Early, J., Grasso, M., &Gray, S. “Understanding FittingPreferences of Female consumers:Development of an Expert system toEnhance Accurate Sizing Selection.”Available online at http//:www.humsci.auburn.edu/ca/html/andentc1.html, July 20, 1999.

JTATMVolume 1, Issue 1, September 2000

14

[9] Kaufmann, K. “Invasion of the BodyScanners.” Circuits & Devices, 12-17,May, 1997.

[10] Paquette, S. “3D Scanning in ApparelDesign and Human Engineering.” IEEEComputer Graphics and Applications, 11-14, September 1996.

[11] Daanen, H., Taylor, S. E., Brunsman, M.A., & Nurre, J. H. “Absolute Accuracy ofthe Cyberware WB4 Whole BodyScanner.” SPIE, Vol. 3023, 6-12, 1997.

[12] Bruner, D. “Application Of BodyScanning Technology.” Bobbin ShowSeminar, Atlanta, GA, October 1, 1999.

[13] Addleman, S. “Whole-Body 3D Scannerand Scan Data Report.” SPIE, Vol. 3023,2-5, 1997.

[14] Potts, L. A. “Government business join tomeasure the human race.” Detroit News,July 30, 1997.

[15] CAESAR. Available:

http://www.sae.org/technicalcommittees/caelaser.htm, December 10,1999.

[16] [TC]2 . “An Introduction to the BodyMeasurement System for Mass CustomizedClothing.” Available online athttp://www.tc2.com/RD/RDBody.htm,May 3, 1999.

[17] Demers, M., Hurley, J.D. & Wulpern, R.C.“Three Dimensional Surface Capture forBody Measurement using ProjectedSinusoidal Patterns.” SPIE, Vol. 3023, 13-25, 1997.

[18] Gray, S. Personal communication, atNottingham Trent, UK, June 9, 1999.

[19] Bye, E. & LaBat K. “Technology: Shapingthe Aesthetic Product”. In Delong, M. &Fiore, A.M. (Eds.) Aesthetics of Textilesand Clothing: Advancing Multi-Disciplinary Perspectives, 28-38, 1994.

[20] Bye, E. “The Influence of Technology onthe Apparel Product.” Proceedings of theWorld Congress on Textiles in theMillennium, 249-255, July 6-7, 1999.

[21] Stefanini, J. “Ink-jet Technology forTextile Printing.” Melliand-Textilberichte,English, Vol. 78 (1-2), E16-E18,January/February 1997.

[22] Fagerquist, R. “Jet, Wave, and DropletVelocities for continuous fluid Jet.”Journal of Imaging Science andTechnology, Vol. 40 (5), 405-411,September/October 1996.

AUTHOR’S INFORMATION

Dr. Cynthia L. IstookCollege of TextilesP. O. Box 8301Raleigh, North Carolina, USA 27695-8301Telephone: (919) 515-6584FAX: (919) 515-3733Cynthia_Istook@ncsu.edu