Communication of pattern data between different CAD/CAM systems is one of the most frustrating problems facing apparel and sewn products firms. But efforts are under way to resolve the complex issue says Kurt Chang of 3D Custom Fit.

One of the major challenges in today's global business-to-business (B2B) e-commerce world is the ability to share and exchange information between computer applications. For the sewn products industry, CAD/CAM data compatibility issues have become especially critical, as more companies increase their outsourcing programs and expand their market share through consolidations and acquisitions.

In such an environment, it is essential that the industry establishes standard methods for presenting CAD/CAM data, in formats that can be easily understood and used by different computer applications. Such standards would allow apparel and sewn product producers to use a choice of machinery and software from different vendors, and to exchange information with contractors and third parties that may not have the same machinery and software.

Richard Downham, customer support manager for Assyst-Bullmer, summed the current situation up well in a presentation earlier this year to a gathering of CAD/CAM users and vendors. If a contractor has different systems than its customer (the producer), the contractor may be forced to re-digitise pattern pieces, buy compatible systems or establish customized interfaces to get the data into its systems, Downham emphasised. At the same time, producers currently are limited in their ability to switch contracting partners if their existing contractors with compatible systems are not performing well.

Moreover, some CAD/CAM vendors provide not only systems for design, pattern making and marker making but also machinery for plotting, spreading and cutting. Generally speaking, a sewn product producer must purchase systems and machinery for all of these functions from the same vendor in order to be guaranteed full connectivity and automation of the processes. This is not in the best interest of the producer, Downham said.

The good news is that significant efforts toward CAD/CAM standardisation are under way, and many participants in the vendor and end-user communities have joined forces to both develop new standards and improve existing ones under the guidance of the American Society for Testing and Materials (ASTM). This article provides background on some of the complex issues of CAD/CAM standardisation, and outlines the latest actions by industry participants working with the ASTM to resolve them.

The ASTM CAD/CAM connection
ASTM is one of the largest voluntary standards development organisations in the world, with 32,000 members from different industries, including producers, consumers, machinery and software vendors and representatives of government and academia. It is a not-for-profit organization that provides a forum for the development and publication of voluntary consensus standards for materials, products, systems and services.

ASTM has more than 130 main committees, including one focused on the sewn products industry: the (D13) Committee "On Textiles." Members of this and all ASTM committees and sub-committees are volunteers.

The D13 committee includes the "Apparel and Sewn Product Automation" (D13.66) sub-committee, which is chaired by Jim Hiegel, president of Blue Sky Ltd and former assistant director and general manager of research and development for Levi Strauss. Within D13.66, there is a task force charged with improving the existing American National Standards Institute (ANSI)/American Apparel Manufacturers Association (AAMA) 292 standard, which is chaired by author Kurt Chang, founder of 3D Custom Fit. This task force is working to clarifying the standards definition for pattern data interchange between CAD systems. (The AAMA is now known as the American Apparel and Footwear Association, or AAFA.)

Determining that the ASTM standards development process was faster and more flexible than ANSI/AAMA methods, the AAMA Apparel Research Committee transferred the sewn products standards development efforts to ASTM.

The D13.66 task force met for three days this past July in Philadelphia, PA, bringing together a variety of representatives of the sewn products industry to discuss issues concerning the AAMA/ANSI 292 standard. The task force also worked to determine the best course of action for continuing the standards development, and to identify specific areas to be addressed.

The open forum meeting was well attended by representatives of both the CAD/CAM vendor and end user communities, including industry participants from: (vendor firms) Assyst-Bullmer, Gerber Technology, Investronica Sistemas, Lectra Systems, Nester Software Technologies, PAD System Technologies, Scanvec Garment Systems (SGS) - Optitex and Wild Ginger Software; and (end user firms) Blue Sky Ltd, Johnson Controls, Lands' End, Levi Strauss, The Limited, Seton, the US Department of Defense, North Carolina State University, Digibits Interactive and 3D Custom Fit.

ANSI/AAMA 292 DFX: what works, what doesn't?
Before taking a closer look at the task force's most recent work, it's important to understand some of the pros and cons of the original ANSI/AAMA 292 pattern data interchange (or ANSI/AAMA 292 DXF) standard.

First approved by ANSI/AAMA in September 1993, 292 DXF was a good first attempt to define a standard. Its development received good participation from industry CAD/CAM vendors, and it was relatively easy to implement.

However, the standard was poorly defined in some respects and lacked meaningful definitions for many pattern data processes, properties and attributes. The result has been confusion, frustration and inconsistent implementation by sewn products industry CAD/CAM vendors.

As Neal Goldberg, business director of software development for Gerber Technology, explained in a presentation to the ASTM task force, the original standard answered the need for a neutral data file format to facilitate transfer of pattern pieces. The standard is based on Autodesk's DXF file format, which is designed to exchange mechanical engineering drawings. DXF is a specially formatted ASCII file, which means it is created using standard characters on the computer keyboard and can be viewed in a word processing application like Microsoft Word.

ANSI/AAMA 292 DXF includes a set of conventions that represent apparel piece geometry, or 2D flat patterns, and associate meanings to pieces, lines and points. The standard defines how the conventions should be organized in a file for exchange of information between CAD systems.

The standard defines and supports the following: turn points, curve points, the grade reference line, the mirror line, the grain line, internal line(s), stripe line(s), plaid line(s), internal cutouts, grade rule identifiers, annotation, the base size name, the cut line, the creation date, drill holes, numerically controlled (NC) cutter instructions, notches, the sew line, etc.

There are 14 "layers" for organising the data inside the DXF file. In another words, the data is divided into 14 categories when it is written into the ASCII DXF file format. Thus, the DXF translation programs know where to go within the file to get the information they need to display and work with the pattern piece.

A major downside of ANSI/AAMA 292 DXF is that its documentation is not very clear in its definitions and leaves room for interpretation, which has created anomalies. Moreover, the standard does not include a grade rule growth specification, and the DXF structure is limiting in the amount of information that can be expressed. Also, it does not relate pieces.

Another major issue is that there are no standards for curve-smoothing algorithms, which leaves the door open for different CAD vendors to individually interpret how the curve line should react to points that control its path. For example, one system may require the curve line to pass through the control points, whereas a different system may only require the same curve line to pass by along the points. This difference in curve-smoothing algorithms can cause accuracy problems when pattern pieces are imported. Most of the time, the difference is very minor. However, when control points are not properly positioned along the curve path, the different algorithms will display the curve line differently between CAD systems.

To sum up the existing ANSI/AAMA 292 DXF standard, there are companies that have tried to use it and failed, and there are companies that are using it in a production environment. The latter typically are firms that have conducted extensive experiments to learn to work around the anomalies between different CAD system DXF translators.

Task force addresses specific pattern data exchange issues
At its July meeting, the task force grouped a list of issues to be resolved related to pattern data exchange standards into the following categories:

  • Problems caused by CAD vendor implementation bugs;
  • Problems caused by lack of end-user computer literacy;
  • Issues to be improved within the ANSI/AAMA 292 standard; and
  • Issues to be addressed in future standards.

Given the need to offer the industry some solutions within a reasonable time frame, the task force initially is working to improve and complete the proposed ANSI/AAMA 292B document, which includes an implementation guide and grade rule table standards. Following are specific issues that the task force addressed (notes in parentheses indicate work that is still under way):

  • Clarification of annotation text (Assyst-Bullmer to provide proper wording);
  • Addition of grade rule standards (need to include tables);
  • Clarification of units, i.e. imperial vs metric;
  • Addition of notch information, including more notch types, grading notches, orientation/size and dependency information (Assyst-Bullmer to submit proposal on thelatter);
  • Addition of orientation line;
  • Clarification of drill holes; and

Quality validation and boundary line accuracy (proposals for solving this issue to be submitted by SGS-Optitex, which will address the use of additional parameters to allow better estimates/understanding of curves; and Lectra Systems, which will address a method that will maintain curve quality though use of a standardized DXF algorithm).

In terms of items to be addressed in future standards, the group identified the following:

  • Matching;
  • Quality or equivalent "zones," i.e. for working with leather, shading;
  • Support for DXF geometric primitives (i.e., arcs, circles, etc.);
  • Style attributes;
  • Cut/sew lines; and
  • 6Piece attributes/properties.

The ANSI/AAMA 292B document - revised and converted into ASTM standards format by Lisa Shanley of Wild Ginger Software - was posted online for comments, and then provided to the ASTM D13.66 sub-committee for review.

At a meeting this month, the D13.66 sub-committee, with the guidance of the task group, will address any objections to the document before sending it forward to the full D13 committee for approval. Provided that any objections can be successfully addressed as of the October meeting, the D13 committee will review the document December 1, and release its decision February 1, 2001.

Ambitious agenda ahead: volunteers wanted
In response to the many issues that have been raised with regard to CAD/CAM standards, the D13.66 task force has formed sub-groups to handle the following challenges.

  • Launching marketing and publicity campaigns to attract end users to D13.66 meetings;
  • Defining pattern piece properties and attributes for future standards development;
  • Gathering sample patterns and establishing "universal" pattern sets that will be available for downloading via the Internet for validation of ASTM pattern interchange standards; and
  • Investigating the best technology (i.e., XML) to use to develop new standards.

In addition, the D13.66 sub-committee plans to develop standards for exchanging product data specifications, marker data, marker plot data and 3D body scanner data and to improve the existing standards for exchanging marker cut/spread data.

XML and standards
The advantages of extensible markup language (XML) are being touted by many branches of the sewn products industry and beyond for use in software development, web development, collaborative communications and supply chain management.

With its almost infinite flexibility and user-friendly attributes, XML also is attractive from a standards development viewpoint. Neal Goldberg, business director of software development for Gerber Technology, discussed the opportunities of XML at this past July's meeting of the CAD/CAM pattern data exchange standards task force of the American Society for Testing and Materials.

With important end uses for XML arising in so many areas of the industry, including the CAD arena, it's important to understand how the language works. Following is an overview of the basics.

While the hurdles to overcome in CAD/CAM data exchange are high, so are the potential rewards for surmounting them. As Gerber's Goldberg concluded in his presentation to the task force: There will always be differences between user needs and systems. The goal is to work together as CAD vendors and users to accommodate and reduce these differences.

Kurt Chang is the chairman of the American Society for Testing and Materials (ASTM) D13.66 Task Force for improvement and development of pattern data interchange standards between CAD systems. Chang, who is a Microsoft-certified systems engineer and apparel manufacturing industrial engineer, also is founder of 3D Custom Fit, which specialises in turn-key 3D solutions for custom "Made to Shape" fashion, going beyond 2D tailor measurements. He may be reached at kurt_chang@msn.com.

What is XML?XML is simply a specification. In other words, it is an agreed upon protocol for how to create certain kinds of documents. It is a new markup language, developed by the World Wide Web Consortium (W3C), mainly to address the limitations in hypertext markup language (HTML). XML is not intended to replace HTML but to complement it.

Most have heard of HTML, the language used to create documents that can be displayed in any browser, regardless of the computer system used. A browser is an application like Internet Explorer or Netscape that reads HTML files, and then follows the HTML instructions to display the enclosed information accordingly. Some studies estimated that there are more than 800 million web pages based on HTML.

Because XML is not intended to replace HTML but to complement it, in order to appreciate XML, we need to understand a little more about HTML.

Defining the different markup languages
First, let us define markup language (ML). The term "markup" originated from the publishing industry. In traditional publishing, prior to the computer revolution, manuscripts were annotated with layout instructions for the typesetter. These handwritten notations are called markup. Markup is a separate activity that takes place after writing and before typesetting. Today we have extended the concept of markup to encode instructions within documents that tell applications like browsers how to handle the information.

Next, let us define hypertext (HT), which was coined in the 1950s by Ted Nelson. Hypertext can be defined as "human readable information linked together in an unconstrained way." Those who use the Internet know the links in the web pages allow information to become related without any constrains, unlike most of the information storage and retrieval methods that we are familiar with. For example, the phone book links its content by names, which are then followed by addresses and phone numbers.

To put it simply, the "HT" in HTML represents the language's ability to allow the creation of unconstrained links between documents located on any computer in the world. "ML" stands for the markup symbols (tags) inserted in front of data sets within the HTML file.

These markup symbols, which many people refer to it as "tags," tell the web browser how to display a web page's words and images for the user. Take the tag <p> for example: <p> tells the browser to display the data proceeding it in a new paragraph. Likewise, the tag <b> tells the browser to display the proceeding text in bold.

XML also uses tags, but differently than HTML. One of the limitations with HTML is that it cannot describe what kind of data it is displaying. This is where XML becomes very useful. Consider the following examples of an address book clip in HTML and then in XML:


From these examples, you can clearly see how XML uses tags in a different manner. The idea behind XML is deceptively simple: its goal is to solve the future conflict demands that the W3C foresees with HTML.

As the web's content becomes richer, web page creators need more tags to make the web pages do more things. For example, chemists want tags for display of formulas, mathematicians want tags for formulas, but they need different tags. However, at the same time, Web page developers wants fewer tags because HTML is getting too complex. In addition, as wireless hand-held Internet access devices become more popular, the need for simpler markup language will be needed.

Advantages of XML
XML predefines no tags, and it is a stricter language. Because there are no predefined tags in XML, the creators can design the tags that they need, and very specialized tags can be created for a specific industry. Each XML file carries a document type definition (DTD) that relates to the set of tags it is using, and identifies the document to the computer and the end user.

XML concentrates on the structure of the information and provides meanings for the data it contains. It does not affect the appearance of the data as HTML does. Instead, XML works with Stylesheet Language to handle data presentation.

Perhaps most importantly, XML presents a new way of working with documents. A document exists to convey information. A web document does this not just through text, but also via multimedia files (pictures, movies, sound, etc) embedded within it, and in the links it has to other resources.

A document can also contain a second kind of information, and this is information about itself. This is known as "meta-information," or data about data. For example, consider a table of contents. The benefit of XML is that it separates semantics, presentation and data. By separating the base elements of a document into structure, information and format, we can treat them independently and perform more efficient and advanced manipulations.

If HTML is about how to display the information, XML is about describing the data. If HTML is the picture of a building, XML is the blue print of the building. After you feed a picture into a computer, the computer can do little with the image, except to display it. However, if the blue print or engineering information is fed into the computer, the system can then generate different 3-D views of the building, perform stress analysis, alter the building layouts and change the design, etc.

HTML was invented with the specific purpose of providing a universal set of tags for displaying of information. The tags only tell the computer how to make the information within them look. XML, on the other hand, defines a standard way of creating markup that is simple and extensible, so users can richly describe any kind of data. Different applications can then process the same data in different ways to produce useful information.

In other words, XML describes data, rather than instructing a system on how to process it. These self-describing documents can provide far more information about their content than their formatting-oriented HTML counterparts. For example, instead of a retail price being marked in HTML as <b> for bold, XML can mark price as .

In conclusion, XML is an exciting emerging technology with great potential. It is a universal format for exchanging information between software components that is legible to both computers and human beings.

For more information about XML, visit www.w3c.org. For specifics about its potential use in CAD/CAM pattern data exchange standards, a sub-group that is researching the matter may be reached via Gerber's Neal Goldberg at tel.: 860-896-2016; fax: 860-875-4350; e-mail: ngoldberg@gerbertechnology.com.

Wanted: industry participation
The ASTM Apparel and Sewn Product Automation sub-committee and its pattern data interchange standards task force are continually seeking new members. No qualifications are required to join, except an interest in participating. The next ASTM D13.66 subcommittee will meet on October 24 at the Rosen Centre Hotel in Orlando, FL, as part of the ASTM D13 Committee "On Textiles" semi-annual meeting. The event - a non-high-tech gathering that is open to the entire industry - will provide sewn products companies and suppliers an opportunity to take an active roll in the standards development process. Both small and large companies are encouraged to participate. For more information on the ASTM and its "On Textiles" main committee or the D13.66 sub-committee, the organisation may be reached online at www.astm.org. Additionally, meeting and registration information for the ASTM "Committee Week" meeting, which takes place October 22-27 in Orlando, is available online at www.astm.org/commit/weekinfo. Contact Bode Hennegan at bhennega@astm.org for additional details.