• Home • Keynote Speakers • Other Speakers • Agenda • Meeting Notes • Presentations

Digital Smart Factory Conference
Hyatt Regency Orlando International Airport
Orlando, Florida | August 15-16, 2000

---

Meeting Notes
pdf format

The notes which follow were taken at the 2000 Digital Smart Factory Conference by Dennis E. Mason of Mason Consulting Inc. These notes are not intended to be comprehensive, and they may not reflect the exact the words or the intent of the presenters. Readers desiring additional information or clarification should contact the presenters. Neither Mason Consulting nor the Research & Engineering Council of the Graphic Arts can be responsible for errors contained herein.

The 2000 Digital Smart Factory conference was sponsored jointly by the Research & Engineering Council of the Graphic Arts Industry and the Graphic Communications Association. The meeting was cohosted by the Graphic Arts Technical Foundation, the Technical Association of the Graphic Arts, and the CIP3 Print Production Format Consortium. The Orlando gathering followed previous meetings–in Chicago in 1999 and Pittsburgh in 1998–both of which dealt with Digital Networked Production Systems (DNPS) in the graphic arts industry. The Orlando meeting was called to order by Ronald L. Mihills, managing director of the Council, and by James E. Harvey, vice president of the Graphic Communications Association . Following a brief overview of the conference, the conference attendees participated in the self-introductions which characterize R&E Council meetings.

The microphone was then passed to conference cochair Stephen Franzino, vice president of technology for the Courier Corporation, and the R&E Council vice president for Digital Networked Production Systems. Franzino began by noting that the meeting–like efforts in the industry to develop the digital smart factory–was a collaborative effort on the part of several trade associations. He observed that the digital smart factory consists of four key elements: process control, industry architecture, E-commerce, and content management. Franzino further said that each of these elements is critical, and must be addressed for the concept to develop successfully. He noted that the conference consisted of separate panels and presentations dealing with each element. Franzino then described how the conference would proceed, and then introduced Mark O. Jones, vice president of technical solutions for Quebecor World.

Jones, the conference keynoter, spoke on Computer Integrated Manufacturing–Managing the Transition. Jones noted that computer integrated manufacturing is a complex problem, and then described how the process is beginning to work at Quebecor World. He noted that change is constant, and recommended Who Moved My Cheese?, by Spencer Johnson, as a handbook for the 21st century. Jones said that his career began with a job in the printing industry doing software programming, and then moved to Scitex, then to Hell, and then to a high-technology startup. Jones then described how the industry had moved from mainframe systems to desktop. Noting how the industry had evolved, Jones commented on the need for standards and the difficulties involved in reading files generated by any number of content developers.

Jones then moved to a description of computer to plate, noting that the process is only now being optimized. He commented that a practical philosophy is: First, you make it work, then you make it work right, and then you make it work fast. Jones said that through all the developments he had described, workflow is the key. Today, as the Internet becomes the conduit for finding and moving information, Jones commented that many new business models are being built on the foundation of the Internet as the transfer medium of choice. Talking about the dot.com phenomenon, Jones once again noted the need for standards to facilitate the communications which are part of all commerce today.

Jones then described the evolution of Quebecor World, noting that a spinoff called QueNet Media has emerged, and the company now has a joint venture with nurun. He said that the company is highly consolidated at the top tier, but highly fragmented and specialized in the middle and bottom tiers. He described the company’s operations–consisting of 162 plants, of which 100 are in the United States. The corporate strategy is to build an organization of diverse, balanced businesses with strong growth potential. Quebecor World seeks to constantly improve its operating cost structure, and to increase efficiency and productivity by investing in advanced technologies and systems. Ongoing benefits are sought from strong operating controls and by matching costs to revenues.

Market sectors served by Quebecor World include:

• Magazines and catalogs–$2.2 billion in 1999.
• Digital services and new media–$125 million in 1999.
• Direct services–$400 million in 1999.
• Commercial services–$575 million in 1999.
• nurun–$30 million in 1999.
• Logistics–8 consolidation and mailing sites.
• Book services–$645 million in 1999.
• Retail–$1.04 billion in 1999.
• International–40 percent of 1999 revenues, with operations in Canada, Europe, Latin America and Asia.

The secret of a successful printing company, said Jones, is to develop strategies built on speed, connectivity, and intangible assets. Convergence, he noted, is the future of communications.

The range of Quebecor World activity runs from broadcast marketing (conventional printing) to target marketing (interactive media and portal distribution), Jones noted. He then discussed brand management, data mining, and communications planning–all as part of integrating the activities of the printing company. He noted three media channels: print media, new media, and content syndication.

Jones then moved to a discussion of nurun, the largest Canadian web solutions integrator, with $50.5 million in 1999 revenues, of which 37 percent are outside Canada. He noted that web solutions are a natural and necessary adjunct to conventional printing services. He also discussed automated publishing solutions which include print, data, web, CD-ROMs and product flyers. Providing a link to enterprise resource planning services is the next step, he said.

Discussing how the marketing process can be expected to evolve, Jones discussed personalized Internet flyers, noting that such projects have been completed for Radio Shack Canada and Boise Cascade. The concept involves the ability to change prices and content at the last possible moment, and tailoring overall content to individual consumers.

Moving to a discussion of best practices, Jones noted that Quebecor is standardizing its back office systems and other computer initiatives. Operational consolidation involves estimating and scheduling. Uniformity encompasses base codes and tables, hardware and software administration, and process control. Hurdles to success in this area exist in executive and business units, as well as in plants, where people often believe their businesses are "different" and thus should not be standardized according to rules developed elsewhere. The strategy is to drive toward a common networked infrastructure, with common administrative systems. The idea is to support a limited number of manufacturing and customer solutions. Business unit commitment is required, as is the willingness to adhere to a common system. Finally, a minimal number of standards must exist to facilitate interface to the outside world. Printing industry benefits sought include those gained by the banking industry with its ATM standardization.

Customer delight is the result, and occurs when the right people are put on the right project. Control must be granted to the project originators. This will occur, said Jones, when the industry learns to work together for the common good.

Moving to a discussion of computer integrated manufacturing, Davison said that CIM is an integration backbone capable of supporting end-to-end enterprisewide production processes. CIM includes rigorously defined information technology and process architectures. It includes a strategy for optimizing automation, and a production engine that is driven by intellectual capital. Finally, it is a system that displays adaptive behavior.

Focusing in on an architecture involves a span which identifies a range of process functions, Davison said. The goal is to concentrate on the range of key value chain activities where your organization holds core competence. Depth is the appropriate level of linkage within functions and applications. Control–with minimal operator intervention–is the goal, with direct communication of data and information between supply chain, sales management chain, administration, and management.

Most large businesses cluster around four processes: supply chain management, management control, selling chain management, and administrative control. The key to a digital smart factory, said Davison, is to tie all these together.

Supply chain management is heavily tied to enterprise resource planning, and involves logistics, production, and distribution. Key, of course, are interfaces with business partners such as suppliers, distributors, and resellers.

Selling chain management consists of customer relationship management, involving customer service, sales, and marketing. Key are the interfaces with customers and resellers.

Administrative control involves business systems, including purchasing, human resource management, and office functions. Here the key interfaces are to employees.

Management control business systems involve finance, accounting and auditing. The principal interfaces are to stakeholders.

In a digital smart factory, all these functions must be integrated. The key is how the front office functions and back office functions are tied together. According to Davison, the concept involves "knowledge tone" applications (analogous to dial tone) which ensure that two parties communicating have a bridge for mutual understanding. Although the logical place for printing companies to focus is in the production sector of the supply chain management piece, Davison said it is important to do business across all business functions.

Franzino then solicited questions from the audience to either of the keynoters. John Werner of Graphic Communications World asked about security issues. Jones commented that the Internet probably involves less risk than handing a credit card to a restaurant waiter. He noted, however, that Quebecor World devotes a great deal of time to security.

Jeff Nytko of Quebecor World asked Davison why so many integration efforts fail. Davison said that these failures occur primarily because the scope of the work is vastly underestimated, and because senior management often assigns responsibility too low in the organization to guarantee success. Participation by top management is critical to success, he added. Davison also noted that the printing industry is populated by people who are hardware focused, rather than software oriented.

Scott Chrisen of AAA asked Davison how best to get people involved in the process. Davison used the example of digital asset management, noting that a staff member who feels the need for hard files must be a key member of any implementation team which seeks to eliminate hard files.

Bill Ray of Group noted that most of the software being sold in the industry is not very good.

Mary Abbott of NPES asked what can be done to ensure that the various individual pieces fit together when the process is complete. Jones discussed the possibility that businesses under stress tend to act in their own best interest.

John Werner of Graphic Communications World then suggested that Jim Harvey of GCA discuss the GCA digital dictionary project.

Steve Franzino noted that the availability of people is an important element in moving toward a digital smart factory. Davison said one way is to use customer interface management to free up time for sales personnel.

Craig Shrader of ImageNet commented that he found concern on the part of sales personnel that their interface with the customer would be reduced through automation. This did not occur, however, because their administrative functions were reduced, and thus sales personnel found that their collateral material could be updated more quickly.

Orlando Boleda of Hart Graphics asked whether the concept of mission criticality is universally understood throughout organizations. Jones noted that specifications must constantly evolve. Davison added that within individual areas, people have a good idea what is mission critical. But the key is to identify those functions and data sets that are mission critical to the overall organization.

Session I–Process Control, was convened by Greg Bassinger, manager of process control at the Graphic Arts Technical Foundation. Bassinger outlined work currently underway at GATF, noting that automation in the printing industry involves more, not less, process control. He commented that as process control improves, the goal is to take people out of the equation.

Bassinger noted that GATF test forms permit diagnosing and measuring printing to yield less waste and reduce makeready time. He discussed the Merritt digital plate control target, a native PostScript target which is a multielement diagnostic and process control device used to monitor the exposure of digital output systems such as platesetters, filmsetters, or proofs. Moving to a discussion of resolution versus screen ruling, Bassinger showed how straight line microlines can be used to detect positive/negative balance. He also discussed microline checkerboards, noting that one-pixel checkerboards are beyond the capabilities of most imaging systems. One-and two-pixel arcs also can be useful in calibrating platesetters, Bassinger noted. Bassinger then moved to a discussion of tone scales and how they can be used to check adjustments made to platesetter linearization curves.

Bassinger then turned over the podium to Virgil K. Horton, Jr., president of The Horton Group, whose subject was HPPM Benchmarks. Horton likened the smart factory concept to that of DNA and the code of life. He observed that just as DNA is the storehouse of the information of life, the smart factory contains the necessary information which permits overall knowledge of how the printing system operates. Just as DNA typing risks perpetuating injustice by focusing an entire case on physical evidence, the power of the smart factory concept lies in its use of statistics. DNA copies itself so faithfully that instructions can be passed between generations. Concepts such as this ensure that the craft is removed from printing and procedures will be conducted in the future according to the numbers.

Moving to a discussion of the smart factory concept and the role of the GCA High Performance Press Management committee, Horton noted that printing is a manufacturing process and the printing plant is a factory. He said that data in the printing factory often is improperly used, or not shared with those who could use it most. Horton said that the first step is to open up minds beyond conventional boundaries. Then, the focus must be on manufacturing and statistical process control, and to determine what information is necessary. Finally, performance information must be exchanged between suppliers and users.

The goal is to obtain the ability to improve the print manufacturing process. In the process, a cooperative and interactive relationship between any two parties who need the information must be developed. Prerequisites include an understanding of what information needs to be developed, plus a vision of how to use the information.

In discussing efforts by the GCA, Horton noted that no single system collects all the data that printers want. Participants include many of the major printers and printing industry suppliers.

Edward M. Granger spoke next, on the subject of Alternative Color Models. He is the president of Ontario Beach Systems, and is the Melbert B. Cary, Jr., distinguished professor in printing at Rochester Institute of Technology. He began with a discussion of IQcolour–a system which promises to improve on-press color.

Granger noted that the print market consists of 8500 high-end printers plus 13,000 enterprise printers, plus 250,000 users of Macs and PCs. Workflow today involves device independence, uniform color space and a number of other factors. He also said that virtually everything that is done is based upon the concept of photomechanical separation methods. Today, the same image may be used on the Internet, on ink jet printers, or on conventional presses. This fact necessitates a new look at color, Granger said.

Color management is difficult, Grainger commented, because four workflow models are used: output-centered (CMYK), monitor-centered (wysiwyg), input-centered (embedded) and workspace-centered (tagless) workflows. The most promising of these, he said, is the workspace-centered workflow, which involves scanning in RGB and then performing transforms as needed, depending on the output device involved.

Granger then moved to a discussion of the different models which are required to successfully reproduce on newsprint and a wide variety of other substrates. Gamut mapping in the IQcolour concept involves always interpreting the RGB input to work most successfully on the output device to be employed. The black model can be made up of any combination of colorants. The amount of each colorant used at any level can be adjusted to build a neutral scale that has 1000 tone values.

In summarizing, Granger noted that spectral data is used to control the output devices. Devices are truly independent. IQrgb is a superset PCS. IQcolour uses fast integer computation. A natural appearance map is output to all devices.

Changes in Process Standards–In-Line and CCD Devices was the subject addressed by David C. Albrecht, product line manager at GretagMacbeth. Albrecht, who also serves as the standards chair for PIMA IT2-28, thanked Lawrence C. Steele, senior technologist at Kodak Polychrome Graphics, who collaborated on his presentation.

Albrecht began by commenting that standards developed for an industry are reviewed every five years. Photography standards are developed by PIMA IT2 and ISO TC42; graphic technology standards are developed by CGATS and ISO TC130. The fact that images may be used in both photographic and graphic applications complicates the standardization issue.

Several densitometry standards exist in the graphic arts industry, and five are being reworked at present. These standards include ISO 5/1:1984, ISO/FDIS 5/2:199x, ISO 5/3:1995, and ISO 5.4:1999, ISO CD 14807, and ISO/DIS 14981. Colorimetry and process control standards such as ISO 13655:1996 and ISO/CD 15790 also exist. Reference standards include CGATS.11/PIMA IT2.11-1999 and ISO/CD 15790.

Critical to the efforts is a joint working group which addresses photographic and graphic arts standards, which are required in both disciplines. But having two standards for the same piece of equipment does not serve either industry group well. Changes now being addressed by the standards group involve the sampling aperture and irradiated area, with more realistic specifications for underillumination and overillumination, as well as addressing translucent blurring (lateral diffusion error) concerns. Spectral conditions and conformance also are being addressed, since key specification areas are missing from ISO 5/4, including calculation of density. The joint working group is concentrating initially on updating ISO 5/3 and ISO 5/4.

Standards and the graphic arts in the digital age are confounded by the fact that printing, in any form, is an analog process. Achieving repeatable and consistent results in the analog process allows the digital data to be reproduced successfully. Standards provide the foothold for consistent and repeatable usage, across a plant or the entire industry. Older standards are now being updated. New standards–especially in graphic technology–are being developed. Websites of interest are www.npes.org and www.pima.net/standards/tc42.htm.

The final speaker of the Process Control panel was Mark O’Connell, technical marketing manager at Graphics Microsystems, whose subject was Development and Issues of In-Line Measurement. O’Connell spoke about the status of closed loop color control in the industry today. He noted that Graphics Microsystems has been producing color control devices since 1984. In 1990, the firm merged with Cosar Corporation, and recently was purchased by Dover Corporation.

Today, it is possible to control printing presses numerically, In the 1930s, Frank Preucil built the first graphic arts densitometer. By the 1970s, Macbeth, Gretag, Cosar, Tobias, and X-Rite were in the business. By the 1980s, color management devices had become robust, and employed onboard computers. Interfaces with RS232 devices had become standard, yielding the ability to gather a vast amount of data. Automation provided the speed necessary for closed loop color control, and interinstrument agreement permitted valid translation among devices based upon ANSI, CIE L*A*B, and CGATS standards. O’Connell noted that it is now necessary to redefine the standards to move to the next level of color operation.

O’Connell commented that SNAP, GRACoL, SWOP, and FIRST standards now have teeth and permit production ‘to the numbers.’ He particularly complimented the Flexographic Technical Association for the manner in which the FIRST specifications were communicated to the industry.

Closed loop color is important, O’Connell said, because it permits standardized printing, on both sheetfed and web presses. The concept permits high speed measuring and reporting of color while the press is running, as well as making adjustments to that color. GMI has more than 75 ColorQuick systems in operation, at publication, book, packaging, and commercial printers. More than 200 presses are now controlled using these systems, which are made by several manufacturers. Press interfaces now available include the GMI MicroColor II, HWS Telecolor II, HWS M600-CPC, KBA Colortronic, Komori, MAN Roland Pecom95, and Mitsubishi. This may be too many ways to do the same job, O’Connell observed.

Use of closed loop color yields more consistent quality, reduced makeready times, and reduced crewing and crew costs. Computer to plate investments can be leveraged through use of protocols such as CIP3. Future opportunities include CIM/E-manufacturing, as well as the development of print quality specifications which will provide the link to E-commerce. E-manufacturing will permit comparing the performance of printing presses and other devices. The technology already exists that will gather the necessary data. Other E-manufacturing opportunities include measuring CIP3 preset accuracy, performing predictive diagnosis and maintenance, and performing enterprisewide color managed workflow. E-commerce applications include conformance to specified print quality, as well as the elimination of proofs for color guidance at the press. Haas Publishing and TV Guide are employing these concepts.

In the question and answer period which followed the presentations, Bill Davison asked how, as a business manager, one could learn all the various parameters that need to be measured. Horton said that the GCA High Performance Press Management committee is attempting to do just that. Davison then asked whether a white paper or specification will result from the committee members’ efforts. Horton answered in the context of the pressroom only, and Davison asked whether some group is taking a higher view, attempting to view the entire business. Bassinger said that GATF is looking to integrate standards beyond just the pressroom. Franzino commented that most of the efforts to date seem to be directed ‘down’ toward the press, but that precious little information seems to be coming back.

John Werner of Graphic Communications World asked Granger how concepts such as the IQcolour can actually be implemented. Granger said that all the necessary targets, definitions, and software will be available by the end of 2000.

Bill Lavelle of Impresse commented that often the part of the process that is left unaddressed is that of changes and whether they have been implemented. He noted that capturing the metadata is critical to the success of the digital smart factory. O’Connell noted that input from key line supervisors must be solicited, because the success of any system will depend upon their cooperation and concurrence.

Bill Davison then commented that the current color paradigm does not permit successfully distributing color–thus a new color paradigm is necessary. Granger said that Davison’s observations are correct, and that the IQcolour concept is contingent upon not having to remanufacture images as different devices are used in the print manufacturing process. The key change lies, Granger said, in how the job actually runs on the press. Using the new system, midtones are controlled in such a way that press variations are less critical. Perhaps most important, these changes can occur in a way that press operators may not even be aware it is happening.

David McDowell of NPES commented that the system really is not an RGB model, but an XYZ model. Granger agreed. McDowell then said that output devices must be fully characterized. Granger said that while this is true, that characterization can be left until the very last moment, and does not involve altering the communication of data. Granger said that no more than 18 patches can correct to the IQcolour paradigm. McDowell suggested that proofing is another matter, and that customers often want a proof which looks like the final output. Granger said that the RGB triplet effectively defines every available color. Thus proofing involves data, not just perception and visual effect.

John Werner of Graphic Communications World asked O’Connell which particular industry has effectively mastered color. O’Connell said that the textiles and plastics industries have made better strides toward the effective control of color than has the printing industry. Printing must become more like these industries, he added.

Franzino closed the session by commenting that it was necessary for the government to get involved in the DNA mapping project, and observed that he hoped this would not be required for the printing industry to reach consensus.

Session II–Industry Architecture, followed a buffet luncheon. The session was chaired by James E. Harvey, vice president of Spectrum Operations at the Graphic Communications Association. Harvey began by noting that the printing business is a $303 billion worldwide market, growing at an annual rate of 12.3 percent. The industry is in a tight labor market, and faces flat or increasing material costs. The printing business is highly fragmented.

Motivators of industry leaders include distribute-then-print, with print buyers demanding standardized Internet access to assets. Advertisers and publishers seek standards to support distributed production. Graphic images and content are now viewed by corporations as digital assets. Printers are attempting to implement supply chain programs. Mergers and acquisitions are changing the landscape. Industry consortia are fractionalizing the industry. Finally, the industry is becoming international in scope.

This industry in flux needs a clear definition of terms and processes, organized in an accessible and useful manner, that will serve as the common basis for standards and the use of standards in business and production systems. A set of Internet practices, including specifications for the interchange of information also is needed.

Object oriented methods of design have evolved over time, Harvey told the crowd, beginning with the simulation of processes. This moved toward a graphics-centric programming concept, with the interaction of graphic objects in virtual space. Markup languages have evolved similarly, and today we have bricks, frames and objects, used in Internet publishing, extranets and intranets. These languages find extensive use in E-commerce and interenterprise process automation.

Looking to the economy, Harvey discussed trends including global networked business. As information moves across international borders, the need for clarity of corporate information is compounded. This causes a move toward knowledge management.

Industry architecture fundamentals now being addressed include physical interface of print systems, a common framework for XML implementations, and print-specific data definitions and elements, as well as system-to-system interoperability standards.

After his introductory remarks, Harvey handed the microphone to Jessica Murphey, Spectrum special projects manager for the Graphic Communications Association. Murphey’s subject was Object Management Group Document Respository Specification. One goal of the Object Management Group is to build consensus in software interoperability. Another is to develop repositories for document management interchange and webDAV (an extension to the http protocol). The core of the effort is the OMG Document Management PSIG.

Interoperability requirements being addressed by the group include providing a full mapping strategy from OMGDR to WebDV and DMA, and back again. Once these requirements are defined, they fit into one of five use cases. The next step is to create the domain model–essentially extracted tables which define the process. Then a scenario is run to test the concept. For example, one test involved converting an Internet page to a PostScript file for printing. Use of a composite pattern increases speed over that realized by calling up tables.

To date, the group has submitted one document repository request for proposal. Next steps include soliciting proposals for interface development, and then testing of the finished concept.

According to Murphey, alliance across vertical industry groups will be necessary for full acceptance.

AMPAC Industry Arthitecture Project and CGATS Coordination was addressed by Lawrence C. Warter, director of new business development for Fuji Photo Film USA. In his talk, Warter reviewed the status of standards, and discussed how the industry should move to improve the standards situation. He began by saying that a hierarchy of standards exists: de facto standards, internationally accredited standards, nationally accredited standards, market and association specifications, vendor supplied standards, and trade customs. Ideally, Warter said, a pathway is established which brings standards to the de facto level as quickly as possible.

Among new factors in the standards situation are E-commerce, with GCA involved for quite some time in paper and magazine order entry. Many of the needs of advertising agencies, publishers, catalogers, list processors, retail distributors, insert printers, and others have been addressed by XML and other GCA concepts. GCA, together with CGATS, is attempting to develop a path to migrate standards; meetings are underway.

CIP3 is another new factor. The goal of CIP3 is to reduce redundancy and eliminate keystrokes. CIP4–the latest version of the protocol–started with the electronic job ticket and has now led to JDF which ties E-commerce more tightly to CIP3. With the new format, Warter said, the industry can deal with any level of complexity.

AMPAC provides an architecture model for total systems, with code descriptions for standard parameters, parameter values and relationships.

William K. Lavelle, director of industry relations for Impresse Corporation, spoke next, on the Industry Architecture Project Subcommittees. Lavelle described UDEX, a GCA industry architecture project which is tightly tied to DNPS and Digital Smart Factory concepts. UDEX stands for Universal Data Element Xcross Reference. From a definitional standpoint, Lavelle told the group that meetings, travel, fragmentation and other time consuming activities mean that a different approach is required to address problems in the 21st century. The idea is to make it possible for every industry participant to use a web-enabled concept which eliminates travel, unifies the industry, and is more time efficient.

The basic idea is to deploy a database (called UDEX Prime). Within UDEX Prime is an area called Base, with subsets called Stores, which in turn have smaller subsets (dictionaries) called Tables. UDEX will be used when one company, which already has a data dictionary and wishes to communicate with another company, submits their data dictionary to the UDEX database, which translates it for use by the second company. Of course, all this is contingent upon the second company also submitting its data dictionaries for translation as well. The concept is expected to achieve a very high user response and to gain popularity quickly.

One feature of the UDEX concept is that different systems maintain data in different areas, so that not all data can be utilized by all other participants. In traditional integration, the sender and received see and recognize the same data. UDEX comes into play when the two systems are unable to deal with the same data, and crossindexing is required.

Demonstrating the prototype UDEX concept, Lavelle showed a data entry screen. He then described numerous potential sources for UDEX data, including AMPAC, CIP3/CIP4, DCS, EMS, GCA B2B, GIF, GRACoL, PrintTalk, and others. The group has identified 17 subcommittee groups or sectors. Initially, the groups will seek to gain input from as many as ten different contributors, after which the committee will use the information to begin building a database. Contributors are being asked to submit data labels, data types, length, value range, description, and object hierarchy. Robert Hu of Collabria is the commercial print chair; Tom Hardy of printCafe is the magazine publication printing chair. Chairs are being sought for packaging, newspapers, and several other groups. The technical activity is being cochaired by Bill Ray and Bill Lavelle. Bill Davison is cochairing the business activity, and is responsible for development of the business plan. The UDEX group is being very aggressive and expects to show results within 12 months.

In introducing the question and answer session, Harvey noted that the UDEX group is seeking $250,000 in support funding.

Chuck Gehman of DAX asked Murphey about potential conflicts between Microsoft and CORBA. She noted that the conflicts will be addressed in a forthcoming meeting. She then said that the principles of document repositories led GCA to look at OMG. Warter added that the number of Macs in the industry led toward OMG.

Bill Ray of Group Infotech noted that big corporations input data from documents and then save the actual documents for legal reasons. Harvey said that participating companies are interested in making a connection down the road. Murphey said that GCA is not involved in an exploratory fashion.

Tom Hardy of printCafe then took the podium to note that Bob Jones of Prograph had initiated the concept of DNPS and the Digital Smart Factory. Hardy said that as the group goes forward, Jones would have suggested that additional work is still necessary. Harking back to the 1996 inception of the concept, Hardy said that integration is now more important than interfacing. Integration involves data created instantaneously and moved across a network. In the case of a retail job, for example, the user would be able to drill down into tables and see information regarding individual stores, warehouses, pallets, and the store counter, without necessitating person-to-person communication. Significant savings can be realized from this concept.

Printing is a ‘push’ business, said Hardy, rather than a ‘pull’ business such as automobile manufacturing, which is driven by the ultimate customer. Tight integration with the customer–such as that achievable with the data interchange concepts being discussed–offer the printing industry the opportunity to increase margins.

Hardy then moved to a discussion of database tables, showing the interaction between the print buyer, customer service at the printer, scheduling, the shop floor system and separate data collection modules. Hardy suggested that successful integration involves determining whether ‘pull’ or ‘push’ concepts are more appropriate for each element of the problem.

Bill Davison then asked where each of the concepts discussed by the panel fits into the overall digital smart factory. Murphey asked for further clarification, and Davison asked whether it is possible to completely define a page and a job. Murphey responded that while still theoretical, such definition is an integral part of the concept. Davison asked if the word ‘document’ is broad enough to encompass web pages and other media. Murphey said that ‘document’ is indeed broad enough. Harvey added that we are now losing history because the Internet often does not retain information once the immediate need has passed.

David McDowell said that the word ‘document’ embodies not only what is said, but the visual means by which it is said. He suggested that perhaps we need a new concept to define the document, since much of the current usage focuses entirely on the content and ignores the aesthetics of the page. Bill Davison responded that most of the Internet people using this information today are ‘top down’ people interested primarily in content, while printing industry personnel are ‘bottom up’ people who focus on type, photography and the constructive elements of the process. The number of books currently done through electrophotography, Davison said, is indicative of the status of the industry. McDowell said that we need a vocabulary which will enable discussion of these issues.

Michael Jahn of Agfa asked whether the UDEX concept would go to the level of defining things such as staple location and orientation. Lavelle said that the implementers of UDEX would be companies such as Heidelberg, Agfa, Quebecor World, and R. R. Donnelley & Sons. Lavelle said that while UDEX may not reduce the number of meetings required, it should make the process work faster and better. Warter said that the idea is that UDEX should work better than anything else currently available. John Warner then commented that CIP3 was criticized as being a child of the hardware vendors, but that the software people appear to be pushing UDEX, and that the pendulum seems to have swung to the far side. Lavelle responded that the integration and translation concepts of UDEX will facilitate translation. Warter said that the UDEX concept is to normalize different languages. Davison said that if Company A is building a system based upon a language and vocabulary that it understands, UDEX will enable communications with Vendor B, which has its own language and vocabulary conventions. A magazine publisher and a newspaper publisher, for example, may have different terms to describe what are essentially identical functions. UDEX would enable these companies to understand the subtleties of differences. Thus three parallel efforts are involved: compiling information, correlating that information, and enabling broad use.

Jahn commented that he is hopeful, but that the process seems like playing poker without ever showing one’s cards. He suggested that vendors like to retain those elements of their technology which they feel give them an edge over competition. Davison noted that while vendors have a need to protect proprietary information, they also have a need to operate in a broader industry context.

Susan Kelly of Raine Consulting asked about the relationship of the concepts to business realities. She noted that the needs of the largest of companies differ significantly from those of the smaller players. Harvey noted that his committees have studied many other industries which have already addressed similar issues. He said that back office issues which are proprietary have no place in the UDEX system, and need not be entered. He also noted that provisions are being made for users who wish to make their data available (even after translation) only to certain other parties.

Session III–E-Commerce was then convened by Susan L. Kelly, managing partner of Raine Consulting. Kelly began by noting that the graphic arts industry does not seem to have in place much of the infrastructure necessary to make the digital smart factory concept a success. The purpose of the E-Commerce session, Kelly said, was to provide additional information around the connectivity issue. She then noted that four separate environments are of concern: the single factory, the entire enterprise, the value chain (end-to-end supply chain), and the intelligent E-business marketplace.

The wide variety of vertical industries which must be served requires a net readiness which embraces enterprise systems, horizontal exchanges, and point solutions. Also of concern are rogue initiatives, put in place by isolated segments of the organization. Kelly said that the net readiness of print customers will be a major determinant of choices, speed and implementation. The Internet is a hybrid of broadcast and point-to-point media. Technology allows people to do more with the same information. The state of the art choices are the legacy systems of tomorrow. Technology choice boards involve perpetual rolling implementations. Connectivity and integration include reach, richness, and affiliation.

GCA B2B Standards–History and Current Initiatives was the subject addressed by Brian Sharlach, manufacturing systems manager for McGraw-Hill Companies. Sharlach began by reviewing the mission of the GCA, and the flow of materials in the industry. From 1985 to 1990, work was done on a complete list of documents in the manifest for a print job. Then, in 1991, the EMBARC standard was updated to conform to ANSI standards, and since the group has further revised the standard to embrace the latest thinking. Problems in the transaction set were addressed throughout the 1990s to streamline the standard further.

Currently, the GCA B2B Standards Committee is heavily involved in the development of XML transactions. Current efforts involve moving the standards toward international adoption, and facilitating electronic processing. Efforts are also underway to broaden usage to smaller printers.

Paul M. Gaboury, director of consulting services for printCafe, spoke on Integration: Goals and Methods. Gaboury began by discussing the printCafe organization–a leading provider of print management solutions, vertical applications, business-to-business web services, and process integration to the printing and publishing industry. The printCafe strategy today is to operate the company at two levels–one that perpetuates the businesses which comprise the collaboration, and another that provides integrated web services to the industry.

The purpose of printCafe PCX is to facilitate integration by enhancing the printing and publishing supply chain. The goals are to provide customers with fully integrated access to third party systems and services, provide those in the marketplace with the ability to flow their precertified data to printCafe management systems, provide the marketplace with a blueprint for the process of integrating third party systems to the printCafe suite of systems, and support existing industry standards and maximize their use in integration projects.

PCX, Gaboury said, is an integration specification encompassing external systems including business-to-business solutions (both horizontal and vertical), e-procurement (commercial and private, supply chain, content origination and management, production system and management interfaces, and other connections. The PCX process is designed to communicate needs and opportunities; coordinate with and participate in standards groups; review with an advisory council and set priorities; establish partner agreements and schedules; develop supplements and enhancements to industry standards; implement, demonstrate and pilot test new interfaces; provide third party certification of PCX partner implementation; and comarket and deploy new interfaces. PCX is standards implementation, not the creation of competing industry standards.

Industry standards support criteria for PCX involve vision, scope, base of support/funding, relevant market value, willingness to cooperate with overlapping standards ventures, commitment to the GCA industry architecture program, use of an industry standard XML version, and existence as part of an industry group rather than an interest group.

PCX involves interbusiness data flow (among third party systems, shared environments, and printCafe web services) as well as intrabusiness data flows (existing and future printCafe print management systems). Interbusiness data flows involve many-to-one-to-many data interchanges; the least cost, most efficient method for all parties; one interface which reaches all printCafe print management systems; and the assurance of data integrity, minimal support and deployment costs, among others.

A PCX white paper will appear soon on the printCafe website. Interested parties may inquire about becoming a PCX certified integration partner by sending an Email to pcx@printcafe.com.

Kelly’s next panelist was Otis H. Johnson, Jr., president of Darby Digital Communications, whose subject was PIA–E-Commerce Council. Johnson opened his talk by commenting that E-commerce is now beginning to impact smaller printers. The mission of the PIA E-Commerce Council is to bring E-commerce products and services to the smaller printer, and to provide a voice for the industry in this period of rapid change. The target audience is midsize printers from $3 million to $30 million; a supporting audience are affiliates, associations, consultants, suppliers, large printers, and print buyers. The initial membership consists of roughly 250 companies. A steering committee provides general governance, and a task force is charged with developing an E-business strategy for PIA.

Immediate offerings are an X-industry analysis, SWOT analysis, and decision matrix; white papers (Raine Consulting); business briefings (January 14 and July 12, 2000, plus GraphExpo); a listserve, and support of industrywide standards development (primarily through GCA). In the future, the group will call for additional white papers and will publish a newsletter. The listserve will be strengthened, and industrywide standards development will be encouraged.

Robert Schaffel, senior product manager for professional publishing systems at Adobe Systems, spoke next, on Using Metadata and JDF to Automate Workflows. Schaffel began by noting print market trends. He commented that print industry growth has been constrained because the cost of doing business is high, because the Internet has changed the value of mass communications, and because future growth must come from workflow automation.

JDF is an open, extensible, XML-based print workflow specification framework initiated by Adobe, Agfa, Heidelberg and MAN Roland. It ties together authoring, production, management, manufacturing, delivery, and MIS control. JDF begins the job definition from the product intent, and ends with process specifications. JDF describes print tasks via processes and resources. JDF consists of nodes, which are collections of hierarchical nodes. A node may represent an entire job, a product component, a group of processes, or a single process. JDF also consists of resources, which are the input and output of nodes–such as PDF files, process parameters, and consumables. Also included are product specifications, which can capture the customer’s product intent, in addition to processing. Product intent allows customers to describe the end product without having to specify the details.

JDF may be created through production management software, through workflow management tools, customer interaction, or print driver. JDF connects more than just workflow. PDF connects the creation and prepress functions, while the press and postpress functions are connected by CIP3. JDF connects all of these functions, and provides links to print business management systems. The print supply chain, Schaffel said, is an end-to-end workflow model which can employ JDF protocol throughout.

Schaffel then moved through a typical job workflow. He began by commenting that arguments about vocabulary notwithstanding, users have already developed their individual definitions for the content they deal with. Because the content owners will wish to maintain control of their content, they will wish to pull information from the data repositories, and have it appear in a form which is familiar. He described a system in which the actual information extracted from the database is the tagging rather than the actual content. This will permit proceeding from page layout, through workflow manager, to project planning. In the process, JDF will permit generating or editing job tickets, and actually scheduling the work. OEMs and systems providers should be able to simply furnish a metadata repository, calling up only the information required, and then pass information on to the manufacturing operation for production planning. Digital assembly then could be automated by simply calling up the raw data.

Schaffel said that this is a future that cannot be built today, but that collaboration such as that being discussed would permit such a scenario. A draft specification is available at www.job-definition-format.org.

The final panelist was Jim Mekis, director of commercial print solutions for Impresse Corporation and elected vice president of PrintTalk, who spoke on PrintTalk Initiative. He began by defining PrintTalk, which consists of a number of founding members including Avanti, Cirqit.com, Collabria, Graphic Arts IT, GraphiTech Computer Systems, httprint, Impresse, MediaFlex.com, Noosh, Pace Systems, Parsec, Press-tige Software, Printable, PrintChannel.com, and others. The underlying goals include easy interoperability, end-to-end connectivity, standards based and open architecture, flexible and extensible operation, and data security. The PrintTalk approach was to review prior work, and to capitalize on the JDF (CIP4) concept by incorporating cXML business logic to obtain broad industry support. cXML was chosen as the preferred logic because it has broad industry support from companies such as Microsoft, Cisco, Oracle, SAP, and other software market leaders. The methodology involved obtaining requirements from print MIS companies, while adhering to industry standards. The PrintTalk concept involved agreeing on data representation (XML) and schema, and separating business logic from the technical specifications. Examples were generated, and encouragement was received from JDF originators.

The initial business objects addressed by PrintTalk are request for quotation, quotation, request for requote, requote, change order RFQ, change order quote, change order acceptance, change order confirmation, purchase order, order confirmation, order status request, order status update, invoice, content push/pull, request for report, and report. The first examples–to be announced at GraphExpo–are request for quotation, quotation, purchase order, and order confirmation.

Mekis then showed some examples of PrintTalk code. The example used in the presentation provided details regarding paper, ink, bleeds, text, proofing requirements, binding specifications, lamination, and all other information required to completely define the job.

The future of PrintTalk involves collaboration with more industry members and organizations. As of GraphExpo, the schema and deployment schedule will be released. New members are being accepted from a large waiting list.

In the question and answer period which followed the presentations, an unknown audience member inquired about the differences between PrintTalk and B2B initiatives. Mekis said that PrintTalk is focused on the commercial print market, while B2B is addressed primarily at the magazine market.

Michael Jahn of Agfa asked about governance issues which will occur in JDF, and asked about the PrintTalk strategy with regard to CIP3. Mekis said that PrintTalk wants to work with the CIP3 group, as well as with printCafe and anyone else who wants to work with the group.

Chuck Gehman of DAX asked for further clarification on where JDF and PrintTalk differ. Mekis responded by saying that PrintTalk is an implementation, not a competitor, of JDF. In PrintTalk, JDF is used intact; cXML is used to develop the implementation.

Horton asked whether work done previously by groups such as those concerned with areas such as direct mail would be embraced by PrintTalk. Mekis denied knowing about any GCA standard for the mailing industry.

John Werner of Graphic Communications World said that his first knowledge of PrintTalk led him to believe it began as an alternative to printCafe. From the conversation that ensued, it became apparent that any early differences between PrintTalk and printCafe are on the verge of being resolved.

Bill Ray commented that the political differences apparent in the room must be eliminated. He said that XML is not a universal solution. Mekis said that the approach of PrintTalk is to move ahead. Sharlach said that GCA seeks to embrace all such initiatives. Davison said that it would be nice if printCafe and PrintTalk could establish the terms under which ultimate collaboration could be achieved. Jung added that the printCafe presentation did not address the cost to others in joining the PCX collaboration. He noted that his company has already moved toward JDF, and is interested in just how much money it might have to invest to move toward PCX. Davison noted that business models are being floated, and no one knows whether they will be successful. Gaboury said that the best way to implement PrintTalk is to access the print business management data now being gathered and retained by 7,000 printers. Horton said that in the recent past, it was necessary to bring 32 different trade organizations to the table to solve environmental issues in the paper industry. Schaffel added that it will be difficult for companies to put the resources into a major industry effort that is not all-encompassing.

David McDowell noted that CGATS already exists, and is willing to take in any group committed to complete openness. He said that while openness of standards is a noble goal, at times groups such as CGATS cannot operate fast enough to be effective. On the other hand, while a closed group can act quickly, excluding key industry members may make ultimate implementation difficult. Unfortunately, documents coming from consortia often do not lend themselves to ultimate adoption by standards groups. Requirements must be clearly separated from desires. Mekis responded by saying that the intent of PrintTalk is to embrace open standards and connectivity.

Bill Lamparter of PrintCom Consulting Group noted that although attending a digital smart factory conference, he had heard little about controlling the functions in the factory. He said that Fuji and Heidelberg talked at DRUPA about entering the E-commerce landscape, but they had not been heard from at the meeting. Lamparter said that he was concerned about doing more in the printing plant than just setting ink keys.

Day two of the conference was opened by conference cochair Orlando Boleda, vice president of manufacturing at Hart Graphics. Boleda reviewed the activities of the day and introduced the second day keynote speaker, Paul Beyer, group product manager for professional publishing at Adobe Systems.

Beyer began by commenting that his background is in the printing industry (at Banta). He said that Adobe Systems corporate policy precludes making product announcements at meetings. He then moved to an outline of the keynote presentation, and began discussing key drivers for the PDF/JDF supply chain. These key drivers include the high cost of doing business, the Internet, network efficiencies, and the pressures of supply chain economics. Beyer noted that the printing industry has not realized the productivity growth which has characterized other capital intensive manufacturing businesses. The Internet is fundamentally changing workflow economics, and the value of printed products is declining.

Beyer then showed a graph of annual output in the printing and publishing sector, as compiled by the Federal Reserve Board. Industry output flattened about 1987 and is expected to remain flat for the next ten years. A subsequent graph showed that the print-on-demand portion of the industry, however, is growing at a much faster rate–a 21 percent compound annual growth rate, versus 4 percent in the overall graphic arts market. Eventually, these two are expected to merge, with the print-on-demand portion becoming equal to the conventional printing portion.

Moving to a discussion of market inflection points, Beyer said that a network solution is needed in the industry, rather than isolated silos of technology. Changing publishing paradigms include the impact of low cost ubiquitous digital infrastructure. Internet trends affecting the industry include X on Tap ASP models, next generation portals, and brokered E-services. All of these are expected to alter process workflow in the graphic arts industry.

Beyer told the group that in the printing market has grown from $91 billion in 1995 to $115 billion in 2000, with a dramatic shift toward the newer digital technologies. According to Beyer, an opportunity exists in a new PDF-centric value layer which will bridge the gap between authoring software and production systems. Today, graphic arts production consists of islands, but in the future, this must become an integrated whole. Beyer predicted that data center printing will grow only at about 5 percent per year, while the commercial printing segment will grow at only about 3.5 percent. But enterprise printing (mostly ink jet and laser) will grow at about 7 percent. The fate of the top ten commercial printers remains an unknown, as does the effect of Microsoft.net and major players such as IBM.

Traditionally, businesses must be built on either growth or efficiency. The Internet plus PDF/JDF as a tool for customer acquisition can equal growth; use of PDF/JDF in internal operations can yield efficiency, Beyer said. Milestones for 2001 on the JDF timeline should include first system implementations, and web based implementations. Acrobat and Extreme may be updated.

Beyer noted that Adobe is now focusing on standardizing processes ranging from creation, through production prepress, and finally in printing and distribution. But these processes vary considerably depending on the medium being addressed by the author or artist, so Adobe processes must have considerable flexibility. This effort requires collaboration throughout the industry, Beyer said.

Workflow today consists of processes connected by physical materials. Multiple revision cycles introduce costs today. In the workflow of tomorrow, physical materials will play a diminished role, as the process definition area is expanded. Beyer called upon printCafe and PrintTalk to share their data dictionaries and collaborate with Adobe for the greater good of the industry.

Beyer then moved to a brief discussion of the Adobe product line, commenting that InDesign, Illustrator, Acrobat, and PhotoShop will form the bases for completing PDF workflows.

Beyer closed with two quotations–one from the Creo IPO prospectus, which characterized the printing industry as slow growing and highly fragmented–and another from the president of Hewlett Packard Imaging and Printing Systems, which predicted that printing would migrate from traditional presses to smaller engines.

In the question and answer period which followed, John Werner asked Beyer to comment about the applicability of JDF to media other than print. Beyer said that to the extent that a web or video job could be defined identically to a print job, JDF could be used, but that JDF is focused on printing–printing of all types.

Another audience member asked whether JDF could be used by authors (rather than designers or publishers). Beyer responded that the author might well have a stake in how the printed product looks.

John Iobst of the Newspaper Association of America commented that authors will indeed have input into the production process.

Asked by Murray Oles of TC Advertising to look into the future, Beyer commented that Adobe seeks to make the commercial printing segment viable in a dynamic market. Goals include improving customer relationships, improving paper purchasing, gaining process efficiencies, and growing the market through new value. Keys are use of the network and a common user interface.

Bill Ray then noted that PDF is nearly ubiquitous, and asked whether the format has a role in new market areas such as E-books. Beyer commented that Microsoft has a number of workable tools. He said that the E-book market must still solve certain form factor and readability problems, but wireless devices are everywhere and could outnumber desktop stations in the near future.

John Werner noted that Microsoft has been criticized for dabbling in content creation, and asked whether Adobe seeks to play a role in this area. Beyer noted that Adobe probably missed any opportunity in this area some time ago.

David McDowell asked about the role of Adobe in package printing and product printing. Beyer commented that Adobe is well aware of the magnitude of these markets. While no vendors are known to be utilizing PDF in the packaging market today, they could well be developing products at this time.

Another attendee commented on the value of JDF in scheduling printing, as a means of predicting job structure. Beyer agreed, saying that vendors must be pressured to include the necessary features.

Asked by Kelly Schultz of Stevens Graphics to discuss potential changes in Acrobat, Beyer demurred. He did suggest, however, that Open Type fonts would be the subject of further development.

Session IV–Content Management and Delivery, was convened by William J. Ray, president of Group Infotech. Ray also is the vice president of TAGA for Papers.

Ray began his talk by predicting that the boundary between pressroom and prepress will disappear. He also said that press operators will become more material handlers rather than machine operators. Prepress will essentially operate the plant as an all digital system. He predicted that standalone prepress is doomed.

Moving to a discussion of workflow models, Ray noted that two exist:

• Process modeled (e.g., Brisque, OPEN, PCC)
• Data modeled

Process modeled workflows are characterized by a large granularity atomic work element–the document. Data modeled workflows are characterized by small granularity atomic work elements–the cell. The later version is intrinsically parallel, and any failure occurs at the cell level.

Ray then moved to a discussion of Page Level Automation, the data automation system developed for Valassis Communications. He showed details of the Valassis database, which is used to generate the myriad versions of inserts printed by the company every week. In the Valassis PLA system, any failure occurs at the cell level, with a signal provided to indicate any problem. An interesting aspect of the Valassis concept is that it is essentially spread out over a number of operating plants. Moreover, the system permits printing coupons with varying "cents-off" designations or other variations, depending upon the needs of the individual marketplace.

The database system described by Ray includes details of the presses to be used, the plants in which the inserts will be printed, paper type, and all other details pertinent to the jobs. Key to success in the system, Ray told the crowd, is that data, once in the system, need never be reentered.

The Valassis system also tracks personnel costs and machine duty cycles, and optimizes resource usage. Prepress workers at Valassis have been reduced by approximately 90 percent as a result of imposing the Page Level Automation system.

Asked by John Werner about how the actual pressroom is managed, Ray responded that the printCafe-Karen system is used.

Paul Beyer asked how and when the EPS or other file is stitched into the system. Ray said that assembly essentially occurs only when it is time to expose the plate (or the assembly can be revised until that time). PDF input is preferred, but PostScript or any other input can be accommodated. At the point of assembly, various price differences are inserted at the last possible moment.

The first panelist to speak was Chuck Gehman, vice president of technology at DAX, whose subject was Using The Internet to Deliver Content and Context. Gehman began by noting that in the recent past, DAX has moved from the role of industry upstart to a pillar of Internet delivery. DAX customers range from the like of Quebecor World and R. R. Donnelley & Sons, to small printing companies, advertising agencies and designers. Newer DAX products address upstream creators.

DAX provides print industry specific E-community, online workflow, supply chain and applications workflow sources. The goal is to take everything off the LAN and move it out onto the Internet. Essentially, DAX is an E-production company.

DAX was founded in 1996 with web.in.a.box, which included job tickets and indicated job status in the printing plant. Early on, it was necessary to type in all the information, so the system did not work especially well. Then, in 1997, DAX introduced the first open standards Internet-based file transfer application, called DAXFTP. That same year, DAX introduced the printing industry’s first PDF forms-based job ticketing application: DAX JOBTIX. Then DAX took over operation of Computer-To-Plate Pressroom, with 200 subscribers. In 1998, DAX and Quebecor announced a connectivity alliance, and the CTPP forum topped 1000 subscribers. Also in 1998, managed Internet access debuted, and the company entered the digital asset management marketplace. In 1999, DAX announced second-generation file transfer, and PrintPlanet.com was launched. In 2000, DAX piloted DAXProof, based on RealTimeImage technology. DAX formed partnerships with Noosh and printCafe.

The DAX strength is in the Digital Courier area. Today 90 percent of digital file transmission is accomplished via Email. FTP and other network methods account for another 9.5 percent, with private networks handling only 0.5 percent. Distribute and Print and Print On Demand/Variable Printing are important new elements in the DAX strategy for supporting the printing industry. Collaboration–via asset management systems, soft proofing, remote proofing, and project management/document sharing–are the final elements in the DSX portfolio of the immediate future.

Gehman then moved to a discussion of DAXAssetManager, the DAX content and context offering. In this role, DAX is essentially an application service provider. DAXProof is being piloted with a number of advertising agencies, who deliver files to the DAX server. Designers upload the file for subsequent review by agency personnel. Network-based preflight, imposition and trapping, and automated page assembly are three new product areas being addressed by DAX.

To achieve the necessary integration, DAX intends to create one single, easy-to-use interface that customers can use. Interfaces to procurement and MIS-centric dot.coms, and to consumable and equipment vendors are envisioned, as are XML-based initiatives.

The DAX PrintPlanet.com portal now has more than 20,000 subscribers. User management tools now in beta test are expected to deliver subscribers a number of applications from a single log-on.

In closing, Gehman noted that no one company can provide everything that customers want, but that integration will make things simpler. Partnerships and alliances–together with Internet and industry standards, will pave the way to the future.

John Werner expressed concern about the majority of Internet bandwidth being usurped by applications such as video and telephones. Gehman responded that recent technology has expanded bandwidth extensively. This bandwidth has been funded, said Gehman, by telephone companies and by companies like DAX. Ray commented that applications will arise to soak up all available bandwidth, however. Gehman then noted that ultimately, a system of bandwidth levels and classes may emerge, with users paying more for increased bandwidth services.

The next speaker was Craig Shrader, president of ImageNet. He spoke on Mechanization of an Ad Process Using Standards. Shrader began by commenting that the print business process is still incredibly labor intensive. He said that we must integrate the computer chip into the business process for focus on customer service. Technology removes the mundane from the workflow. Success in digital workflow involves balancing cost, value and complexity. In collaborative workflows, however, Shrader noted that complexity usually is introduced by people.

Shrader then discussed how production and business workflows differ in the advertising agency and publication model. He proposed standards that would incorporate the various elements of the workflows and enable all involved to immediately find the information they require.

But, he said, significant differences exist between the largest customers and the smallest customers, as well as between the various media types involved. The ImageNet Commerce Software and Tracking Network involves placing Constable servers at various customer sites to ensure that files remain proprietary and behind firewalls.

The Universal Ad ID Project that Shrader described creates a single identification number unique for each ad insertion. The system is based on SPACEX.12 insertion orders. It is an ASP application.

According to Shrader, E-commerce reduces the cost of doing business in a number of industries. The media and advertising segment is expected to realize a 10 to 15 percent reduction in the cost of placing advertisements through the use of E-commerce. An example from two publications, as projected by Goldman Sachs, demonstrated a potential 15 percent savings.

The system is evolving as agency of record relationships are handled effectively. Electronic insertion orders save five minutes per insertion order over paper, an annual savings of 312 man-days at a typical agency. Since differences of opinion generate discrepancies, Shrader proposes to use technology to alert people about discrepancies in the process. Finally, infrastructure improvements create vital links in customer service and other basic elements in the process.

Shrader then compared a manual and a digital insertion order, showing how, with the electronic insertion order, changes can be clearly denoted and flagged for attention by production personnel. This information also can be repurposed into a format that is standard for the individual publication and/or agency.

Shrader then demonstrated a website used by The Wall Street Journal to view insertion orders. The website includes a thumbnail file of the proof. But individual websites for each publication is not in the best interest of any advertising agency. A second demonstration involved electronically sending an insertion order, with both sending and receiving software shown. The demonstration showed several insertion orders, destined for several different publications. A copy of the actual insertion order document is an integral part of the browser image, but a mechanism is included which permits making last minute changes to the insertion order, which then is flagged as a change in the receiving software. The software–because every transaction is databased–logs every transmission, and denotes those that are not successful, as well as those that are completed without error.

Asked how permissions are controlled, Shrader said these are done at two levels–the application level and the local level.

Bill Lavelle asked whether one advertisement can be sent to multiple publications. Shrader said that this requirement can be accommodated.

John Werner asked what GCA has licensed to Shrader’s company. Shrader said that GCA owns the format, and he has licensed it for use in the advertising community. Asked whether an advertisement running in 12 magazines has 12 different suffixes, Shrader said that is up to the advertising agency.

The final speaker was Martin Bailey, senior technical consultant at Harlequin, whose subject was CIP3? CIP4? PPF? JDF? Alphabet Soup! He also chairs the CIP4 interim advisory board, and is the Harlequin representative to CGA IAP. Bailey began by commenting that every conference he has attended in 2000 has addressed computer integrated manufacturing, and that 2000 appears to be the year of standards. Most printers, Bailey observed, seem to have little idea what computer integrated manufacturing really encompasses; many seem to think in terms of simple automation. Bailey noted that in many other industries, automation has been proceeding for many years, but this is not true in the printing industry.

Moving to a discussion of CIP3, Bailey noted that the group was formed in 1995 and now has more than 40 members, from North America, Europe, the Middle East and Asia. Most members are vendor companies; MIS and E-commerce companies are just beginning to join. Most vendors sell worldwide.

The Print Production Format concept promoted by the CIP3 consortium is intended to reduce the need for repeated data. Plate scanning and manual cutter alignment with cut marks are two examples. PPF is now in its fourth iteration–Version 3.0 was released in 1998. PPF supports job/administration data, ink key presetting, register and color controls, cutting and folding data, collecting and gathering information, sewing and stitching and saddlestitching requirements, endsheets and gluing and binding, and trimming.

One element of the PPF format is a standard for interoperability, which includes the standard, a generator application, and a consumer application. But the problem with CIP3, Bailey said, is that not enough heavy iron involvement has been seen. In today’s print world, shorter print runs mean more makeready, and shorter makeready means more print runs.

Looking to the future, Bailey said that computer integrated manufacturing is insufficient. Manufacturing, he said, includes how the job is received from the customer, and how the printed product is sent to the customer, and thus needs to be rethought in the printing industry.

JDF was first formulated in 1999, and required restructuring CIP3. The concept was renamed CIP4, to become the International Cooperation for Integration of Processes in Prepress, Press and Postpress. This new model is designed to attract users, consultants, systems integrators, and many more. But, JDF still may be bigger than the overall concept envisioned in CIP4.

The target area for JDF includes sales, estimation, and billing (E-commerce); concept, design, creation, preflight, prepress, conventional press, digital press, finishing, and shipping (customer to production); planning, supply chain management, resource tracking, and costing (MIS); strategic planning (board); and (potentially) nonprint publication. Bailey said, however, that JDF may not be all things to all people. Instead, a well order construct of standards must be developed that can be used by a wide variety of people for a wide variety of uses. The core competence of those companies already in CIP4 is sufficient to begin the effort, but those not now involved may not join the CIP4 effort quickly enough. Thus CIP4 must work with many other standards organizations to effectively bridge the gaps in the printing industry. One benefit is that other standards organizations will be able to influence the ultimate CIP4.

Noting challenges to the future, Bailey noted that it will be necessary for the various participants to "move fast, and play nice." Interoperability and portability are key. The business model is as important as the technicalities, Bailey said, adding that the exact meaning of graphic arts is not clear.

In closing, Bailey invited audience members to join CIP4 at any of several levels. Further information is available at www.cip3.org.

In the question and answer period which followed, Bailey was asked about potential conflicts between JDF and CIP4. He responded that, with only a few exceptions, CIP4 embraces JDF.

Another audience member asked about the timeline for CIP4 to encompass JDF. Bailey said that the first schedule will be published in September 2000, and that input is being earnestly solicited.

John Werner asked about the ongoing role of the Fraunhofer Institute. Bailey said that the institute is acting as the interim secretary for CIP4, and that the group will elect an advisory board at GraphExpo 2000 in Chicago.

Amit Sharma of QTI commented that CIP3 and CIP4 currently has far more support in the postpress area than Bailey’s talk may have implied. He added that work is well along in areas to support web press operations. In several cases, register data, stacker data, and other information is being correlated by several vendors and announcements will be forthcoming.

David McDowell commented that by virtue of reduced fees, the CIP4 process now is more open, and asked whether all members will have an equal vote. Bailey said that the advisory board probably will have final authority, but that input from the broader membership is expected to be weighed heavily. The advisory board is expected to have a dozen or so members, and no company holds a permanent or designated seat. Partner members, however, may have an effectively permanent seat.

Another audience member asked whether the CIP4 process is essentially on hold until the organizational details are worked out. Bailey responded that in certain areas–including the web subgroup–things seem to be on hold pending precise definition of CIP4. Sharma commented that by March 2001, much of the web work should be complete.

Bruce McLean from RIT asked about other projects in the works, noting activities being conducted at RIT and by Xerox and others. He asked whether JDF will be used to automate postpress operations, or whether work such as that at RIT should continue. Bailey observed that ideally, other postpress efforts should have been conducted with input from CIP3 from the outset. Of course, issues regarding PPF have complicated the matter, but too many organizations seem to be committed to going their own ways.

It was noted that the installed base of legacy equipment is great, and that work being done in CIP4 may not have broad impact for perhaps ten years.

Rick Littrell of Polaroid noted that, in general, legacy systems cannot be affected, and that in most cases a few manufacturers will collaborate in a closed environment to develop the future. Bill Ray took issue with Littrell’s comments, noting that Heidelberg and MAN Roland are not enough to make it happen. Littrell said that including Agfa and printCafe might be enough. Both agreed that trying to focus efforts in the printing industry is a great deal like herding cats.

Bailey added that PPF apparently will be used by several vendors for some time, and that many may not embrace JDF for some time. Thus it will be necessary to accommodate PPF.

Bill Lamparter observed that large chunks of the printing industry simply "didn’t get" what is going on in the printing industry. Lamparter said that if something cannot be done with legacy systems, we are doomed to fail, and that legacy equipment must be included in efforts to develop the digital smart factory. He then asked who is doing the actual work to develop CIP4. Bailey said that work in the past was essentially divided between the members and the Fraunhofer Institute. In the future, the methodology is not clear, he said, but he expects the process to be rationalized soon. Bailey said that he agreed that legacy equipment must be accommodated.

Michael Jahn of Agfa observed that with the fragmentation of the industry, and with relatively few industry suppliers participating in CIP3 and CIP4 efforts, Agfa has found implementation quite difficult. He said that the industry now seeks a process control mechanism (probably JDF) that will do what PDF has done to page communication. He said that many people, who were attempting to connect at Print City at DRUPA, simply did not have the proper connection mechanisms available.

Bill Lamparter asked whether he could attend the GraphExpo 2000 meeting, if he joined CIP4. Bailey said that he could.

David McDowell said that many small islands of automation remain to be completed, before the larger process can be accomplished.

A user, who said that he deals with legacy systems all the time, noted that it must be possible to implement CIP4 or any such system throughout a plant, but that it must be possible to get started and to add capable systems through time. He suggested that JDF may be the way to get going.

The final portion of the conference, Session IV–Industry Round Table, was chaired by William T. Davison, president of Point Balance. Davis began by congratulating the R&E Council and GCA for putting the conference together, but said that the true test lies in perpetuating the idea in the future. He said that value can be added only if more attendees can be attracted.

Moving then to the conference wrapup, Davison said that just talking about the digital smart factory is too limiting, in and of itself. Noting that when the process is viewed as beginning with design and origination, the smart factory concepts is insufficient. Front end customer relationships must be managed, for example. More appropriate would be a discussion of the complete E-business enterprise–or, at least, the locus which extends from design through delivery. Of course, everyone at the conference comes at this problem from a different direction and with a different set of boundaries. At the moment, no one seems to be driving the smart factory concept upstream into the design area.

The ability to navigate color across boundaries may be a seriously limiting gate in the smart factory concept. From a page assembly standpoint, the fundamental unit of work is the page. In the Valassis paradigm described by Ray, the page is no longer the base unit. Thus working at the cellular level is much more conducive to implementing the digital smart factory concept. Perhaps, Davison said, we need to move more toward the cell concept, and to address the color navigation issue directly.

Moving to standards, Davison commented that these are usually driven by one or a few vendors. The leader in the standards area, said Davison, is JDF, which probably will emerge as the true standard. Davison noted that "the network is the computer" and that as a result, the actual control of many mission critical systems is being diffused across the network. Intellectual property issues will only become more important.

Finally, Davison observed that for this group to be influential in how the digital smart factory unfolds, more effort must be put into a hypothesis of how the system would work. A number of dimensions are important–some are broad and others are deep. Those issues include:

• Infrastructure
• Process
• Data Management
• Application Architecture
• Management Control and Decision Making
• Internal Resources
• Supplies

Responding to Davison’s postulates, Mark Jones commented that the industry is still craft based. Talking about the page-centric concept, Jones disagreed, saying that the next level may be the database, and whatever output the database drives. Jones commented that he needs to make money until JDF emerges as a practical solution. And the customer does not care about JDF, Jones said, adding that it remains to be seen what actual products will include JDF. He closed by saying that the digital smart factory begins on the customer’s desktop.

Jim Harvey noted that color has various levels of meaning, depth and detail. As the ‘cells’ discussed previously move through the production process, it is necessary to include additional information, such as: Who approved this? Do we have the necessary copyright approvals? Who revised this? Thus it will be necessary to automate many of the business elements before the digital smart factory can be realized.

Bill Ray then weighed in, commenting that, eventually, the concept of cells will revolutionize the book industry. Digital paper ‘books’ are now being developed, and the cellular approach will permit ‘binding’ information in the optimal form for the use intended.

David McDowell observed that color is not the gating issue, but merely evidence of a greater issue. Since the originator and the user want to understand how the ultimate output will look (a cultural problem) color is only one element. Today, the art director will not live with a ‘best efforts’ concept.

Davison responded by commenting that a cultural change thus is necessary before the ultimate result can be realized. Steve Franzino noted that "more, better, faster" is quickly becoming "more, more, faster." He noted that while publishers hate printers, authors also may hate publishers.

Bill Ray commented that the concept of publishing will change.

John Iobst of NNA expressed concern about the Lemelson patents, and how they may impact the concept of the digital smart factory. At the request of Davison, Iobst described the problem centered around the machine vision concepts. Ray observed that someone will break the Lemelson patents. Dennis Mason commented that the fact that certain manufacturers have settled with the Lemelson estate only serves to strengthen the patent.

Davison then threw open to floor to a discussion of the conference, asking what would add value to proceedings such as these. Murray Oles said that while the content was good, he was disappointed in what he was able to take away from the meeting. Oles said that it is necessary to produce something that participants can take away. Littrell noted that it is difficult for a smaller company such as his to rationalize participation, adding that some areas of the digital smart factory are being implemented today. Unfortunately, said Littrell, the conference content was 95 percent future. Larry Warter noted that while he agreed with Littrell, the conference discussions must be broadcast to a much wider audience, with different needs than the participants. Littrell noted that if users attend the meeting, vendors will follow. Jim Harvey asked whether a meeting to discuss technical issues should have preceded the meeting. Not too many thought so. Bill Lamparter commented that while the meeting was good, it contained little of interest to the smaller printer. Lamparter suggested a related conference dealing with real world issues of interest to the printer. Printers must be educated more than is being done now, he said.

Steve Franzino noted that the R&E Council has struggled for years with the digital smart factory concept. He said, however, that it is necessary to conduct discussions at a very high level at this time. Orlando Boleda suggested that the R&E Council might be able to bundle the digital smart factory concept with other Council seminars and conferences. Ron Mihills noted that, unfortunately, chief technology officers and MIS personnel should attend, but either do not exist or are not interested.

Mark Jones commented that the value of the conference lies in its high level nature, adding that he would like to bring the people who do the work to the conference in the future. Jones said that, despite the motto of the R&E Council, ‘not too much was heard here first’ at this meeting.

Bill Lavelle then weighed in, commenting that print is the outcome of a ‘money spend.’ He then observed that the attendees are the problem, by virtue of training and experience. Thus the digital smart factory may have more to do with management. It may be of more interest to the ‘chief strategy officer’ than to the ‘chief technology officer.’

The conference was closed by Ron Mihills. No date was set for any subsequent meeting.

Co-Produced by:

Research and Engineering Council 816 Rappahannock Dr. PO Box 1086 White Stone, VA 22578-1086 Phone: (804)436-9922 Fax: (804)436-9511	Graphic Communications Association 100 Daingerfield Road Alexandria, VA 22314 Phone: (703)519-8167 Fax: (703)548-2867