SPECTRUM 2001 CONFERENCE NOTES

Back to Index Page

Best Practice Session - Process Automation Discussion Group

Led by:

Alan Darling
COO
Quality House of Graphics

Alan pointed out that standards using XML are being developed, but one of the problems is that it up to the vendors to implement and the users are not involved in the process yet. With all these tags being put into the system, Alan asked, where are they going to go? What are we going to do with them? If we are integrating systems from different vendors and the system breaks down, how do we recover? In short, how do we make our systems "fault tolerant?" As an example, Alan pointed out that different systems could interpret similar concepts differently. Overprint of white type, when it hits a rip may be interpreted as an error, it may be dropped, or it may be set.

The first implication of total process digitization is the quality control must move upstream. Other key factors include digital rights management, system implementation, software upgrades. One ques-tion is who starts this process? The printer, the publisher, the prepress standard, the agency? Jerry D'Elia said that this time, since we are dealing with business issues, the company CFO and executive management must be involved, not just production people. Jerry believes that whomever supplies the file must take responsibility for its quality and suitability for the production process. Betty Maul pointed out that it used to be the creative director who was the file source person, but now they even get files from outside sources such as photographers.

Perhaps there should be a file certification process? This raises the issue, how do we get this infor-mation back to the users who aren't readers of the specifications and don't attend conferences such as Spectrum. The whole food chain must be involved in the process of migrating to process automation.

Alan pointed out that if you automate your systems, you couldn't expect to staff your plant with cheap, untrained labor; rather you'll need smarter more highly trained employees who can fix the process when it fails or breaks down. This is the double-edge sword of process automation, you may reduce labor by head count, but you will need much more expensive labor where labor remains necessary.

John Sweeney of GMI, a provider of process automation systems, asked the question, "Who pays the bill for process automation?" Michael Wienglass of EZ Rider magazine said that you can't imple-ment all process automation that is available at once, but must pick and chose from budget year, to budget year.
Point for the book: an objective must not be to go to a dumb work force, and because it can be done (or automated) doesn't mean that it should be automated. Michael also said that he would like to see the processes be easier to understand.

The user must be able to articulate what they want "automation" to achieve. According to Richard Browne, the objective may be best reached by a method other than electronic or computerized automa-tion. Everybody felt that process automation is going to happen, but nobody felt that it was going to happen in the next 12 months. Consensus is that automation will probably occur in 3 to 5 years.

TOP