Commentary & Analysis
FREE SPECIAL: Snap Functionality On and Off: Building Workflow with Backbone
By WhatTheyThink Staff
Published: April 23, 2002
Alan Sullivan, IBM Printing Systems consulting & integration services practice leader (www.ibm.com/printers), works on the front line evaluating digital workflow requirements and developing automated processes for output management. He's seen a lot of approaches -- some good, some not so.
"Customers often evolve their workflow automation by implementing individual products to handle various pieces of their workflow, like composition, sorting or insertion management," he says. "Some vendors also try to hardcode components together, such as tying a composition engine to the print manager. People need the ability to modify business processes, but it's hard to change when they have four or five vendor-specific wired applications together."
A Model of Modularity
In order to assemble a workflow solution that is flexible and practical for the output industry, Sullivan suggests an approach that has a workflow "backbone" at its core. The solution is based on the integration capabilities of a workflow engine, rather than just products glued together with multiple interfaces and databases for operators.
With a workflow backbone, users can "snap" functionality on and off as needed, and these tools don't necessarily have to be all from the same vendor. The backbone model might include Doc Sciences or Exstream software for composition or personalization, Sefas products for print stream manipulation, IBM InfoPrint Manager for print management, and Pitney Bowes tools for insertion control.
"Most other vendors create solutions by hard gluing applications together, but this is where IBM differs. We use a workflow backbone to map any number of products together, from composition tool to print manager to inserter control. Where the process needs specific functionality, the engine will invoke other vendors' products to provide it." The IBM InfoPrint workflow model gathers information and presents it to a component, such as a composition tool, the way that component wants to receive information, yet there is just one consistent user-facing interface for operators to learn.
A workflow backbone can use the same engine for multiple applications, such as POD processes to produce digital books from a digital library, or to create 1:1 marketing materials, or to generate personalized benefit statements.
Information Exchange Drives Digital Workflow
Sullivan says that output management has been one of the last areas to think about digital workflow, but it represents a large opportunity, especially for operations with high volume output requirements, such as telecommunications or financial services. "They are very interested because the manual processes in their applications can be extremely expensive. Some operations might have half a dozen operators tracking data manually across multiple printers and inserters. Or when a file involves a million statements, and the inserter mangles 300, someone has to manually figure out which ones were destroyed and go back to the MVS host to rerun the job." It's no wonder workflow automation is getting noticed in the output marketplace.
Good, fast communication is essential for good, fast workflow. A sophisticated tool must be able to rapidly describe to the workflow engine what process is occurring. The object is not just to control data, but to efficiently and automatically manage equipment and resources. IBM Printing Systems’ backbone-based workflow model focuses on processing and tracking information about applications, data, hardware and operators. "If an operator is having trouble, stats compiled by the InfoPrint Workflow engine will reveal the problems. If a printer is down, the backbone knows it is unavailable and routes output to another device."
In upcoming columns, we will continue talking with IBM’s Alan Sullivan about the costs of digital workflow, the importance of your staff in workflow development, and other useful ideas.