Editor’s note: Jonathan H. Markey is president and chief executive officer of the North Jersey Media Group in Hackensack, N.J., publisher of The Record and other New Jersey papers. In late 2002, Markey spoke at Ifra’s Printing for Profit conference in Munich, Germany. His presentation focused on NJMG’s use of computer modeling and simulation tools for operations analysis. The following is an edited version of his presentation.

We have recently discovered a simulation tool, or computer-modeling concept, that can be quite beneficial in planning successful capital projects. The application also helps us evaluate performance and measure costs relative to expected results.

Integrating multiple functions

First some background: Several years ago we embarked on an ambitious program to move our systems infrastructure to a more technologically advanced status.

Our mission was to integrate multiple functions into a more stable and integrated platform, thus enabling us to manage our business more effectively. We also wanted our data to be available in real-time and also have it serve as a foundation for future modeling.

In looking at our goal, we understood that in many areas we would probably be forced to take technology and software from other industries and adapt it to our needs.

Our first major effort involved the replacement of our entire business system, replacing several vendors with a single-enterprise resource planning system from SAP. That effort put a floor under our ability to effectively do cost analysis for all operations.

It is within this progressive framework that The Record started using computer modeling. We, like many other newspapers in the United States, were looking into the increased zoning needs of our products and knew we needed to improve our existing mailroom operations at our main facility.

Measuring performance

At that time, we had also started to look at replacing our legacy buffer system. Our mailroom inserting operation also needed to be overhauled. We felt we could improve our overall operation by decoupling the press from the mailroom operation and wanted to see which vendors would be able to provide a solution for our needs.

We wanted to be sure we could measure the performance characteristics of the latest storage and inserting equipment on the market. We sent out a request for proposal that asked how each individual inserter or storage system would match up with the existing press and distribution equipment. We needed to know the real-world performance characteristics that would distinguish each system.

Above is a test model of the simulation possibilities supported by Extend software from San Jose, Calif.-based Imagine That!
Photo: Ifra

Before the overall mailroom capital request could be submitted, we wanted to test the concept of decoupling the press and mailroom completely. And just as importantly, we wanted to test the how well vendors’ equipment performed.

Simulation helped

To do that, we decided to simulate our mailroom. We first built a model of our existing operation. At the same time, we began researching which vendor we would use to supply computer-modeling software. Our requirements were straightforward. The software had to be:

• relatively easy to learn and implement;

• run on standard PC and Mac platforms;

• be inexpensive to purchase; and

• not need support from the IT department to run.

We looked at several simulation software packages before choosing Extend, an application developed by Imagine That!, a vendor based in San Jose, Calif.

Extend, a general-purpose package, is disarmingly simple to package together. No computer programming is needed. It is an icon-based system. And it is as simple or complex as you have time to make it. Finally, Extend is affordable. It cost us about $10,000 to run the application on a single workstation.

To represent a press, we used a generator icon and set the output at 35,000 papers per hour. To represent an inserter we used a machine icon set to insert papers at an average throughput speed of 12,000 papers per hour.

Between the two we placed a storage system capable of collecting papers at an 80,000-per-hour clip.

As I already mentioned, our goal was to see what happened if we decoupled the press from our existing mailroom. The first operation we studied was buffering.

What we needed

We looked at three vendors — Heidelberg, GMA/Muller Martini and Ferag. From these suppliers we received the specs on the machines we would put in, the speed that a system could handle straight off the press and what the storage system could hold.

In the first round of information gathering, most of the vendor-supplied specification information was inadequate to ensure system performance and success. The speeds were averages and/or peak performance. What we needed to know was how well the machines would perform day in and day out. This was a critical point of discussion and debate and created new challenges for the vendors.

We needed effective data to serve as the baseline for the simulation process using the Extend engine. Without empirically developed data, the use of simulation tools has diminished value.

Unfortunately, most of the information available from vendors is based largely on engineering projections and theory. As a result, it makes the outcome of any simulation much less reliable. We were able to offset that, however, by feeding into Extend our manufacturing history. Fortunately, we maintain several years of manufacturing data; consequently, we were able to loan an entire year’s worth of data into the modeling software.

‘Totally honest’

Our initial computer modeling showed we still had far to go. For one thing, it was quite evident that the data from the vendors lacked sufficient credibility to enable us to create a stable model environment.

We were able to go back to the vendors and help them better understand the limits of their performance metrics.

This discussion was one of the main benefits of this process. We were able to engage in a totally honest discussion to define what we needed. The vendors, in response, were able to pinpoint their performance metrics, making simulation possible.

Armed with the new performance specifications, we ran the simulation. Initial gating factors were solved, but new, unforeseen, challenges began to emerge. Being aware of these challenges let us solve performance issues before contracts were signed.

A side benefit: Since we had the ability through SAP to look at real-time costs we were able to couple the simulation runs with cost information to establish operational economic models that have proven reliable.

The new mailroom was installed and has performed to our expectations.

Looking ahead

We are working on examining our current product mix to compare new press technology with our current infrastructure.

This takes us into new and somewhat uncharted territory.

To do this, we are currently working with major press vendors Heidelberg, MAN Roland and WIFAG. Our efforts are focused on validating manufacturers’ specifications and seeing how new press capabilities might support new businesses.

In the United States, most newspapers only buy new press equipment because their current presslines are aging or obsolete.

Rarely would a newspaper anticipate that a new press might impact its operating structure. Nor would a paper assume that less equipment could produce more products.

Quite frankly, this is logical. Our industry and its capital capabilities are appropriately defined as “mature” and as such, they rarely offer opportunities for dramatic innovation.

Freedom from constraint

This is where simulation can open up your thinking. Simulation lets you examine previously unexplored concepts. At the same time, you can find out how much each step of the concept will cost.

This freedom from constraint is leading us toward answers that previously would have been unthinkable.

When we put the information into the model it will allow us to create realistic virtual-world press runs, complete with planned stops and restarts, unplanned production interruptions and various process flow concepts.

As we run the virtual press runs we can simultaneously operate costing models that yield the operational economics of the various iterations.

As the two models progress we are able to build a very solid foundation for decision-making. One of the greatest values of this approach is that in many cases constraints almost always turn out to be only part of the answer.

This pushes your horizon back in thinking about the problem. Bottom line: Managing expectations and making implementation happen as predicted.