January 1996 € Volume 6 € No. 1

The Customer Connection

Finite Is Fine

By Tom Wallace


Many of you APICS members work in a field called--variously--logistics, planning, materials management, production control, etc. A long time ago, when I transferred into this field, I called it an exercise in futility. Nothing worked right. We were always wrong. We had lousy tools to help us do our jobs.

In those days this entire function was in its infancy. Now it's in its late adolescence, and in another 10 years it'll be almost completely grown up. Working in this area no longer has to be an exercise in futility because solid, proven, effective tools exist to help us succeed in the complex task of keeping demand and supply in balance.

One of the signs of this maturity is an ability to assimilate new concepts and approaches into the existing body of knowledge. The technique called finite capacity scheduling or finite loading comes to mind.

Back in the old days of the field's infancy, intense debates raged between the "merpers" (MRP believers) and the "finiters" (advocates of finite loading). The finiters said "MRP assumes infinite capacity and that's absurd. Nobody has infinite capacity. Besides, there's a need to have computer support for decision-making regarding changeovers, sequences, alternate routings and resources, etc."

The merpers responded: "MRP doesn't assume finite anything. Capacity requirements planning (also called infinite loading) calculates the workload irrespective of capacity, and displays the information to the people involved so they can make the right decisions on overloads, bottlenecks, offloading, alternate routings, etc.

"Besides" they went on, "finite scheduling messes up the priorities set by MRP, so how do you ever get matched sets of parts into the finishing operations? Where does this stuff work in the real world?"

Well, they were both right and both wrong. Yes, the early versions of finite scheduling did distort the priorities needed for matched sets of components. Nor did it seem to generate real operational results once it got past the pilot stage.

But, these early attempts were trying to do an important thing: help human beings make better decisions about scheduling the complex environment of the plant floor with so many variables-people, machinery, tooling, material, priorities, and on and on. The problem was that they went too far. They "helped" the human beings too much; they tried to make the decisions for them.

This approach, referred to as "rule reliant," attempts to capture all the plant floor variables and build them into mathematically-based rules. The result: enormous complexity. Complexity violates the principle of obvious logic, which states that if people don't understand the underlying logic of the program, they will either follow it blindly or ignore it completely. Both are bad. The bottom line on rule reliant software: It didn't work.

There's a new breed of finite scheduling software called "decision support." It doesn't make decisions and then tell the people what to do; rather, it calculates the "best" scheduling solution to the situation presented to it. Then it displays this solution and the relevant factors-stockouts, cost, inventory, changeovers, output, etc.-to the person controlling the software. Capture that: The human being controls the software rather than the software dictating the schedule.

Today's finite capacity scheduling software runs on PCs; it's interactive; it's graphical; it's very user friendly; and it works. When used intelligently, it enables plant schedulers to develop better schedules. It helps them to balance demand and supply at the most detailed, finite level.

These packages work well when they're tied to the company's main resource planning system. In this environment, the near-term master schedule is downloaded into the finite scheduler and the scheduling person begins the interactive simulation process. The scheduler selects the solution that he/she prefers, and this schedule is then uploaded back to the master schedule or down to the plant floor system for execution.

If you'd like to learn more about this new breed of finite scheduling software, I suggest you start with an article titled "Interest In Finite Scheduling Is Growing ... Why?", which appeared in the August 1994 issue of this publication. The author, Andrew Gilman of Waterloo Software, states that decision support scheduling software "allows the scheduler to balance delivery needs against efficiency based on today's conditions and today's customer orders, not simply according to some predefined rule."

Since better plans and schedules mean better execution, it's logical to expect better results from finite scheduling. Successful users have reported dramatic improvements in on-time delivery, lead times, inventory turns and plant efficiency.

The message: Don't tune out finite scheduling just because you had a bad experience 15 years ago--or because you heard that it didn't work. Today's new breed of finite scheduling software is helping more and more companies do a better job of balancing demand and supply--meeting customer needs with the most efficient use of resources.
Tom Wallace is an independent consultant based in Cincinnati. He is the author of Customer Drive Strategy: Winning Through Operational Excellence (1992) and editor/author of The Instant Access Guide to World Class Manufacturing (1994). Tom is co-director and a Distinguished Fellow of the Ohio State University's Center for Excellence in Manufacturing Management.
Copyright © 2020 by the American Production and Inventory Control Society Inc. All rights reserved.
Click here to return to the table of contents.