Intelligent Manufacturing € July € 1995 € Vol. 1 € No. 7

Is MRP Dead?
No, But Experience Is

By Bob Turek

Over lunch with two good friends recently, I posed the question: "Is MRP dead?" This question never fails to provoke a good discussion. The subject is also considered whenever a new buzzword acronym (BRP, JIT, QRM) is introduced; I read many articles very carefully for new thoughts and ideas - they are few and far between. Usually a rehash of "old" features and benefits is replayed as a new and different approach. So, I thought you might be interested in a personal roundtable conducted with two individuals - Ray June and Butch Rey-Hipolito - who have a combined 40+ years of experience managing and implementing materials functions in a variety of companies.


The Importance of MRP Basics
June and Rey-Hipolito, manufacturing managers at JBL (Northridge, Calif.), a $350 million manufacturer of audio equipment, laughed as they exchanged "Didn't we just talk about this?" looks in response to my original query. Ray June, a successful materials manager, half-jokingly said that he thinks, "Just-in-Time is dead." June has successfully implemented what I would call JIT disciplines in aerospace, engineer-to-order and consumer electronics environments. Of course, I requested a clarification of his JIT death pronouncement with an intelligent "Huh?"

June said that if more people understood basic MRP principles (and he meant really basic - things like bill of material construction, average lead time accuracy and inventory accuracy), and then attempted to implement MRP, they would be far beyond any attempt at implementing buzzword technologies. Often JIT is "dead on arrival" because of this lack of experience. June emphasized that people lose sight of the fact that JIT is a philosophy and MRP is a tool.


"To Order" Companies Are Different, But...
This sounded too easy, too elementary for me. So I pursued bigger game, and asked, "What about 'to order' environments where there is very little commonality between components, where tools like 'work order' or 'sales order line' MRP (i.e., taking the specific demand from a sales order and letting it generate multi-level requirements, often thought to be a quick way to generate project-specific work plans and purchase plans) are required?"

June looked at me as if to say, "You just answered your own question!" You see, you still have to know the basics and the ramifications of not taking advantage of common parts planning before retooling MRP for your benefit. "Sales order" MRP is really just another version of MRP.


Forward Finite Simulation for Customer Order Promising
What about promising product delivery for a customer using material and capacity simulation by proposing a start date and letting the system finite forward schedule a customer promise date that you can live with? This time Butch Rey-Hipolito, a man as equally experienced as Ray June, patiently explained that people who don't know how basic MRP works through hard-won experience cannot have any idea how a complicated simulation like I described works, let alone how to use the output of such information intelligently. He made too much sense.


Agile MRP
I know I appear to be inventing another buzzword, but I'm only mentioning agile MRP to make a point: You can't be agile before you're MRP literate. MRP literacy requires training and, most of all, experience - experience struggling with the creation of data for MRP, strategizing about the application of it, and experience carefully measuring the accuracy of the output in your dynamically changing environment.

For example, what are the appropriate lead times to complement your business objectives? This is one of the toughest questions to answer as you move toward an agile environment.


MRP as a Purchasing-Only Tool
You may decide that MRP in your environment is more valuable as a purchasing tool. Patiently, Rey-Hipolito said, "Usually, people say they don't have the discipline to apply MRP to their manufacturing planning." He went on to talk about how these types of companies end up misusing advanced simulation tools by expecting miracles and end up purchasing all their material in the wrong amounts and at the wrong times. The net result is poor inventory control, heavy reliance on previously "more reliable" manual systems, and cursing of the "new" computer system for its lack of results. Believe me, Rey-Hipolito has been there.

The flip side of this is a company that is largely engineer-to-order and understands MRP and the degree of reliance to put on the output of advanced finite forward-scheduling tools. "Degree of reliance" just means that there is a healthy dose of wisdom in regard to knowing when the human being must take over from the computer.


The Tale of Two Companies
I talk to manufacturing companies every day - systems sophisticated ones and naive ones that have no manufacturing software experience. Let me ask you a question: Which type of company recently told me that they loved the idea of doing finite forward simulation of every customer demand? It was the naive company.

Contrast that with another experience I had in talking about the same concept. The director of operations, who is very experienced with managing manufacturing operations in computerized environments, expressed reservations about the simulation concept because she realized the underlying sophistication of the tool. It is actually assessing current capacity and inventory levels and, given a proposed start date, simulating a "real" start and due date with all of the constraints identified. While realizing the value of the tool, she doubted whether she had the resident expertise to utilize it - today. She sees that she needs to educate and train some of her people in the basics, and limitation, of MRP in her company.

The second company realizes what the tool is and what it takes to get up to speed to use it effectively. The first company will probably false start their implementation until they realize they need to import some experience and train key individuals.

The message is this: MRP, or some form of it, must be the basis for manufacturing planning. The different forms it takes, through first understanding MRP completely and only then its limitation in specific company environments, are wonderfully supported by many software offerings in the market. Don't make the mistake of trying to implement a buzzword technology before its time.

Bob Turek is a Los Angeles-based senior consultant for Visibility (Wilmington, Mass.), a developer of business information systems for agile manufacturers. He can be reached at (805) 268-8808.