Under Control


Manufacturing execution systems are a kind of middleware for manufacturing. A term originally coined by Advanced Manufacturing Research, Boston, an MES is the means to bridge the gap between process control and production management systems. In addition to being the means to better integration, the MES includes specific functionality for scheduling, quality control, document management, and others.

The history of MES began with applications developed by specialists aimed at tackling specific problems. Tracing operations, in-process materials management and scheduling are some of these areas. Coupling this functionality with relational database management systems, for both real-time and archival applications is crucial to effective utilization of MES.

Users, quoted in the trade press, cite productivity improvements of up to fifty percent for some operations. But it's no secret that while some enterprises are getting real improvements, others have struggled to realize benefits using the "traditional" MES approach.

As middleware, an MES must interface with shop-floor process control applications as well as with enterprise resources planning systems (ERP). AS most readers of this magazine are well aware, ERP is a transactional system that typically includes functionality for financials and accounting, human resources, manufacturing, and distribution. ERP systems are increasingly involved in supply-chain management.

My experience has been with the factory floor proper. But I, and you, need to understand the management aspects of the floor, factory and enterprise.

As middleware, an MES must interface well with the floor as well as the upper mainframe functions. Historically, companies patched together execution systems without even knowing what they were. Then came the promise of a single, integrated system. But costs, and the inclusiveness of the functionality provided with the system, raised questions.

We expect the future will bring a matrix style of applications suited to the MES concept. In this way, an overall framework allows each vendor to work within the matrix, offering fully compatible components to an open system. But still allowing for innovation in implementation.

Why is the title of the column "The emperor's new clothes"? It all started with an aluminum-tube journey from New York City to Manchester, New Hampshire. My seat mate was a high-paid consultant with a blood alcohol content above the legal limit. He lauded the existing structure of many information systems, since it assured consultants of an obscene profit for a long time. Need for customization, difficulty of installation, and excessive procedural code were all included in his praises. Installations that took up to a year seemed a glorious thing to him.

Listening to brought to mind the story of the Emperor's New Clothes. In that story, the emperor was conned into believing that the finest clothes could only be seen if the viewer had a pure heart. Otherwise, the fabric was invisible. The con men took the money and skipped. When the emperor traveled in a parade, wearing nothing, a child asked, "Why is the emperor wearing no clothing?" Since the heart of a child is pure (except in LA), it's obvious that all could see that the emperor was truly traveling in the buff.

In the Angel investment business, we always ask about exit plans. The future is hard to foresee, and we are interested in the ability of the company to change business vectors. Plant software should have the Darwinian aspect of survival and quick adaptation. Software that can only feed off of Eucalyptus leaves is vulnerable to cusp events. Flexibility is not a key feature of complex, consultant ridden systems.

Venders are smart and working hard to respond to these criticisms. The new silicone and the ADSL/WWW communications will certainly enable systems in a way unimaginable today. Web and Internet coupling are, so far, only now being recognized in the manufacturing applications space. The Oscillating fashions of distributed/centralized architecture continues. Work at the Santa FE Institute suggests that the technical answer is a hybrid of both the distributed and centralized approaches. We shall see. remember Unix?

So watch out for interpretability, Web compatibility, costs of install and change, and buzz-word decisions. Fast and flexible response to the business environment requires a behavior business model, not a procedural model. The future framework concept and hardworking vendors are our ultimate solution. If the business fits the model then go for it. Otherwise, keep your heart pure

As appeared in Manufacturing Systems Magazine October 1996 Page 136

Manufacturing Systems Magazine Articles
References - Table of Contents

Send mail to rmi.info@barn.org for more information.
Copyright © 1996, 1997, R.Morley Inc. All Rights Reserved
Last modified: October 1997

Please send mail to webmaster@barn.org regarding web site structure.

R. Morley Incorporated
614 Nashua Street, Suite 56
Milford, NH 03055-4992 USA
Tel: 603-878-4365 FAX: 603-878-4385