Printer Friendly

Overcoming CIM hurdles.

At IMTS this year, one of the underlying concerns voiced by many was interconnectivity. The concern that moderates the enthusiasm of many potential buyers is how well a new machine or system will integrate into the user's existing system.

The problem is that in most companies multiple computer systems have been installed in production engineering and on the factory floor. Each is well adapted to its task and may even be able to communicate electronically with the next downstream system, but that's as far as anything seems to go. CIM has essentially been stopped in its electronic tracks by the willy-nilly implementation of computerized systems.

Companies have connected CAD to CAM and CAM to DNC. This constitutes computer integrated manufacturing (CIM) to some. But as any of its users will tell you, it isn't doing their bottom line any good because, at best, it merely matched competitors' moves instead of providing a competitive edge.

The way it is

There is no technological reason why this state of affairs should persist. The "hardware" and connectivity problems have been solved in various ways by UNIX workstations, by networked systems, and by the increasing openness of the major proprietary operating systems and databases from companies such as IBM and DEC.

Yet in most companies today, the environment in production engineering and on the factory floor consists of at least one of every type of computing device ever developed: PLCs; NC, CNC, and DNC systems; cell and factory controllers; CAD and SPC on workstations and PCs; CAD and CAPP on minicomputers; CAD and MRP on mainframes; CAM on PCs; plus electronic on-line gaging, digital micrometers, etc. Not to mention the truly obsolete: tape-controlled machine tools and hard-wired PLCs and NCs.

Much of this hardware and software is rightly castigated as a dinosaur on the verge of extinction. But snug in their glass houses and bureaucracies, "legacies" among these systems hang on.

Each department uses what it likes (or what it was stuck with by some forgotten interdepartmental horse trade). Never mind whether its output can be used by other departments needing its data. This is why data is entered and re-entered, created and recreated again and again in manufacturing companies.

There is, too, the lament of the overly multi-vended. The problem would not exist, the refrain goes, if companies would just standardize on one system from one vendor. True, but no one system has yet been devised to do CAD, CAM, MRP, SPC, text processing, general ledger accounting, billing, receivables, and track sales orders, inventories, shareholders, and accounts payable.

Options and alternatives

Paper-based systems--the status quo: print/plot it out, mark it up, send it on, followed by manual re-entry of the data downstream, over and over again. The medium of transmission is the interoffice mail--the US Post Office in miniature, but without the incentives.

Technical documentation and word-processing systems: treating the symptom--the paper--as if it were the disease, which it is not. The mediums of transmission are networks designed for PC-level tasks with a hodge-podge of bridges and routers as links to the real world.

Imaging systems: taking pictures of the symptoms comes no closer to treating the disease. The networks are better, but the images can't retain the intelligence underlying the document; for example, the vector graphics data in a rasterized CAD file.

What is needed is a software solution that takes advantage of what is already there. What is there may not be "standards" in the objective sense; however, the prerequisites for successful automation already exist. All that's required is a fresh look.

The information resides in known places. It is gathered with known--the repetitive--procedures. There are standard formats for information presentation. The underlying hardware technology is available in today's workstations, established communications, and Ethernet/DECnet. What is needed is a way to bring it all together into a useable framework.


The situation is actually an excellent opportunity for automation. Turn the paper forms, since they are already "company standards," into an electronic format. Capture the present-day business procedures by which these forms are filled out and script them to individual boxes in the forms. Build electronic links between the scripted boxes and the databases containing the information that goes into those boxes. The resulting forms, in essence, become the graphical user interface, or GUI. Does it work? It does, and the common denominator is multimedia work instructions in all the amazing varieties:

* Shop-floor travelers

* Inspection reports

* Assembly instructions

* Materials management documents

* Tool setup sheets

* Factory data collection forms

* Requests for expenditures and justifications

* Methods and process sheets

* Job tracking sheets

Successful users

Cummins Engine Co, Columbus, IN, a manufacturer of diesel engines, uses this approach for computer-aided process planning systems for both of its major new programs for the 1990s: high-pressure fuel injectors and a new mid-range engine factory. It is also a key element in complying with Cummins' Corporate Production System, which mandates standards for process documentation company wide.

Eldec Corp, Lynnwood, WA, producer of complex electromechanical systems for aircraft, has taken this approach to tie together its on-line SPC with its product data management system and, through that, all of its CAD systems.

General Dynamics has used this approach to develop a "hypergraphics" capability linking engineering schematics with CAD drawings used in monitoring space shots during Cape Canaveral countdowns.

General Electric Co's Power Generation Business also uses these concepts and architectures. In their Gas Turbine manufacturing operations in Greenville, SC, GE is bringing its process documentation into compliance with ISO 9000 standards. In their main engineering operations in Schenectady, NY, the approach is being used to develop a variant-type process planning system.

Raychem Corp's Advanter machining operations, Richmond, BC, Canada, started out making electrical adapters and connectors in a plant with just one piece of paper--the shipping tag. When that business fell off, the plant was acquired by a manufacturer of optical "jukeboxes," primarily because of the plant's built-in information automation.

All of these firms are using Linkage Multimedia Work Instructions software from Cimlinc Inc, Itasca, IL. All these firms also have, at one time or another, tried information-integration applications such as these with a CAD system, word-processing and technical-documentation systems, and imaging systems. All failed at previous attempts, but learned valuable lessons:

* CAD systems have not text tools and are poor at handling non-CAD data.

* Word-processors have limited vector-graphics handling capabilities, and little or no graphics (vector or raster) editing capabilities.

* Imaging systems require building huge raster-image databases, creating yet another set of problems in managing product configuration and engineering revisions.

* None of the methods they tried had efficient macro-programming languages for scripting procedures necessary to automate data retrieval.

* None of the systems had 4GL interfaces to relational, hierarchical, and flat-file databases.

Finally, none of the previously tried systems provided a way of returning shop-floor and production engineering information upstream. Without such feedback, even highly valuable information such as the difference between "as built" and "as designed" are lost. Without an upstream flow of product and process information, no manufacturing organization can hope to improve at anything more than an incremental pace--a pace we all recognize as too slow for the 1990s, let alone the 2000s.


Forgoing the benefits of information integration is a lot like putting a million-dollar machine tool in the same spot in the plant where it replaced three older NC machine tools. This obvious placement creates a bottleneck for a very practical reason. Suddenly, every job in the plant has to be routed across the new machine tool. Otherwise, the managers who bought it risk not reaching their justification target. The work flow suddenly jams. The effects on productivity can be catastrophic unless the root causes are addressed.

Similarly, new machine tools have a voracious appetite for data and, consequently, their operators will generally be scrambling for information. Typically, a 1990s model CNC machine tool will require two to five times as much data as each of the machine tools it replaces. If this kink in the information system isn't addressed, there will be a jam-up of information, differing from the jam-up of parts only in its invisibility.

In addition, the operator also has a huge need for information. Setups are different and perhaps simpler, but they will also be more frequent because the machine is more productive and more flexible. If it didn't complete jobs sooner, it wouldn't have been purchased.

What is required here is a combination of data, information, knowledge, and understanding. All of it must be supplied accurately and timely to keep the factory running more productively than competitors' factories. In today's manufacturing companies, all this information is already on line somewhere--inspection results, NC tape images, CAD drawings, and process plans. The movement of this information is what virtual reality seer Brian Park of Flogiston Corp terms "laminar information flow."

"Linkage's range is so wide that it could be used as the standard horizontal application throughout a company. This is a tool for the 1990s, and it patches right in with concurrent engineering practices, empowered workers, and horizontal power structures."

The system does not replace existing databases "but lies lightly across all of them," Mr Parks adds. "The existing data dinosaurs can have their springs wound up each morning without being upset by this upstart."

For more information on electronic linkage systems from Cimlinc Inc, circle 308.
COPYRIGHT 1992 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:computer-integrated manufacturing
Author:Stovicek, Donald R.
Publication:Tooling & Production
Date:Dec 1, 1992
Previous Article:Innovative probe upgrades CMM.
Next Article:Transmission plant scores 6 sigma.

Related Articles
Computer integrated manufacturing: a new look at cost justifications.
Automation offsets labor shortage.
Mazak bets big on future of CIM.
New CIM software lets you design plant automation networks yourself.
To invest in technology or not? New tools for making the decision.
CIM connectivity: freeing the information flow.
CIM brings it to you 'live.' (computer-integrated manufacturing; blow molders monitoring)(includes related article)
Blow molder plunges into CIM.
Programmable automation in manufacturing systems.
Human issues in technology implementation.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters