Our experience with computer aided design goes back to 1978, when our workstation was the size of a volkswagen (and almost as noisy). Software updates came with a mimeographed copy of the programmers notes, signed “Dave”. And board fabricators were charging us EXTRA to use our computer generated NC drill data. Huh?
 

It was design by the seat of the pants.. will SOMEBODY tell us how to do this?!! In some ways our design quality suffered; we were preoccupied with just getting through the PROCESS, and found there could be a lot going wrong without our knowing it. And why couldn't we drive the factory floor like the magazines said we could?

We gained invaluable experience in our corporate environment, serving several divisions with conflicting design requirements, manufacturing systems, and of course, schedules. We concentrated on our own productivity, and were soon cranking out PWB's of all shapes and sizes. That part started working pretty well. 

We felt that our productivity and the expertise we had gained through it was good reason to start our own design service. And we wanted to focus on how we organize, catolog, store and retrieve the information needed to do our job; to drive our CAD tools with an information system that is flexible, accurate and repeatable.

We have been using Zuken-Redac EDA systems since our first customer's design (and before), which has allowed us to build on proven library resources and focus on our expertise. It has always been a workhorse system with a relatively smooth migration path, tracking (and contributing to) the revolution in electronics and software tools. We’ve seen many CAD systems come and go, or morph from good systems to bad, and feel that the Redac systems' longevity is proof of a good choice of tools. 

The early Redac workstations sported a banner reading “the interactive systems people”; they've always put a high priority on providing visually rich, responsive graphics, and a fast user interface. Designers are, after all, using their visual cortex more than anything else. Their Route Editor product can actually be FUN to use, besides being one of the best autorouters around. 

Under the hood, the Redac system makes heavy use of relational data bases (ask any programmer what “pointers” can do for you!). Relational data bases can provide a tremendous amount of flexibility, and are what makes parametric design work for you.

When we moved from light table to “data table”, we learned early on to dot our I's and cross our T's. CAD systems could duplicate bad data just as easily, no, easier than they could duplicate good data.. and usually across the board. Our electronic templates, our building blocks, now represented a lot more information, and took on a new, literally hidden, meaning. Before long every designer worth their salt began building their own “standard” libraries, indexed with cryptic naming schemes that would baffle even the NSA. It was every designers' private Idaho. 

When we started Lumagraph in 1985, we recognized that a key to our productivity and accuracy was in how we organized and managed this kind of information. Reducing the possibilty of error is largely a function of how well you can re-use data that is proven, and that depends on how disciplined you are at storing and retrieving it. And we knew there was the added challenge of being an independant service; we HAD to have a high degree of design flexibility while maintaining all the benefits of a data base that we could link to our clients' data.

Our libraries are structured using parametrics, allowing us to “overlay” current design technologies and specific customer specifications on even the oldest library models (we are still using the first footprints we created back in 1985). And we’ve developed our own library cataloging and retrieval software, giving us the accuracy and consistency of proven designs.

We've put a large amount of thought and effort into “designing” our design system. Our goal is to provide a system that can “talk” to engineering and manufacturing systems with as little human intervention as possible. Beyond giving us the freedom to concentrate on producing the highest quality design, fulfilling our niche in your product development behooves us to become a part of YOUR information system, and hopefully to enhance it.

So, where IS this factory of the future?..

Incorporating human expertise in the virtual world is not as easy as the salesmen would have you believe. How we “process” information is more inferential and subjective than we realize. It's not just a matter of having a computer on everyones desk, real time.. it's why corporations have departments, that solve problems using things like forms, procedures, and mEEtings.

Part of the solution becomes one of process engineering, defining what information means in each context, where decisions are made, what data needs to be exchanged, and having a structure in place that can routinely assimilate new definitions. A big old data base from your local CAD vendor would be nice, but forget it. Electronics is all about creating new and better widgets; building this data base is part of everyones daily routine. The point is that the NEXT time that chunk of data comes down the pipeline, it flies on its' own...

Our obsession with process stands behind our consistent and accurate, state-of-the-art services. We know better than anyone that the moment you go into manual override, the chain is broken and the next time around something is likely to bite you.. or more likely, somebody else. Ultimately process provides an information bridge between our clients engineering and manufacturing systems. Isn't that what those magazines were talking about?

We’ve provided support for just about every size and shape of electronics firms, from individual consulting engineers to large corporate customers:

• Cross Systems
• Loral
• Gray Labs
• Scientific-Atlanta
• Digital Wireless Corporation
• American Megatrends

• Electronic System Products
• Digital Video Systems
• Wegener Communications
• SCI Systems
• Texas Instruments
• Contel-Executone