Part IV: In-House Computers
Ten years after we installed our first computer, our turnaround of truss designs remained unacceptable. But one evening in the late Seventies, I saw daylight when a red, white, and blue van pulled up to our St. Louis office. I recall my wonderment back then, while handing the FedEx Man the overnight pouch at 7pm, that our customer would receive his drawings 15 hours later. For once we had achieved 24-hour turnaround! Rarely, though, could we replicate that effort: receiving a truss request via facsimile machine, completing a truss input form, keying it into a computer terminal, pasting up the output, and me sealing it. Contrast that fire drill 40 years ago with the near-automatic process that dispatches a million sealed drawings a month to CMs in 2018!
The key element of our improved turnaround was the file-cabinet sized box we installed at a CM’s plant; the mini-computer shown here. Our customers, like Heart Truss in Michigan (still owned today by the Butcher family) had no less of a drill: drawing a sketch of the trusses, getting the trusses to run successfully on their computer, sending their sketch and printed output to us via fax machine, and waiting for sealed drawings to be returned. Since Heart’s computer contained the same truss design program as ours in St. Louis, we were assured that we could reproduce the identical truss output, but we were required to rerun it.
For the first time, truss designers at Heart and select truss plants across the country began reaping the rewards of interactive design. Most importantly they derived nearly instantaneous results. If a truss “crashed,” they could alter the input and rerun it immediately; for instance, raising the top chord pitch of a scissor truss incrementally until it worked. While this iterative process consumed valuable computer time, its optimized outputs were unbeatable.
However, these few pioneers faced considerable challenges. Their mini-computers allowed only a single user, they were unable to analyze complex structures, and their computers were too expensive to replicate. But their impact was felt across the construction industry, as other CMs scrambled to match their success. Fax machines proliferated. Toll-free telephone lines stimulated communication. And all of us worked diligently on better in-house computer solutions.
Finally, at a truss industry event in Chicago in 1982, the first affordable, self-contained desk-top engineering computer, the Hewlett Packard 9836 (or 9845), was marketed to CMs under the “Forest Products Inc.” (FPI) label. It contained the truss program originally developed for Carroll Sanford’s now dormant firm. The 9836/9845 sold for about $15,000, and its truss program for $5,000. Its self-contained printer produced sealable output instantly on heat-sensitive paper. Its total investment of about $50,000 (in today’s dollars) appealed to a large number of CMs.
By 1983, the stage was set for the coming wave of hands-on truss design. In our home and thousands across America, Apple IIe computers had been giving us a taste of the future for several years. Also that year, our engineering office was given a promising preview of Apple’s graphic-savvy Lisa computer. And most critically, our industry craved better tools to satisfy a hot housing industry, once again approaching two million starts.
Next Month:
“Hands-on Design”