Not every company has to undertake a furious set of high-tech changes to survive

Tom Smith Industries Inc. (TSI), a 15-year-old molder of thermoplastic material in Englewood, Ohio, servicing the automotive and other industries, used to be a company trapped by tradition. It enjoyed the benefits but was buffeted by the liabilities of standard operating procedure.

On the benefits side, the company could plunk along without much effort, trying to lowball competitors when bidding for molding jobs that came already fully designed and engineered. All the winning bidder had to do was "shoot the parts," that is, manufacture them in injection-molding machines and ship them.

But that passive and reactive approach was sinking the ship. A $2-million company in 1981, then employing about 40 people in Dayton's rustbelt, TSI was just scratching by on the lowest of low-margin products--like those little plastic cigarette-filter inserts R. J. Reynolds Tobacco Co. used in its Doral brand. The margins were so low, in fact, that the company seemed to be giving its money away.

"We ran millions of those things a week," recalls 61-year-old owner Tom Smith, who bought the company in 1981, "but they had very little margin. It was as though we were shipping away dollars with the trucks when they went out." Still, he continued to make those filters for eight more years because "it gave us a little bit of cash flow until we were able to get some other work in." In those postacquisition years, Smith was carrying 22% interest on his buyout loans; he needed all the work he could get.

At the same time, Smith knew that upgrading the company's technology--to improve how the business was managed and how it developed and produced products--would come closer to guaranteeing TSI a future.

Smith has a degree in electrical engineering. He knows polymers from his schooling and his years of work in the plastics industry. (Smith cofounded and then managed Tomco Plastic, in Bryan, Ohio, for 14 years before he bought TSI.) He figured that TSI's salvation would come from the value it could add to products, largely through its engineering expertise. By taking some of the design and engineering burden off its customers, TSI could help them improve their products by converting to plastic where appropriate. Making plastic parts to replace metal ones demands far greater care and precision in the design and manufacturing phases, and is more difficult than producing commodity plastic parts. But it would give TSI a specialty, something to distinguish it from the 15,000 other molding operations across the country at the time.

Ultimately Smith realized he had no choice. In 1989 during a state trade mission to Japan and Korea, he had a kind of revelation. In Japan he saw large manufacturers integrating a few chosen suppliers into cohesive, cooperative enterprises. He believed that the model would soon become standard for the large U.S. manufacturers he served. When it did, he wanted TSI to be among the elect. "I decided that if we wanted to stay around, we were going to have to buy technology, even though we really couldn't afford it at the time."

As a first step, Smith decided to use technology to upgrade the company's internal-management systems. TSI was like many molders: facility-wide tasks--job scheduling, for one--were handled manually, by moving cards around on a master board.

The company's first foray into computerization, however, ended in disaster. In 1989 TSI tried out an internal-management program called Plastic Advisor, which ran on PCs. But the program didn't mesh with the company's established business practices: it could not perform certain necessary functions. For instance, the billing department wasn't able to factor in price breaks for larger orders; those still had to be figured manually. And the system was unreliable: "Every time we turned around, we were doing something to fix it," says Tom Smith Jr., executive vice-president of sales.

The second try worked better. In 1992 Smith brought in an IBM AS/400 computer to run manufacturing-resource-planning (MRP) software that was designed to streamline and integrate the company's major administrative chores. By doing so, TSI positioned itself ahead of its competitors, according to Mike Mucciarone, president of Intrac Systems Inc., in Erie, Pa., which supplied the basic MRP package (Molder's Office Plus, starting at $6,400; 800-944-9625). Mucciarone says that TSI has what fewer than 20% of molders have in business automation. "Most of the plastics companies say they're using computers," he says, "but they're using them only as islands of information." The program saves a lot of duplicated effort by allowing staff members in different departments--billing, production scheduling, shipping--to refer to the same order information in the same database. Access to data can be limited when the software is being customized.

Now TSI receives customers' orders on a PC that's plugged into the outside world and dedicated to electronic data interchange (EDI), an on-line exchange of information with the company's customers about orders, shipping, and billing. After orders are entered manually into the AS/400, the software generates work orders for the manufacturing department, working backward from the ship date. The program sets several processes in motion: raw-materials purchasing, press scheduling, and billing (once the work is shipped). By eliminating human intervention--including tasks like redundant data entry and paper passing--the system reduces the possibility of error. It even sends out an alert if production falls behind schedule anywhere along the line.

Life runs smoother not only between departments but also between the company and its customers. If someone suddenly needs a delivery date moved up, the software not only reworks the schedule but can run simulations to determine how far other orders will be set back by the change.

There are still hurdles to clear. For example, TSI's computer-systems administrator Roger Weller is in the initial stages of linking the PC that takes orders via EDI with the central AS/400. The connection will allow orders to enter the MRP system automatically.

Yet overall, slow and steady is working for TSI. In fact, to anyone who follows what's possible in this hyperconnected, PC-enabled era, TSI, with its 125 employees, may seem woefully behind the times. The company still operates only eight PCs, and the AS/400 it uses, though popular, is a midrange machine that centralizes computing tasks--an approach much maligned by proponents of client-server computing. TSI is far from the sexiest, showiest, most state-of-the-art engineering and manufacturing installation going.

But Smith isn't ashamed of that. On the contrary, he's purposely embarked on the go-slow approach.

In fact, TSI has found that waiting until customers have upgraded their computing technology before upgrading its own is not only economical; it can be good business. For example, TSI's PCs still run on Windows 3.1, even in the face of the newer, more capable, and heavily promoted replacements Windows 95 and Windows NT. It turns out that Windows 3.1, and the software that's written for it, match what's being used by the clients with whom TSI exchanges engineering and administrative files. "We still have a lot of older programs that simply wouldn't work with Windows 95," says Weller. "Our customers use them, and it would throw everyone off if we were to change."

And Tom Smith hesitates to demand that his tool designers and machinists--who create the high-precision molds used in the company's injection machines--change the way they do things just to accommodate more automation. So the company has been introducing design software and computer-driven metal-cutting tools gradually, letting the company's craftspeople figure out on their own how they want to use the technology. Right now, TSI's designers and machinists are using technology to create the rough cuts in the mold-making process; they continue to do the precision detail work by hand. And Smith stands completely behind them.

For instance, about a year ago TSI installed its first computer-numeric-controlled (CNC) machine tool, a tool that cuts each mold component according to a computer program--one program per component. Each program is created by one of the company's toolmakers at a PC that runs computer-aided manufacturing (CAM) software. TSI uses a package called SurfCAM, from Surfware Inc., in Westlake Village, Calif. (The price ranges from $2,500 to $17,500; 800-787-3927.)

Toolmaker Rick Bruner admits, "When we first got it, I just took a side interest in it." Now his interest, and that of his colleagues, has grown to the point where TSI is ready to sidle to the next step: building a full-blown machining center that will use a highly versatile CNC machine tool requiring almost no human intervention. With features like the ability to change cutting heads automatically, a machining center can be programmed to run around the clock.

BUT PERHAPS the greatest contribution technology has made to TSI is helping the company go after its chosen niche: that is, converting metal parts to plastic alternatives.

TSI actually started courting that business in its precomputer days. In 1986 it converted a new metal diaphragm-pump design to a plastic model for a company called ARO (now a division of Ingersoll-Rand), in Bryan, Ohio. Knowing what it could get out of its processes and recognizing ARO's urgent need to come up with a lower-cost pump, Smith reworked a die-cast metal model into an all-plastic version that dominates its market segment today, according to Nick Kozumplik, senior development engineer at ARO. According to Smith Jr., the pump, which has medical, chemical, and industrial applications, saved ARO about $1 million in tooling funds.

But by 1989 it was clear to Smith that creating product designs with pen and ink on a drafting table was too slow. Worse yet, that old-fashioned process didn't allow TSI to participate in the design process with its customers, who were increasingly well versed in computerized design and engineering. So Smith went shopping for a computer-aided design (CAD) package and in a no-turning-back, lock-the-door-behind-me gesture, dropped the cigarette-filter business. "I went to R. J. Reynolds and said, 'You've got to find another molder; this is really not our type of business.' "

Today, with two CAD stations networked together so that engineers can job-share, TSI still uses its first choice for CAD: AutoCAD, designware from Autodesk Inc., in San Rafael, Calif. (CD-ROM: $3,750; disk: $3,995; 800-964-6432). Because it runs on PCs rather than on higher-powered workstations, both the AutoCAD software and the hardware to operate it generally require a lower investment in time and money than other CAD programs. Ease of use was Smith's number-one criterion in originally selecting the software: "You don't have to be a Ph.D. to run the thing."

The conversion to technology, delayed as it might have been, has paid off. Smith has locked up a supplier partnership with Goodyear Tire & Rubber Co.'s Air Spring Division, in Green, Ohio, which makes the balloonlike cushions that soften the ride for big trucks. (Air springs are actually the next step in shock absorbers--air and plastic replacing metal altogether.) TSI's contribution: the cake-size plastic drum parts that are key components in air-spring assemblies. The contract followed from the plastics maker's initiative in designing the parts, which formerly were made from machined aluminum.

Today TSI is running at 85% capacity. Sales are up tenfold from 1981, to $20 million, largely because the company's technological growth has led to more exclusive work for both Ingersoll-Rand and the automobile industry. (TSI's customer list now includes Toyota, Nissan, Chrysler, Ford, and GM.)

The 70,000-square-foot facility TSI built only two years ago already has three shifts of 20 injection-molding machines working five days a week. And the tool shop where metalworkers sculpt the molds for those 20 machines are on 12-hour days. So Smith is shopping for an acquisition. "The companies we're looking at would double our capacity almost immediately, and we'd use it," he says. "My biggest dilemma now is do I manage the growth, or do I slow it down?"

Jeffrey Zygmont is a freelance writer based in Salem, N.H.