In the spring of 1983, Richard Sebastian, armed with a PhD in physics, dreaming entrepreneurial dreams of lucrative contracts with the Pentagon, journeyed to Cleveland to see how he might be useful to a venerable machine-tool manufacturer, founded in 1916, called Acme-Cleveland Corp. It was a speculative trip, to put it mildly. Sebastian, after all, had little in common with his hosts. He was a James Bond outfitter from outside Washington, D.C., who could sell you a gizmo that could detect a security guard nodding off to sleep, or the presence of passengers hidden in the trunk of a car.So what was he doing in Cleveland, talking up his wares to a roomful of Rust Bowl managers and engineers who, for all he knew, were one step from the scrap heap of history?

As it happened, he was there precisely because Acme-Cleveland, like so many other machinery companies, was in very serious trouble. The company needed to retool its manufacturing operation, and its executives thought that Sebastian's spy devices might help. They wondered, for example, whether the sensor signal-processing technology that could alert you to a body in a car trunk might not also be able to tell you whether metal tools were cutting correctly, and when the blades were wearing out. The meeting lasted for around two hours. When it was over, the Acme-Cleveland executives invited Sebastian to come back soon, this time bringing a business plan with him.

Over the next few weeks, the discussions grew earnest, focusing on just what sort of deal was most beneficial to both parties. Sebastian wanted money -- to get his new company off the ground, to do research -- but he also wanted to retain his independence. His ambition encompassed more than a lifelong devotion to refurbishing the Rust Bowl. He remembers being somewhat surprised that Acme-Cleveland appeared to understand this. "It seemed enough to them that we were interested in their problems," he says. "They said they didn't want to control us; they didn't want to be our Sugar Daddy." So a deal was struck: Acme-Cleveland put up $250,000 in seed money, and another $100,000 for a research and development contract to be spent during the following year; Sebastian agreed to help the company apply sensor technology to factory processes.

It was not the first such deal between a high-technology start-up and an old-line manufacturer, but the idea was new enough that both entered into the arrangement with a sense of experimentation. It was a venture that offered something to both companies -- a partnership, really, based on mutual needs. For lack of a better term, it was called a "strategic alliance."

All across the heartland of American industry, such antique companies as Acme-Cleveland are scouring the countryside for business innovators that can help them bring technological changes to their markets and retool their factories for the future. Their motive -- survival -- is as basic as their industries. The pulse of the heartland, pressed by low-cost foreign competition, is slowing down. Plants are out of date; products are being rejected as irrelevant or inferior; layers of management, once thought indispensable, now stand revealed as so much cotton batting. Competition has turned the gaze of these giants outward and downward -- to the world of small companies, sometimes to copy their procedures, sometimes to harness their genius, always to find the spark that will enable the large companies to live and grow.

The effort takes many forms. One of the most talked-about is intrapreneurship, as when such industrial behemoths as IBM Corp. and 3M Co. discover that product innovation thrives best when employees are turned loose to function like independent entrepreneurs. Another is the corporate start-up, as when such companies as Control Data Corp. and Tektronix Inc. offer capital and support to restless employees eager to go out on their own. The trend can be seen as well in the changing relationships between large manufacturing companies and their small suppliers. And even General Motors Corp. is rejiggering its legendary bureaucracy in hopes of fashioning a more nimble, decentralized operating system. Indeed, GM chairman Roger B. Smith himself has appeared on "The Phil Donahue Show," explaining to the nation's housewives why executives of the new Saturn subsidiary will have total freedom to invent new ways of manufacturing automobiles.

All these maneuvers reflect an apparent change in big-company attitudes toward small enterprise. The giants, to be sure, have long recognized that smaller companies offered things they wanted. Historically, however, their approach has resembled that of a sultan in search of a new favorite for his harem. Acquisition was the goal, whether in the short run or the long, and as often as not the acquired company didn't protest. But funny things happened on the way to the seraglio. For one thing, the small company founders tended to take their money and run, leaving little more than the shell of a business behind. Then there were the cultural conflicts -- notably the attempts by large companies to impose highly developed management systems on entrepreneurial ventures. A case in point was Exxon Corp., which bought or spawned several new businesses in the late 1960s and early '70s, smothered them with management, then watched them expire one by one. Other big companies learned similar hard lessons, which put a temporary halt to their search for small company "windows on technology."

But along about 1980, a new kind of relationship began to appear, primarily in the computer, telecommunications, and pharmaceutical industries. In these relationships, the large company would buy a minority stake in a small company without seeking to acquire or even control it. Thus, for example, Memorex Corp., a leader in computer-memory storage, bought 4% of a young disk-drive company called DMA Systems Inc. General Instrument Corp., a big force in the cable TV equipment industry, settled for less than half of Sytek Inc., a computer networking company. Such companies as Abbott Laboratories took minority interests in young, avant-garde medical research companies.

These deals had two things in common. In each, both parties were intimately involved in research and technology, and the focal point was new products. What the large company wanted was a leg up on products and markets it felt would be important to its future. The alliance was a way of extending its internal research and development effort at relatively low cost, while protecting its approaches to potential new markets. What the small company wanted was money, without giving away its freedom and incentive to create a viable business.

And the newfangled arrangements worked, or at least some of them did. They certainly became very popular. During 1984, the number of minority investments by big companies in small companies financed by venture capitalists soared to 200, almost three times the level in 1981, according to Venture Economics Inc., in Wellesley Hills, Mass. There was also a wave of deals between young technology companies in the United States and more established European and Japanese companies, and this was paralleled by another wave that linked big American companies and smaller ones abroad. Meanwhile, some growing companies began to explore alliances with each other. Cullinet Software Inc. and Lotus Development Corp., for example, worked out an agreement to develop software links that will provide users of 1-2-3 and Symphony wih access to mainframe-computer databases. Then there was the deal last December between Jack Eckerd Corp., the $2.6-billion Florida drug retailer, and a California start-up called HomeClub Inc., which is developing a chain of home-improvement centers. In addition to acquiring about 20% of HomeClub's stock, Eckerd agreed to have its chairman, Stewart Turley, serve on the smaller company's board.

But perhaps the most startling development has been the appearance of such alliances in the industrial heartland, where some of the woolly mammoths of American enterprise have begun teaming up with hot, young technology companies in an effort to rejuvenate their operations and adapt to changing environments. The incongruity of such partnerships aside, these new industrial alliances represent a significant broadening of the trend, for their goals are quite different from those of the earlier technology alliances. Unlike Memorex, General Instrument, and Abbott, the large companies involved in these heartland deals are generally not looking for new products. Rather, they are seeking ways to streamline and upgrade their manufacturing processes, in hopes of competing better in international markets.

General Motors is perhaps the foremost example of the heartland's new outreach policy. For generations, the auto giant chugged along on a proud tradition of internal R&D. Almost overnight, it has struck up operating alliances with six small companies, all but one of them in the fast-emerging area of machine vision, which uses computer-controlled cameras during manufacturing for such activities as quality- and process-control. Besides providing the companies with equity capital (it owns about 10% of each of them), GM has signed multimillion-dollar R&D contracts.

Another example is Caterpillar Tractor Co., which is looking to alliances to restore its overseas competitiveness. Last February, its venture capital arm put $2 million in a three-year-old Arlington, Tex., company that develops systems for factory robots. Caterpillar is now considering other, similar deals.

So, too, is Rockwell International Corp., the $9.3-billion defense and aerospace giant. Last November, it paid $1.2 million for a 5% interest in Micro Linear Inc., a young San Jose, Calif., company with expertise in the design and manufacture of semicustom linear integrated circuits. Rockwell expects to be one of Micro Linear's major customers during its first year of operation.

It remains to be seen, of course, how these relationships will work out, but already new questions are arising and new issues emerging:

* How, for example, should a deal be structured to maximize the chances of success? How much ownership is too much, or too little? Companies are experimenting with different equity percentages, development contracts, and licensing deals. A few big-company partners have discovered, to their dismay, that the wrong deal can present entrepreneurs with the same irresistible temptations to maximize short-term over long-term objectives that the big company executives are so often accused of falling into.

* How do you cope with the inevitable cultural differences between industrial giants and small, entrepreneurial companies?Robert J. Eaton, a vice-president at General Motors, remarked recently that he had been meeting with guys in jeans and sneakers who rode to work on 10-speed bicycles. "They don't wear suits," he said. "I'm not sure some of them even own suits." On a more serious level, skeptical observes wonder whether big companies will be able to restrain themselves from imposing their own decision-making structures, their compensation policies, their very language, on smaller, more fragile partners.

* What about the independence that lies at the heart of the relationship? To what extent should the small company be allowed to set its own agenda? How much latitude should it have? As a minority owner, the big company can try to persuade; it can't order. On the other hand, its persuasion can be mighty persuasive, and this has already made for some interesting discussions between large and small partners. The chief executive officer of a GM-allied machine-vision company, for example, recalls a dispute over the responsibilities of the man whom GM wanted to sit on his board. He insisted that the director's duty, first and foremost, was to foster the small company's interests, not GM's. "We went round and round on that issue, before they saw our point of view," he says.

* What sort of commitment is needed from the big-company partner? Will deals initiated by high-level corporate managers be accepted in the operating divisions? Maybe not, or so some companies have discovered. Their experience suggests that the most successful alliances are those that speak to the particular needs of line managers.

Questions like these drive home the point that alliances are no magic bullet, no instant cure for the ills of basic industry. Rather, they constitute an experiment to try to tap the very different resources and cultures of large and small companies. Like most experiments, they proceed by trial and error, on a case-by-case basis, through the gradual accumulation of knowledge and experience. And therein lies the advantage to companies that have begun acquiring such knowledge and experience -- companies like Acme-Cleveland.

In the summer of 1982, the principal question was whether or not Acme-Cleveland had a future at all. The recession was spreading through the Midwest like the plague, and the 66-year-old machinery maker was in serious jeopardy. Only a year before, Acme-Cleveland had been a profitable $400-million company. But when the slowdown hit the shores of Lake Erie, everything unraveled at once. Orders suddenly collapsed; sales went into a dramatic free-fall; profit margins were deteriorating rapidly. Acme-Cleveland quickly shrank to less than one half of its former size. The work force -- once 6,500 strong -- was slashed to 2,600. To be sure, the compan's competitors were also hurting -- the total market declined from $5.5 billion in 1981 to $1.5 billion in 1983 -- but none quite as badly. And nobody knew where the bottom would be.

None of this was especially surprising to B. Charles Ames, Acme-Cleveland's tough-minded 59-year-old CEO and chairman. He had anticipated the difficulties almost from the day he joined the company in January 1981. Chuck Ames, a veteran consultant with McKinsey & Co., had been the CEO of Reliance Electric Co., a large manufacturer of electric motors and components, until shortly after its acquisition by Exxon. He had spent his early months at Acme-Cleveland skeptically touring the facilities. Nothing he saw pointed to a very optimistic future.

The problems, in Ames's view, went beyond the specifics of any economic cycle. Rather, it seemed that the company was all geared up to conquer a world that didn't exist. Many of its products were based on outdated technologies that took no account of the changes its major customers were going through. The most profitable division, for example, made dedicated transfer systems that the Detroit auto companies were in the process of abandoning. Ames thought they were "absolute dinosaurs," but Acme-Cleveland hadn't developed newer, more flexible systems. Another division made high-speed steel drills and other tools for cutting through metal, but had no tools for cutting holes in the increasingly popular plastics and ceramics.

But it was the central research facility that most disturbed Ames. To begin with, the building itself was several miles from the nearest production plant. He soon learned, moreover, that there was very little dialogue between the research people and the marketing staffs over what customers really wanted. "Walking around that R&D center," recalls Ames, "I got this sick feeling in my gut. We had a bunch of guys in lab coats doing showcase research. And my reaction was, 'Holy Christ, is this a commercial operation?"

Ames quickly concluded that some of Acme-Cleveland's operations needed to be reorganized. He had too many managers at corporate headquarters: He got rid of some and dispersed the rest. Production facilities were poorly designed: He reduced the working floor space by more than 1 million square feet with no change in productive capacity.But beyond that, the company had to acquire new kinds of expertise. In factories all over America, new technologies were revolutionizing the way products were designed, assembled, and inspected. They involved robotics, lasers, and other dazzling applications of computers. Bigger machinery manufacturers, like Cincinnati Milacron Inc. and Cross & Trecker Corp., were moving in those directions, but Acme-Cleveland knew almost nothing about them. All the company knew, notes Ames, was old-fahsioned metal-bending and -cutting techniques. "If I had brought our very best machinists to IBM or Xerox," he says, "they wouldn't have had a thing to talk about."

But how could Acme-Cleveland explore a lengthy list of new areas quickly enough? The company didn't have much money for acquisitions. What's more, Ames had learned at Reliance that acquiring 100% ownership of smaller businesses didn't always work; it had purchased several owner-managed service businesses during the 1970s, only to see the spark go out once the owner had the money. So over the next two and a half years, Ames embarked on an ambitious course, negotiating one strategic alliance after another.

With one deal, he sought to introduce Acme-Cleveland to the powers of robotics. The corporate planning staff looked at prospective partners and settled on a robotics consulting firm in Detroit. At the same time, Ames also wanted to align himself with a partner that understood laser applications. The research led to a company in Santa Clara, Calif., named XMR Laser Systems Inc.

Sometimes, Ames simply identified an area of expertise that looked promising. He knew, for example, that computer-aided design and manufacturing (CAD/CAM) was becoming important to manufacturers all over, so an alliance was formed with a CAD/CAM time-sharing business in Dayton. There was yet another deal with a company in St. Paul, Minn., that licensed a Soviet technology for surface coatings on metal, using a chemical process that made tools and parts more durable.

Just as Ames was hustling to get the company vinvoled in new technologies, he was experimenting with new forms of deal-making. The structure of the alliances varied greatly, depending on the circumstances. With the robotics company, for example, Acme-Cleveland invested $500,000 in exchange for 35% of the equity and a seat on its board. But the arrangement with the CAD/CAM company was different: In addition to investing $2 million for 40%, Ames negotiated an option to buy the rest, based on a multiple of earnings five years down the road. In light of his experiences with acquisitions, he thought this approach was less risky than buying the whole company at once. With the coating business, there was still another approach: Acme-Cleveland didn't invest in the company itself, but agreed to become a 50/50 partner in a new joint venture. Together, they would operate metal-coating service centers all over the country. Acme-Cleveland's sales force would push the new service.

There were other types of partnerships to fill other types of needs. A Rochester, N.Y., factory automation software company had a series of quality-assurance products in the works; Acme-Cleveland became a 10% owner. There was a small industrial-consulting and software house in West Lafayette, Ind., named Pritsker & Associates Inc., which Ames hoped would help on big-ticket development contracts; its co-founder, Alan Pritsker, a leader in computer simulation of manufacturing processes, agreed to sell around 25% of the business.

And then, of course, there was Richard Sebastian's Springfield, Va., company, Digital Signal Corp. Ames hoped that Sebastian, 42, and his colleagues could apply their laser sensors to such elements of manufacturing as noncontact gauging and inspection.

To be sure, Ames didn't know what the future held with any of the alliances. The road ahead had nothing but blind turns. It was almost as if Ames were attempting to create a different sort of company.

To date, Acme-Cleveland has invested more than $10 million in eight partnerships and joint ventures. It has been three and a half years since Ames negotiated his first alliance. Sitting in a leather wing chair in his office about 15 miles from downtown Cleveland, Ames, a square-jawed, balding Illinois native, says he never expected all of the relationships to result in benefits for Acme-Cleveland, and indeed the process has been a series of trials, errors, and surprises.

The robotics consulting business in Detroit, for instance, never really got off the ground. The founders weren't very good at marketing their expertise, says Ames, "They had too much substance and not enough veneer." Acme-Cleveland suggested possible customers, but to no avail. The investment was sold.

The experience with the quality-assurance software company was just as frustrating. New programs haven't been developed, and Ames laments that the founders frittered away their money on lavish overhead expenses. "I think it's been a simple case of poor management," he says. "But when you own a minority, there's not much you can do about it."

Even when the young companies have done well, Ames has learned some important lessons about the subtle art of structuring deals. The CAD/CAM company, for instance, is growing and making money. But Ames thinks that it would be growing faster if there were no buyout formula. He says that the purchase option has had a chilling effect on the way the founder has managed the business. He plans to rethink his approach to purchase options in the future.

Other relationships, however, have taken interesting turns. The deal with XMR, the California laser company, has spawned a whole new laser business within Acme-Cleveland. About two years after the alliance was formed, Ames sold the XMR interest and set up his own laser subsidiary south of Cleveland. The new subsidiary licenses technology from XMR, which was offered a 10% stake; it has recently sold several large systems, including one to General Electric Co. "We're not in the dark ages anymore," offers Ames, "which is more than we could have said a couple years ago."

There are other sources of optimism for Acme-Cleveland, too. In the year and a half since the metal-coating venture was formed, it has become an $8-million business. Within the decade, Ames thinks it could be 10 times that big -- rivaling Acme-Cleveland's entire metal drill division. Acme-Celveland's systems-design people, moreover, have recently been working more closely with Pritsker & Associates. And Richard Sebastian's company is in the final phases of testing for a new laser measurement device. Ames is hopeful that the technology (which Acme-Cleveland would license from Digital Signal) will provide the basis for some breakthrough inspection products. "I think we're going to be one or two years ahead of our competition," he says.

Not that Acme-Cleveland has always agreed with its partners about the best way to do business. The managers of the metal-coating venture, for instance, recently wanted to hire about 10 new sales specialists to drum up more business. Acme-Cleveland people argued against it, saying that the added expense would be too harmful to profits. In another instance, Pritsker wanted money to develop an elaborate $3-million factory-control system. Ames advised him to seek outside funding, which Pritsker is doing.

Such back and forth has been common in Acme-Cleveland's relationships. Last year, for example, Sebastian approached the company for a guarantee for a computer lease, which his bank would otherwise have refused. Ames had told him at the time of the companyhs founding that Acme-Cleveland would guarantee only up to its percentage of ownership. "There's a real danger in being too helpful . . . in being seen as a deep pocket," Ames believes. Thus, last spring when Acme-Cleveland indicated its willingness to make a bridge loan of $260,000 to a Digital Signal subsidiary, Ames insisted on a tough penalty. If the loan isn't repaid on time, Acme-Cleveland will take a majority interest.The threat, says Sebastian, is more than sufficient to assure that it won't happen.

While many of the more blatant clashes have been over money, the most fundamental difficulties have involved communication. Ames confesses that several of the relationships have suffered from lack of attention by him and other managers throughout Acme-Cleveland. Most of the alliances were initiated from the top. With all the turmoil the machinery company has experienced, it has been difficult to set aside the time or the people to make things happen. In fact, there has been an almost constant swirl of restructuring and reshuffling of people.

Yet some of the difficulties, Ames feels, are living proof of how hard it can be for any old-line company to learn new approaches to doing business. He says he had no idea how important it is to be comfortable with entrepreneurs as people, as well as with their businesses. In the future, he will set aside more time with founders before deals are signed, to assess whether the respective interests seem compatible. "I'm not looking for a best friend," he says. "But it's important to talk about values and goals."

The company is beginning to change in other ways, too. Over the past three years, several dozen newly hired managers are hastening what Ames calls a cultural transition. No one is expecting overnight miracles, but there are signs that the new people are willing to approach things with more open minds. Last spring, for instance, the company was considering three new partnerships, all of which were being championed by divisional managers, not by Ames or his corporate staffers.

For all of the question that Acme-Cleveland still has about its future -- the company still hasn't seen any profits, although the losses have been abating -- there are other questions being asked about strategic alliances by large and small companies all over the United States. It has yet to be seen how many of the larger companies will succeed in their attempts to become more in tune with the marketplace. Nor is it clear how many of the smaller companies will find the advantages they hope for -- or what the longer-term consequences will be. There are no guarantees that the relationships will continue or, for that matter, that the large companies won't ultimately try to control their partners or compete with them.

But if the stakes are high for companies, they are just as high for the nation's basic industries and for the economy at large. Alliances, to be sure, offer promises of greater efficiency, and even survival. But for anything to happen, independent companies will have to learn new forms of interaction and interdependency. Corporate managers and entrepreneurs will have to stub their toes before they can do new dances. The long-term benefits of this painful learning process may be a new ability to compete.