(FORTUNE Magazine) – The Baikonur launch complex in Kazakhstan is not a good place to be a human being. The winters are as cold as Siberia, the summers as hot as Death Valley, and often Baikonur's inhabitants will enjoy the singular pleasure of getting both extremes in one day. From midnight to noon, the temperature can fluctuate by more than 100 degrees. The soil in Baikonur is too acidic to be used to make concrete, rendering the complex a desert enclave into which even sand has to be imported. There is nothing in Baikonur. There is nothing near Baikonur, except for small mining towns that even in the sun look like black-and-white photographs.

But Baikonur's forlornness and its geographical location make it a perfect place to launch rockets. That was true during the Cold War, when Sputnik was sent hurtling into orbit from Baikonur, as was the rocket that carried Yuri Gagarin, the first man in space. It's still true, and in the past year, Baikonur has launched five rockets carrying satellites for Globalstar, a company that has spent $3.5 billion and eight years trying to build a global phone system that might actually turn a profit.

Unfortunately, only four of those rockets made it into space. Last September, a Ukrainian-made Zenit 2 took off from Baikonur, headed over the hills of Kazakhstan, and almost immediately vanished. (Technically speaking, the rocket didn't "blow up." Explains one Globalstar exec: "It crashed. Then it blew up.") The people back at Globalstar headquarters in San Jose and New York didn't know anything for a few hours, because the Russians tracking the rocket decided that their instruments weren't working and kept insisting that the rocket was doing fine. It wasn't.

In the wake of the explosion, Globalstar's stock collapsed, falling 40% in one day. Since Globalstar needed 32 satellites in orbit to start operations and 48 to provide full service, the loss meant that its debut would be pushed back until who knew when. The 12 satellites that the rocket was carrying when it went down cost $100 million. All in all, says one Globalstar executive, "it was the worst day of my professional career."

Now, days like that have to be expected when you embark on a project as grandiose and complex as Globalstar, which aims to create a satellite-based phone system that offers phone service to anyone, anytime, just about anywhere around the world. But Globalstar's real nightmare was a management one: how to push along a fractious alliance that had to woo more than 100 government regulators, appease a host of strong-willed partners, weather assorted currency crises (and the occasional explosion), build 52 satellites (including four spares) and 36 land gateways to transmit and receive calls, and persuade investors to pony up billions of dollars on the strength of little more than a good idea. It's all enough to account for the white hair on the head of Bernard Schwartz, the CEO of Globalstar and of Loral (the defense contractor turned satellite provider that owns 45% of Globalstar). True, Schwartz is 73, but that's beside the point. "This isn't a project anymore, it's a real business, and I have great confidence," says Schwartz, sitting in his office overlooking Third Avenue in New York a month before Globalstar's debut. "But when I look back, I think about this quote from the movie Out of Africa: 'God made the world round so you can't see too far down the road ahead of you.' If we'd known what was coming, we might never have started."

Of all the things that were ahead, there's one that Schwartz could never have foreseen: the implosion of Globalstar's competitors. Normally when your competitors take a fall it's welcome news. But in Globalstar's case, more problems than possibilities were created by the debacles at Iridium, the Motorola spinoff that spent about $5 billion to launch its 66-satellite service only to file for bankruptcy protection in early August, and at ICO, a consortium led by Hughes Electronics that planned a $4.6 billion service but followed Iridium into Chapter 11 later that month.

True, Iridium committed so many marketing and sales mistakes that its experience could form the basis of a textbook on how not to sell a product. Its phones started out costing $3,000, were the size of a brick, and didn't work as promised. They weren't available in stores when Iridium ran a $180 million advertising campaign. And Iridium's prices, which ranged from $3 to $7.50 a call, were out of this world. Globalstar has some of the same problems: Its phones are smaller and cheaper, but they are still bigger and more expensive than the tiny cell phones business people slip into their jacket pockets, and although the company charges less per call, the cost is still much more than that of the average cell-phone call. But Iridium has hurt Globalstar most because it has raised serious questions about whether there is any market for satellite phone service. Which means that on top of everything else it must do, Globalstar must spend its first six months proving to the world that it hasn't invested $3.5 billion in pursuit of customers who don't exist.

Assuming those customers do exist--and we'll get to that later--Globalstar has one fundamental advantage over Iridium in the quest to reach them. Iridium's management adopted a top-down, one-size-fits-all style that kept the company isolated from its customers. It seemed to want to run a global partnership with a management manual written by Henry Ford. By contrast, Schwartz abandoned the command-and-control model when he started Globalstar. He constructed the company as an alliance between Loral and a host of telecom equipment makers and service providers around the globe. Everyone who's building any part of the Globalstar system is invested in the company, and all the major telcos that will be selling the service are invested in Globalstar too. This virtual corporation outsources nearly everything to take advantage of its partners' core competencies, which makes it more responsive to customers, more adaptable to changing markets, and more efficient at using capital. But it also makes Schwartz's job of holding the company together a perpetual high-wire act. On the eve of Globalstar's October debut, it's fair to say that he has successfully walked the tightrope, but he's still working without a net.

Running Globalstar has required many of the skills Schwartz developed in 27 years of running Loral, a defense contractor whose success, in large part, depended on Schwartz's talent for navigating regulatory thickets, cajoling skeptical partners, and sweet-talking investors and government types. But Globalstar has been a learning experience for Schwartz. "I've been accused of being an autocratic leader," he says crisply, "and it's true." But when he talks about Globalstar, it's clear that he's had to come to terms with the limits of his power.

The partnership would never have worked otherwise. Globalstar started in the late 1980s as a small operation co-owned by Loral and Qualcomm. It applied to the FCC for a license to use frequency spectrums that had just been made available. It commissioned a market-research study that suggested that a satellite phone system could be built for a reasonable cost and that there was a market for its services. But it was not a real company.

Schwartz believed the study's findings, but not enough to want Loral to bear the full weight of the multibillion-dollar investment needed to build the system. Having worked at the company for so long--Schwartz came on as president in 1972, when the aerospace giant itself was close to bankruptcy--he knew that Loral did not have the in-house expertise or the flexibility to create the system from the ground up. So he went looking for partners--not merely investors but strategic partners that would be involved in the business. From the beginning, then, he was conceding that Globalstar could not function as a typical top-down corporation. To get regulatory approval from more than 100 countries, many of them just beginning to open their telecom markets, would require the assistance of companies with experience in those countries. To build, install, and service a system of 52 satellites and dozens of gateways at a reasonable price required the kind of innovation and price competition not normally associated with the in-house manufacturing of a defense contractor. Most important, to sell a global phone service required the skills that only telcos could offer.

"You don't sell hardware to a telephone user--you sell a service," says Schwartz. "The continuity of the relationship between the supplier of that service and the customer is not a one-time event but goes on every time someone picks up the phone. So we wanted to make sure that we worked with people who knew what that relationship required. AirTouch, for example, would never have allowed us to come out with an inferior product, so they imposed very strict requirements on us with regard to the quality of the service. But now we're reaping the benefits."

In other words, Schwartz sought out partners who could do things Loral couldn't do alone. He made two important decisions that made it easier to find topflight partners. First, the key players would have to have equity in the operation. Second, Globalstar would not be a competitor to cellular service. Subscribers wouldn't replace their existing service with Globalstar but supplement it. Cell phones would be fine for customers' everyday use, but Globalstar's partner-providers could distinguish themselves from competitors by offering this anytime, anywhere service without worrying that it would cannibalize existing business.

Schwartz's plan worked. In March 1994 nine different telecom and aerospace companies invested $37.5 million apiece in the new firm: From the U.S. came cellular provider AirTouch, while from abroad came telcos Elsacom (Italy), Dacom (Korea), France Telecom, and Vodafone (with which AirTouch is now merging), as well as Alcatel (the French telecom equipment maker), Alenia (part of an Italian aerospace giant Finmeccanica), DASA (Daimler-Chrysler's aerospace division), and Hyundai (the Korean industrial conglomerate). Loral and its subsidiaries still own 45% of Globalstar and have operational control. In other words, all of Globalstar's strategic partners have an equity interest, so the common good has a chance of winning out over individual concerns. But in the end what Globalstar says goes.

That's the theory, anyway. In truth, since Schwartz chose to have the strategic telco partners get out there and push the service in their regions rather than to centralize marketing, a la Iridium, he doesn't have as much power over Globalstar as, for example, Jack Welch does over GE's far-flung operations. That probably works to Globalstar's advantage, since the partners will end up selling, in their own chosen styles, a service that they feel they helped design. But it wasn't easy. "I'm not sure I would say I like alliances," Schwartz says. "They're useful, but they're also painful. They're costly, and they're time-consuming."

The fact that this partnership was international just added to the pain, cost, and time consumption. Early meetings of the partners had all the clarity of a session of the U.N. General Assembly, since everyone insisted on bringing their own translators rather than trusting one another enough to rely on a uniform, if unfamiliar, language--English. "You'd have to wait for translation," says Tony Navarra, Globalstar's president, "and that's incredibly difficult when you know the position you want to take. I've become very good at sitting politely and listening while knowing exactly what I'm going to do after the speeches stop."

Eventually everyone gave up the translators and felt comfortable enough to fumble through in English. But other problems are ineradicable, like the fact that there's no time when all of Globalstar's partners are relaxing. Navarra sometimes will start days at 6 a.m. to hook up with Eastern European partners, spend the day dealing with Europe and the U.S., and as early evening rolls around field calls from Australia and Asia. "There was a time," he says, "when we would never let the Asian partners know we were in the office late, because we'd be deluged with calls."

Of course, Navarra and Schwartz would rather have to hide from their partners than have to chase them down. One defining characteristic of the company is that it has chosen no wallflowers as partners. Jean-Bernard Legarde, for instance, is CEO of Tesam, a joint venture of France Telecom and Alcatel that is offering service in much of Europe and all of South America except Brazil. Legarde looks like a young Vincent Gardenia. And it takes him little time to point out all the things that he thinks Globalstar and its partners have done wrong. "Globalstar didn't analyze carefully and realistically the market before it was launched. And there are lots of things about the way the system was designed that represent pure stupidity from a technical point of view," he says. But Legarde is also convinced that the system will work. "It is a niche market, but it's a big niche. Globalstar is an incredibly cost-effective way of reaching those parts of the world --which is most of the world--that still don't have any cellular service at all." Legarde's deeply skeptical yet oddly hopeful outlook has helped create in Tesam a certain us-against-a-foolish-world mentality. That gives the company the feel of a startup even though many of its employees are veterans of huge enterprises formerly owned by the state. But it also makes Legarde the most nettlesome of global partners.

"That's Jean-Bernard," laughs Luigi Gasparollo, the managing director of Elsacom, Globalstar's service provider in Italy and Eastern Europe. Like Legarde, Gasparollo has been involved in the project almost since its inception, and he too has tried to give his European company--Elsacom is a division of Finmeccanica, the Italian defense conglomerate--a Silicon Valley feel, complete with flat organization, no timecards, and rewards (if not stock options) rather than sneers for individual initiative. He speaks of Legarde as one might of a younger brother. "Tesam always makes things look negative, so then they can say, 'Hey, look, we solved the problem.' They come to every meeting and say, 'We're never going to make it,' so you say, 'Jean-Bernard, then why do you keep going?' and he says, 'No, we've already committed to this. We'll follow it through.' But he always gets everything done."

With the likes of Gasparollo and Legarde, Schwartz has never been dealing with shy subordinates. So when he has made decisions that have seemed in Globalstar's long-term interest but have run contrary to the wishes of some partners, he's heard about it. Globalstar sells its phone time wholesale to the partners, who then sell those minutes to customers. AirTouch pays a lower wholesale price than the European partners. Now, that's either because AirTouch is buying more minutes (which is what AirTouch told me) or because Schwartz decided he was willing to give AirTouch a break to get a really established service provider in the key markets of Mexico and Canada (which is what a couple of other partners told me). The most expensive prices, by the way, are in the developing world. That doesn't sit well with the Europeans. "We're selling cheapest to the country that can afford it best," says Enrique Fernandez, president of Tesam. "That's Globalstar stupidity: 'Let's sell the most expensive time to the place where the people are poorest.'"

Similarly, no one is too thrilled about the way China Telecom became a partner. When the Chinese government telco finally came on board in April of 1998, it did so on the condition that it could buy its equity stake at the 1994 price. Since Schwartz needed a service provider in China, he assented. He also conceded to China Telecom the unique right to buy wholesale satellite service from suppliers other than Globalstar.

"It was unfair, but we gave something to get something," Schwartz admits. "We've made concessions everywhere. The Chinese also said it was against the law for them to be exclusive. Now, I don't believe that. But that's what they said, so that's what we went with. No one has ever accused me of being a soft negotiator, but sometimes you have to shape the animal differently to get to where you want to go."

Just as he balanced the concerns of his service partners, Schwartz had to balance the desires of his aerospace partners. So each has a healthy slice of the business of building Globalstar's satellites. Alcatel built the antennae, while Hyundai made the electronic sub-systems. Aerospatiale built the structures that house the satellite's innards. Daimler-Benz Aerospace created the propulsion system. Alenia put everything together at a factory in Rome.

For the most part, however, Schwartz and the Europeans have faced one another with the wariness of a father and strongly independent children. The Europeans banded together to demand earlier access to and lower prices for the gateways (they paid Globalstar for these structures, each of which has a handful of satellite dishes) and to lower the number of phones they were committed to buy. The relationship has been most tested by Schwartz's support of Qualcomm, the San Diego phone and equipment manufacturer.

There are several different technologies for wirelessly transmitting the sound of your voice. For years, Europeans have designed their systems around one called GSM (global system for mobile communications). But way back in 1991, before it had any founding partners, Schwartz committed Globalstar to a different standard, called CDMA (code division multiple access). He did so by agreeing to have Qualcomm provide all the essential telephone and satellite technology; besides manufacturing many of the phones, Qualcomm is designing and building each of the 38 gateways around the world. Qualcomm CEO Irwin Jacobs has staked his company's future on CDMA; Qualcomm pioneered it, and if it spreads, the company is guaranteed a giant revenue stream for years to come. But Jacobs' championing of CDMA as a superior system has rubbed many Europeans the wrong way. Legarde decries the "CDMA imperialists" on this side of the Atlantic and has little patience with what he calls Jacobs' insistence that "everything else is bullshit."

Qualcomm has also been a source of contention because it is the only Globalstar equipment manufacturer that has a cost-plus contract. That was inevitable, since back in the early 1990s there weren't any other manufacturers with equal expertise in CDMA technology. All indications are that the gateways and the phones work exactly as they're supposed to (which was not the case with Iridium when it launched). But that doesn't mean that the other partners have to be happy about it.

"Through this whole experience, Qualcomm has not behaved like a partner," says Legarde as we drive through the French countryside outside Toulouse. "Qualcomm does excellent work, and they have great technology, but if we can avoid being in partnership with them, we will. They're an extremely arrogant company, and they will roll over anybody in front of them. Even their partners."

Gasparollo, as always, takes a more temperate line. "Certainly I wish we had been able to have a different kind of contract with Qualcomm, but when I see the performance of Iridium, I'm happy we have Qualcomm," he says. "While they have perhaps made more money than we had planned, they've also done more work. Maybe we would have saved one, two, three hundred million dollars, but the quality of their work has been superb, and they've responded to changes very quickly."

Despite all these wrangles, the Globalstar "family" has been able to stay intact. How? Well, the wrangles themselves point to one reason. With the key players all equity investors, the alliance was structured to allow for both argument and accommodation. A group called the Council of Service Operators (CSO) helped ease tensions. It meets monthly, and after a slow start it eventually led to some camaraderie among its members, who began to identify with Globalstar as well as with their individual companies. Says Navarra: "I can chair a meeting now and know exactly what every person in the room is going to say. That's how well we know each other."

But the key factor in holding everything together is Schwartz. He won't take credit, preferring to attribute the group's success to "the personal dynamics of everyone involved." And Globalstar has benefited by having a core group of impressive leaders who recognize the value of Globalstar to their core businesses. Legarde at Tesam and Gasparollo at Elsacom are headstrong anti-bureaucrats. Sam Ginn at AirTouch and Jacobs at Qualcomm are honest-to-God pioneers of cellular communications. Bringing together that kind of star power was Schwartz's idea. In the hands of a less confident director the coalition might have fallen apart under the weight of so many egos. "I think I understood intuitively when we were negotiating with him that having someone with his credibility was a requirement," says Ginn. "But now I really know it."

No matter how harmonious, the partnership will have been all for naught if there is, as the Iridium and ICO failures suggest, no market for satellite telephony. Many industry analysts still believe that the market does exist. Merrill Lynch has estimated that by 2007 there will be 32 million people getting satellite phone service, paying a total of $31.6 billion a year. Sixty-five percent of the world's households are still without telephones, and penetration in most of the developing world is still remarkably low. Even in the U.S. and Europe, large chunks of territory have no cellular coverage.

So there may be openings for Globalstar--even if the partners can't agree on what they are. "The primary customers," opines Schwartz, "will be people who do not have phones, who live in areas where there is no cellular service but who are still integrated, through commerce or other relationships, with urban centers and the rest of the world. I think there are a lot of VIPs who are going to use this and a lot of people who have cell service but travel to regions where there is none. But the real market are people who live in a semimodern world that's missing only instant communication." Schwartz thinks that the key markets for Globalstar are Mexico, Canada, Brazil, India, China, Indonesia, and Russia. Gasparollo, on the other hand, says, "I think the United States will be Globalstar's single biggest market." Meanwhile, Andy Radlow, marketing director of AirTouch's Globalstar operation, stresses the initial importance of vertical markets: natural-resource companies, offshore oil rigs, construction companies.

What seems at first to be an irredeemable flaw (doesn't a big corporation selling a product need a single focus?) may be a strength. Perhaps the only way to sell a global phone service is to sell it regionally. In Europe, where something like 80% of the population--but only 50% of the geography--is covered by cellular, it makes sense to think of Globalstar as an extension of cellular service, and the same is probably true in the U.S. Globalstar's dual- and tri-mode phones will let customers switch easily between cellular and satellite, so they may be able to get full coverage while paying the cheapest price available. In Brazil or India, where large villages are often hundreds of miles apart, satellite phones may be important business tools. And on fishing vessels off the coast of Singapore, Globalstar may be a cheap and portable alternative to radio communication.

You have to say "may be" about all these things because question marks do hang over much of Globalstar's business. For one thing, when Globalstar launches in October, the "global" part of its name will be only metaphorical. Globalstar has arrangements with service providers to cover essentially the entire world, with the exception of the Poles and the deep ocean. But the telcos are taking their time about going into unstable or less lucrative markets, which is why much of Africa and parts of Asia, including Indonesia, are going to have to wait. By the end of this year, 45% of the planet will be covered, and a year from now, says Schwartz, effectively all of it will be.

Then there are the parts of Globalstar that seem just a bit too similar to Iridium. Globalstar phones will be cheaper, but they'll still start around $1,250; a high-end StarTac goes for $200. The phones are not brick-sized, but they have a giant antenna. That's a far cry from Nokia's darling cigarette-lighter phone, and with the ongoing consolidation of the international telecom business, who's to say that you won't soon be able to use that little handset anywhere in the world at any time? And although Globalstar's price-per-minute will be cheaper than Iridium's, at $1.25 to $1.50 in most places it will be much more expensive than the dime a minute Americans can get from Sprint PCS.

But Schwartz is not selling Globalstar just to Sprint PCS customers. He's also selling it to people who would be Sprint PCS customers if they could be but who live or work where phone service doesn't exist. Even in the age of globalization, one assumes that anyone who lives where there are no telephones can't afford one. But Globalstar's premise for existence regards that assumption as a fallacy. Instead it sees an opportunity in the uneven development of the world's communications facilities. For businessmen 60 miles outside of Kuala Lumpur, looking at a site for a new disk-drive factory; or 100 miles outside Seoul, contemplating a new semiconductor plant; or retailers at stores in Morocco who want to talk to their metropolitan suppliers; or construction crews out in the field, the question isn't 'How much more than your cell phone is Globalstar worth?' It's 'How much more than nothing is Globalstar worth?'

Sure, it may be a reach. At the very least, it's far more ambitious than just firing up another cell-phone provider for yuppies. But it's a vision in which investors keep believing. Since the explosion at Baikonur, Globalstar's stock is up almost 200%, near its all-time high. Over the same 12 months, investment bankers have poured billions of dollars more into the project, even after Iridium's problems had become well known and the rest of the market for satellite financing had dried up. That's a rather clear profession of faith, and it's a profession of faith not only in Globalstar but also in Bernard Schwartz. "The first rule if you're going to do a project where you have to invest a lot before you get a dollar of revenue is that you have to have someone in whom the investor community has immense confidence," says Ginn. "And Bernard came out of Central Casting for that role."