by Ben R Rich
Both companies were preparing quarter-size models of their designs for a “shoot-out” on an Air Force radar range in New Mexico to determine stealthiness. Wind tunnel tests would follow to determine the best lift-over-drag ratios and other aerodynamic characteristics. The winner would claim the big bomber contract, a high-stakes twenty-year B-52 replacement program.
Although the two designs were very similar, the big difference between them was that John Cashen was getting advice from a three-star general at the Pentagon to make the airplane as large as possible to extend its range, while I was listening to a three-star general at SAC headquarters in Omaha, who urged me to stay as small as I could get away with while still meeting the basic Air Force requirements for the new bomber. “I’m telling you, Ben, that small will win over big, because budget constraints will force us to go with the cheaper model in order to buy in quantity.” His strategy made sense. But it also created a few significant structural differences in the two models. Because our airplane was designed to be smaller, the control surfaces on the wing were smaller, too, which meant we needed a small tail for added aerodynamic stability. Northrop had larger control surfaces and needed no tail at all. So they had a slight advantage in lift-over-drag ratios, which meant a better fuel efficiency for extended-range flying.
In May 1981, we and Northrop contested on the Air Force radar range. Our results were spectacular; through the grapevine I heard that we beat John Cashen across the board, on all frequencies. A few weeks later I received a classified message from Wright Field questioning the figures we had submitted on aerodynamic wing efficiency. The message was addressed to Northrop, but mistakenly routed to me. So I saw that Northrop’s team was claiming an efficiency 10 percent greater than our own. Frankly, I would question that, too.
Our quoted price to the Air Force per B-2 was $200 million. I heard that Northrop’s quote was significantly higher, so I was shocked when we received formal notification, in October 1981, that Northrop had been awarded the B-2 project “on the basis of technical merit.” I was so outraged that I took the unprecedented step of trying to challenge the ruling. Lockheed’s CEO, Roy Anderson, agreed with me and marched on Verne Orr, then secretary of the Air Force, to protest. The two had an angry confrontation. Orr pounded on his desk and said, “Goddam it, not only was Northrop better than you, they were much better than you.” Anderson, barely in control of his own temper, just looked Orr in the eyes and responded, “Well, Mr. Secretary, time will tell.”
Yes, indeed. Truer words than Roy Anderson’s were seldom spoken.
A blue-suiter called me to explain that the Air Force had determined that Northrop’s B-2 had better payload and more range and therefore would be the better buy. While it was true that Northrop’s B-2 was more visible in most radar frequencies than our airplane and therefore more vulnerable, it would need to make fewer sorties because it carried more bombs than our model. Therefore, fewer sorties evened out our advantage in being less visible. “The bigger the bomber, the fewer the missions over hostile territory. Their loss rate would be no worse than yours, and might even be better,” Secretary Orr had insisted to Roy Anderson. I figured Orr must have had Jesuitical training.
Lockheed’s management was, of course, disappointed, but no one blamed me for the loss. To a man, we knew we deserved to win that contract. But the toss of the dice was out of our hands, and we found solace in our leadership in stealth technology, which had made the Skunk Works a billion-dollar enterprise for the first time in its history.
In the end, the government’s B-2 decision cost the taxpayers billions. Northrop was supposed to build 132 B-2s at a cost of $480 million each—more than twice what we had originally estimated per airplane. But as those projected costs mounted drastically, Congress lowered the number of bombers to be built to seventy-five and the cost per airplane leaped to $800 million. The fewer the airplanes, the higher the cost is a reliable rule of thumb and a painful lesson about the awful cost of failures in the expensive defense industry business. Now the number of B-2s authorized by Congress is only twenty, and the American taxpayers are spending an incredible $2.2 billion on each B-2 being produced, making the B-2 the most expensive airplane in history. When one crashes—and new airplanes inevitably do go down—it will be not only a tragedy but a fiscal calamity. Northrop’s management is in large part to blame for all the delays and cost overruns, but so is the Air Force bureaucracy, which has swarmed over this project from the beginning. When we began testing our stealth fighter, the combined Lockheed and Air Force personnel involved totaled 240 persons. There are more than two thousand Air Force auditors, engineers, and official kibitzers crawling all over that troubled B-2 assembly building in Palmdale. What are they doing? Compiling one million sheets of paper every day—reports and data that no one in the bureaucracy has either the time or the interest to read.
The Air Force now has too many commissioned officers with no real mission to perform, so they stand around production lines with clipboards in hand, second-guessing and interfering every step of the way. The Drug Enforcement Agency has 1,200 enforcement agents out in the field fighting the drug trafficking problem. The DOD employs 27,000 auditors. That kind of discrepancy shows how skewed the impulse for oversight has become both at the Pentagon and in the halls of Congress.
Under the current manufacturing arrangements, Boeing makes the wings, Northrop makes the cockpit, and LTV makes the bomb bays and back end of the B-2 airplane, in addition to four thousand subcontractors working on bits and pieces of everything else. Because of the tremendous costs involved, this is probably a blueprint for how big, expensive airplanes will be built in the future. For better or worse, this piecemeal manufacturing approach—rather than the Skunk Works way—will characterize large aerospace projects from now on. With many fewer projects, the government will have to spread the work across an ever broader horizon. What will happen to efficiency, quality, and decision making? At a time of maximum belt-tightening in aerospace, those are not just words but may well represent the keys to a company’s ability to survive.
Other Voices
Zbigniew Brzezinski (National Security Adviser to President Carter)
When our administration canceled the B-1 bomber program, we knew we would be attacked by political opponents who were unaware of tremendously promising breakthroughs there on stealth technology. Both developments rendered nearly obsolete everything about the B-1, and we in the administration saw that they represented the way of the future, that they were viable, and given what we knew about the state of play in stealth projects and the record of performance by the Skunk Works, that they were really going to perform as advertised. But because of national security, we were unable to reveal to the public the existence of stealth and exploit the strategic facts about it that influenced the decision we made to cancel the B-1. Planning had already begun on a whole new series of stealth bombers and fighters that would revolutionize aerial warfare. So we bit the bullet and just took the heat. This was similar to a political problem faced twenty years earlier by President Eisenhower, who was unable to reveal the U-2 overflights of Russia to answer the charge of a so-called missile gap made against him in the 1960 election campaign.
Leapfrogging technology was the name of the Skunk Works’ game, and that occasionally created political problems for both the administration and Congress. That amazing Skunk Works organization was unique in the world in its ability for stretching far beyond that which was thought to be feasible and enjoying a success rate unprecedented for advanced technology projects. Now, in the post–cold war era, we are likely going to be involved in a variety of future conflicts in which overflights for intelligence purposes and for military operations will be of enormous importance. That access is going to have to be surreptitious and undetectable, and clearly the Skunk Works will be continually called upon to keep leapfrogging technology in behalf of the national security. But how we will be able to maintain the tremendously high standards of the Skunk Works during a new
era of downsizing defense and intelligence appropriations is really outside my realm of expertise. What is clear is the nation’s need to keep this kind of unique operation intact and thriving far into the foreseeable future. Downsizing notwithstanding, it simply must be done.
16
DRAWING THE RIGHT CONCLUSIONS
IN MY FORTY YEARS at Lockheed I worked on twenty-seven different airplanes. Today’s young engineer will be lucky to build even one. The life cycle of a military airplane is far different from the development and manufacturing of anything else. Obsolescence is guaranteed because outside of a secret, high-priority project environment like the Skunk Works, it usually takes eight to ten years to get an airplane from the drawing board into production and operational. Every combat airplane that flew in Operation Desert Storm in 1991 was at least ten to fifteen years old by the time it actually proved its worth on the battlefield, and we are now entering an era in which there may be a twenty- to thirty-year lapse between generations of military aircraft.
Whether we as a nation will develop intelligent military planning and spending policies in the post–cold war era, I will leave to futurists and politicians to argue about. My interest is the future survival of the Skunk Works as a widely adopted concept, and understanding the key reasons for its unsurpassed success. Will its future be anywhere as bright as its past? Will the concept itself finally receive broad acceptance across the industrial landscape at a time when development dollars are as sparse as raindrops on the Sahara?
To my mind, the leaner and meaner Washington becomes in doling out funding for defense, the more pressing the need for Skunk Works–style operations. Any company whose fortune depends on developing new technologies should have a Skunk Works in operation; in all, there are fifty-five or so scattered around various industries, which isn’t very many. But if Lockheed’s Skunk Works has been a tremendously successful model, why haven’t hundreds of other companies tried to emulate it? One reason, I think, is that most other companies don’t really understand the concept or its scope and limitations, while many others are loath to grant the freedom and independence from management control that really are necessary ingredients for running a successful Skunk Works enterprise.
Unfortunately, the trend nowadays is toward more supervision and bureaucracy, not less. General Larry Welch, the former Air Force chief of staff, reminded me recently that it took only two Air Force brass, three Pentagon officials, and four key players on the Hill to get the Blackbird project rolling. “If I wanted an airplane and the secretary of the Air Force agreed,” the general observed, “we had four key congressional committee chairmen to deal with and that was that. The same was true of the stealth fighter project—except we had eight people to deal with on the Hill instead of four. But by the time we were dealing with the B-2 project, we had to jump through all the bureaucratic hoops at the Pentagon and on the Hill. So it is harder and harder to have a Skunk Works.”
To buck smothering bureaucratic controls inside or outside government takes unusual pluck and courage. Smallness, modest budgets, and limiting objectives to modest numbers of prototypes are not very rewarding goals in an era of huge multinational conglomerates with billion-dollar cash flows. There are very few strong-willed individualists in the top echelons of big business—executives willing or able to decree the start of a new product line by sheer force of personal conviction, or willing to risk investment in unproven technologies. As salaries climb into the realm of eight-figure annual paychecks for CEOs, and company presidents enjoy stock options worth tens of millions, there is simply too much at stake for any executive turtle to stick his neck out of the shell. Very, very few in aerospace or any other industry are concerned about the future beyond the next quarterly stockholders’ report.
Yet if times stay tough and the New World Order evolves without any new big-power confrontations, the need for innovative, rapidly developed, and relatively inexpensive systems that are best supplied by a Skunk Works will be greater than ever. Which is the main reason why I remain optimistic about the future viability of the Skunk Works. By its very definition as a low-overhead, advanced development operation for crash production of hot items—prototypes representing cutting-edge technologies that the customer eagerly needs or wants to exploit—the Skunk Works is needed more than ever. The beauty of a prototype is that it can be evaluated and its uses clarified before costly investments for large numbers are made.
Extremely difficult but specific objectives (e.g., a spy plane flying at 85,000 feet with a range of 6,000 miles) and the freedom to take risks—and fail—define the heart of a Skunk Works operation. That means hiring generalists who are more open to nonconventional approaches than narrow specialists. A Skunk Works is allowed to be less profitable than other divisions in a corporation only if its projects are not financial back-breakers and are limited to producing about fifty units or so. Going “skunky” is a very practical way to take modest risks, provided that top management is willing to surrender oversight in exchange for a truly independent operation that can make everyone look good if its technology innovations really catch on, as with stealth. By keeping low overhead and modest investment, a Skunk Works failure is an acceptable research and development risk to top management.
Frankly, in today’s austere business climate I don’t think a Skunk Works would be feasible if it could not rely on the resources of the parent entity to supply the facilities, tools, and workers for a particular project and then return them to the main plant when the task is completed. But given the right project and motivation, even the usually rigid corporate controls dominating Detroit automakers can benefit enormously from a Skunk Works–style operation for new product research and development, as Ford Motor Company proved recently in producing a new model of its classic Mustang automobile.
For many years the idea of reopening that once popular line of cars was rejected by company executives as being prohibitively expensive. Development costs were projected at more than $1 billion. But in 1990, management put together an ad hoc Skunk Works operation called Team Mustang, composed of designing and marketing executives and expert shop people, swore them to secrecy, then instructed them to design and produce a new Mustang for 1994. Most important, management allowed Team Mustang to do the job with a minimum of second-guessing and management interference. The result: the group took three years and spent $700 million to produce a new vehicle that was extremely well received and became one of Ford’s hottest sellers. That represents 25 percent less time and 30 percent less money spent than for any comparable new car program in the company’s recent history.
General Motors is now following Ford and has started a separate and secret development group of its own for future projects. Four or five aerospace companies now claim to have a Skunk Works. McDonnell Douglas calls its group Phantom Works, and it apparently emulates what we tried to do at Lockheed. Overseas, the Russians and the French have evolved the most sophisticated Skunk Works operations modeled on Kelly Johnson’s original principles. The French aerospace company Dassault-Breguet probably has the best operation in Europe. The concept is beginning to spread, and the ground is certainly fertile because there are only a few thousand days remaining before twenty-first-century technology becomes a reality.
At our own Skunk Works nowadays we are investigating development of vertical landing airplanes and the feasibility of hypersonic airplanes to carry space stations into orbit because rockets are so horribly expensive. We are looking into dirigibles as the ultimate heavy-lift cargo transports, perhaps a better and safer way to ship crude oil around the world than vulnerable tanker ships. We could build dirigibles double-hulled for additional environmental safety and fly them high above violent weather systems. We are researching new uses of composite materials used in our stealth airplane and ship prototypes. These materials never corrode or rust and might be used in new bridges. We are also mulling over better methods to speed up the handling and transportation of large volumes of products.
O
n the military side, the end of the cold war has brought some welcome new thinking into the military development and procurement areas. Avoiding casualties is now a political imperative for any administration, and the search is on to find new nonlethal disabling weapons that can knock out hardware but leave troops relatively unharmed. There are techniques, for example, to disable a tank without killing its crew by using sprays that crystallize metals or supersticky foams that clog traction. Research is underway to perfect powerful sound generators tuned to frequencies that incapacitate humans without causing any lasting harm. This wall of agonizing sound stops troops in their tracks. So can stunning devices and nonlethal gases being researched, as well as laser rifles firing optical munitions that explode dazzling light brighter than a dozen flashbulbs, causing temporary blindness. Another possibility is an immensely powerful jammer that puts out so much energy that it shuts down all enemy communications or missile defense systems.
But schemers never sleep and there are always counters to every new technology. Currently, the French and Germans are trying to create a missile that can shoot down our stealth fighter. It might well take them twenty years to succeed, but ultimately they will find a way. And then we will find a way to counter their way, and on and on—without an end.
Lockheed’s Skunk Works will find useful and productive ways to stay in business for years to come. But a Skunk Works is no panacea for much that ails American industry in general or the defense industry in particular. I worry about our shrinking industrial base and the loss of a highly skilled workforce that has kept America the unchallenged aerospace leader since World War II. By layoffs and attrition we are losing skilled toolmakers and welders, machinists and designers, wind tunnel model makers and die makers too. And we are also losing the so-called second tier—the mom-and-pop shops of subcontractors who supplied the nuts and bolts of the industry, from flight controls to landing gears. The old guard is retiring or being let go, while the younger generation of new workers lucky enough to hold aerospace jobs has too little to do to overcome a steep learning curve any time soon. I’ve recently seen young workers install hydraulic lines directly over electric wires—oblivious to the dangers of a hydraulic leak that could spark a fire. We are not producing enough airplanes for workers to learn from their mistakes.