Experimentation Is the Way to Transformation?
Experimentation Is the Way to Transformation?
By Chuck Myers
If the Department of Defense really wants transformation, especially in its acquisition process, it should revisit Kelly Johnson’s SR-71 Blackbird and Colonel John Boyd’s Lightweight Fighter programs and examine the environment that permitted such accomplishments.
In defense acquisition, there were times when a concept for a new flying machine progressed from idea to experimental hardware in less than a year, and sometimes even into production within three or four years, all without the Joint Requirements Oversight Council or turbulence or breaking the bank.
The Quadrennial Defense Review says we need a DoD-wide reform of the planning, programming, budgeting, and acquisition processes. It also calls for new operational concepts, new capabilities, and an emphasis on experimentation and training. Is anyone listening?
I remember experimentation as something we used to do a lot of, and without great fanfare or a proclaimed national commitment to procurement. In the mid-1950s, there always were at least a half-dozen military aviation experiments in process. As a young test pilot, I flew the Navy XFY-1 vertical take-off/landing (VTOL) Pogo Stick. At Cornell Labs, my colleagues were flying a machine with rotating engine nacelles; at Convair, Sam Shannon was shooting “splash and gos” in San Diego harbor in a delta-winged sea-based jet fighter on skis, while Don Germerad was struggling with the big contrarotating turboprop R4Y amphibious Navy patrol airplane. At Pax River, Victor Utgoff was testing the Martin amphibious “supersonic” mine-laying patrol aircraft as Pete Gerard of Ryan was showing off the jet-powered delta-winged VTOL Vertijet in front of the Pentagon River entrance.
None of these aircraft was procured, but neither were they considered “failed” experiments, because from them we learned a lot. During this era, Kelly Johnson’s Skunk Works at Lockheed was designing and flying the Mach 2 F-104 and the high-flying U-2, and beginning on the SR-71 Mach 3 reconnaissance Blackbird, all without an SPO, a MENS, a ROC, or approval by a DSARC or a DAB.
The diversity and level of activity kept a number of young design teams spun up, sharp at their trade, and competitive. Termination of an experimental program (many never progressed beyond that stage) was not a traumatic experience for the sponsoring service or the industry—both learned and pressed on.
The cost of these projects was less than what we spend today on program definition or issuing and responding to a request for proposal. The reporting obligations were minuscule compared to today. And our engineers generally were limited to slide rules for calculations. So, logically, one might assume that with today’s technology, we should outperform those ancient design cadres, that is, we should get it done quicker, better, and for less cost.
At this point, it is fair to ask: What the hell happened?
In 1962, I hung up my test pilot “g-suit” and came to Washington as an aerospace contractor. Little did I know I was entering the scene at the beginning of a period of “unintended transformation” of the defense acquisition process.
In those “good old days,” I could ride a cab to the entrance of the Pentagon and walk in without an ID, usually getting to visit whoever it was I wanted to see, and without an appointment. The popular contractor martini and crab salad lunch was five minutes away at the Windjammer Club atop the Marriott. Military and Defense guests abounded; everybody seemed to have time for fun as well as serious discussion, even the big guys.
With little warning, acquisition management became less fun. Secretary of Defense Robert McNamara was tasked by President John Kennedy to alter the process and product of defense acquisition. His spending guidance was to ensure the taxpayer got the most bang per buck—an invitation for “cost effectiveness (C/E) analysis,” a nightmare to military and industrial leaders who had grown up in a “requirements/procurement by the seat of your pants” process. Along with C/E came life-cycle costing and the Programming, Planning, and Budgeting System criteria for program definition and funding justification. The services’ and industry’s response was to create their own systems analysis organizations to reconfigure “requirements” into acceptable C/E format and vernacular. The comfortable “gut feel” expressions of operational requirements began to reappear as irrevocable formulas that could not comfortably be altered once supported by a ton of computer printout.
In the research-and-development “playpen” of the 1950s, do you think we knew how our favorite project was going to turn out? Did we have a clue about the probable life-cycle cost of an envisioned system or could we explain where it was going to fit within the total force capability for the next 30 years? Are you kidding?
One reason we were able to perform those fun experiments for not much money was that there were relatively few government acquisition management regulations or progress reports to be presented through multiple channels and then reviewed by congressional factions, the Government Accounting Office, the Congressional Budget Office, etc.—activity that creates overhead, bogs down project managers, and tires/discourages creative talent.
Behind the gates of the Lockheed Skonk Works, Kelly generally lacked even a contract or a systems program office, and he had his own version of security (the most effective/lowest cost I have ever witnessed; the SR-71 was flying over the desert for three years before it was discovered). Visitors were not welcomed. Even board members could not “drop by.” He had no customer relations staff; he ensured minimum distraction to his staff and insisted on low overhead. Progress reports were viewed as a security risk. He gave no assurances about equal opportunity employment, asked for no special consideration for management format. Is it any wonder his aeronautical achievements are unlikely to be equaled?
Imagine what it might be like to work in an environment that invites innovation, encourages freedom of design, promotes competition, and rewards initiative. Now, abruptly impose an approval process that demands precise predictions of performance, effectiveness, and future procurement and support costs as the criteria for contract award. Might such a change invite prevarication? We can’t know how our project is going to progress or what it is going to cost, so the only way to keep it alive is to concoct optimistic technical and fiscal projections. And when challenged by the oversight analysts, we will spend valuable time creating an analysis that mitigates or shrouds our system’s potential weaknesses. In addition, we may feel compelled to modify the design to be responsive to the whim of a powerful operational/technical dilettante, either civilian or military. Such responses will increase the cost and/or reduce the utility of the product, all in the name of program survival.
Assuming a military service can convince Congress (often a willing partner) to buy the product that has evolved from this government-directed research and development, and if it operates reasonably well in the hands of the user, it might possibly be procured in large enough quantities to yield a profit. In a worse case scenario, the producer privately hopes it will never be exposed to a war environment, which could reveal flaws introduced by the compromises made to keep the program alive.
After a number of years, industry and the sponsoring services learned how to ensure continuation of a program in spite of gross cost growth or marginal performance. The DoD program manager, with cooperation from the prime contractor, distributes contracts to companies located in selected congressional districts so as to garner support from Congress. It soon becomes unthinkable to cancel an incumbent program merely because a better idea has emerged or the threat that supported the original need evaporated. And there is no threat of cancellation for minor blemishes such as poor performance or cost overruns. There is a myth that defense is a zero sum game, but there are numerous examples where additional funds have magically materialized in the nick of time to launch an overlapping program via insightful congressional intervention.
During 40 years in defense research and development, I have been party to spurts of success in outmaneuvering our ponderous acquisition process: the A-10, which evolved from the Air Force AX program; the F-16/F/A-18 from the Air Force Lightweight Fighter (LWF) program; and the F-117 from Project Harvey—none of which were initially in the “plan” or supported by the acquisition cadres of the services.
The original system program office for the LWF (an experiment) was an obscure colonel, two majors, and four other employees. The LWF request for proposal, guided by Colonel John Boyd and the "Fighter Mafia", was 25 pages, and the industry response was limited to 50. Two prime contractors were selected from a field of five. For less than $104 million (FY 75$) total project cost, General Dynamics and Northrop each created and flight tested (in head-to-head competition) two prototype fighters that exhibited leading edge design, material, and control technology. The demonstrated aircraft performance and cost were so attractive that the NATO countries became heavily involved and the Navy eventually was badgered into participating—major unintended accomplishments. It was a classic example of the development and acquisition of a prime Free World weapon system for which there was no military “requirement.” The required operational capability document for the F-16 was not written until after production had begun.
In response to the call for transformation, I suggest a return to better days. Those who would transform the DoD acquisition process should revisit the SR-71, LWF, AX, and Harvey programs. Perhaps those experiences contain the key to restoring the environment that permitted such accomplishment.
RETURN TO HOME PAGE