General Marsh, an Indiana native, entered the Army Air Forces in 1943, then completed aircraft mechanic and aerial gunnery training on B-17 and B-24 bombers before his appointment to West Point in 1945. Graduating in 1949, he began a series of increasingly responsible assignments in nuclear and electronic projects and units. Later duties covered space, reconnaissance, strike, and electronic warfare in Washington and in the field. Before assuming command of AFSC in February 1981, he commanded its Electronic Systems Division for nearly four years.
* * * *
AFM: How do you decide exactly what you need without making something too complex or so complex that it meets everybody’s requirements but costs a lot and may not be supportable? It’s the old question of quantity vs. quality, or simplicity vs. complexity.
General Marsh: I feel, as most of the leadership in the Air Force does today, that the very first and fundamental prerequisite is that everything has to focus on superiority. If you can’t win, if you can’t defeat the enemy with a given system, no matter how simple it is or how easy it is to operate or maintain, it’s not worth building.
The “day fighter” argument says, “Wouldn’t it be neat to go back to the F-86 days?” The simple fact is you can’t do that against an enemy that has good long-range radar capability, fine high-quality air-to-air missiles, and good maneuvering ability. In that sense, the old dogfight days are gone. We simply must equip our airplanes with sufficient systems so they are superior to those of the enemy. That’s a simple word—superior—but it may drive systems to complexity or sophistication. You don’t refer to your pocket calculator as sophisticated or complex. You probably refer to it as a rather simple tool. But it gets the job done better than any way you’ve ever done it.
It seems to me that that’s the sort of view one has to hold in my business. Given that we know the performance required, the name of the game is to do that job with the least sophistication possible, the least complexity. We must make it the most reliable within the constraints and the most maintainable, and at the least cost. But we can’t sacrifice that first and fundamental requirement of superiority.
Look back in history. I like to consider the long bow. Obviously it required fletchers, long arrows, and had more moving parts than a sword. But it certainly annihilated the French knights at the battle of Crecy. That’s the real bottom line of all this . . . winning. Nearer our own time, look at radar. When introduced in England, radar was really a key of winning the Battle of Britain in 1940. Radar was far out in terms of people’s understanding in those days. It was complex and it was sophisticated, but it tipped the balance of the fight. The introduction of precision guided munitions in Southeast Asia had a profound effect in that conflict. Initially they were sophisticated and complex, but they did the job. The point is that to outflank the enemy in performance, you require some degree of sophistication.
AFM: Hasn’t it happened in cockpit displays and other displays where the actual information that a crew member has a deal with is simpler than it was before? The circuitry behind the panel might be considered wizardry, but the person is dealing with simpler displays than somebody twenty years ago.
General Marsh: Absolutely. I don’t think any of the “simplicity vs. complexity” arguers have problems with the externals, such as the displays or the controls. But as soon as we encounter a reliability problem behind those, then they blame lack of readiness on complexity. We need to work these reliability problems in the very worst way, but we also had to do that back in the supposedly simpler days.
AFM: With the F-86, for example?
General Marsh: Sure we did. The problem’s been with us forever, and will be with us forever. Given a certain degree of complexity, there will be reliability, maintainability, and supportability considerations. We sincerely believe, and we have the statistics to prove that we are not less supportable, less ready, less able to generate our sorties than before. In fact, we are more capable in all of those areas.
Our problem is very simple. We needed to modernize the Air Force when we came out of Vietnam. We had insufficient resources to do it, so we make conscious decisions to constrain supportability dollars in order to lay those keels and get those new airplanes into production. We knew we were going to have supportability problems, and problems with logistics support and spares. I think the decisions were right, and we brought these new systems on line in the right sequence. If we had had to go to war with these new systems—the F-15, F-16, A-10, Wild Weasel, EF-111, AWACS—we might have held a different view of whether that was the fight strategy. In my view, it was the fight strategy. Now we really need to focus very hard on raising the level of readiness and supportability of those systems.
AFM: Do the people you deal with in industry and the Congress and elsewhere in the government understand this, and are they receptive?
General Marsh: A few hold views to the contrary, but I don’t think that’s true of the majority. It’s certainly not shared by the responsible leadership of the Air Force. Ask General Creech. He’s charged with fighting the tactical air forces, and he clearly perceives the critical need for performance and capability. When we bring our spares support up to the wartime levels we have specified, we’ll really be in good shape to fight with a sustainable, highly capable force.
AFM: What are some of the things going on within the AFSC system to improve sortie generation, or advances to get more sorties out of aircraft, for example?
General Marsh: One of the most important is in the air base survivability management team at Eglin AFB, Fla. We’re looking at the full range of things that affect our ability to keep the facilities operating and supporting sortie generation during and after attack. That includes rapid runway repair. We have had an aggressive effort under way in that area for sometime. It includes trying to understand how rough our runways can be so that, to the degree possible, we can assure realistic wartime requirements for aircraft operations are clear. In addition, this testing will help determine the degree of roughness that an existing aircraft can tolerate, as well as what changes we could make in landing gear to make them more capable of rough runway conditions.
AFM: Design limitations?
General Marsh: Yes, trying to realistically assess limitations of the airplane as it is now. We’re looking at other survivability measures for our air bases. This includes active air defense. We’re concerned about this, and we want to make darn sure that we have planned and supported security measures in case of attack. This ground attack threat, of course, also is of concern to the ground attack threat, of course, also is of concern to the Army.
AFM: And you are working with the Army on that.
General Marsh: Yes, we are. And our air base defense against an attack will depend upon systems like Rapier. Now then, that’s in the base survivability area . . . sortie generation is dependent upon the ability of the air base to support it. As to the aircraft itself, one thing we’re trying to get is more flexible munitions, so that you won’t have the tailoring requirement for every task. The low-level laser-guided bomb is a good example of this. It’s a very versatile weapon. So are Antiarmor Cluster Munitions (ACM) and more versatile cluster munitions. We want to reduce the number of unique loads to worry about. More versatile weapons are certainly one of our major goals.
One very important area where we have a lot more work to do is the BIT. . . [the Built-in Test] capability. Built-in Test hasn’t really fulfilled all of our hopes and expectations. It’s generally true that the Built-in Test features of our avionics, our stores management sets, and our weapon release systems are more sophisticated generally than the item itself. The ability to automatically test and isolate faults is really a difficult engineering undertaking.
Getting an unambiguous indication of what’s wrong with a radar is frequently a more difficult design problem than the radar itself. We are working very hard at that so BIT can quickly tell the flight-line mechanic which box to swap. We want BIT to be able to say that “this particular box is bad; replace it, and the aircraft will be back in commission.” Today we remove a lot of boxes we think are bad, but they retest okay. Sometimes we get a BIT indication that says that one of several boxes may be faulty. We have to perfect BIT so that it becomes a lot less ambiguous, more reliable, and a lot more capable.
AFM: Does that mean that you have to train the people on the flight line and the immediate backup maintenance people to make up for the present shortcomings?
General Marsh: Yes. The flight-line mechanic’s job today is tougher than it ought to be in that regard. We should be able to tell him with high confidence what Line Replaceable Unit needs to be replaced. It doesn’t happen that way in too many cases.
AFM: Are the units and the people of the flight line, the young men and women who are dealing with it, involved with your engineers and designers on improvements?
General Marsh: We have a tremendous “lessons learned” program, and we work at getting the information directly from the field. All of our feedback does come from the field. If we don’t get it ourselves, then our contractor tech reps do and it comes to us for review.
We’ve been working very hard on a BIT improvement program with the AWACS, for instance. We undertake BIT design at the outset of the development in the system, but it also has to have equal importance throughout the whole development test cycle.
AFM: Can you do that during the development process or do you have to build it in later?
General Marsh: Yes, and it must be done. Obviously, in complex software systems, to postulate every possible failure mode is an almost endless undertaking. To test against every failure mode and prove to yourself that you can isolate it in every case is tough. But with modern computer technology, you can run through a failure mode on LSI chips with complex test programs that will check every circuit in that system. So the answer is, yes, you can do it if you spend a lot of time on it at the outset. We have learned that testability of complex electronics circuitry is a major undertaking, but a vital one.
AFM: Does that lead into the need for engineers and designers, and the fact that the supply is less than the demand?
General Marsh: You’re into one of my biggest problems. Getting the qualified, experienced engineers we need to carry out our responsibilities is vital. We had a big exodus of experienced engineers out of the command back in the ’70s. Today about forty percent of all my officers are lieutenants. Now, they’re fine, capable young people, but they naturally lack experience. Unless one’s been through this BIT problem, or lived with one like it for awhile, he can’t attack that problem in an aggressive way until he spends some time building the necessary framework.
AFM: Can you compress that experience in a very brief time, or do you just have to let them mature?
General Marsh: It’s tough. We’re taking many initiatives o try to accelerate the experience. We have all kinds of “lessons learned” information passing through the system. We have accelerated courses, and the Air force Institute of Technology is helping us formulate courses that we can get out via our electronic blackboards and videotape courses to pass along lessons. We’re working in every conceivable way that we know to make up for that lack of experience. You probably can’t expect to make up for it completely; all you can do is minimize the problem. As I say, this is one of the toughest problems have at the moment.
AFM: Are AFSC’s civilian engineers a corporate memory and a corporate pool of experience for you?
General Marsh: Yes. And they help to a great degree with that problem. Some of our organizations depend more on civilian engineers than others. For example, our Electronic Systems Division and the Space Division don’t have the civilian engineers to the extent that we have them out at Aeronautical Systems Division. We also rely on federal contract research centers. But it’s that blue-suit engineer we count on for understanding the real problem on the flight line. The officer brings a real understanding of the operator’s problem, having operated the equipment. We need a guy who has maintained it, who’s been in the maintenance squadron. He understands what it’s really like. And that’s what we’re lacking today.
AFM: That’s operational experience brought to bear on the design or the design issue?
General Marsh: That’s just part of it. The other part is having people who have worked on several systems, been in the air, learned the lessons from other systems, and then being able to transfer them to a new one. To a great extent, we’ve lost that capability.
AFM: During a symposium at the AFA Convention the participants concluded there are no quick and easy answers. You can’t suddenly create a generation of new engineers. Will you be faced with this for a generation, do you think?
General Marsh: It will be around for quite some time. It’s a larger problem than just Systems Command, as you point out. It’s a national problem. There are various 16,000 engineers a year for some years to come. Unless something is done in a major way, that will tend to even get worse because demographics are such that schools will likely be producing fewer engineers.
The problem has many dimensions. The faculty and plant capacity of the universities are particular problems. Universities are losing their faculty to the lures of industry. Their facilities are frequently out of date and will require great investment to update them. These kind of problems will face us for some years to come. That means that the Air Force may have to bear a disproportionate share of the shortage because we can’t compete financially with industry. I think there are fairy tough times ahead. But we must try and retain those young engineers that we do get. And this will require some very special attention. The retention bonus for engineers that Congress recently passed will help. I just don’t want to predict right now what we can do about retention, but that’s one of our major objectives.
I think we can solve the accession problem with the many things I’m sure you’ve heard about—the ROTC scholarship program, the College Seniors program, and so on. We may be able to meet our recruiting objectives within the next three or four years, although I’m somewhat skeptical of that. Personnel thinks we can. If we succeed at that, then the question is how are we going to retain them. That’ll be the major challenge.
AFM: Do you find that giving them bona fide engineering challenges tends to retain them?
General Marsh: Sure. We have a unique challenge in the command in that engineers coming out of school, just be the nature of their training, look forward to “hands-on” engineering efforts. That is what you’re trained to do in school—go out and design something. Now we don’t do much of that. We do some in our laboratories, but the major part of our engineers’ duties is engineering management—to manage and oversee engineering, review engineering, and so on. The role of our engineer is to oversee the contractors’ engineering. And that’s important—very important—it’s proven to be an absolute necessity. To some young people, that’s not what they want to do. So, initially, they are somewhat frustrated, and that’s a problem we have to cope with.
We have to persuade them that engineering management is challenging and rewarding. And the responsibility they soon are given is at a much higher level than that of their civilian contemporaries. We have to get them over the shock that engineering in the Air Force is not what they thought it was going to be. We’ve tried to do everything we can back in the ROTC, with films and all, to try to explain what an engineer does in the Air Force. We need to tell them that the engineer does engineering management, and that it’s a very rewarding and important challenge.