McNair Paper Number 52, Chapter 9, October 1996

9.

"SITUATION AWARENESS" IN AIR-TO-AIR COMBAT AND FRICTION

The last three chapters sought to build indirect arguments supporting the proposition that the potential for, if not the actuality of, general friction is likely to persist more or less undiminished in future wars despite technological advances. This conclusion should not, however, be construed as implying that friction is impervious to technological manipulation. The weapons of war influence how friction will manifest itself, and superior weapons can broaden the potential for manipulating the relative balance of friction between opposing sides in one's favor.

To demonstrate technology's ability to manipulate Clausewitzian friction, it will be necessary to focus on tactical interactions. Tactical interactions and effects, unlike those at the operational and strategic levels of war, are amenable to quantification and statistical analysis, at least up to a point. The reason for separating the aspects of war that are quantifiable from those that are not along the imprecise boundary dividing tactical interactions from the operational level of war lies in the degree of penetration by political-strategic objectives. In the author's experience, the concrete, specific political-strategic objectives that pervade the conduct of actual conflicts so influence strategy and operational art as to render both inputs and outputs at those levels irredeemably qualitative in character. Only at the level of tactical interactions do political-strategic aims become sufficiently remote to allow a fair degree of quantification of overall results.(Note 1)

The choice of what type of tactical interactions to examine is not, presumably, of critical importance beyond the fact that the small numbers of participants typically involved allow certain patterns to be seen more readily than would be possible in ground engagements involving hundreds or thousands of combatants. (Note 2) That said, air-to-air combat has been selected for a couple of reasons. Not only has an extensive body of air-to-air combat experience been accumulated since 1914, but a number of test evaluations have been flown on instrumented ranges and in simulators for the express purpose of providing statistically relevant data. In addition, air-to-air combat is as dependent on and pervaded by advanced technology as any area of late 20th-century warfare. Land and naval warfare reach back to the beginnings of recorded history. Powered flight, by comparison, was only achieved in 1903 as a result of aeronautical and engineering advances achieved by Orville and Wilbur Wright, and the heavier-than-air fighter is a phenomenon that dates only from World War I. Without the airplane, there would have been neither airmen nor air forces, and so strong has been the psychological attachment of American airmen to the planes they happened to fly that, to this day, many pilots identify themselves first and foremost as the "drivers" of specific aircraft types, often down to the model. (Note 3)

What factors have tended to drive engagement outcomes in air-to-air combat? As suggested in chapter 6, surprise has been linked to general friction. Air combat experience going at least back to the Second World War suggests that surprise in the form of the unseen attacker has been pivotal in three-quarters or more of the kills. In writing about his experiences flying long-range escort missions over northern Europe with the U.S. Eighth Air Force during World War II, P-38 pilot Lieutenant Colonel Mark Hubbard stressed that "90% of all fighters shot down never saw the guy who hit them." (Note 4) Hubbard was by no means alone in observing that friction in the form of the unseen attacker from six o=clock played a dominant role in engagement outcomes. The American P-47 pilot Hubert Zemke (17.75 air-to-air kills in World War II) stressed that "few pilots are shot down by enemies they see." (Note 5) Similarly, the German Me-109 pilot Erich Hartmann, whose 352 kills during World War II made him the top scorer of all time, later stated that he was "sure that eighty percent" of kills never knew he was there before he opened fire." (Note 6)

Subsequent technological developments in the means of air-to-air combat did not change the basic pattern observed by Hubbard, Zemke, and Hartmann during World War II. These developments include the shift to jet fighters for air superiority during the Korean War, the advent of infrared air-to-air missiles by the mid-1950s, and the appearance of radar-guided air-to-air missiles in time for American use in the Vietnam War. In Southeast Asia, American fighter crews experienced around 600 air-to-air engagements from April 1965 to January 1973. These engagements produced some 190 kills of North Vietnamese fighters against 75 American losses. (Note 7) Detailed engagement reconstructions under Project Red Baron revealed that over 80 percent of all aircrews downed, friendly as well as enemy, either were unaware of the attack, or else did not become aware in time to take effective defensive action.8 In the jargon of American fighter crews, such failures to be sufficiently cognizant of what is taking place in the air combat area around one to avoid being hit by enemy fire have come to be described as a breakdown of "situation" (or "situational") awareness. In an air-to-air context, situation awareness (or "SA") can be understood as the ability of opposing aircrews to develop and sustain accurate representations of where all the participants in or near the air combat arena are, what they are doing, and where they are likely to be in the immediate future.

While it seems plausible to those experienced in the air combat arena that situation awareness provides at least an indicator or marker of frictional imbalances at the tactical level, the two concepts are not coextensive. Generally speaking, high SA for a given side implies low friction for that side and vice versa. However, the better than 20 percent of fighter crews shot down by enemy fighters in Southeast Asia who became aware of the attacker just before ordnance impacted their aircraft-but too late to take effective defensive action-arguably had high SA during the final seconds before being hit. Yet it is equally clear that they also confronted high friction in those last seconds. After all, their option sets in possibility space had abruptly been reduced to the point where few, if any, desirable courses of action remained. True, aircrews in these dire circumstances could still choose among such "possibilities" as initiating futile "last-ditch" maneuvers, reaching for the ejection-seat handles, or simply waiting to see how much damage their aircraft sustains from the opponent's bullets or missiles. But since none of these possibilities is very desirable, their real options were few-to-none.

With this distinction between situation awareness and general friction in mind, how is the ebb and flow of the former related to the waxing and waning of the latter? Does the imbalance of situation awareness between adversaries over the course of an engagement provide a sense of the magnitude of frictional imbalances? Building on the notion of option sets in possibility space, the answer involves two interwoven but opposing threads or tendencies. The first concerns the disappearance of options over time, the second the creation or emergence of new ones. One is dissipative, the other is creative. In air-to-air combat, the unexercised options that may disappear or recede over the course of a tactical evolution are more concrete than the high-level or "strategic" options affecting whole campaigns such as the Germans' Blitzkrieg of May 1940. If, for example, the pilots on one side elect to enter the fight in a formation so tight that their opponents are easily able to acquire all of them at the outset, then a host of more dispersed, difficult-to-deal-with options are quickly foregone by the side embracing the tight formation. The potential for that side to generate ambiguity, confusion, or surprise later in the engagement is, of course, by no means closed. Still, some options have been lost at the outset with consequences that may or may not eventually prove fatal, and others will fall by the wayside with each successive move or action.

At most, however, the falling aside of options at successive branch points in an air-to-air engagement represents only the negative or dissipative aspect of how option sets can wax and wane during an aerial encounter. The positive or creative aspect is the generation of new possibilities by dint of one's own initiative, creativity, quickness, and, above all, interaction with the opposing side. For instance, at long ranges an abrupt maneuver by one pilot to move his aircraft outside the effective envelope of an enemy's air-to-air missiles introduces a rapid transient that the opponent is unlikely to grasp and assimilate instantaneously. If the move is successful, then the defender's options widen while the attacker's narrow, at least momentarily. Or, equivalently, the defender's friction lessens while the attacker's does not, thus producing a shift in the relative balance of friction in favor of the defender. Although changes in each opponent's situation awareness may sometimes briefly diverge from changes in the balance of friction, the evolution just described can be understood in terms of a quantitative shift in SA in favor of the defender. Hence, with brief exceptions, there appears to be some link between the two sides' levels of friction and their levels of SA. (Note 9)

Even without the evidence from subsequent tests like ACEVAL (Air Combat Evaluation) and the AMRAAM (Advanced Medium Range Air-to-Air Missile) OUE (Operational Utility Evaluation), combat data from Europe in World War II and Southeast Asia during 1965-1973 not only confirm the contention in chapter 8 (Proposition I) that general friction can dominate combat outcomes, but indirectly quantify what the term "dominate" has meant in historical air-to-air combat. If some 80 percent of the losses have resulted from aircrews being unaware that they were under attack until they either were hit or did not have time to react effectively, then a relative deficit of "situation awareness" (Note 10) has been the root cause of the majority of losses in actual air-to-air combat. A deficit in SA accounts for four out of five losses. While this statistic may not measure frictional imbalances directly, it does reflect the influence friction has had on outcomes over the course of large numbers of air-to-air engagements.

After the Vietnam War, SA's dominance of fighter-versus-fighter combat was extensively confirmed by large-scale simulations of air-to-air engagements. In the late 1970s, two major air-to-air tests were flown on an instrumented air combat maneuvering range north of Nellis Air Force Base, Nevada: the Air Intercept Missile Evaluation (AIMVAL) and ACEVAL. These tests pitted "Blue Force" F-15s and F-14s against "Red Force" F-5Es, chosen to simulate the Soviet-built MiG-21; Cubic Corporation's air combat maneuvering instrumentation (ACMI) system provided a combat area some 40 nautical miles in diameter as well as "realtime" data on the engagements. (Note 11) The Blue fighters were "armed" with guns, short-range infrared (IR) missiles, and the medium-range, radar-guided AIM-7F Sparrow; Red ordnance was limited to guns and IR missiles.

AIMVAL sought to assess the operational utility of five existing and proposed infrared missile concepts. (Note 12) ACEVAL explored the factors affecting engagement outcomes when multiple aircraft are involved, with force size, force ratio, and initial GCI (ground controlled intercept) condition (Red advantage, neutral, or Blue advantage) as the primary test variables. (Note 13) AIMVAL's test matrix included Blue-versus-Red force ratios of 1-v-1 (one F-15 or F-14 versus one F-5E), 1-v-2, 2-v-2, and 2-v-4, and called for 540 valid engagements involving 1,800 sorties. (Note 14) ACEVAL's test matrix added 2-v-1, 4-v-2, and 4-v-4 trials to the four force ratios used in AIMVAL and required a total of 360 valid engagements involving 1,488 sorties. (Note 15) Needless to say, many additional engagements and sorties were flown preparing for the actual trials. In fact, perhaps the most famous single engagement of both tests was the ACEVAL "Towering Inferno" 4-v-4 in which all eight participants were shot down after a minute and 52 seconds, was not a valid trial. (Note 16)

By the end of 1977, there was growing debate over the implications of these tests. ACEVAL, for example, was originally designed on the premise that when Blue fighters employed radar-guided missiles to shoot from beyond visual range (BVR), engagement outcomes would depend "primarily upon the performance characteristics of the avionics and weapons systems employed" and, with proper testing, could be predicted on that basis. (Note 17) If this premise had been borne out, it would have represented a sharp departure from historical combat experience in which human factors such as inferior situation awareness had consistently proven to be the dominant factor in engagement outcomes. At the time, there were a number of observers who viewed AIMVAL and ACEVAL first and foremost as a basis for choosing between hardware alternatives. Even some seasoned fighter pilots insisted on reading the tests as showing that hardware factors like long-range identification and missile performance would drive engagement outcomes in the future. (Note 18) On the other hand, there were also informed interpretations that fit much better with historical combat experience. In particular Lt. Col. "Shad" Dvorchak, who had been directly involved in the analysis of AIMVAL/ACEVAL data at Nellis, noted that in AIMVAL incremental hardware advantages had tended to wash out in the long run as opponents adapted, and that in ACEVAL human interactions had been five times as influential on outcomes as test variables like force ratio or the initial GCI condition. (Note 19) Suffice it to say, however, that by 1979 there was a wide range of opinion on the potential of technological advances to preempt the frictional factors that had dominated historical air combat.

In retrospect, the AMRAAM OUE, conducted in McDonnell Douglas flight simulators starting in 1981, resolved the issue in favor of Clausewitz. The AMRAAM OUE test matrix called for over 1,200 engagements involving around 10,000 simulator sorties. (Note 20) Instead of a small cadre of specially selected aircrew, AMRAAM OUE participants were drawn from operational units in the U.S. Scenarios included fighter-sweep situations (2-v-2 and 2-v-4) as well as trials in which the Blue fighters faced Red fighters escorting strike aircraft (2-v-4 + 4, 2-v-2 + 6, and 4-v-4 + 4). About half the trials were excursions from the standard sweep and combat air patrol scenarios.

Throughout AIMVAL and ACEVAL, visual identification prior to weapon employment had been a mandatory rule of engagement and the only radar missile allowed had been the AIM-7F Sparrow. (Note 21) Inevitably these constraints biased both tests against effective BVR employment. In the AMRAAM OUE, by comparison, the Blue force was given the medium-range, radar-guided AMRAAM and half the non-excursion trials were run with BVR rules of engagement. The natural expectation was that in the BVR trials at least, Blue hardware advantages would drive engagement outcomes. The bottom line from the test, however, turned out to be otherwise. Situation awareness proved to be "the single most important factor affecting engagement outcomes." (Note 22) For both sides, being aware of adversary weapons envelopes and keeping outside them to avoid being "shot," while trying to maneuver adversaries into their own weapons envelopes, proved as important and dominant as it had been in ACEVAL. Especially for Red fighters facing the AMRAAM, the role of situation awareness in scoring "kills" and avoiding being "shot down" tended to hang on even smaller differences than before.

To preempt misinterpretation, the statistical dominance of AMRAAM OUE engagement outcomes by situation awareness should not be construed as implying that hardware, including aircraft performance, avionics, and missile capabilities, counted for nothing. To the contrary, superior Blue hardware conferred building blocks or baseline elements of advantage that the Red side had to work hard to overcome and, in the aggregate, Blue hardware advantages were reflected in superior Blue exchange ratios. Statistically, though, the outcome of any particular engagement most often hinged on very small differences here or there across a large set of interrelated human and hardware factors, and the dominant of these factors was situation awareness.

This test result obviously reinforces historical air-to-air combat data rather than contradicting them. It also supports the Clausewitzian hypothesis that friction is a structural feature of combat interactions with humans "in the loop." Finally, it lends concrete empirical supportCat least at the tactical levelCfor the proposition that eradicating friction in some permanent way through hardware improvements is, at best, unlikely.

Does the bottom line from the AMRAAM OUE further imply that technological manipulation of friction is difficult or impossible? Not at all if the technology is focused on the root problems. Since the earliest days of air-to-air combat in World War I, sorting in the sense of the timely and effective targeting of opponents has been a sine qua non of positive results (downing enemy aircraft as opposed to merely avoiding being shot down oneself). While sorting is a trivial problem in isolated 1-v-1 engagements, and manageable with voice communication in 2-v-2s, it becomes vastly more difficult in 4-v-4s or more complex engagements. Obviously sorting in complex engagements requires situation awareness, and this awareness has to be maintained while maneuvering one's own aircraft, manipulating its sensors, and making the critical targeting decisions that produce kills. Since today's top-performing fighters can sustain G-loadings as high as nine times the earth's gravity, as well as high pitch and roll rates, the environment inside the cockpit is both physically and mentally demanding. All-aspect, high-G missiles, including some able to be fired from well beyond visual range, further complicate the problems of developing and maintaining high SA because the missile envelopes themselves are highly dynamic. Last but not least, the time spans participants have for observing what is going on, orienting themselves, deciding what to do (including sorting or targeting), and trying to execute split-second decisions are highly compressed. Thus, even the slightest misstep in sorting during 4-v-4 or more complex engagements can quickly lead to disaster, as the Blue side's failure to target two of the Red fighters at the outset of the one-minute-and-52-seconds "Towering Inferno" trial from ACEVAL illustrates. (Note 23)

Small wonder, then, that sorting efficiencies in complex engagements have been quite low. By 1984, Billy R. Sparks, a former U.S. Air Force F-105 "Wild Weasel" pilot with combat experience over North Vietnam, had been involved in analyzing or running three major humans-in-the-loop tests: AIMVAL/ACEVAL, the AMRAAM OUE, and the Multi-Source Integration test (also conducted in simulators). Yet, in reflecting on all that experience, Sparks felt that he had not once witnessed perfect sorting in 4-v-4 and more complex engagements. (Note 24) "You're only about one-third as efficient as you think you are [at sorting in complex engagements], which is why you go out with a sexy missile and lose your ass anyway." (Note 25) As Clausewitz wrote, in war "the simplest thing is difficult" and it is hard for normal efforts to achieve even moderate results. (Note 26) Such observations go far to explain why even small deficits in SA relative to the opposition have been statistically more dominant in engagement outcomes than differences in aircraft, weapons, force ratios, or other conditions such as having help from GCI. It also strongly suggests that friction's influence on outcomes in air combat during World War II was not noticeably different in Korea's "MiG Alley," the Vietnam War, the Middle East in 1967, 1973 and 1982, or even in Desert Storm. In this sense, general friction's "magnitude" does not appear to have diminished noticeably over the course of all the technological advances separating the P-51 from the F-15.

Could information technology be used to mitigate this longstanding pattern of very low sorting efficiencies in complex engagements arising from seemingly small lapses in situation awareness? Early experience in 4-v-4 and more complex engagements with the recently fielded Joint Tactical Information Distribution System (JTIDS) indicates that the answer is "Yes." JTIDS not only provides integrated, all-aspect identification of friendlies and hostiles based on available information but even displays targeting decisions by others in one's flight. The aggregate gains in air-to-air effectiveness resulting from these improvements in SA and sorting have been nothing less than spectacular. During Desert Storm, U.S. Air Force F-15Cs, aided in most cases by E-3A Airborne Warning and Control Systems (AWACS) aircraft, downed 28 Iraqi fighters without a single loss, including fifteen kills from engagements that began with BVR shots. (Note 27) When JTIDS-equipped F-15s flew against basically the same fighter/AWACS combination that had done so well in the 1991 Gulf War, the JTIDS "information advantage" enabled them to dominate their opponents by exchange ratios of four-to-one or better.(Note 28) Hence technology, properly applied, can certainly manipulate the differential in friction between opposing sides to one's advantage at the tactical level.

This conclusion suggests a corollary concerning non-technological stratagems for gaining tactical advantage that, while deserving mention, will not be argued at length. If advances in information technologies alone, when focused on root problems such as SA, can provide a four- or fivefold advantage in air-to-air exchange ratios, then it seems plausible to expect that superior tactics, training, employment concepts, or even organizational arrangements would also provide substantial margins of advantage at the tactical level as well. Perhaps improvements in tactics or training could even provide margins of superiority comparable in magnitude to the initial experience with JTIDs-equipped F-15s, although this extension of the hypothesis would have to be confirmed empirically.

The implication that cannot be drawn from JTIDS experience, however, is that friction has been permanently eliminated. If adversary forces fielded a system comparable to JTIDS, then the burden of achieving superior situation awareness and sorting in air-to-air would fall back on the "manual" abilities of human brains to absorb, interpret, and act upon automated information about who is where and doing what more quickly or more effectively than the opponents (or both). Exactly how frictional imbalances might ultimately manifest itself in this "technologically altered" set of conditions is hard to anticipate. What can be said with confidence, though, is that by reducing the aspects of friction we have been discussing with improved information systems, friction will probably manifest itself in other ways or in areas that we may not even be able to predict. There are two reasons for this conclusion. First, new technology amounts to introducing novelty into the combat area, and the indirect and second-order consequences of novelty within the context of human interactions are seldom, if ever, fully predictable. Second, if both sides have access to the novelty or innovation, in this case JTIDS, then transforming the resulting SA and targeting data into knowledge and action better or quicker than the opponents will still ultimately be taking place in the same sort of "gray matter" that members of our species have been carrying around in their skulls for the last 45,000-90,000 years. Both sides will have improved compared to where they were without JTIDS, but the relative margin of advantage will fall back to differences between the men in the machines.