OPINION ON LIABILITY

JAMES TURNER, Senior Judge

Hughes Aircraft Company owns U.S. Patent No. 3,758,051, entitled "Velocity Control and Orientation of a Spin-Stabilized Body," which describes an apparatus for controlling the attitude of a spin-stabilized spacecraft. Hughes brought this action pursuant to 28 U.S.C. § 1498, seeking just compensation for the unauthorized use or manufacture by or for the government of spacecraft containing an embodiment of the patented apparatus. Hughes seeks compensation in excess of four billion dollars. This opinion addresses all previously unresolved liability issues; an opinion addressing damages will follow.

I A

Hughes filed its complaint on November 13, 1973. Since then, the parties have engaged in a lengthy and costly litigation in both trial and appellate courts.

A trial — confined to liability issues pertaining to the 14 spacecraft then accused — was conducted in 1976-77 and resulted in a decision by a trial judge of the Court of Claims that the patent was invalid under 35 U.S.C. § 102(b). Hughes Aircraft Co. v. United States, 205 U.S.P.Q. 381 (Ct.Cl. Trial Div. 1979). On appeal, a panel of the Court of Claims rejected that conclusion and remanded the case for a determination of the other issues presented. Hughes Aircraft Co. v. United States, 226 Ct.Cl. 1, 640 F.2d 1193 (1980). On remand, the trial judge determined that the patent was valid; he also concluded that a group of spacecraft referred to as "real time" spacecraft infringed and that another group of spacecraft referred to as "store and execute" spacecraft did not infringe. Hughes Aircraft Co. v. United States, 215 U.S.P.Q. 787 (Ct.Cl. Trial Div. 1982). On appeal, the United States Court of Appeals for the Federal Circuit reversed-in-part, concluding that the store and execute spacecraft did infringe, and affirmed in all other respects. Hughes Aircraft Co. v. United States, 717 F.2d 1351 (Fed. Cir. 1983). The Federal Circuit remanded the matter to this court for a determination of the damages due for the 14 spacecraft adjudged to infringe. Id.

Based on this history of the case, the portion of this long-running litigation extending from initial filing through the 1983 decision of the Federal Circuit has been frequently referred to as the "liability" phase, and the portion occurring since the 1983 remand to this court has been referred to as the "accounting" or "damages" phase. However, in addition to the 14 spacecraft adjudged to infringe in the "liability" phase, Hughes has accused 94 spacecraft during this "accounting" phase, for a total of 108 accused spacecraft. (The appendix to this opinion lists each of the 108 spacecraft accused during any stage of this litigation.) Thus, this so-called "accounting" phase has entailed the determination of liability concerning the 94 newly accused spacecraft as well as a determination of damages with respect to any of the 108 spacecraft found to infringe.

Trial of issues involved in this current "accounting" phase was accomplished in two major parts, one occurring before expiration of the Williams patent on September 11, 1990 and the other occurring thereafter, it having been found more efficient and otherwise appropriate to defer final proceedings concerning 24 specific spacecraft until after patent expiration. The first part involved receipt of evidence in Pasadena, California and Washington, D.C. over a period between February 2 and December 16, 1988, and a series of arguments over a period from February to September 1989. The second part involved receipt of evidence in Pasadena, California between January 15 and 24, 1991 and final argument in May 1991.

During this post-1983 "accounting" phase, the court heard the testimony of 61 witnesses, reviewed the deposition testimony of approximately 95 additional witnesses and received into evidence more than 4500 exhibits; documentary exhibits alone comprise approximately 142 linear feet. The transcript of proceedings over the period from February 1988 through May 1991 includes 20,812 pages.

B

The matter is now matured for final decision on all issues. We assume familiarity with the Federal Circuit's 1983 liability decision, 717 F.2d 1351.

Part II of this opinion briefly describes the Williams patent. Parts III, IV and V address the three remaining groups of liability issues, which can be summarized as follows:

1. Whether the Galileo or Atmosphere Explorer spacecraft infringed the Williams patent under the doctrine of equivalents.

2. Whether the Global Positioning System spacecraft have been manufactured or used with the authorization or consent of the government so that they constitute an infringement of the Williams patent under 28 U.S.C. § 1498.

3. Whether any of the "foreign" spacecraft have been "used" by the government sufficiently to constitute infringement of the Williams patent under 28 U.S.C. § 1498.

II

The Williams patent discloses an apparatus for controlling the attitude of the spin axis of a spin-stabilized spacecraft. The patent consists of three independent claims. Because claim one is the broadest of the three, the parties agree that a decision concerning claim one will determine whether the spacecraft in issue infringe claims two and three. Claim one reads as follows:

Apparatus comprising:

a. a body adapted to spin about an axis;

b. fluid supply means associated with said body;

c. a valve connected to said fluid supply means;

d. fluid expulsion means disposed on said body and coupled with said valve and oriented to expel said fluid substantially along a line parallel to said axis and separated therefrom;

e. means disposed on said body for providing an indication to a location external to said body of the instantaneous spin angle position of said body about said axis and the orientation of said axis with reference to a fixed external coordinate system;

f. and means disposed on said body for receiving from said location control signals synchronized with said indication;

g. said valve being coupled to said last-named means and responsive to said control signals for applying fluid to said fluid expulsion means in synchronism therewith for precessing said body to orient said axis in a predetermined desired relationship with said fixed external coordinate system.

PX 1.

In a typical application, the Williams patent generally functions as follows. Sun sensors located on one side of a spinning spacecraft transmit electronic pulses to ground control. Those sensors consist of two slits in the shape of a "V". One of the slits is vertical, and the other slit is canted. Each slit is backed by a photocell. Each of the slits sends an electronic pulse to the ground each time it detects the sun. Those pulses enable ground control to determine the spacecraft's spin rate, orientation with respect to the sun and instantaneous spin angle. The spin rate is measured by the time difference between successive pulses from the same photocell. The ground controllers calculate the orientation of the spin axis by measuring the sun angle and the strength of radio signals emitted toward the earth by sensors on the spacecraft. The sun angle is the angle between the spin axis of the spacecraft and a line drawn from the spacecraft to the sun; this is calculated based on the time difference between pulses from the canted and vertical slits. The instantaneous spin angle is the spacecraft's location in its 360 degree revolution at any given instant.

In the Williams invention, the spacecraft is "dumb"; the logic function is performed by ground controllers. Using the data just described, ground controllers calculate the time delay between the occurrence of a sun pulse and the times at which the jet should be fired to accomplish the desired precession.

Ground control transmits this information to the spacecraft in "real time," i.e., ground control sends a control signal each time it wants the spacecraft's attitude thrusters to pulse, resulting in an instantaneous firing of the thrusters. Ground control continues to send these control signals for each successive spin revolution of the spacecraft until the spacecraft has completed the desired maneuver.

The prior art surrounding this patent and the prosecution history of the patent application have been discussed extensively elsewhere. See Hughes, 717 F.2d at 1354-56, 1362-63 (discussing prosecution history); id. at 1359-60, 1362 (discussing prior art).

III

In patent litigation, often one of the major issues is whether a patent claim "reads on" the accused device — either literally or by application of the doctrine of equivalents. In this action, that issue has been resolved with respect to all but four of the accused spacecraft. See Appendix to Opinion (table describing the infringement determinations for the various spacecraft at issue).

This portion of the opinion addresses whether the attitude control systems of those four spacecraft — Galileo and three Atmosphere Explorer spacecraft (AE C, D and E) — infringe the Williams patent. Hughes concedes that neither Galileo nor AE literally infringes the Williams patent, but alleges that both spacecraft infringe the patent under the doctrine of equivalents. After describing the attitude control systems of those spacecraft in some detail, we discuss the scope of the doctrine of equivalents and then apply that doctrine to Galileo and the three accused AE spacecraft. We conclude that the attitude control systems of the Galileo and the AE spacecraft do not infringe the Williams patent under the doctrine of equivalents.

A 1. Galileo

Galileo is a one-of-a-kind, interplanetary explorer with a mission to explore Jupiter. For most of its mission, Galileo operates as a dual-spin spacecraft with a spinning section and a despun section. A high gain antenna (by which the spacecraft communicates with the earth) and star scanners, among other things, are mounted on the spinning section, and a scan platform and gyroscope assembly are mounted on the despun section. The sections are connected by a spin bearing motor assembly that controls the relative position of the sections. While operating as a dual-spin spacecraft, Galileo uses two modes — a cruise mode, during which the velocity thrusters are firing, and an inertial mode, when they are not. Tr. 7305-07, 15,462.

The attitude control system of Galileo performs three different types of maneuvers — a sun acquisition maneuver, a high gain antenna maneuver and a commanded turn maneuver (Tr. 7783-96). Hughes's allegation of infringement is limited to the commanded turn maneuver.

Following the lead of the parties, we think the relation between the Williams patent and the commanded turn maneuvers performed by Galileo is adequately illustrated by consideration of a particular commanded turn maneuver referred to as the "turn — burn — re-turn" maneuver. This is the maneuver Galileo uses to change its translational motion, whether the change is a change in the speed of that motion or a change in the direction of that motion. In order to make such a change, Galileo must fire its velocity thrusters in the direction opposite the direction toward which the change in motion will occur. Because Galileo normally travels with its high gain antenna pointing toward earth, it can accomplish this maneuver only if it changes its attitude so that its velocity thrusters are pointing in the correct direction for the desired firing (the "turn" maneuver). Once the velocity thrusters have fired (the "burn"), the spacecraft returns to the original attitude, with its high gain antenna again pointed toward earth (the "re-turn" maneuver). Tr. 7448.

The attitude control system of Galileo performs both the "turn" and "re-turn" procedures described above by controlling the firing of the spacecraft's precession thrusters. The attitude control system of Galileo performs the commanded turn maneuver as follows. Ground control sends Galileo two pieces of information: (1) two coordinates specifying the direction relative to the celestial sphere system that it wants Galileo to point before firing its velocity thrusters (referred to as EME-50 coordinates) and (2) the desired start-time of the turn maneuver (Tr. 15,483, 15,510). With this information, Galileo's on-board computers perform all the calculations required to turn the spacecraft.

Galileo has V-slit star scanners that ascertain the earth-pointing orientation of the spacecraft with respect to a fixed external coordinate system (the celestial sphere system) by comparing the light received from various stars with star maps stored in its on-board computer memory (Tr. 7688-90). Because the on-board computer knows where the celestial bodies are with respect to the fixed coordinate system, it can calculate the spacecraft's orientation with respect to that system. Then the on-board computer calculates a "turn plane," which is the geometric plane between the spacecraft's present orientation and its desired orientation through which the spin axis will move as the spacecraft precesses. It does this by referring to the coordinates of the present spacecraft orientation as determined from the star maps and the coordinates of the desired orientation as sent from the ground (Tr. 7481-83, 15,473).

Then, an on-board gyroscope is initialized with the present orientation of the spacecraft. Once initialized, the gyroscope acts as a reference to determine if the spacecraft is correctly turning from its earth-pointing orientation to its desired orientation. Tr. 7765.

The physics of spinning objects dictates that thrusters on a spinning spacecraft must pulse at an angle of 90 degrees with respect to the direction in which the spacecraft will precess (Tr. 7476). Calileo's precession thrusters are fixed so that they fire at a 90 degree angle to the axis of the scan platform. Accordingly, Galileo's scan platform — which, as indicated, is at a 90 degree angle to the precession thrusters — must be aligned with the turn plane (Tr. 7482). The scan platform which is mounted on the despun portion of the spacecraft is connected to the spinning portion of the spacecraft by a spin bearing assembly with an optical encoder containing 360 tick marks, reflecting the 360 degrees in a circle. Galileo's on-board computer rotates the scan platform so that it is aligned with the turn plane by reading those tick marks. Once the scan platform has rotated into the turn plane, the appropriate precession thruster is fired. It is important to note that Galileo does not operate by calculating a time or angle delay from a sun pulse (Tr. 15,637) but instead always fires at the same angle — 90 degrees from its target.

Finally, after each thruster pulse, the onboard computer and the gyroscope determine if Galileo has completed the turn by comparing its precessed orientation with the EME-50 coordinates transmitted from ground control. If Galileo has not completed the turn maneuver, the thruster will pulse again on its next revolution. Once the on-board computers determine that the turn is complete, the precession thruster will cease pulsing. This verification process is referred to as "error nulling." It is critical that Galileo's attitude control system has an error nulling feature because the mission of Galileo requires that its orientation be accurate to at least a tenth of a degree. Tr. 7450, 7458.

It also is relevant that Galileo precesses in what is called a "great circle path," which is the shortest distance between two points on a sphere. By doing so, Galileo is able to minimize fuel use. Galileo is able to precess in this more direct path because it calculates a turn plane that contains the target direction and then always fires its precession jets at a 90 degree angle to the turn plane. Tr. 7637-41.

After turning, the spacecraft's velocity thrusters fire. As explained, the velocity thrusters are not part of the spacecraft's attitude control system but instead control the translational motion of the spacecraft. Once the velocity thrusters have fired, Galileo's on-board attitude control system is now ready to turn Galileo to its initial earth pointing orientation. The attitude control system functions the same way in performing the "re-turn" as it did in performing the "turn." To perform the "re-turn," Galileo's on-board computer recalls the coordinates of Galileo's earth-pointing orientation. The on-board computers calculate a turn plane. The gyroscope is again used as a reference. The scan platform is aligned with the turn plane by referencing the tick marks on the spin bearing motor assembly. The precession thrusters fire until the gyroscope and on-board computers verify completion of the "re-turn" maneuver. Once the spacecraft has completed its "return" maneuver, the high gain antenna is pointing towards the earth and Galileo has totally completed its "turn — burn — re-turn" maneuver. Galileo is then able to receive a new set of EME-50 coordinates from ground control for its next desired "turn — burn — re-turn" maneuver.

In sum, the only commands ground control sends Galileo are the EME-50 coordinates, which contain the direction Galileo must point before firing its velocity thrusters, and the start time of the "turn" maneuver. The calculations of the turn plane, initialization of the gyroscope, positioning of the scan platform, firing commands for the thrusters, determination of when the spacecraft has reached its desired orientation, and all other commands necessary to perform the "turn" correctly are performed by Galileo's sophisticated on-board computer using a program installed before launch. All similar calculations necessary to perform the "re-turn" maneuver, i.e., to reorient Galileo in its original earth-pointing position, are also performed by the onboard computers, with no additional aid or direction from ground control.

During the entire commanded turn process, communication with ground control is broken because Galileo's high gain antenna is no longer pointing towards the earth (Tr. 7506). Ground control is not able to restore communications with Galileo until after the spacecraft has "re-turned" to its original earth-pointing orientation.

2. Atmosphere Explorer

The Atmosphere Explorer ("AE") space program consisted of five spacecraft. The final three of those spacecraft, designated AE C, AE D and AE E, are at issue in this litigation. The attitude control systems of those three AE spacecraft were identical for purposes of determining infringement. Accordingly, all three AE spacecraft will be referred to as a single satellite.

The AE spacecraft was a relatively low-cost vehicle that could be launched in a short period of time. The mission of this spacecraft was to measure constituents of the earth's atmosphere at altitudes as low as 130 kilometers. The purpose of those measurements was to further the study of the interaction between ultraviolet rays from the sun and the earth's atmosphere. In order to accomplish its mission, this spacecraft traveled in an elliptical orbit around the earth.

AE was equipped with three methods of attitude control. The system used most often was a magnetic torquing system (practical because of the near-earth orbit of AE). The other methods involved the use of the yaw thrusters in either a despun or spinning mode. Only the spinning yaw mode is accused of infringing the Williams patent. That maneuver was a backup mode to be implemented only if the other modes failed. Also, the spinning yaw maneuver would not have been useful for placing the spacecraft in its permanent orbit; it was designed only to alter the attitude of the spacecraft after it had been placed successfully in its permanent orbit by the other attitude control systems. Although non-use does not avoid infringement (it is uncontested that the accused back-up system was available for use), it is worth noting that the backup yaw precession system accused of infringing the Williams patent was never used to perform a precession maneuver for any of the launched AE spacecraft.

The yaw maneuver was a 180 degree precession of the z-axis (or spin axis) of the spacecraft. The purpose of the maneuver was to allow the earth sensors to keep the earth within their field of view as the spacecraft orbited, while at the same time keeping the side of the spacecraft that carried certain louvers used for temperature control out of direct sunlight. Every few months, AE had to be rotated to prevent those louvers from overheating.

Unlike the commanded turn maneuver that Galileo performs, the yaw maneuver was quite limited. In performing this maneuver, AE turned 180 degrees either clockwise or counter-clockwise. The spacecraft could not reorient itself to any orientation that ground control desired. AE's mission to explore the earth's atmosphere did not require the ability to reorient the spacecraft in multiple directions. The only purpose of the yaw precession system was to flip the spacecraft 180 degrees so that the thermal louvers were kept out of the direct sunlight. Because the mission requirements were much simpler, the attitude control system of the AE spacecraft was likewise much simpler.

Although never used for any of the three launched AE spacecraft (Tr. 8399-8400), ground control could implement the backup yaw precession system by transmitting one set of commands to the AE spacecraft approximately one half-hour in advance of the start-time of the desired maneuver. These commands would consist of a start-time and a stop-time. The spacecraft's on-board computer would have performed all remaining functions necessary to execute the precession maneuver (Tr. 8392). The following description of such a yaw maneuver assumes implementation.

The yaw maneuver was a precession of the z-axis about the y-axis. The maneuver depended on the knowledge of the precise location of the earth sensors and thrusters with respect to the imaginary axes of the spacecraft. The earth sensors passed the earth once during every revolution of the spacecraft. As the body of the spacecraft passed the earth, the sensors generated a series of pulses; one pulse was generated when the sensors made the sky to earth transition, and another pulse was generated when the sensors made the earth to sky transition. That information allowed AE to determine its spin rate.

AE's attitude control system operated with little aid from ground control. AE synchronized pulsing with the instantaneous spin angle position by using two counters. Those counters were initiated as the earth sensor made the sky to earth transition. The first counted the earth crossing time; the second, driven by a clock ticking 360 degrees per revolution, counted until the spacecraft's y-axis was aligned with the horizon. The first counter then counted down twice as fast as it counted up, reaching zero when the y-axis was directed toward nadir, i.e., when the negative y-axis was pointing at the earth. At that instant a nadir pulse was generated. Accordingly, AE could determine its orientation with respect to the earth. Tr. 8384-86, 8400.

There were four fixed regions or windows during the spacecraft's revolution during which firing could occur; two of the windows were used for yaw maneuvers, and the other two were used for delta-V maneuvers. Each window corresponded to 60 degrees of the spacecraft's rotation (30 degrees of each quarter revolution were not used). Each window was derived by the use of a counter generating 360 ticks during the spacecraft's revolution. After selecting the proper window (provided by ground control), pulsing was automatic and continuous throughout that window. The thruster firing continued during that portion of the revolution for successive revolutions until the stop-time occurred. Tr. 8400-03.

B

Hughes brought this action under 28 U.S.C. § 1498(a), contending that the government's unlicensed use of the Williams patent constituted a taking. The theoretical basis underlying a patent suit under § 1498 is that the government takes by eminent domain a compulsory, compensable license to use a patented invention whenever the invention is manufactured or used by or for the government. Motorola, Inc. v. United States, 729 F.2d 765, 768 (Fed. Cir. 1984). Nonetheless, the legal standard underlying the term "infringement," as used in patent actions under Title 35, generally is the same as the standard in an action against the government for "unauthorized use." See Lemelson v. United States, 752 F.2d 1538, 1548 (Fed. Cir. 1985) ("[P]rinciples of claim construction and reading claims on accused devices and methods are the same for either type of action.").

A patent infringement can be established in one of two ways. Hughes acknowledges the absence of literal infringement for Galileo and AE because elements e, f and g of the Williams patent are not found on the accused devices. Cf. Hughes Aircraft, 717 F.2d at 1365 (noting that the "store-and-execute" spacecraft did not literally infringe the Williams patent because they did not send an indication of the instantaneous spin angle to the ground). Hughes predicates its case for infringement on the doctrine of equivalents. Hence, this portion of the opinion addresses the law pertaining to the doctrine of equivalents.

1.

The doctrine of equivalents permits the protection accorded to a patent to exceed the scope of the patent's literal claim language where the accused device is essentially the same as the patented device. See London v. Carson Pirie Scott Co., 946 F.2d 1534, 1538 (Fed. Cir. 1991). The doctrine of equivalents "recognizes a fact of the real business world: words are not misappropriated; claimed inventions are." Laitram Corp. v. Cambridge Wire Cloth Co., 863 F.2d 855, 857 (Fed. Cir. 1988), cert. denied, 490 U.S. 1068, 109 S.Ct. 2069, 104 L.Ed.2d 634 (1989).

Whether a patent is alleged to be infringed literally or by application of the doctrine of equivalents, analysis of infringement entails two inquiries: "determination of the scope of the claims, as a matter of law" and "the factual finding of whether properly construed claims encompass the accused structure." Texas Instruments, Inc. v. United States Int'l Trade Comm'n, 805 F.2d 1558, 1562 (Fed. Cir. 1986). The first step — interpreting the scope of the claims — is a question of law. The second step — determining whether a patent claim reads on the accused devices — is a question of fact to be decided under the three-part test set forth in Graver Tank Manufacturing Co. v. Linde Air Products Co., 339 U.S. 605, 70 S.Ct. 854, 94 L.Ed. 1097 (1950). In that case, the Supreme Court held that "[a] finding of equivalence is a determination of fact" to be made if the accused device "`"performs substantially the same function in substantially the same way to obtain the same result"'" as the patented device. Graver Tank, 339 U.S. at 608-09, 70 S.Ct. at 856-57 (quoting Sanitary Refrigerator Co. v. Winters, 280 U.S. 30, 42, 50 S.Ct. 9, 13, 74 L.Ed. 147 (1929) (quoting Machine Co. v. Murphy, 97 U.S. (7 Otto) 120, 125, 24 L.Ed. 935 (1877))).

For purposes of this case, three judicial doctrines regarding the scope of the doctrine of equivalents have particular importance. First, courts have held that the breadth of equivalents that will be found to infringe a patent depends on the degree to which the patent represents an advance over prior art in the field. A "broad breakthrough invention" or "pioneer invention" — which the Supreme Court has described as one that performs "a function never before performed, a wholly novel device, or one of such novelty and importance as to mark a distinct step in the progress of the art" — merits a broad scope of equivalents. Westinghouse v. Boyden Power-Brake Co., 170 U.S. 537, 561-62, 18 S.Ct. 707, 718, 42 L.Ed. 1136 (1898); see Perkin-Elmer Corp. v. Westinghouse Elec. Corp., 822 F.2d 1528, 1532 (Fed. Cir. 1987). On the other hand, an "improvement [patent] in a crowded art" merits only a "narrow range of equivalents." Hughes, 717 F.2d at 1362.

The second relevant doctrine in the law of equivalents is the doctrine of prosecution history estoppel, which "precludes a patent owner from obtaining a claim construction that would resurrect subject matter surrendered during prosecution of his patent application." Hughes, 717 F.2d at 1362. "[W]henever the doctrine is evoked, `a close examination must be made as to, not only what was surrendered, but also the reason for such a surrender.'" Loctite Corp. v. Ultraseal Ltd., 781 F.2d 861, 871 (Fed. Cir. 1985) (quoting Bayer Aktienge-sellschaft v. Duphar Int'l Research B.V., 738 F.2d 1237, 1243 (Fed. Cir. 1984)).

Third, the Federal Circuit has emphasized that a patent owner must show the presence in the accused device of each claim element, or its equivalent. Pennwalt Corp. v. Durand-Wayland, Inc., 833 F.2d 931, 934-35 (Fed. Cir. 1987) (en banc), certs. denied, 485 U.S. 961, 108 S.Ct. 1226, 99 L.Ed.2d 426; 485 U.S. 1009, 108 S.Ct. 1474, 99 L.Ed.2d 703 (1988). Accordingly, even if the overall function, way and result of the patented and accused devices are the same, the absence of an element of the patent claim, or its equivalent, in the accused device means the device cannot infringe. Id.

2.

In 1983, the Federal Circuit upheld the validity of the Williams patent and concluded that the patent covered not only the original invention — the "real time" spacecraft — but also a variation developed later — the "store and execute" (S/E) spacecraft. The attitude control system in the S/E spacecraft functioned as follows. Sun sensors sent pulses to an on-board computer that calculated the spin rate and sent that information to the ground. The computer also sent information so that ground control could calculate the orientation of the spacecraft. In most of the S/E spacecraft, ground control did not know or need to know the ISA position. The ground, knowing the spin rate and orientation, then calculated a time delay between the occurrence of the sun pulse and the firing of the thruster. Then, ground control sent spaced signals indicating the position in the spin cycle in which jets should be fired, how many firings should be made and when to begin firing.

The Federal Circuit held that the S/E spacecraft did not literally infringe the Williams patent because of the absence from the S/E spacecraft of a means for providing ground crew an indication of ISA position (element e) and because the precession jets in the S/E spacecraft were not fired in "real time." Hughes, 717 F.2d at 1361. The court then proceeded to address infringement under the doctrine of equivalents.

The court noted that the prosecution history of the Williams patent application limited the equivalents analysis. Id. at 1362. Because Williams chose specific words of limitation to distinguish the patent he sought from a prior invention known as the McLean Space Vehicle (discussed in Hughes, 717 F.2d at 1354-55), the court concluded that the Williams patent could not be construed to encompass all structures in which a pulsed jet is used to precess the spin axis of a spin-stabilized body. Id. at 1362. In addition, the court concluded that the patented invention was "not of such `pioneer' status as to entitle the invention to the very broad range of equivalents to which pioneer inventions are normally entitled." Id. The court indicated, however, that the patent merited some range of equivalents because it was not merely an "improvement [patent] in a crowded art." Id.

Turning to the Graver test, the court concluded that the S/E spacecraft had an equivalent to the limitation in clause (e) of the Williams patent (which, as noted above in Part II, generally required a means on the spacecraft for indicating to an external location the spacecraft's ISA and orientation) because "retention of the ISA position in an on-board computer, while transmitting sufficient information to enable the ground crew to use that computer-retained information to control the satellite, is the modern-day equivalent of providing an indication of ISA to ground as taught by Williams." Id. at 1365. The court also found an equivalent to clauses (f) and (g) (which, as noted above in Part II, generally required a means on the spacecraft for receiving signals directing how thrusters should be fired to precess the spacecraft) because the "use of modern memory circuits on-board S/E spacecraft to store commands for later use" is merely "substitution of an embellishment made possible by post-Williams technology." Id.

In closing, the court concluded that the S/E spacecraft and the claimed Williams spacecraft

perform the same function (receipt of and response to command signals from an external location to accomplish precession), in substantially the same way (jet firings synchronized, albeit later and internally, with ISA position) to obtain substantially the same result (controlled precession of spin axis in a predetermined direction to orient a hovering satellite).

Some judges have suggested that there is a conflict between the Hughes "invention as a whole" approach and the element-by-element approach set forth in the Federal Circuit's sharply divided en banc decision in Pennwalt Corp. v. Durand-Wayland, Inc., 833 F.2d 931 (Fed. Cir. 1987). Pennwalt involved a patent on an apparatus that sorted items, such as fruit, by color or weight. The patent owner alleged that sorters manufactured by defendant infringed its patent under the doctrine of equivalents. See id. at 933-34. The Federal Circuit held that defendant's sorter did not infringe the patent because the plaintiff had failed to establish the "`presence of every element or its substantial equivalent in the accused device.'" Id. at 935 (quoting Lemelson v. United States, 752 F.2d 1538, 1551 (Fed. Cir. 1985)). In particular, the Federal Circuit rejected the argument that computer technology contained in the accused devices was equivalent to hardwired circuitry described in the patent. The court relied heavily on limitations in the original patent requiring the device to track the position of the object being sorted, coupled with its conclusion that "the accused machine had no component which satisfied either of [those] limitations." Id. at 937.

The government — characterizing the analysis of the Hughes panel as an "invention as a whole" approach — contends that Pennwalt's emphasis on the need to consider whether the accused device satisfies each limitation (or element) of the original patent dramatically undercuts the analysis of the Federal Circuit's earlier analysis of the Williams patent in Hughes. Hughes, on the other hand, argues that Pennwalt is fully consistent with the Hughes opinion. For two reasons, we doubt that the tension between the opinions is as significant as the government suggests. First, the Hughes opinion itself based its finding of infringement in large part on its conclusion, discussed above, that the store and execute spacecraft contained the substantial equivalent of each of the limitations in the Williams patent. Second, the majority opinion in Perkin-Elmer Corp. v. Westing-house Electric Corp., 822 F.2d 1528 (Fed. Cir. 1987), on which the Pennwalt majority relied heavily for its articulation of the element-by-element approach, expressly stated that its analysis was consistent with the analysis of the Hughes panel. See Perkin-Elmer Corp., 822 F.2d at 1532-33 (noting that the Hughes "infringement as a whole" analysis was a "recognition that, in applying the doctrine of equivalents, each limitation must be viewed in the context of the entire claim"). The government does not suggest that any decision of the Federal Circuit has expressed a contrary understanding of the current viability of the Hughes decision. In any event, as the remainder of this part of the opinion shows, it is our view that neither Galileo nor the AE spacecraft infringes the Williams patent even under the allegedly more lenient analysis of the Hughes panel. Accordingly, the government suffers no prejudice from our reluctance to conclude that the Federal Circuit's more recent decisions have rejected the analysis of the Hughes panel.

C

Although Hughes acknowledges that Galileo does not literally infringe the Williams patent, Hughes contends that Galileo infringes the patent under the doctrine of equivalents. In Hughes' view, the only difference between Galileo and Williams is that Galileo performs calculations on board that Williams taught to be performed on the ground. Accordingly, Hughes contends, the difference between Galileo and Williams is only an increased level of sophistication made possible by advances in computer technology.

According to the government, Galileo does not contain even the substantial equivalent of clauses (f) and (g) so that the patent is not infringed whether the accused device is considered as a whole or under the element-by-element analysis.

At the outset, we note that the proper inquiry is whether claim 1 of the Williams patent reads on Galileo. That inquiry does not involve a comparison between Galileo and the adjudicated infringing devices, namely the "store and execute" spacecraft. See Amstar Corp. v. Envirotech Corp., 823 F.2d 1538, 1545 (Fed. Cir. 1987). We conclude that the differences between Galileo and claim 1 of the Williams patent are not limited to the location at which certain operations are performed. Instead, the function, way and result in which Galileo precesses are fundamentally different from those of the Williams patent.

As noted above, the determination of the scope of the claims made by a patent is a matter of law. Accordingly, we start in this case from the Federal Circuit's description in Hughes of the function, way and result of the Williams patent:

The S/E spacecraft and the claimed Williams satellite reflect the precise circumstance envisaged in Graver, supra, for they perform the same function (receipt of and response to command signals from an external location to accomplish precession), in substantially the same way (jet firings synchronized, albeit later and internally, with ISA position) to obtain substantially the same result (controlled precession of spin axis in a predetermined direction to orient a hovering satellite).

Function. — Focusing first on the function, we note that Hughes contemplated the sending of a "command signal" from ground control. In performing the commanded turn maneuver, the only information sent from ground control to Galileo that is relevant for attitude control is the desired orientation. All subsequent calculations related to attitude control are performed by the on-board computers that are programmed and installed before launch. In fact, once the maneuver begins, ground control loses contact with the spacecraft. In contrast, it is basic to the operation of Williams that ground control generate signals that tell the spacecraft where in the spin cycle and how many times the jet must be pulsed to precess. Ground control does not lose contact with the spacecraft once the maneuver begins, and, in fact, constantly monitors the progress and sends additional firing signals.

In our view, the Hughes opinion teaches that the function of Williams is the receipt and response to specific firing commands, not directional coordinates. That function is not reproduced on Galileo. The mission of Galileo requires it to be largely autonomous, and, as part of that autonomy, computers generate all jet firing command signals on board the spacecraft. It is true that both require some sort of signal from the ground to initiate precession. The Federal Circuit has made clear, however, that the range of equivalents to be accorded to Williams is not broad enough to cover all ground controllable spacecraft. Hughes, 717 F.2d at 1363. If the general command signal required by Galileo were covered, then Williams would extend in practice to any type of ground control signal. Because we find that result inconsistent with the Federal Circuit's analysis in Hughes, we find that the type of commands transmitted by ground control substantially distinguishes the functions of Williams and Galileo.

Way. — The differences between Williams and Galileo are perhaps most marked concerning the "way" in which each accomplishes precession. We conclude that there are numerous conceptual differences between Williams and Galileo that are far greater than a simple change of the location at which various calculations are performed. We summarize three somewhat overlapping differences here.

First, the use of a different type of reference to fix the place in the spin cycle in which the precession jets will fire shows that the manner in which Galileo and Williams accomplish precession is not substantially equivalent. Galileo, unlike Williams, does not calculate a time delay or angle delay from a reference. When discussing whether angle-based systems infringed Williams, we noted that "the essence of the precession system [in both the time-based and angle-based systems] is the firing of the thrusters at a certain point after the sun pulse" (Tr. 17,191). Instead, Galileo uses a gyroscope and tick marks, both of which are mounted on the despun section of the spacecraft, to align the scan platform with the turn plane. Although we are not persuaded that a gyroscope-based system cannot be covered by the Williams patent (because it substitutes an internal reference point for the external reference point described in the Williams patent, ) we do conclude that the way in which the gyroscope is used here makes Galileo fundamentally different from the device described by Williams.

The laws of physics dictate the direction in which a spinning spacecraft will precess given the point in the spin cycle at which the precession jet is fired. The corollary of this is that in order to precess in the proper direction, a spinning spacecraft must fire its precession jets at a particular point in its spin cycle. Because prior art (the McLean Space Vehicle) had included a method of using an angle between the target and the spin axis as a reference to trigger firing, the Williams patent cannot cover all devices that rely on that principle. As construed by the Federal Circuit in Hughes, the particular way covered by the Williams patent is a way of precessing by use of a time or angle delay from an external reference, namely a sun pulse. By relying on physical marks on the spacecraft coupled with a gyroscope to determine the appropriate firing point, Galileo reflects a different way of accomplishing the desired precession.

Second, the ability of Galileo to see its target enables it to travel the shortest route. Both Williams and Galileo precess at a 90 degree angle from the direction the pulsing occurs. Galileo, which knows where its target is, always pulses at a 90 degree angle to the target direction. By doing so, Galileo is able to precess directly in its target direction, i.e., it travels a great-circle path (Tr. 7637-38, 15,467). In contrast, in Williams where the system cannot see the target and the sun is used as the reference, the time delay following the sighting of the reference is variable and must be computed by ground controllers. As a result, the Williams system travels in a spiral path, or rhumb line. Accordingly, Galileo travels the shortest possible route and minimizes fuel use (Tr. 15,467).

Finally, the two processes of determining that a spacecraft has precessed to the desired orientation are different. In Williams, ground control precomputes the position in the spin cycle in which the jets should be fired and the number of times the jet should be fired. Ground control then monitors the precession by using an oscilloscope. If the spacecraft does not reach the desired orientation, ground control sends additional command signals. In contrast, Galileo's error nulling system does not depend on the precomputation of these factors, in which there is always a degree of uncertainty.

Result. — The last part of the Graver tripartite test, that is, whether the results of Galileo and Williams are equivalent, is a closer question. Hughes defines the result of Williams as the "controlled precession of [the] spin axis in a predetermined direction." Hughes, 717 F.2d at 1366. We note that there are at least two differences in the results achieved by Galileo and Williams. First, the error nulling function of Galileo permits it to achieve a greater degree of accuracy (Tr. 7478). Second, the great circle path traveled by Galileo permits greater fuel efficiency. Although those differences are not dramatic, they also tend to support a finding of nonin-fringement.

In sum, we conclude that the fact that Galileo is essentially a closed-loop system that only receives a target direction from ground control renders its function fundamentally different from the function of Williams. And further, even if the movement of all functions on board except for the transmission of a target direction could be considered insufficient to avoid equivalence with Williams, the evidence shows that the way in which precession occurs on Galileo is quite different from — and thus not at all equivalent to — the way in which precession occurs on a Williams spacecraft. Lastly, we conclude that even the results achieved by Williams and Galileo are so different that Galileo cannot be considered equivalent to Williams.

In closing, we reject Hughes' general contention that even if Galileo contains improvements that may be patentable in themselves, it still relies on basic teachings of Williams and thereby infringes under the doctrine of equivalents. It is true that Galileo is much more sophisticated than Williams and performs on board, by use of computers, many functions that Williams performed on the ground. Yet, for the reasons summarized above, we conclude that Galileo's attitude control system, rather than being an improvement on Williams, is an entirely different system. At bottom, Galileo's design does not merely replace features described in Williams with improved features; it eliminates them entirely as unnecessary in light of the wholesale differences in the overall design of the attitude control system. Accordingly, Galileo does not infringe the Williams patent.

D

Hughes also contends that the AE spacecraft infringed Williams under the doctrine of equivalents. The government has articulated three reasons for its contention that AE does not infringe. Two of these reasons have been decided in Hughes' favor. First, we rejected the notion that AE's 63 degree canted thruster was not substantially parallel as required by clause (d) for purposes of infringement under the doctrine of equivalents (Tr. 17,-153-57). Second, we rejected the suggestion that angle-based systems could not infringe Williams (Tr. 17,189-91).

The government's remaining argument is that AE was unique among angle-based systems because the pulsing was not keyed to occur at a given angle relative to an external reference. The system was not as autonomous as Galileo, and thus the decision regarding equivalence is a closer one. In particular, the evidence at trial did not establish any of the pronounced differences in the way and result discussed above with respect to Galileo. Nonetheless, in our view the function of the AE spacecraft differs sufficiently from that of the device invented by Williams to avoid infringement.

As stated above, the Federal Circuit in its 1983 decision in Hughes held that the function of the Williams spacecraft was to receive and respond to command signals from an external location to accomplish precession. Significantly, the only information sent from the ground controllers to the spacecraft was the start and stop time for the yaw maneuver. With that information, the spacecraft performed all functions required to precess. Other than the fact that AE used earth sensors as an external reference to accomplish precession, and in that respect is much more similar to Williams than is Galileo, the attitude control system of AE operated quite differently from both Williams and the store and execute spacecraft that have been adjudicated to infringe the patent. In both Williams and the store and execute spacecraft, ground controllers use information obtained from sun sensors to determine the appropriate number of pulses and in what portion of the spacecraft's spin cycle the jets are to be activated in order to achieve a desired new orientation. Those operators then issue commands directing the firing of the spacecraft's precession jets. In Williams, the spacecraft receives real-time firing signals that ground control has coordinated with the spin cycle. In the store and execute system, the spacecraft receives a signal to fire after a given time or angle delay from a sun pulse that is also coordinated with the spin cycle. In contrast, in AE, the only information sent from ground control to the spacecraft — the start and stop time — had nothing to do with the position in the spin cycle in which pulsed firing should occur (Tr. 8393). The synchronization in both the Williams "real time" spacecraft and the "store and execute" spacecraft is accomplished by ground personnel, who calculate when the jet pulses must be fired relative to the sun pulses in order to achieve the desired precession. In AE, that calculation was performed by spacecraft's on-board computer.

Admittedly, the Federal Circuit's decision regarding the store and execute spacecraft held that the advances in computer technology presented there did not allow the spacecraft "to escape the `web of infringement.'" 717 F.2d at 1365 (quoting Bendix Corp. v. United States, 220 Ct.Cl. 507, 537, 600 F.2d 1364, 1382 (1979)). Essentially, the difference between AE and Williams is one of advancements in computer technology. Nonetheless, the Federal Circuit made clear that the range of equivalents to be accorded to the Williams patent was limited. The patent does not cover all ground controllable spacecraft nor all spin-stabilized spacecraft that precess by way of jet pulsing. In particular, the court emphasized that the attitude control system invented by Williams depended on some type of ground control participation. In our view, the computer technology at issue here is different, because it so limits the role of ground control in determining the nature of the attitude thruster firings as to make it implausible to suggest that the spacecraft falls within the limitations of the invention described by Williams.

This is particularly evident when one refers to the precise claim made by Williams. A central focus of clauses (e), (f) and (g) (quoted in Part II) is the delivery of information to the ground, coupled with the receipt from the ground of signals directing firings at a ground-determined point in the spin cycle. The AE spacecraft do not contain anything that fairly can be characterized as the substantial equivalent of those limitations.

As the Federal Circuit has explained, although "a `non-pioneer' invention may be entitled to some range of equivalents, a court may not, under the guise of applying the doctrine of equivalents, erase a plethora of meaningful structural and functional limitations of the claim on which the public is entitled to rely in avoiding infringement." Perkin-Elmer Corp. v. Westinghouse Elec. Corp., 822 F.2d 1528, 1532 (Fed. Cir. 1987). The Federal Circuit emphasized:

Though the doctrine of equivalents is designed to do equity, and to relieve an inventor from a semantic strait jacket when equity requires, it is not designed to permit wholesale redrafting of a claim to cover non-equivalent devices, i.e., to permit a claim expansion that would encompass more than an insubstantial change.

IV

This portion of the opinion addresses the liability issues associated with the alleged unlawful use or manufacture for the government of 14 Global Positioning System (GPS) spacecraft, specifically GPS 22 and 24 through 36.

Defendant acknowledges that the GPS spacecraft are designed to contain an attitude control system that uses the Williams patent. See Appendix to Opinion n. 12. Defendant maintains, however, that certain of the GPS spacecraft were not manufactured or used by its contractor with its authorization or consent within the life of the patent and therefore did not infringe the Williams patent.

A

The GPS ("Global Positioning System") family of spacecraft is part of a navigation system designed to provide users with precise three-dimensional position, velocity and time information. In addition to their use for navigation, the GPS spacecraft carry a payload capable of detecting nuclear detonations.

The GPS system consists of 18 spacecraft orbiting the earth simultaneously in six different orbits, with three spacecraft in each orbit. The GPS spacecraft use both the Williams' spin-stabilized attitude control system and a full-body attitude control system. The spin-stabilized system is used in reaching final orbit. Once the spacecraft has reached final orbit, the spacecraft is despun and the full-body system is employed thereafter to orient the spacecraft. Nonetheless, the spin-stabilized system is always available as a backup attitude control system. Thus, if the full-body attitude control system fails during the life of the spacecraft, the spacecraft can be spun up again and the spin-stabilized system used to reorient the spacecraft.

B

Rockwell International Corporation designed and built the GPS spacecraft under a procurement contract with the government. Rockwell's organizational structure for the GPS program divided responsibility for the project among eight program components. Two of those components were Systems Installation and Testing. Systems Installation was responsible for installation of the electronics and physical assembly of the spacecraft. Once the spacecraft was completed and inspected, the spacecraft was sent to Testing. Testing assured that the spacecraft functioned properly in normal conditions and in a simulated outer-space environment (Tr. 18,988).

The sequence of testing that an assembled GPS spacecraft experienced is as follows. The first test, the Ambient Functional Test, verified the functionality of the individual subsystems and the functionality of the entire spacecraft assembly. The second test, the Thermal Cycle Test, exposed the spacecraft to extreme temperatures to simulate the environment of outer space. The third test, the Acoustics Test, exposed the spacecraft to different levels of vibration. The fourth test, the Thermal Vacuum Test, exposed the spacecraft to extreme temperatures and a vacuum. The fifth test, the Final Factory Processing Test, included a spin test to verify the balance of the spacecraft, a post-thermal vacuum test to verify the spacecraft's oper-ability after the thermal vacuum test and a final factory functional test to ensure that the entire spacecraft operated properly. Tr. 19,005-06. All of these tests were performed in accordance with Military Standard 1540, which specifies that a "full-up" space vehicle must be the object of these tests (Tr. 19,066).

The spacecraft were assembled and tested in incremental stages. The GPS assembly sequence began with assembly of the various structural elements that form the core of the spacecraft. Next, the wire harness and reaction control system were installed. Then the other systems, including electrical power, telemetry tracking and command, attitude control, velocity control and payload, were installed. Finally, the orbit insertion system was installed. Tr. 19,185-87. Before each subsystem was installed, it was tested individually (Tr. 19,-200).

Between Systems Installation and Testing, the Defense Contract Manufacturing Area Office, acting on behalf of the government, conducted an "ITI 3201 Inspection" to ensure that systems installation was complete. Government representatives, using aids such as flashlights, mirrors and magnifiers, inspected the spacecraft's structure, as well as a variety of the individual components such as the electrical wire harness, thermal blanket insulation, optical sensors, coaxial cables, black boxes, RCS propellant tanks and lines, L band antenna mast, RCS thruster assembly and reaction wheel assemblies (Tr. 19,067-74). At that point, the Testing group subjected the assembled spacecraft to the sequence of tests described above. Each of those tests was designed to ensure that the assembled spacecraft could perform its mission.

As of September 11, 1990 (the date the Williams patent expired), each of the 14 GPS spacecraft at issue had completed Systems Installation and passed ITI 3201 Inspection (PX 3018.417); thus, each of those spacecraft had been assembled and inspected and was ready to be extensively tested. Further, as of September 11, 1990, each of the 14 GPS spacecraft at issue had begun the testing sequence described above (PX 3018.417; PX 3018.6). In fact, 13 of those spacecraft had completed the ambient functional test (PX 3018.417). (GPS 36 was undergoing that test when the patent expired (PX 3018.6)). As described above, that test verifies the reliability of both the individual subsystems and the spacecraft assembly. The spacecraft's power is turned on, and the spacecraft then proceeds through a series of possible commands. Each of the subsystems is tested to make sure that it functions individually as well as in connection with the entire system. The attitude control system is thoroughly tested to ensure that it functions properly. Upon successful completion of the ambient functional test, engineers know that the spacecraft has been assembled securely and is functioning correctly. In addition, several of the spacecraft at issue had undergone some of the environmental tests (PX 3018.417).

C

Section 1498(a) of title 28, United States Code, provides in part:

Whenever an invention described in and covered by a patent of the United States is used or manufactured by or for the United States without license of the owner thereof or lawful right to use or manufacture the same, the owner's remedy shall be by action against the United States in the United States Court of Federal Claims for the recovery of his reasonable and entire compensation for such use and manufacture.

For the purposes of this section, the use or manufacture of an invention described in and covered by a patent of the United States by a contractor, a subcontractor, or any person, firm, or corporation for the Government and with the authorization or consent of the Government, shall be construed as use or manufacture for the United States.

(Emphasis added). The words "manufacture" and "use" in § 1498 (as well as the words "make," "use" and "sell" in Title 35) have never been defined by Congress and have become a matter of judicial interpretation.

Although the government acknowledges that each of the 14 GPS spacecraft in issue were designed to incorporate the Williams patent, the government contends that none of these spacecraft were used or manufactured during the patent term within the meaning of § 1498(a). As part of this argument, defendant maintains that any manufacture or use by Rockwell was not with the authorization or consent of the government. We conclude that each of the spacecraft in issue was manufactured by Rockwell for the United States during the patent term with the authorization of the government. Consequently, we find that each of the 14 spacecraft should be included in the compensation base.

1.

First, we address the government's contention that none of these spacecraft were manufactured or used during the term of the patent. The leading Federal Circuit case addressing the meaning of the terms "make" and "use" in this context is Paper Converting Machine Co. v. Magna-Graphics Corp., 745 F.2d 11 (Fed. Cir. 1984). Paper Converting involved a patent for an automatic rewinder used in the manufacture of rolls of densely-wound industrial toilet tissue and paper towels. According to the Federal Circuit, the machine operated as follows:

First, from within a 72-inch long "cutoff" roll, a 72-inch blade ejects to sever the continuous web of paper which is wound around the bedroll. Then, pins attached to the bedroll hold the severed edge of the web while pushers, also attached to the bedroll, transfer the edge of the web towards the mandrel (the roll on which the paperboard core is mounted).

The Federal Circuit held that the testing of significant assemblies to the extent that an infringer would have confidence in the operability of the completed patented device constituted infringement. Id. at 19-20. In effect, the court held that defendant's actions constituted a "use" of the patent. The court framed the issue as "the extent to which a competitor of a patentee can manufacture and test during the life of a patent a machine intended solely for post-patent use." Id. at 16 (emphasis in original). The court recognized the potential unfairness that may arise if liability cannot attach before a device is completely assembled and tested. In concluding that neither complete assembly nor complete testing was necessary to find infringement, the court declared that the fact "[t]hat the machine was not operated in its optimum mode is inconsequential: imperfect practice of an invention does not avoid infringement." Id. at 20. The court reasoned that "a contrary result would emasculate the congressional intent [of preventing] the making of a patented item during the patent's full term of 17 years" and create a climate in which "[n]othing would prohibit the unscrupulous competitor from aggressively marketing its own product and constructing it to all but the final screws and bolts." Id. at 19.

The Federal Circuit distinguished the Supreme Court's decision in Deepsouth Packing Co. v. Laitram Corp., 406 U.S. 518, 92 S.Ct. 1700, 32 L.Ed.2d 273 (1972). Deepsouth involved a patent covering a shrimp deveining machine. The defendant had partially assembled the device in the United States and then shipped the partial assemblies overseas for final assembly. At that time, the relevant patent statute, 35 U.S.C. § 271, provided that "whoever without authority makes, uses or sells any patented invention, within the United States during the term of the patent" infringes the patent (emphasis added). The Supreme Court held that selling subassemblies of a combination patent with the intent that the buyer would combine the parts in a foreign country did not constitute infringement. Id. at 532, 92 S.Ct. at 1708. The Court reasoned:

We cannot endorse the view that the "substantial manufacture of the constituent parts of [a] machine" constitutes direct infringement when we have so often held that a combination patent protects only against the operable assembly of the whole and not the manufacture of its parts.

The Federal Circuit in Paper Converting argued that its decision finding infringement based on incomplete assembly was consistent with Deepsouth for two reasons. First, it argued "that Deepsouth was intended to be narrowly construed as applicable only to the issue of the extraterritorial effect of the American patent law." 745 F.2d at 17. Second, it asserted that the Supreme Court's reference to an operable assembly of the whole "probably [would include] something short of a full and complete assembly." 745 F.2d at 18.

There unquestionably is a significant tension between the analysis of the Federal Circuit in Paper Converting and the analysis of the Supreme Court in Deepsouth. As judges and commentators have noted, it is plausible to understand the Court in Deepsouth as holding that the term "make" was not satisfied by the manufacture of three separate assemblies that would infringe a patented apparatus when they were assembled in a process that would take less than an hour. Although the briefs of the parties debate this question in some detail, we have no occasion to resolve that tension. We conclude that the state of the 14 GPS spacecraft in issue when the Williams patent expired was such that the spacecraft would be found to have been manufactured on that date under either the Paper Converting analysis or the Deepsouth analysis.

A few preliminary observations illustrate the boundaries of the problem. The issue of whether a device has been "manufactured" for purposes of the patent laws is a mixed question of law and fact. A patentee cannot lengthen the term of his patent by receiving compensation for devices that are not in fact "manufactured" or "used" during the patent term. Conduct falling short of manufacture, even if it capitalizes on the patent in some sense, does not constitute infringement. See Tektronix, Inc. v. United States, 216 Ct.Cl. 144, 152, 575 F.2d 832, 837 (holding that the execution of a contract to manufacture infringing devices did not constitute infringement), cert. denied, 439 U.S. 1048, 99 S.Ct. 724, 58 L.Ed.2d 707 (1978). On the other hand, the question of whether a device has been manufactured cannot depend on whether the alleged infringer deems the accused device to be ready for use. See Paper Converting, 745 F.2d at 19-20. To interpret the term "manufacture" that narrowly would subject the patent laws to manipulation and unfairly deprive the patentee of the benefits of the 17-year term of its patent.

The government contends that the spacecraft in issue were not manufactured as of the expiration date of the patent for two separate reasons. First, the government maintains that the spacecraft were not physically assembled by the expiration date because certain flight components had not been integrated into the spacecraft either because the details of the spacecraft's design had not been finalized or because the parts were unavailable. Second, the government maintains that even if the spacecraft were completely assembled as of the expiration date they had not been "manufactured" in the statutory sense because they had not been completely tested by that date.

On the first point, the government does not disagree with Hughes' allegation that each of the spacecraft in issue had completed "systems installation" as of the expiration date. As we noted above, systems installation is the final assembly stage during which all of the components, subsystems and systems ordinarily would be installed. Nonetheless, components of the spacecraft frequently were missing even after the completion of systems installation. One of Hughes' witnesses accurately summarized the true state of affairs when he explained that by the completion of systems installation "the spacecraft has been assembled to the extent of the available components" and is considered to be "in an acceptable configuration to start the testing process" (Tr. 19,005). When a spacecraft that was missing a component proceeded to testing, that component would be added during one of the testing phases. Hence, we accept the government's point that the fact that each of the spacecraft in issue had completed systems installation as of the expiration date does not necessarily mean that the spacecraft were completely assembled.

The government contends that the missing components prevented the spacecraft from being bodies adapted to spin about axes, as required by clause (a) of claim 1 of the Williams patent. The record indicates, however, that as of the expiration date of the patent, the contractor had dealt with each of the parts missing at the time systems installation was completed either by inserting a final part, or by inserting a test part. The test parts would not have been used when the spacecraft was launched and, in many cases, were inoperable "dummy" parts included only to give the spacecraft the appropriate mass and balance during the testing process. Tr. 19,234-42, 19,-113-23; PX 3018.417. Hughes points out that the missing components did not prevent the spacecraft from being tested for operability and contends that the use of test components satisfied the requirements of the patent because they provided the required mass. The government argues that the use of "dummy" parts does not suffice because Deepsouth requires that all elements of the claim be constructed into an "operable assembly." Because dummy parts are not intended to be launched into space, the government reasons, their inclusion on the spacecraft renders the spacecraft inoperable.

We are not persuaded by the government's argument. Even though test parts were not intended to be launched with the spacecraft, the fact remains that the spacecraft, including the attitude control system, had been entirely assembled to the extent feasible at the time. In our view, that complete physical assembly readily satisfies Deepsouth's requirement of an "operable assembly of the whole," 406 U.S. at 528, 92 S.Ct. at 1707. This holding is buttressed by our earlier ruling that the manufacture of a qualification test vehicle can constitute infringement. We stated that

the construction of that qualification test vehicle, whether that's all it was ever intended to be and all it ever became up to the present or whether it was later refurbished and became a flight vehicle, my ruling is that a separate taking can occur under [ 28 U.S.C. §] 1498 when a qualification test vehicle has been manufactured that incorporates the patented device.

Tr. 16,969. In our view, the most important point is that by the time the patent expired the spacecraft with test components had become bodies adapted to spin about axes that satisfied all the claims of the Williams patent. Accordingly, those spacecraft were "manufactured" for purposes of § 1498 by that date.

The government attempts to avoid that conclusion by noting that even if each of the spacecraft had been assembled to the extent to which they were intended to be as of completion of systems installation, there were still some components that had not been integrated. Specifically, the government argues that because the fuel and flight batteries were added at later stages, and because the thrusters were not wired until later stages, the spacecraft were not manufactured as of the expiration date of the patent. We disagree. With respect to fuel, the government's construction reads too much into the word "manufacture" because it almost equates "manufacture" with "use." Even Deepsouth's construction of "make" does not suggest that an item such as fuel would be a part of the making. Under the government's interpretation, a radio would not be made until batteries were added or a power cord was plugged into a socket. Similarly, we do not view the use of a test battery, which is not actually mounted on the spacecraft, as opposed to a flight battery, or the fact that the thrusters were not wired until right before launch, as undermining our conclusion that the 14 spacecraft had been manufactured as of the expiration date of the patent. Even without those final touches, the spacecraft constituted completely assembled objects that incorporated the invention reflected in the Williams patent.

The government's final contention pertaining to its incomplete assembly argument is that some components were found to be defective during the testing process and had not been replaced as of the expiration date of the patent. We reject that contention because it assumes that a device has not been manufactured at any point in time when one of its components is not functioning properly. Under that reasoning, a spacecraft that malfunctioned after launch would at that time become a spacecraft that had not yet been manufactured. Such reasoning neither comports with the ordinary meaning of the word "manufacture" nor finds support in any of the case law. Accordingly, we conclude that the spacecraft were physically assembled by the expiration date of the patent to the point that they constituted operable assemblies as contemplated by Deepsouth.

The government's second basis for asserting that the spacecraft had not been "manufactured" relies on the fact that the full testing process (i.e., the series of five tests to insure readiness for a space environment) was not complete on any of the spacecraft in issue by the expiration date. The importance of testing to these spacecraft cannot be overstated. There were four purposes of the GPS testing program: the minimization of infant mortality, the detection of workmanship and material deficiencies, verification that performance meets specified requirements and demonstration that the vehicle is acceptable for delivery (DX 4607H). Testing represented a substantial portion of the total time expended in performing the contract and the total cost of the contract (Tr. 19,721, 20,-185). Further, it is almost, if not entirely, impossible with current technology to build an apparatus as complex as a GPS spacecraft without some manufacturing defects (Tr. 19,714). The record shows that components failed during each phase of the testing process (Tr. 19,705-07). Moreover, once a spacecraft has been placed in orbit the cost of repairing it at that stage is prohibitive. Accordingly, based on each of those factors, the government argues that a GPS spacecraft is not an operable assembly before the testing is completed because before that time it is highly unlikely that it would operate as intended.

This point has the most force with respect to the spin test; the government argues that until that test has been completed, the spacecraft is not adapted to spin about an axis and thus does not satisfy clause (a) of the Williams claim. The government points out that the spin test does more than simply confirm that the body spins around the appropriate axis. Instead, the test typically results in the addition of balance weights that perfect the balance of the spacecraft. DX 4607H.

Despite the importance of testing, we do not view the function of the tests in issue (all post-assembly and post "ITI 3201" inspection) to be a part of the manufacturing process. The ordinary usage of the term "make" within this context is "to create by putting together component parts." Webster's II New Riverside University Dictionary 718 (1984). (Similarly, "manufacture" is defined as "to make or process (a raw material) into a finished product," id. at 725.) Synonyms for "make" are "assemble," "build," "construct" and "fabricate." Id. at 719. See also Bauer Cie v. O'Donnell, 229 U.S. 1, 10, 33 S.Ct. 616, 617, 57 L.Ed. 1041 (1913) ("The right to make can scarcely be made plainer by definition, and embraces the construction of the thing invented."). To define the term "manufacture" as "make ready to accomplish the intended mission" nearly obliterates the distinction in § 1498 between "use" and "manufacture." These tests — however important — are not part of the process of assembling or fabricating the accused object; that process is completed, at the latest, when the spacecraft complete systems installation. The tests at issue here are part of a later process for perfecting the ability of the object to perform its intended use. Those tests simply do not fall within any fair usage of the term "manufacture." In sum, we conclude that each of the 14 GPS spacecraft in issue was "manufactured" as of the expiration date of the patent.

2.

The government argues that an authorization and consent clause contained in its contract with Rockwell limited the government's authorization or consent until it "accepted" the spacecraft. That clause provides, in part:

The Government hereby gives its authorization and consent (without prejudice to any rights of indemnification) for all use and manufacture, in the performance of this contract . . . of any invention described in and covered by a patent of the United States (i) embodied in the structure or composition of any article the delivery of which is accepted by the Government under this contract, or (ii) utilized in the machinery, tools, or methods the use of which necessarily results from compliance by the Contractor or the using subcontractor with (a) specifications or written provisions now or hereafter forming a part of this contract, or (b) specific written instructions given by the Contracting Officer directing the manner of performance.

(Emphasis added.) As the language suggests, clause (i) addresses apparatus patents, and clause (ii) addresses method patents. Accordingly, clause (i) is relevant here.

The government asserts that this clause means that the government has not authorized use of the patent until it accepts the object, and that acceptance does not occur until the execution of the final DD 250. In response, Hughes contends that the government authorized or consented to the "manufacture" upon executing the contract that required Rockwell to infringe the Williams patent. According to Hughes, the authorization and consent clause in the government's contract with Rockwell can-not be used to limit the government's preexisting authorization. In any event, Hughes contends that the government authorized use of the Williams invention in each GPS spacecraft upon its acceptance and launch of the first spacecraft manufactured under the contract. Finally, Hughes relies on various other provisions in the government's contract with Rockwell to establish that acceptance of each particular patented device occurred before the patent's expiration date.

Authorization or consent by the government can be expressed in many forms, including "by contracting officer instructions, by specifications or drawings which impliedly sanction and necessitate infringement, [or] by post hoc intervention of the Government in pending infringement litigation against individual contractors." Hughes Aircraft Co. v. United States, 209 Ct.Cl. 446, 465, 534 F.2d 889, 901 (1976). In addition, authorization or consent can be implied. TVI Energy Corp. v. Blane, 806 F.2d 1057, 1060 (Fed. Cir. 1986). An express authorization or consent will be found where the government requires the private contractor to use or manufacture the allegedly infringing device, even if the government does not know that the device infringes a patent. TVI, 806 F.2d at 1060.

Accordingly, because the parties agree that the government's contract with Rockwell required Rockwell to manufacture spacecraft that embodied the Williams patent, the issues are whether an "acceptance" is required to establish "authorization or consent" and whether a final DD 250 is required to establish such an "acceptance."

Because Congress has not defined the terms "authorization" and "consent," we start with the premise that they must be given their ordinary meanings. Putting aside for the moment the authorization and consent clause, it is clear that the government "authorized" the manufacture of the accused spacecraft when Rockwell became contractually bound to build spacecraft embodying the Williams patent. See TVI, 806 F.2d at 1060; Hughes, 209 Ct.Cl. at 465, 534 F.2d at 901. As mentioned earlier, Rockwell was obligated by its contracts with the government to construct spacecraft using the Williams design; the relevant obligations arose before Rockwell constructed the spacecraft, and thus long before the patent expired. The authorization and consent clause cannot change that result. Even if the terms "authorization" and "consent" had a particular contractual meaning between Rockwell and the government, it is the statutory meaning of those terms that controls this case under § 1498, between a patentee and the government. Accordingly, we conclude that the requisite authorization or consent was given no later than the time at which Rockwell became contractually bound to manufacture a spacecraft containing an embodiment of the Williams patent.

In any event, even if the authorization and consent clause was binding on Hughes and precluded liability until the government accepted the spacecraft, we would find the government liable because we conclude that the government had accepted the accused devices before the patent expired. First, we note that a number of other GPS spacecraft (including GPS 15, 18, 20, 21 and 23) that were manufactured pursuant to the same contract had in fact been the subject of a final DD 250 and had been launched. Surely whatever justification there is for requiring government acceptance before imposing liability, it is satisfied once the government accepts the first of a series of the accused spacecraft, which is identical in all relevant aspects to the remaining accused spacecraft.

Second, the fact that each of the accused spacecraft had been completely assembled, had been subjected to an inspection by the government, and had become government-furnished property by the time the patent expired suggests that the government had accepted the patented devices. The significance of the final DD 250 was that it indicated that Rockwell had satisfactorily performed the testing component of its contract with the government for that particular spacecraft. We believe that the acceptance of the spacecraft as government-furnished property is sufficient to constitute acceptance of the design of the spacecraft for purposes of § 1498.

Accordingly, we conclude that Rockwell manufactured each of the 14 accused GPS spacecraft for the government pursuant to the government's authorization or consent. See Appendix to Opinion for a list of the precise taking dates for the GPS spacecraft in issue.

V

In accordance with one of the stated objectives of the United States space program — to cooperate "with other nations and groups of nations," 42 U.S.C. § 2451(d)(7) — the United States has undertaken joint projects with other nations that have enabled the sharing of the costs and benefits. This portion of the opinion addresses defendant's liability for five spacecraft associated with four international programs. Those spacecraft are: Helios A and B, ISEE B, AMPTE UKS and Ariel 5. We have determined elsewhere or the parties have stipulated that each of the spacecraft in issue in this part contain an embodiment of the Williams patent. See Appendix to Opinion (listing citations to the bench rulings that pertain to these spacecraft). Although it is undisputed that the government had some association with each of these spacecraft, defendant contends that its participation did not constitute use or manufacture by or for the government within the meaning of 28 U.S.C. § 1498(a). Each of the spacecraft in issue was manufactured in another country and funded by another nation. Accordingly, we conclude that none of the spacecraft was manufactured "by or for the United States" within the meaning of § 1498(a). See Part V.A.3. for a discussion of the extraterritorial application of § 1498. Based on a consideration of the government's participation in the various spacecraft programs, we conclude that Helios A and B and ISEE B were used by the government within the meaning of § 1498(a) and that ARIEL 5 was not used by the government within the meaning of § 1498(a). With respect to AMPTE UKS, we find use by the government within the meaning of § 1498(a) but conclude that the temporary presence doctrine, as contained in 35 U.S.C. § 272, proscribes imposition of liability.

A

Section 1498(a) of title 28, United States Code, imposes liability on the government "[w]henever an invention described in and covered by a patent of the United States is used . . . by or for the United States without license. . . ." The language of the statute itself imposes two requirements — there must be a use and the use must be by or for the United States — and for reasons discussed below we infer a third requirement — that the use be within the United States. The first three sections of this subpart discuss these requirements in turn. The final section of this subpart discusses the temporary presence doctrine, as codified in 35 U.S.C. § 272, which precludes liability for patent infringement if a vessel, aircraft or vehicle of another country carrying an infringing device enters the United States "temporarily" or "accidentally."

1. Use

As stated above in Part IV, the words "manufacture" and "use" (as well as the words "make," "use" and "sell" in title 35) were not defined by Congress but do have an established judicial interpretation; that interpretation is a broad one.

In the first major case to interpret this language, Bauer Cie v. O'Donnell, 229 U.S. 1, 10-11, 33 S.Ct. 616, 617, 57 L.Ed. 1041 (1913), the Supreme Court declared that the term "use," as used in a predecessor to title 35, was "comprehensive" and included "put[ting] into service any given invention." See Webster's II New Riverside University Dictionary 1271 (1984) (defining "use" as "[t]o bring or put into service or action"). Although few subsequent cases have addressed the meaning of the term "use," those that have consistently followed the Supreme Court's lead in giving the term a broad interpretation. In Olsson v. United States, 25 F.Supp. 495 (Ct.Cl. 1938), cert. denied, 307 U.S. 621, 59 S.Ct. 792, 83 L.Ed. 1500 (1939), the Court of Claims concluded that the United States "used" howitzers (a type of cannon) that contained an embodiment of a patented invention even though the howitzers were never actually fired and had been disassembled for storage:

Although held in storage in a convenient form for proper preservation the guns were, after being brought to this country, employed for the accomplishment of a purpose, i.e., the purpose of national defense. In these circumstances we think it is clear that there has been a "use" in the United States of plaintiff's invention for which he is entitled to recover compensation.

We glean from those precedents that the question of what constitutes "use" is a mixed question of fact and law to be determined on a case-by-case basis. A device may be "used" in many different ways, and all uses that rely on the teachings of a patent constitute infringement.

The broad interpretation reflected in those general principles already has been applied in this case. In Hughes Aircraft Co. v. United States, 215 U.S.P.Q. 787 (Ct. Cl. Trial Div. 1982), the court addressed the meaning of the term "use" with respect to spacecraft that were launched before the patent was issued, and thus were already in space when the patent term began. The government argued that it did not "use" spacecraft that had not actually activated the Williams attitude control system during the patent term; in those spacecraft, the Williams attitude control system was intended to be merely a back-up system. In rejecting that non-use argument, the court determined that actual firing of the precession thrusters was not necessary to constitute "use" of the invention because "there is no requirement in any of the claims that a control signal synchronized with the ISA be provided." Hughes, 215 U.S.P.Q. at 815. Instead, the court found that "[o]nly the capability to receive such a signal is claimed." Id. Hence, the court's holding regarding "use" of the Williams patent did not require an activation of the attitude control system.

Another decision in this case that has some bearing on the "use" issue is Hughes Aircraft Co. v. United States, 226 Ct.Cl. 1, 8-9, 640 F.2d 1193, 1194-97 (1980). In that case, the Court of Claims held that the Williams patent did not cover the ground control device. In doing so, the court, relying on clause (a) (claiming a "body adapted to spin about an axis"), held that the patent covered a spacecraft with certain specified features. Id. at 8-9, 640 F.2d at 1197 ("The invention was the satellite."). In our view, that case establishes the rule that because the patent covers the spacecraft as a whole (i.e., a body adapted to spin about an axis) rather than simply the features of the attitude control system, the "use" question must focus on whether the spacecraft as a whole has been used.

For purposes of this case, it is important to consider whether launching a spacecraft constitutes a use of the patent. As the foregoing discussion suggests, we think it clearly does. Hughes Aircraft, 215 U.S.P.Q. 787, makes clear that the availability of the attitude control system on the spacecraft at a time when the spacecraft is being operated constitutes a use of the patent.

The government argues that because the attitude control system cannot be activated until the spacecraft separates from the launch vehicle in space, the launch cannot constitute a use of the patented invention. We think that argument is foreclosed by the decision of the Court of Claims in Olsson, which held that howitzers, in a disassembled state for purposes of storage, were used for purposes of § 1498(a) and by the fact that it is the spacecraft as a whole whose use constitutes a use of the patent.

2. "By or for the United States"

Next, we address the meaning of the phrase "by or for the United States." Although in most contexts the meaning of that phrase is relatively clear, the issue is complicated in this case by the association of the accused devices with cooperative arrangements between the United States and other governments. Again, as with the use question, much of the law on this point is found in earlier cases involving the Williams patent, in particular Hughes Aircraft Co. v. United States, 209 Ct.Cl. 446, 534 F.2d 889 (1976) (a decision issued in the case at bar, No. 426-73), and Hughes Aircraft Co. v. Messerschmitt-Boelkow-Blohm, GmbH, 625 F.2d 580 (5th Cir. 1980), cert. denied, 449 U.S. 1082, 101 S.Ct. 868, 66 L.Ed.2d 807 (1981). Both of those cases addressed the issue of whether United States involvement in international spacecraft programs constituted use by the government. Although both decisions purported to turn on the jurisdictions of the respective courts to entertain the action, both decisions directly addressed the meaning of the phrase "by or for the United States" in the context of specific international spacecraft programs. Accordingly, we consider those cases in some detail. Hughes Aircraft Co. v. United States, 209 Ct.Cl. 446, 534 F.2d 889 (1976), involved Skynet II, the product of a joint program between the United States and the United Kingdom. A long-term cooperative aim of the two governments was to introduce spacecraft with interoperable characteristics so that each country "could effectively augment its own defense communications capacity by direct and reciprocal utilization of the other country's capacity." Id. at 460, 534 F.2d at 898. Skynet II was intended to be the United Kingdom segment of a joint defense satellite communications system.

In 1970, the United States and the United Kingdom outlined the respective responsibilities of each country in a "Memorandum of Understanding" (MOU). The MOU provided that the United Kingdom was responsible for the design, development and procurement of the spacecraft. In addition, following the spacecraft's positioning in final orbit, the United Kingdom was to have exclusive communications and operational control over Skynet II. The United States was responsible for providing technical assistance in the design of the spacecraft, procuring the launch vehicles and launching the spacecraft. In addition, in the event of an emergency, the United States could assume operational control of the spacecraft. Each government was to provide the communications and earth control stations required for its own use. Costs incurred by the United States on behalf of the United Kingdom were to be reimbursed by the United Kingdom.

The Court of Claims determined that Skynet II had been manufactured for the United States and that it had been used by the United States. In addressing whether Skynet II was manufactured for the United States, the court focused on the cooperative nature of the program. Quoting a statement by the Department of the Air Force, the court noted that

US involvement in the Skynet II program cannot be viewed as a mere sale of satellite launch services by an uninterested party. Rather, the furnishing of the launch equipment and services was directly related to US National Security as an integral step toward expanding the US defense satellite communications system via the shared use cooperative program set forth in the aforementioned MOU's.

Unlike the Skynet project, NASA's only role following launch of the Helios spacecraft was to provide tracking and data acquisition services. German stations had exclusive control of the spacecraft after the launch. Hence, although the court did not identify a particular activity that constituted a "use," and instead took a totality of the factors approach, the decision itself is consistent with our view that launching a spacecraft constitutes a "use" of that spacecraft.

Those cases stand for the principle that United States involvement in a joint international space program will be sufficient to make any use of the spacecraft a use "by" or "for" the government within the meaning of § 1498 if the project is a cooperative one with the potential of substantial benefits to the United States. Our view of those decisions is buttressed considerably by Congress' enactment in 1981 of 42 U.S.C. § 2457(1). That provision states in pertinent part:

The use or manufacture of any patented invention incorporated in a space vehicle launched by the United States Government for a person other than the United States shall not be considered to be a use or manufacture by or for the United States within the meaning of section 1498(a) of title 28, unless [NASA] gives an express authorization or consent for such use or manufacture.

The Senate Report associated with this provision explains its purpose as follows:

The launch capability of NASA is available to persons other than agencies of the United States on either a reimbursable or cooperative basis. Generally, a launch performed under a cooperative arrangement entails mutual benefits to the United States and the other person, such as the sharing of scientific data from the space vehicle. The courts have held that where a launch is part of such a cooperative arrangement that the use of any patented inventions incorporated in the space vehicle furnished by such other person is a use by or for the Government within the meaning of 28 U.S.C. 1498(a). On the other hand, there are many instances where NASA provides launch services to another person on a reimbursable basis without any tangible benefits flowing to the Government other than monetary reimbursement for such launch services. In such cases, the Government should not be held liable under 28 U.S.C. 1498(a) since any use of a patented invention contained in the space vehicle for which launch services are provided is generally for the sole benefit of the person procuring the launch services.

S.Rep. No. 97-100, 97th Cong., 1st Sess. 52 (1981); see H.R. Rep. No. 32, 97th Cong., 1st Sess. 222 (1981). The legislative history strongly suggests that Congress understood the preexisting case law, just as we have explained it above, to impose liability for cooperative ventures, and enacted the statute to ensure that the government would not be liable when it acted merely to accommodate other nations. Hence, although the statute does not apply to all the "foreign" spacecraft at issue in this case, we view it as supporting our interpretation of the "by or for" requirement in § 1498.

3. Application of § 1498 to Activities in Outer Space

It is generally recognized that the activities of nations in outer space are governed by public international law. Glenn H. Reynolds and Robert P. Merges, Outer Space: Problems of Law and Policy 25-47 (1989) (discussing international space law). The United States is a signatory to several treaties governing the exploration of outer space. The Treaty on Principles Governing the Activities of States in the Exploration and Use of Space, Including the Moon and Other Celestial Bodies (Outer Space Treaty), 18 U.S.T. 2410, T.I.A.S. No. 6347, 610 U.N.T.S. 205, was the first treaty to be adopted, and provides the basic framework for international space law. Article III of that treaty establishes that all activities in outer space must be conducted "in accordance with international law, including the Charter of the United Nations, in the interest of maintaining international peace and security and promoting international co-operation and understanding." Article II generally prohibits any nation from claiming sovereignty over areas of outer space, but Article VIII provides an exception, under which a State "on whose registry an object launched into outer space is carried shall retain jurisdiction and control over such object, and over any personnel thereof, while in outer space or on a celestial body." The record indicates that ARIEL 5 and AMPTE UKS were registered as United Kingdom spacecraft, but the record does not indicate the state of registry of the other spacecraft in issue. The government argues that under the international law governing outer space, as established by the Outer Space Treaty, we may not apply § 1498 to claims arising in outer space relating to non-United States registered spacecraft. We need not decide whether international law prohibits the extension of our patent laws to activities in outer space on foreign spacecraft because we conclude that Congress has not extended § 1498 to cover those activities.

The Supreme Court has set forth a "presumption that Acts of Congress do not ordinarily apply outside our borders." Sale v. Haitian Centers Council, Inc., ___ U.S. ___, ___, 113 S.Ct. 2549, 2560, 125 L.Ed.2d 128 (1993) (applying presumption to prevent application of § 243(h)(1) of the Immigration and Nationality Act to actions in the high seas); see Smith v. United States, ___ U.S. ___, ___, 113 S.Ct. 1178, 1183, 122 L.Ed.2d 548 (1993) (holding that the presumption prevented application of the Federal Tort Claims Act to a claim arising in Antarctica). Hence, those cases suggest that, absent some textual indication to the contrary, we should construe § 1498 not to apply to claims arising beyond United States territorial limits.

The only portion of § 1498 that directly addresses extraterritorial application states that the provision "shall not apply to any claim arising in a foreign country." 28 U.S.C. § 1498(c); see Decca Limited v. United States, 210 Ct.Cl. 546, 550, 544 F.2d 1070, 1072 (1976) (stating that § 1498 does not apply to "acts in foreign countries that would be infringements at home"), cert. denied 454 U.S. 819, 102 S.Ct. 99, 70 L.Ed.2d 89 (1981). That provision, by itself, does not prevent the application of § 1498 to activities in outer space, because outer space is not a "foreign country" in the ordinary meaning of that phrase.

The analogous provisions of title 35, however, strongly support a reading of § 1498 limited to activities in this country. First, section 154 of title 35, United States Code, provides that a patent grants "the right to exclude others from making, using, or selling the invention throughout the United States." Section 100(c) of that title provides that "[t]he terms `United States' and `this country' mean the United States of America, its territories and possessions." Hence, because the patent itself grants rights only against use in the United States, it would be quite odd to interpret § 1498 as imposing liability on the government for use outside this country. See also 35 U.S.C. § 271(a) (defining an infringer as "whoever without authority makes, uses or sells any patented invention, within the United States during the term of the patent" (emphasis added)). The Supreme Court's holding in Deepsouth that our patent laws have no extraterritorial effect, 406 U.S. at 531, 92 S.Ct. at 1708, bolsters our view that § 1498 in its entirety should be construed consistently with title 35 as limited in application to United States territory and thus as not applying in outer space (absent a specific enactment extending the reach of patent laws to uses in space).

Finally, and perhaps most dispositively, the Supreme Court's recent interpretation of a similar phrase constrains us to limit § 1498 to activities in this country. Smith v. United States involved the Federal Tort Claims Act (FTCA), which contains language similar to that in § 1498(c) providing that the statute did not apply in a foreign country. The Court held that the FTCA did not apply to claims arising in Antarctica. Although the decision rested in part on the Court's view that the phrase "foreign country" meant a region of land outside the borders of the United States, without regard to whether it is a sovereignless region or a sovereign state, id., ___ U.S. at ___, 113 S.Ct. at 1181, the Court went further, holding that "[t]he applicability of the presumption is not defeated here just because the FTCA specifically addresses the issue of extraterritorial application in the foreign-country exception," id. ___ U.S. at ___, 113 S.Ct. at 1183. Instead, the Court held that the presumption should apply unless there was affirmative evidence of congressional intent to apply the FTCA to claims arising in Antarctica.

In our view, Smith also prevents the assertion of § 1498(a) over claims arising in outer space. See Smith v. United States, ___ U.S. at ___, 113 S.Ct. at 1184, 122 L.Ed.2d 548 (1993) (Stevens, J., dissenting) (noting that the case "will surely have its parallels in outer space"). Accordingly, we conclude that § 1498(a) does not apply to activities in outer space and thus we confine our analysis to activities occurring within the territorial limits of the United States.

4. Temporary Presence Doctrine

The temporary presence doctrine presupposes that there has been some use of an invention within the United States that would otherwise constitute infringement of a United States patent. The government contends that there has been no such use within the United States with respect to any of the spacecraft in issue. Nonetheless, as an alternative defense, the government argues that liability for any such use would be excluded by the temporary presence doctrine. Because we view the launch of a spacecraft containing an embodiment of the patented invention from the United States as a "use" of that invention within the United States, we briefly address the effect that this doctrine has on our analysis.

The temporary presence doctrine was enunciated by the Supreme Court in Brown v. Duchesne, 60 U.S. (19 How.) 183, 15 L.Ed. 595 (1856), and, on an international scale, was the subject of a convention to which the United States is a party. In 1952, Congress codified the doctrine at 35 U.S.C. § 272.

Brown v. Duchesne involved an action brought by the owner of a United States patent on a gaff for a sailing vessel, against a French national who sailed a French schooner that allegedly contained the infringing gaff from St. Peters, to Boston, and back to St. Peters. The Court phrased the issue as

[W]hether any improvement in the construction or equipment of a foreign vessel, for which a patent has been obtained in the United States, can be used by such vessel within the jurisdiction of the United States, while she is temporarily there for the purposes of commerce, without the consent of the patentee?

Section 272 of title 35, United States Code, provides:

The use of any invention in any vessel, aircraft or vehicle of any country which affords similar privileges to vessels, aircraft or vehicles of the United States, entering the United States temporarily or accidentally, shall not constitute infringement of any patent, if the invention is used exclusively for the needs of the vessel, aircraft or vehicle and is not sold in or used for the manufacture of anything to be sold in or exported from the United States.

Hughes contends that launches occurring before 1981, when Congress added 42 U.S.C. § 2457(k) (providing that "[a]ny object intended for launch, launched, or assembled in outer space shall be considered a vehicle for the purpose of section 272"), are not subject to the doctrine because spacecraft, at that time, were not vessels, aircraft or vehicles. We agree with plaintiff's position. We emphasize that in the absence of said 1981 enactment (in effect declaring a spacecraft brought here for launch into outer space to be a "vehicle" within the meaning of the temporary presence doctrine) such spacecraft would properly be deemed cargo brought to this country for use (i.e., launch), not a vehicle or vessel within the contemplation of the temporary presence doctrine. Among the five foreign spacecraft addressed in this opinion, only AMPTE UKS could be affected by the 1981 amendment; the other four were launched prior to 1981.

The only published decision discussing 35 U.S.C. § 272 of which we are aware, Cali v. Japan Airlines, Inc., 380 F.Supp. 1120 (E.D.N.Y. 1974), aff'd without op., 535 F.2d 1240 (2nd Cir. 1975), involved foreign transoceanic passenger and freight air carriers that contained a modification to a jet engine that was covered by a United States patent. The plaintiff contended that the regular and extensive presence of the carriers in the United States could not be considered "temporary" within the meaning of the temporary presence doctrine. In concluding that the word "temporary," as used in § 272, could not be construed so narrowly, the court stated that the word "could not sensibly mean any less than entering for the purpose of completing a voyage, turning about, and continuing or commencing a new voyage." Cali, 380 F.Supp. at 1126. The court compared

a[n] [aircraft], manufactured in France and powered with such an engine, delivered here for use by an airline in this country for domestic traffic, even though manufactured and sold in France, and a foreign aircraft arriving here on an international flight only to unload, turn about, reload and depart.

Our case is substantially different from Cali and Brown v. Duchesne. In those cases, the entries of the vessels and aircraft (operating as means of conveyance) were for purposes of depositing cargo and passengers and initiating new trips elsewhere. When a spacecraft is delivered to the United States for the purpose of allowing the United States to launch it, the spacecraft is the cargo brought here for an essential use, not a "vessel" or "vehicle" which enters the United States as a means of conveyance. Thus, we conclude that § 272 has no application to spacecraft launched before 1981.

B

1. Helios

The Helios program (named for the Greek god of the sun) was a joint project between the Federal Republic of Germany and the United States. Hughes alleges that the government is liable for use of two Helios spacecraft, Helios A and B. In April 1969, following an extensive cooperative effort, a joint United States/German Mission Definition Group issued a report discussing the feasibility of a joint program. The report defined three objectives: (1) to provide German and United States experimenters an opportunity to conduct integrated experiments aimed at investigating properties and processes in interplanetary space near the sun, (2) to "advance the technical and technological expertise of German industry" and (3) to allow Germany to develop "more effective and more sophisticated management experience" (DX 2514B at 1-2). Further, the report contained design recommendations for numerous aspects of the spacecraft, including a detailed discussion of attitude control.

On June 10, 1969, shortly after the mission definition report declared the project feasible, NASA and the German Ministry for Scientific Research executed an MOU outlining the details of the Helios program. According to the MOU, the "primary objective of Project HELIOS [was] to investigate the properties of and processes in interplanetary space in the direction of and close to the sun by developing, launching, and operating automated spacecraft" (PX 1033C at 1). The MOU required NASA to provide its experiments and the portions of the spacecraft and instruments peculiar to those experiments, to provide a launch vehicle and launch the spacecraft, to provide tracking and data acquisition services during certain phases of the mission, to reduce, analyze and distribute certain data and to provide training and consultation to the Germans as needed. On the other hand, the MOU required the German agency to provide its experiments and related instrumentation, to provide both of the spacecraft, to integrate all of the experiments with the spacecraft, to deliver the spacecraft to the launch site, to supply tracking and data acquisition services during certain phases, to control the spacecraft during all phases, to reduce, analyze and distribute certain data and to cooperate with the training of German personnel. Each party to the MOU was responsible for paying the costs incurred in discharging its responsibilities; neither was to be reimbursed by the other for any of its costs.

Messerschmitt-Boelkow-Blohm (MBB), a contractor hired by Germany, designed, constructed and tested Helios A and B in facilities near Munich (Tr. 11876-78). The Helios spacecraft contained ten experiments, three of which were contributed by the United States (Ousley deposition at 9-10). NASA launched Helios A and B from Cape Canaveral, Florida on December 10, 1974 and January 15, 1976, respectively (Tr. 11888, 11891); the attitude control systems were first activated in space shortly after separation from the launch vehicles (Tr. 11950). Thereafter, NASA and the German agency cooperated in performing the mission substantially as planned, with German personnel assuming complete responsibility for controlling the spacecraft (Tr. 11952-55). NASA's main role following launch was to provide tracking and data acquisition services during some phases of the mission.

Tracking and data acquisition is accomplished by the use of a worldwide network of NASA ground stations. These facilities are interconnected by ground communications lines, undersea cables and communications circuits. The interconnections provide the communications capability needed among the various facilities, spacecraft and the control centers in the United States from which the flights are directed. During the years in question, NASA used two types of facilities — the Spaceflight Tracking and Data Network (STDN), supporting orbital missions, and the Deep Space Network (DSN), supporting planetary and interplanetary missions. S.Rep. No. 100, 97th Cong., 1st Sess. 41-42 (1981). The types of support provided by those networks are as follows:

(1) tracking to determine the position and trajectory of vehicles in space; (2) acquisition of scientific and Earth applications data from onboard experiments and sensors; (3) acquisition of engineering data on the performance of the spacecraft and launch vehicle systems; (4) transmission of commands from ground to spacecraft; (5) communication with astronauts and acquisition of biomedical data on their condition; (6) communication of information between the various ground facilities and central control centers; and (7) processing of data acquired from the launch vehicles and spacecraft.

In 1974, Hughes filed an action against MBB in the United States District Court for the Middle District of Florida, alleging that MBB infringed the Williams patent. As a result of that litigation, the Administrator of NASA issued an authorization and consent letter dated March 18, 1977, which provided:

The United States Government hereby gives its authorization and consent, without prejudice to any rights of indemnification, for all use and manufacture of any invention described in and covered by U.S. Patent Nos. 3,758,051 [the Williams patent] and Reissue 26,887 in the performance of the obligations of the Government of the Federal Republic of Germany under the Memorandum of Understanding (MOU) of June 10, 1969 and in the performance of the obligations of MBB under its contract with the Government of the Federal Republic of Germany for services to be performed in the United States relative to the Helios satellites. This authorization and consent shall apply retroactively to all Helios-connected activities that have taken place.

PX 1033C-1. As we discussed in Part V.A.2., the district court ultimately concluded that the allegedly infringing devices were used by the United States and that the infringement action therefore properly belonged in the Court of Claims; the Fifth Circuit affirmed. Hughes Aircraft Co. v. Messerschmitt-Boelkow-Blohm, GmbH, 437 F.Supp. 75 (M.D.Fla. 1977), aff'd, 625 F.2d 580 (5th Cir. 1980).

Based on these facts and the three requirements for liability under § 1498(a) discussed in Part V.A., we conclude that Helios A and B should be included in the compensation base. First, NASA's launch of both of the spacecraft constituted a use of the patented invention. Second, that use was "by or for" the government because the project was a cooperative one resulting in a considerable benefit to the United States government. A review of the MOU signed by the Federal Republic of Germany and the United States demonstrates that both parties considered their project a cooperative arrangement for mutual benefit. According to the MOU, the primary purpose of the mission was to study the properties of interplanetary space near the sun by sharing the costs and responsibilities associated with building, testing, launching and controlling the spacecraft. Nothing about the execution of that plan suggests that the parties considered the project anything other than a cooperative one. Further, although we base our decision on use by the government rather than use for the government, we view the government's issuance of an authorization and consent letter as persuasive evidence of the joint nature of the Helios project. Third, the use was within the territorial borders of the United States because it was launched from Cape Canaveral, Florida. Hence, we conclude that the government used Helios A and B for its benefit within the United States and therefore is liable under § 1498(a).

2. ISEE

The International Sun-Earth Explorer Program (ISEE) was a joint project between the United States and the European Space Research Organization, now referred to as the European Space Agency. (The current and predecessor organizations are referred to as "ESA.") This portion of the opinion addresses Hughes' entitlement to recovery for ISEE B. On March 17, 1975, NASA and ESA executed an MOU outlining their respective roles in the project.

According to the MOU, the mission of the ISEE spacecraft program was to make a joint contribution to the International Magnetospheric Study (IMS) by focusing on solar-terrestrial relationships in the earth's magnetosphere (PX 1102 at 1). The IMS was a "coordinated international cooperative program to study key problems of our nearby space environment" (PX 1037JJ-1). The ISEE spacecraft were "core" or "reference" spacecraft for the IMS. The ISEE program involved three spacecraft, ISEE A, B and C (also known as "Mother," "Daughter" and "Heliocentric," respectively). Two of those spacecraft, ISEE A and B, equipped with similar experiments, were to travel at a controlled distance from each other making similar simultaneous measurements. The third spacecraft, ISEE C, was to act as a monitor for the solar wind.

The MOU required NASA to contribute ISEE A, to integrate ISEE A and B and to test the dual spacecraft, to provide a launch vehicle and deliver ISEE A to the launch site, to provide tracking and data acquisition services as agreed and to provide technical advice as needed. On the other hand, the MOU required ESA to contribute ISEE B, to deliver it for integration with ISEE A and then to the launch site, to provide tracking and data acquisition as agreed and to provide consultation as needed. Each party was to pay the costs associated with fulfilling its responsibilities under the MOU.

ISEE A and C were designed and constructed by NASA at Goddard Space Center (PX 1029K; PX 1029N). ESA funded the construction of ISEE B (PX 1029K; Tr. 13793-96). ISEE A and B were launched on October 22, 1977 from Cape Canaveral, Florida (PX 1029K), and ISEE C was launched on August 12, 1978 from the same site (PX 1029N). Thereafter, the mission was executed in accordance with the division of responsibility outlined in the MOU, with NASA providing tracking services for all three spacecraft and controlling ISEE A and C and ESA controlling ISEE B (Tr. 13791). The MOU between the United States and ESA was modified by letter dated December 17, 1976, to include an authorization and consent clause because of the ESA's concern about the Williams patent litigation (PX 1103).

Based on these facts, we conclude that ISEE B should be included in the compensation base; the three requirements for finding liability under § 1498(a) are present. First, NASA's launch of ISEE B constituted a use of the patented invention. Second, that use was "by or for" the government because the ISEE program was a cooperative one. The applicable MOU demonstrates that NASA and ESA considered the mission of the ISEE program to be making a joint contribution to the International Magnetospheric Study. That mission was executed as planned. Put simply, all of the evidence suggests that NASA launched ISEE B as its part of a joint program and not as a disinterested party. In fact, this program, even more so than Helios, involved a great deal of cooperation, with the United States being required to contribute a spacecraft that would travel at a controlled distance from ISEE B. In addition, the authorization and consent letter issued by NASA demonstrates that the government considered the ISEE program a joint project. Third, because the ISEE B was launched in Florida, the use occurred within the territorial borders of the United States. Thus, we conclude that Hughes is entitled to recovery with respect to ISEE B.

3. AMPTE

The Active Magnetospheric Particle Tracer Explorer (AMPTE) program involved two international agreements — one between the Federal Republic of Germany and the United States and one between the Federal Republic of Germany and the United Kingdom. The AMPTE program included three spacecraft: the Ion Release Module (IRM) contributed by Germany, the United Kingdom Subsatellite (UKS) contributed by the United Kingdom and the Charge Composition Explorer (CCE) contributed by the United States. This portion of the opinion addresses Hughes' accusation of AMPTE UKS.

On October 15, 1981, the United States and the Federal Republic of Germany executed an MOU that detailed their respective duties. That MOU described the mission as follows:

The primary scientific objective of the mission is to study the entry of solar wind ions into the magnetosphere and the processes by which particles are energized in the magnetospheric tail. Tracer ions (lithium and barium) will be released and measured in the solar wind and within the distant magnetosphere. The AMPTE mission will also seek to obtain comprehensive measurements of the composition and dynamics of the natural charged particle populations within the Earth's magnetosphere. Furthermore, it is intended to study, by in-situ measurements and by remote optical observations, the interactions of released and ambient plasmas.

In addition, the mission will serve to establish values of physical parameters necessary for planing and execution of active experiments on Shuttle/Spacelab missions.

PX 1027E at 2-3; see Tr. 2579-80, 9002-03. The MOU between the United States and Germany contemplated the use of two spacecraft — the CCE and IRM. That agreement required NASA to contribute the CCE spacecraft, to launch both spacecraft, to provide tracking for IRM as needed and for CCE and to process AMPTE CCE data. Likewise, the MOU required Germany to contribute the AMPTE IRM spacecraft, to deliver it to the launch site, to conduct tracking operations for AMPTE IRM as agreed with NASA and to process AMPTE IRM data. The MOU also provided that NASA and the German space agency would each bear the costs of discharging their respective responsibilities.

The AMPTE CCE and IRM were originally planned to be launched in a stacked form, with the CCE on top of the IRM and with an interface adapter inserted between the two spacecraft. Then, Germany and the United Kingdom developed a novel design for a spacecraft which could be built on and around the interface adapter. Tr. 9003-04, 15079-80.

Subsequently, by separate agreement, the United Kingdom and the Federal Republic of Germany agreed to launch an additional spacecraft, AMPTE UKS, to be provided by the United Kingdom. The MOU stated that the spacecraft would be a subsystem of the AMPTE IRM and was intended to "improve the in-situ study of the interaction between the released [barium] and [lithium] plasmas with the natural environment" (PX 1027II). In that agreement, the United Kingdom agreed to provide the AMPTE UKS spacecraft, to control and provide tracking services for the spacecraft and to process data from the spacecraft; in return, Germany agreed to provide various services related to integrating UKS with the IRM.

AMPTE UKS was built in the United Kingdom and funded by the Science Research Council (Tr. 15079). The spacecraft was registered with the United Nations as a United Kingdom spacecraft (Tr. 15081). The spacecraft contained five experiments, all contributed by the United Kingdom (Tr. 15077).

NASA launched the three AMPTE spacecraft on a Delta/Thor vehicle on August 16, 1984 (Tr. 9004). The three spacecraft were stacked with the IRM on the bottom, the UKS in the middle and the CCE on top. The spacecraft remained attached in the launch phase and separated in space, with the CCE in one orbit (within the earth's magnetosphere) and the UKS and IRM in a different orbit (outside of the earth's magnetosphere) (Tr. 2582-85, 9007-08). The UKS kept a position relative to the IRM so that it provided a second reference for studying the release of chemicals by the IRM (Tr. 9080). Although NASA STDN stations were not used to send commands to the UKS (Tr. 15100-01), for two days NASA's Deep Space Network was used to receive data from UKS (Tr. 15104).

Based on these facts, we conclude that the three requirements imposed by § 1498(a) are present. First, NASA's launch of AMPTE UKS constituted a use of the patented invention. Second, that use was "by or for" the government. AMPTE CCE and IRM were part of a cooperative program between the United States and the Federal Republic of Germany similar to the Helios and ISEE programs discussed above. Even though AMPTE UKS was contributed by the United Kingdom pursuant to a separate agreement between the Federal Republic of Germany and the United Kingdom, we consider it to be part of the cooperative program between the United States and Germany because it was considered a subsystem of the AMPTE IRM and because the United States consented to the addition of AMPTE UKS and benefitted, from its inclusion. Under these circumstances, we conclude that the government's actions were not that of a disinterested party providing launch services, but rather those of a member of a joint venture launching a spacecraft in connection with a joint project resulting in considerable mutual benefit. Thus, the government's launch of AMPTE UKS constituted a use "by or for" the government. Third, because AMPTE UKS was launched in Florida, the use occurred within the territorial limits of the United States.

Nonetheless, we conclude that AMPTE UKS may not be included in the compensation base because 35 U.S.C. § 272, which sets forth a defense to certain claims of patent infringement, applies to AMPTE UKS. Section 272 of title 35, United States Code, provides a defense to claims of patent infringement if five conditions are met. First, the invention involved must be a part of a "vessel, aircraft or vehicle." Second, the presence of the vessel, aircraft or vehicle in the United States must be "temporar[y]" or "accidental." Third, the vessel, aircraft or vehicle must be from a country that "affords similar privileges to vessels, aircraft or vehicles of the United States." Fourth, the invention must be used "exclusively for the needs of the vessel, aircraft or vehicle." Fifth, the invention must not be "sold in or used for the manufacture of anything to be sold and/or exported from the United States."

As the discussion in Part V.A.4. makes clear, spacecraft launched after 1981 are deemed to be "vehicles" within the meaning of 35 U.S.C. § 272. See 42 U.S.C. § 2457(k) (providing that "[a]ny object intended for launch, launched, or assembled in outer space shall be considered a vehicle for purposes of section 272"). Because AMPTE UKS was launched in 1984, the requirement that the invention be a part of a "vessel, aircraft or vehicle" is satisfied.

Second, the presence of AMPTE UKS was "temporar[y]" within the meaning of § 272. AMPTE UKS entered the United States one time for the sole purpose of being launched into outer space. In Cali v. Japan Airlines, Inc., 380 F.Supp. 1120, 1126 (E.D.N.Y. 1974), aff'd without op., 535 F.2d 1240 (2nd Cir. 1975), the district court held that the presence of air carriers in the United States on a regular basis as part of an transoceanic passenger and freight service was temporary so that the carriers were not subject to the patent laws of the United States. Unlike the systematic presence of the air carriers in Cali, AMPTE UKS was in the United States one time for a short duration. Accordingly, we conclude that the presence of AMPTE UKS was temporary within the meaning of § 272.

Third, the United Kingdom extends "similar privileges" to "vessels, aircraft and vehicles" of the United States. The United Kingdom is a signatory to the Paris Convention, which is a treaty containing a temporary presence doctrine provision. Further, it is clear that the domestic law of the United Kingdom has had some version of the temporary presence doctrine since 1852. See DX 2658 (British Patents Act of 1852); DX 2661 (British Patents Act of 1949); DX 2665 (British Patents Act of 1977). Accordingly, we conclude that the third requirement — that the vessel, aircraft or vehicle be from a country which extends "similar privileges" to the United States — is present.

There can be no dispute as to the two final requirements imposed by § 272. The invention — the attitude control system — was used exclusively for the needs of the vehicle, and the invention was not sold or used for manufacture within the United States. Accordingly, we conclude that the temporary presence doctrine, as codified in 35 U.S.C. § 272, applies, establishing a complete defense to defendant's infringement of the Williams patent by using AMPTE UKS.

4. ARIEL

The ARIEL program was a cooperative project between NASA and the Science Research Council of the United Kingdom. This portion of the opinion addresses Hughes' claim for compensation under § 1498(a) for the ARIEL 5 spacecraft.

In 1970, the United States and the United Kingdom executed an MOU that outlined the responsibilities of each country and which described the scientific purpose of the project as "identifying and examining cosmic X-ray sources other than the sun" (PX 1036A). The MOU required the Science Research Council to contribute the spacecraft, to supplement NASA in providing tracking and data acquisition services and to transform data into an acceptable form and analyze the data. Similarly, the MOU required NASA to provide a launch vehicle, to "direct" the launch, to provide tracking and data acquisition services, and to cooperate with Science Research Council in analyzing the data. Further, the MOU provided that each agency would pay the cost of discharging its responsibilities. Subsequently, by letters dated December 27, 1973 and January 28, 1974, NASA and the Science Research Council agreed to revisions of the MOU, including, among other things, a change in the launch date and launch site, and a provision that NASA would be responsible for "directing" the launch instead of "conducting" the launch (PX 1036A).

The ARIEL 5 was built in Portsmouth, England and was funded by the United Kingdom Ministry of Defense (Tr. 15027). The spacecraft was registered with the United Nations as a United Kingdom spacecraft (Tr. 15028). In accordance with the revised MOU, the spacecraft was launched by a team of Italian engineers on a Scout launch vehicle from the San Marco launch platform off the coast of Kenya, Africa on October 15, 1974. NASA supplied the launch vehicle, provided some assistance in training the Italian personnel and directed the launch of the Ariel 5 spacecraft. After launch, the United Kingdom assumed responsibility for controlling the spacecraft, and NASA's role was limited to providing a data communications link using the NASCOM and STDN systems.

The spacecraft carried six x-ray experiments — five from the United Kingdom and one from the United States. Generally, these experiments were used to conduct a systematic survey of x-ray sources both within and beyond our galaxy (the Milky Way), from above the earth's atmosphere which filters out x-rays from space and thus makes it impracticable to study x-ray sources from the ground. In particular, the experiment sponsored by the United States, the All-Sky Monitor, was an x-ray imaging device capable of monitoring the entire sky continuously for transient x-ray phenomena and, at the same time, monitoring all the strong sources in the sky for long-term effects (PX 1036D).

Based on these facts, we conclude that ARIEL 5 should not be included in the compensation base. Although the government "directed" the launch, the launch was actually conducted by a team of Italian engineers off the coast of Africa. Hence, we must address whether ARIEL 5 was used within the territorial boundaries of the United States. The only nexus with the physical territory of the United States was that NASA STDN systems located around the world, but headquartered at Goddard Space Center, Greenbelt, Maryland, were used. In our view, that fact, by itself, is insufficient to establish liability for use within the United States.

If the United States government had actually originated the commands within the United States and then transmitted those commands to the satellite through its STDN system, we would find "use" within the United States. The fact that those functions were separated, with the government providing only tracking and data acquisition services, undercuts plaintiff's argument that the patent was used within United States territory. It is important to note that unlike the other four "foreign" spacecraft in issue, the ARIEL 5 spacecraft never entered the United States. That fact, combined with the absence of direct control from United States territory, suggests that any connection with the United States was attenuated. More important, though, is the fact that the so-called "home territory" or "master station" was not in the United States.

The Court of Claims recognized that if an invention is an instrumentality that necessarily extends beyond the United States with a control point in the United States, the "use" is deemed to occur here. In Decca Limited v. United States, 210 Ct.Cl. 546, 552-53, 544 F.2d 1070, 1074-75 (1976), the court addressed the problem of applying § 1498 to an allegedly infringing system that necessarily operated in more than one country. That case involved a worldwide radio navigation system with two stations in the United States and one station in Norway. Because the "master" station was in the United States, the court found that the system was "used" in United States territory:

At any rate, the whole Omega system must be deemed, at least for purposes of litigating the patent here involved, to be a unity and the location of that unity must be deemed to be in United States territory. Here it has planted several of its feet, and use of United States territory is indispensable to it. The location of facilities in some foreign countries is also essential to the plan, but the selection of any single other country is, apparently, not essential. Any one such country could readily be abandoned for another.

Accordingly, even though the ARIEL program was similar to the Helios and ISEE programs in that it was a cooperative joint project between the United States and other governments, the fact that any use by or for the United States occurred beyond the boundaries of the United States renders the patent laws inapplicable and thus obviates any liability under § 1498.

5. Conclusion

Part V has addressed the government's liability under § 1498(a) for five spacecraft associated with four international spacecraft programs. The primary issues were whether the government used each of the spacecraft, whether that use was "by or for" the government and whether the use occurred within United States borders. We conclude that the government is liable under § 1498(a) for three of the spacecraft, Helios A and B and ISEE B, and that the government is not liable under that provision for two of the spacecraft, AMPTE UKS and ARIEL 5.

VI

The foregoing portion of this opinion resolves all heretofore undetermined liability issues in this litigation. The appendix to this opinion lists all of the 108 spacecraft accused at any stage of the case and provides references to the liability disposition for each spacecraft.

There will follow an opinion addressing all previously unresolved issues pertaining to costs, compensation base, royalty rate and delay damages. Judgment shall be withheld pending resolution of all damages issues.

APPENDIX TO OPINION

This appendix contains a complete list of the 108 spacecraft accused at any stage of this litigation, together with (1) an indication for each accused item whether liability for a taking under 28 U.S.C. § 1498 has been found, (2) a citation to the record for the liability determination, (3) a statement of the taking date (where applicable) and (4) an indication of the major (but by no means exclusive) issue concerning liability. SPACECRAFT CITE MAJOR ISSUE LIABILITY DATE