Opinion
Penda Corporation (“Penda”) owns a patent on a plastic pallet, U.S. Letters Patent *542No. 4,428,306 (“the ’306 patent”). Defendant, acting through the United States Postal Service (“Postal Service”), procured 640,000 pallets from third-party defendant Cadillac Products, Incorporated (“Cadillac”). Under its contract, Cadillac supplied the Postal Service with two types of pallets (“the accused pallets”). Penda contends that its patent reads on both of these pallets.
Penda claims that by this procurement the Postal Service used the ’306 patent without providing just compensation as required by 28 U.S.C.A. § 1498(a) (West 1973 & Supp.1993). The Postal Service and Cadillac (“defendants”) counter that the ’306 patent is invalid as obvious under 35 U.S.C. § 103 (1988). Additionally, defendants contend that the ’306 patent is invalid for failing particularly to point out and specifically claim the subject matter the inventors regard as their invention as required by the second paragraph of 35 U.S.C. § 112 (1988).
I. Background.
A. The Patent.
The ’306 patent covers a design for a pallet. A pallet is an object used in material handling to provide an interface between bulk loads and a mechanized method of moving them. The ’306 patent concerns a pallet made of two sheets of thermoforma-ble plastic, which are molded and fused to form a single structure as set forth in the patent claim. Figure 1 is a perspective view, taken from the patent, of the pallet in suit. Pallets may be made of many materials, such as various types of wood and plastic. They have either one or two load-bearing surfaces. The pallet claimed in the ’306 patent and the accused pallets have one load bearing surface. In addition, the pallets claimed have feet on which the pallet rests. The pallet feet are formed from the same sheets of material as the load bearing deck.
Thermoforming is a process by which sheets of plastic are molded into useful objects through the application of heat and differential pressure. In twin-sheet ther-moforming, the manufacturer simultaneously molds the product out of two sheets of plastic, which the thermoforming process fuses together.
1. Claims.
The '306 patent contains seventeen claims. Penda contends that the first 200,-000 pallets that Cadillac sold the Postal Service infringe claims one through six and nine through eleven of the ’306 patent. Penda also contends that the 440,000 modified pallets that Cadillac sold to the Postal Service infringe claims nine through twelve, fifteen, and sixteen of the ’306 patent. The court has not found it necessary to address all the claims. If even a single claim is valid and infringed, plaintiff is due compensation. Because claims nine through eleven are the broadest in the patent, the court directs its attention toward them. These claims read:1
9. A pallet (10) comprising:
*543a substantially planar load-bearing member (12) formed from an upper sheet of material (13) and a lower sheet of material (15) located generally in fixed parallel relationship to each other; and
a plurality of feet (32) depending from the load-bearing member (12), the feet (32) being formed as a downwardly depending portion of each of the two sheets of material (13, 15) forming an upwardly opening pocket (36) on the interior thereof, each of the feet (32) including integral solid wall portions (40) formed of deformed portions of both of the sheets of material (13, 15) fused together.
10. A pallet (10) as claimed in claim 9 wherein there are also hollow wall portions (34) separating the solid wall portions (40) in each of the feet (32).
11. A pallet (10) as claimed in claim 9 wherein the solid wall portions (40) of each of the feet (32) are formed from portions of both of the two sheets of material (13, 15) joined together and wherein each of the feet (32) terminates in a foot floor (42) which is an integral, planar horizontal sheet of solid material also formed of portions of both of the two sheets of material (13, 15) fused together.
(the ’306 patent, col. 10, 11. 16-40) Defendants admit that claims nine and ten read on the accused devices, but assert that they are invalid. Those claims describe the foot design in detail, but only describe the deck as consisting of two sheets in a fixed parallel relationship. Claim nine describes the feet as having integral solid wall portions. The limitation of integral hollow wall portions separating the integral solid wall portions is introduced in claim ten.
Claim eleven depends from claim nine and thus incorporates its limitations. It does not, however, incorporate the additional limitation of integral hollow wall portions added by claim ten. Claim eleven adds the limitation that each foot “terminates in a foot floor (42) which is an integral, planar horizontal sheet of solid material also formed of portions of both of the two sheets of material (13, 15) fused together.” (’306 patent at col 10, 11. 33-40.) Defendants argued that claim eleven does not read on the accused devices because it limits the design of the foot floor to one containing a floor of single thickness generally parallel to the surface of the load-bearing member.
2. Prosecution History.
The inventors filed the application for the patent in suit on October 9, 1981. The application names William Dresen, Harlon Breezer, and Bill Price as coinventors. Also at that time, they filed their oath of inventorship as required by 35 U.S.C. § 115 (1988), stating, among other things, that they regarded as their invention the subject matter of the claims.
In January 1983, the examiner rejected initial application claims one through three, six through ten, and twélve through nineteen as being either anticipated by prior art or obvious to one of ordinary skill in the art. The examiner objected to the remaining claims as being dependant upon rejected claims, but found them otherwise allowable.
In response, on July 26, 1983, the applicants amended claims one, ten, twelve, and sixteen of the application to incorporate all the limitations of the application’s original claim four. They amended application claim eighteen to depend from application claim sixteen. They canceled claims four and seventeen, and renumbered the remaining claims. The examiner allowed the revised claims on September 25, 1983. The patent issued January 31, 1984.
3. Specification and Lexicography.
The specification describes the pallet feet:
It is another object of the present invention to provide such a pallet in which the legs elevating the load-bearing member of the pallet off of the surface have a particularly high compressive strength so that the pallet has a large capacity.
(’306 patent at col. 1, In. 66 through col. 2, In. 2.)
*544It further described the preferred embodiment:
The feet 32 of the pallet 10 are also particularly constructed so as to be extremely strong and extremely resistant to deformation or collapse. As can be seen by referring to FIGS. 5 and 6, each of the feet 32 has alternating solid portions 40 and hollow wall portions 34 spaced about the periphery of the foot 32. The fact that the solid portions 40 alternate with the hollow portions 34 makes the feet 32 stronger and more crush resistant than would be possible if the entire periphery of the foot was constructed from either solid or hollow material. The solid portions 40 serve in essence as stiffening ribs for the hollow portions 34; while the hollow portions 34 break up the continuity of the solid portions 48 so that0 the overall strength of each of the feet 32 is stronger than would be if the entire foot was constructed of solid material and the resistance of the foot 32 to flexing under a load is greatly increased.
(’306 patent at col. 7,11. 36-52.) Figure 2, a drawing from the ’306 patent, illustrates this description.
Each of the alternative embodiments also contains the alternating hollow and solid structure. Not all of the claims, however, are limited to legs in which hollow portions separate the fused solid portions. Defendant contends, and plaintiff disputes, that the patent does not teach the advantage of a solid foot without hollow portions. The patent does indicate that one aspect of the present invention is that the feet are stronger than they would be if the foot were completely solid. (’306 patent col. 7, lines 46-52.)
The specification describes the foot floor in the preferred embodiment as being “formed along the bottom of the foot 32, with the floor 42 of each foot being a planar, single-thickness section generally parallel to the load-bearing member 12.” (’306 patent, col. 4, 11. 31-35). In an alternative embodiment, the specification describes the foot floor as being “provided with a canted bottom surface so as to drain toward the drain hole 143 provided in the center thereof so that fluid received inside of the foot 132 would drain out the drain hole 143.” (’306 patent col. 8, 11. 45-49.) In this described alternative embodiment, the drain hole in the foot is centrally located.
*5454. The Inventors and the Invention.
William L. Dresen is President of Penda and has worked for it since 1978. Before joining Penda, he was Director of Research and Development at Glasfab, a company engaged in thermoplastic molding as well as reinforced fiberglass and urethane molding. Penda purchased Glasfab in 1978, and Dresen became General Manager of the combined companies. At that time, the company had $10,000,000 in total annual sales, of which thermoplastics accounted for one fifth. The company had between 50 and 60 employees.
Among articles that Penda manufactured before it designed the pallet at issue were single sheet thermoform pallets. These pallets had certain disadvantages. Most notably, they tended to flex under stress and could not hold heavy static loads, being limited to a load range of 1,000 to 1,500 pounds. Comparable wooden pallets could carry as much as 20,000 to 30,000 pounds. Wood pallets have the disadvantage, however, of being less durable. They are brittle and fracture easily. Also, they are more difficult to clean. Consequently, Pen-da’s customers wanted a heavier weight plastic pallet. Dresen sought plastic pallets that met the following criteria: they had to be stronger than existing pallets, that is, they could not flex; they had to be comparable in weight-bearing capacity to wooden pallets; and they had to have the durability of plastic. He found no such pallets to be available in the market.
It occurred to Mr. Dresen that Penda could use its twin-sheet thermoforming machine to build a stronger pallet. To develop a twin-sheet pallet, Mr. Dresen organized a small group of Penda employees— Harlon Breezer, vice president of sales at Penda; himself; and Bill Price, a consultant to Penda in thermoplastics. The group started meeting and brainstorming in early 1980. Dresen was in charge and served as the group’s facilitator.
Mr. Price was working as a consultant for Penda. He testified that he had also earlier sensed the need for a stronger plastic pallet. In part, these observations occurred while he was traveling with Mr. Dresen. After deciding that traditional thermoform pallets were inadequate, he approached Mr. Dresen who, it turned out, had the same idea. Mr. Price discussed his ideas with both Mr. Dresen and Mr. Breezer.
Mr. Price did the drawings during the development of the invention. His expertise is in designing parts and sketching parts and tools. Mr. Price testified that the inventors’ first drawing was a rough engineering sketch of a construction he felt would overcome some of the problems of single-sheet pallets. The pallet leg in this first drawing had two distinct sidewalls. Over time, the inventors developed this structure into the alternating solid and fused sidewalls disclosed in the patent.
After the inventors had a design for the pallet foot that they thought would work, they made a wooden model of that section of the plastic part. Mr. Price testified that the design objective of the inventors was to combine the strongest legs with the strongest deck. To this end, they considered the pallet legs separate from the pallet deck. Their intention was then to integrate the leg into the deck design. If they could design an appropriate foot, it was felt that the deck design would be easy. Mr. Breezer confirmed that the inventors worked out a leg design before making a wooden pattern, then added a deck design after they had a workable foot. Mr. Price testified that although the use of wooden models was common, it was unusual to test part of a design separate from the whole. The inventors did so because they initially considered the risk of failure in the pallet foot too great to justify the expense of making a model for the whole pallet. They made the model of the foot section when they felt they had a formable design.
After several months of work, the inventors had the design they wanted. Fusing together portions of the leg sidewalls imparted to the pallet sufficient compressive strength to withstand heavy loads. In addition to this strong pallet foot, the inventors had a design for the pallet deck that provided good rigidity.
*546B. The Accused Pallets.
The accused pallets are nestable, twin-sheet, thermoform items. They have a deck portion in which the two sheets are generally parallel to one another, and downwardly depending legs in which portions of the sheets in the sidewall are fused together. In each of the accused pallets, the foot floor has ridges on the interior surface that are not fused to the outer sheet. Although the accused pallets have many other features, these are the only ones relevant to an analysis of infringement with respect to claim eleven. As explained below, the court deems it unnecessary to discuss the other claims of the patent. Defendants contend that because of these ridges in the foot floor, the feet in the accused pallet do not satisfy claim eleven’s limitation of “an integral planar horizontal sheet of solid material.” All the pallets designed by Penda had a solid foot floor. None showed hollow spaces in the foot floor. Although Cadillac actually produced two different models of its pallets, the differences between these models lay mainly in the pattern of channels appearing in the deck. The differences that did exist between the pallet feet in the two runs are not significant in discussing infringement of claims nine and eleven.
C. The Procurement.
Marshall Weiss of the Postal Services Engineering and Development center first learned of Penda’s pallet through a design firm. He contacted Penda and visited its facility to learn about the company’s products. In January 1987, the Postal Service purchased 500 Penda pallets for evaluation. The Postal Service purchased 20,000 additional Penda pallets in October 1987, in conjunction with the purchase of other pallets made of structural foam. After testing the structural foam and thermoform pallets under field conditions, the Postal Service concluded that the thermoform Penda pallet stood up better than the structural foam pallet.
The Postal Service contracted with Shuert/Oakland to design a Postal Service pallet and supply prototypes. Shuert/Oak-land supplied the prototypes in April 1988. At that time, however, a contract dispute arose, and the Postal Service was not able to follow through on its plans with Shuert/Oakland. The Postal Service decided to procure pallets in production quantities. It created a pallet specification describing how the pallet was to function. The specification was derived from an earlier specification used to purchase structural foam pallets and pallets molded from compressed wood fibers. It contains requirements as to the gross shape of the pallet: a single deck with nine feet. The Postal Service knew such pallets were feasible because of Penda’s commercial product and the Shuert/Oakland contract. The contract required the contractor to supply a design that the Postal Service would have an unlimited right to use.
Cadillac Products is a plastics company located in Troy, Michigan. It was founded in 1946 to make packaging materials. About fourteen percent of the company’s business is thermoforming. Its largest product area is automotive parts, which it makes by thermoforming and injection molding.
Cadillac first learned of the Postal Service’s interest in a plastic pallet at a vendor orientation seminar conducted by the Postal Service. The Postal Service issued its Request for Proposals on April 4, 1988. When the proposals were opened on May 20,1988, Cadillac was the low bidder. Marshall Weiss and other Postal Service employees conducted a pre-award survey of Cadillac’s facilities. The Postal Service awarded Cadillac the contract on July 20, 1988. About the time of bid opening, Cadillac began to prepare a design. For that purpose it hired WolPac Co. Mr. Jim Waun, a WolPac employee, did the work.
In June, 1988, Mr. Dresen and Mr. Mar-kos, Penda’s chief executive officer, approached Mr. Michael Williams, a Cadillac officer, at the National Plastics Exhibition. They suggested he review the '306 patent for possible infringement. Shortly thereafter, Cadillac began to redesign the pallet with the purpose of avoiding infringement. In July 1988, Cadillac notified the Postal *547Service of Penda’s patent and requested additional time to prepare a design. Cadillac decided to incorporate different deck characteristics, and told WolPac to modify the design. Mr. Weiss of the Postal Service was aware of Cadillac’s effort, but did not recommend any design changes. The Postal Service refused to give Cadillac relief under the contract or to grant Cadillac additional time to redesign the pallet. After the redesign, Cadillac built prototypes from wood patterns and submitted them to the Postal Service for approval. Beginning November 28, 1988, Cadillac delivered pallets under the contract.
II. Discussion.
A. Infringement.
Plaintiff contends that the accused Cadillac pallets read on claim eleven of the ’306 patent. Defendants dispute this contention, but admit that claim nine, from which claim eleven depends, reads on the accused devices.2 Moreover, there is no question that the accused devices contain a foot floor which is solid, integral, and formed of portions of the two sheets fused together. They contain a set of ribs on the upper, interior surface of the foot floor. There may or may not be a space between the upper sheet of material and the lower sheet of material at the ribs. The same material that forms the ribs composes the rest of the pallet. Defendants contend that this foot floor is not “an integral, planar horizontal sheet of solid material also formed of portions of both of the two sheets of material fused together.” The point of contention is whether the foot floor is “planar horizontal.”
Plaintiff maintains that the accused devices literally infringe claim eleven of the patent because the ribs in the foot floor are merely additional structure, by which defendant does not avoid infringement.3 The plaintiff bears the burden of proving infringement by a preponderance of the evidence. Z.M.I. Corp. v. Cardiac Resuscitator Corp., 844 F.2d 1576, 1582 (Fed.Cir.1988); Lee v. Dayton-Hudson Corp., 838 F.2d 1186, 1187 (Fed.Cir.1988); Mannesmann Demag Corp. v. Engineered Metal Prod. Co., 793 F.2d 1279, 1282 (Fed. Cir.1986). The inquiry comprises two distinct steps: first, the court interprets the claims of the patent without reference to the accused device; second, it compares the properly interpreted claims to the accused device to determine whether it infringes. Hormone Research Found., Inc. v. Genen-tech, Inc., 904 F.2d 1558, 1562 (Fed.Cir. 1990), cert, dismissed, 499 U.S. 955, 111 S.Ct. 1434, 113 L.Ed.2d 485 (1991); S.R.I. Int’l v. Matsushita Elec. Corp., 775 F.2d 1107, 1118 (Fed.Cir.1985).
1. Claim Interpretation.
Claim interpretation must precede the ultimate determination of infringement, because the claims of the patent delineate the extent of the property right conferred. Moeller v. Ionetics, Inc., 794 F.2d 653, 656 (Fed.Cir.1986); A.B. Dick Co. v. Burroughs Corp., 713 F.2d 700, 702 (Fed.Cir. 1983), cert. denied, 464 U.S. 1042, 104 S.Ct. 707, 79 L.Ed.2d 171 (1984). In general, the meaning given to terms in a patent is the meaning that one of ordinary skill in the art would ascribe to them.4 Intellicall, Inc. v. Phonometrics, Inc., 952 F.2d 1384, 1387 (Fed.Cir.1992). In interpreting the claims of a patent, the court refers to the claim itself, other claims, the specifications (including the drawings), the prosecution history, and extrinsic evidence presented at trial. Hormone Research Found., 904 *548F.2d at 1562; LaBounty Mfg. v. United States Int’l Trade Comm’n, 867 F.2d 1572, 1574 (Fed.Cir.1989).
The dispute between the parties about the proper interpretation of the language of claim eleven centers on the phrase “each of the feet (32) terminates in a foot floor (42) which is an integral, planar horizontal sheet of solid material also formed of portions of both of the two sheets of material (13, 15) fused together.” Similar language appears in claim fifteen of the patent: “terminating in a horizontal, integral foot floor (142).” Claim fourteen describes the foot floor as “planar,” while claim fifteen does not.
A determination of the ordinary meaning of a term is not dispositive without reference to the specification and prosecution history. Z.M.I. Corp., 844 F.2d at 1580. In the case at bar, the drawings and descriptions in the specification do suggest the meaning intended by the applicants. The specification is significant because the applicant must use language consistently in the claims and the specification. Z.M.I. Corp., 844 F.2d at 1580; McGill, Inc. v. John Zink Co., 736 F.2d 666, 674 (Fed.Cir.), cert. denied, 469 U.S. 1037, 105 S.Ct. 514, 83 L.Ed.2d 404 (1984). Figure six of the specification illustrates the patent foot floor. The upper surface of that foot floor is clearly not flat; it slopes to the middle. The reference number in the drawing for the foot floor is forty-two. This same reference number appears in claim eleven. One of ordinary skill in the art referring to the specification to interpret claim eleven would be drawn to figure six as an illustration of what the patentee meant by the phrase “integral, planar horizontal sheet,” and would see that the upper surface need not be everywhere parallel to the lower.
Similarly, figure nine of the patent illustrates a foot floor on which the upper surface slopes to the middle. In fact, none of the patent’s drawings illustrate a foot floor in which the upper, interior surface is level, horizontal, and everywhere parallel to the bottom of the pallet foot. The reference numbers for the foot floor in figure nine do not appear in claim eleven, but they do
appear at the corresponding place in the similar language of claim fifteen. The specification describes a foot floor with a “canted” surface. (’306 patent, col. 8 11. 47.) A review of other claims may provide significant evidence regarding the scope of a claim in issue. Fromson v. Advance Offset Plate, Inc., 720 F.2d 1565, 1570 (Fed.Cir.1983). A person of ordinary skill in the art seeking to interpret claim fifteen of the patent would refer to figure nine for an illustration of what the patentees meant by the term “horizontal, integral foot floor.” Applying that knowledge in interpreting the phrase “integral, planar horizontal sheet” in claim eleven, that person would conclude that the upper surface need not be parallel at all points to the lower.
Representations made during the prosecution history, including amendments and arguments, are also relevant. Jonsson v. Stanley Works, 903 F.2d 812, 818 (Fed. Cir.1990); Loctite Corp. v. Ultraseal Ltd., 781 F.2d 861, 867 (Fed.Cir.1985). The prosecution history limits claim interpretation by precluding interpretations disclaimed or disavowed during the prosecution. LaBounty, 867 F.2d at 1574; Z.M.I. Corp., 844 F.2d at 1580. The prosecution history of the ’306 patent indicates that the applicants amended claim eleven after the first office action. The plaintiff added the word “integral” to indicate that the foot floor was made of one material fused together, and replaced the word “joined” with the word “fused” to distinguish prior art references teaching a pallet with two sheets attached to one another by an adhesive. Neither of these changes, by itself, prohibits the plaintiff from contending that the claim does not require that the upper, interior surface of the foot floor be smooth, if the language of the claim is sufficiently broad .to encompass such a structure.
Courts may refer to general and specialized dictionaries in interpreting terms. See Fromson, 720 F.2d at 1571. “Planar” means “1: of or relating to a plane: lying in one plane 2: having a flat, two dimensional quality,” Webster’s Third New International Dictionary of the English *549Language, Unabridged 1730 (Philip B. Gove ed., 1976), or “belonging to, situated in, or related in some way to a plane.” 9 The Oxford English Dictionary 959 (J.A. Simpson & E.S.C. Weiner eds., 1989). The court, however, would not rely on these definitions if it appeared that one of ordinary skill in the art would have interpreted the claim differently.
To interpret the claims, courts may also refer to extrinsic evidence. Texas Instruments Inc., v. United States Int’l Trade Comm’n, 988 F.2d 1165, 1171 (Fed.Cir. 1993); Hormone Research Found., 904 F.2d at 1562. Several expert and factual witnesses were called on the question of infringement. Of particular note, plaintiff called Dr. John E. Johnson and Mr. John W. Hofeldt, Esq., as expert witnesses, and .defendant called Mr. Larry Stiller and Dr. Paul Sher Singh.
Dr. Johnson is an emeritus professor in the College of Engineering at the University of Wisconsin. He received a B.S. in civil engineering from Gonzaga University, an M.S. in structural engineering and structural design from Stanford, and a Ph.D. in mathematics and structural engineering with a minor in engineering mechanics from Purdue University. In the early sixties, Dr. Johnson developed a structural system of using plastics for buildings and roofing systems. He has done much research through the years both in an academic and a corporate setting. His handbook on the structural use of plastics has become an industry standard. As a full professor, he taught most of the design and analysis courses his department offered, as well as many seminars on the structural use of plastics.
Mr. Hofeldt obtained a bachelor’s degree in religious philosophy from the University of Wisconsin in 1943 and, after serving a stint in the navy, a law degree from the University of Wisconsin in 1947. He practiced law from 1955 until 1990, principally in the litigation of patent cases. At the time of trial, he was of counsel with the firm of Woods, Philips, Van Santen, Hoffman and Ertel. He has taught patent litigation and infringement courses at John Marshall Law School, which courses included within their ambit claim interpretation and infringement proofs.5
Mr. Stiller learned the art of thermo-forming principally through apprenticeship. Before graduating high school in 1960, he worked in his father’s prototyping shop. He learned other aspects of the trade from other shops through his father’s contacts.
Dr. Singh is an assistant professor at Michigan State University in the School of Packaging, where he has been on faculty since 1985. He received a B.S. in chemical engineering in 1982, an M.S. in packaging in 1983, and a Ph.D. in agricultural engineering in 1987.6 He worked at Michigan State in 1982 and 1983 as part of his degree program. He worked briefly at the University of California, Davis until he ceased his Ph.D. studies there, then worked for IBM in Rochester, Minnesota for a year. About six years ago he started a conference on *550palletization unitization, which now meets annually. Also, he chairs the task group on a new pallet standard within American Standards for Testing and Materials, a professional society that develops standards and procedures to test and evaluate different types of systems.
None of the evidence indicates that one of ordinary skill in the art would assign the terms at issue meanings different from the common meaning given in the dictionary. Dr. Singh’s use of the term to mean “in one plane, flat,” is consistent with the dictionary definition. Similarly, Mr. Stiller’s interpretation of the term as meaning “on one plane” comports with the common definition. The court concludes that the term “planar horizontal” in claim eleven refers to the quality of being flat, in one plane, or of horizontal orientation. From the testimony of Dr. Johnson and Mr. Hofeldt that they consider the foot floor of the accused pallets planar, it would seem that they do not adopt a more restrictive definition of the term. Inasmuch as Dr. Johnson’s and Mr. Hofeldt’s testimony inseparably involves applying the claim to the accused device, current Federal Circuit precedent prohibits its use in interpreting the claim language. Therefore, the court refers to their testimony only by way of comparison. The court does not construe claims in the light of the accused device, nor does it construe them to “cover” or “not cover” the accused device. S.R.I. Int’l, 775 F.2d at 1118.
Based on all the forgoing considerations (the claim language, the specification, the prosecution history, the dictionary definitions, and the extrinsic evidence that the court may consider), the court concludes that in the claim the phrase “planar horizontal” means of or relating to a plane and generally flat. It does not mean that the two surfaces are everywhere parallel, as urged by defendants.
2. Application of the Claims to the Accused Devices.
With this background understanding of the meaning of the terms in
claim eleven, we proceed to determine whether these claims read on the accused devices.7 To prove literal infringement, the plaintiff must show that the accused device contains every limitation of the claim. Intellicall, Inc., 952 F.2d at 1389; Becton Dickinson & Co. v. C.R. Bard, Inc., 922 F.2d 792, 798 (Fed.Cir.1990); Hormone Research Found., 904 F.2d at 1562; Johnston v. IVAC Corp., 885 F.2d 1574, 1577, 1580 (Fed.Cir.1989); Pennwalt Corp. v. Durand-Wayland, Inc., 833 F.2d 931, 949 (Fed.Cir.1987), cert. denied, 485 U.S. 961, 108 S.Ct. 1226, 99 L.Ed.2d 426 (1988). In determining infringement, the court may only compare the accused devices with the patent, not with a commercial embodiment of the patented invention. Loctite Corp., 781 F.2d at 870. The court has disregarded any evidence that compared the accused pallets to the actual Penda pallet.
Dr. Johnson testified that he considers the foot floor of the accused pallet to be planar. Similarly, Mr. Stiller, a witness for defendant, when asked if he considered the foot floor of the accused pallet planar answered “On one plane? Yes.” When counsel for third party defendant specifically directed him to consider both the upper side and lower side of the foot floor, Mr. Stiller replied that the bottom was flat, but that the top was beaded. Plainly, however, his initial reaction was that the foot of the accused device did terminate in a planar foot floor. Mr. Hofeldt testified that he considered the accused pallets to contain a foot floor made of an integral, horizontal sheet of material because he regarded the “bubbles” as additional structure irrelevant to the claim.
Only Dr. Singh, defendant’s expert, testified that he did not consider the foot floor planar. He spoke of the ribs as imparting to the foot floor a “vertical component.” Yet on direct examination, Dr. Singh also spoke of the Cadillac pallet as having “vertical ribs in addition to the integral floor.” *551On cross-examination he spoke of “the addition of these ribs” and stated that he considered the ribs an “addition” to the foot floor.
As Mr. Hofeldt noted, claim eleven is an open claim because it depends from claim nine, which is an open claim using the term “comprising.” The addition of new elements does not avoid infringement. Uniroyal, Inc. v. Rudkin-Wiley Corp., 837 F.2d 1044 (Fed.Cir.), cert. denied, 488 U.S. 825, 109 S.Ct. 75, 102 L.Ed.2d 51 (1988); Mannesmann, 793 F.2d at 1279; Amstar Cory. v. Envirotech Corp., 730 F.2d 1476 (Fed.Cir.), cert. denied, 469 U.S. 924, 105 S.Ct. 306, 83 L.Ed.2d 240 (1984). The proper inquiry is whether the claim and specification prohibit the addition of the structure. Stiftung v. Renishaw P.L.C., 945 F.2d 1173 (Fed.Cir.1991). Here, they do not. Dr. Singh believed that the patent required the top of the interior sheet and the bottom of the exterior sheet in the pallet foot to be everywhere parallel. The court finds that the hypothetical “person of ordinary skill in the art” would not have so understood the patent.
Dr. Singh’s reading is also inconsistent with the drawings, which show a slanted interior surface, and with the specification language, which speaks of a canted interior surface to aid in draining the pallet foot. Even if there were evidence that the interpretation adopted by Dr. Singh is the interpretation one of ordinary skill in the art normally would assign the phrase “planar horizontal,” a patentee is free to be his own lexicographer. Intellicall, 952 F.2d at 1387; Jonsson, 903 F.2d at 819; Z.M.I. Corp., 844 F.2d at 1580; Loctite Corp., 781 F.2d at 867; Lear Siegler, Inc. v. Aeroquip Corp., 733 F.2d 881, 888 (Fed.Cir.1984); Lemelson v. United States, 14 Cl.Ct. 318, 322 (1988). From the specification of the patent, it is clear that the patentees intended this claim to encompass a foot floor structure in which the interior and exterior surfaces were not everywhere parallel.
3. Doctrine of Equivalents.
Because the court finds that the accused pallets literally infringe claim
eleven, it is unnecessary to consider the doctrine of equivalents. Assuming for the sake of argument, however, that one of ordinary skill in the art could disregard the specification and read the phrase “planar horizontal” in the restrictive sense urged by Dr. Singh, the court finds that the accused devices nonetheless infringe under the doctrine of equivalents. The same interpretation of the claims used in determining literal infringement governs a determination of infringement under the doctrine of equivalents. Senmed, Inc. v. Richard-Allan Medical Indus., 888 F.2d 815, 818 (Fed.Cir.1989). An accused device infringes under the doctrine of equivalents if it performs substantially the same function as the patented structure, by substantially the same means, to provide the same result. Graver Tank & Mfg. Co. v. Linde Air Prod. Co., 339 U.S. 605, 609, 70 S.Ct. 854, 856, 94 L.Ed. 1097 (1950). The accused device must contain every limitation of the claim either literally or by substantial equivalent. Intellicall, 952 F.2d at 1389; London v. Carson Pirie Scott & Co., 946 F.2d 1534, 1538 (Fed.Cir.1991); Jurgens v. McKasy, 927 F.2d 1552, 1560 (Fed.Cir.), cert. denied, — U.S.—, 112 S.Ct. 281, 116 L.Ed.2d 232 (1991); Becton, 922 F.2d at 798. This tripartite formulation is also appropriate in evaluating whether the accused device contains the substantial equivalent of a particular element of the claim. Malta v. Schulmerich Carillons, Inc., 952 F.2d 1320, 1326 (Fed. Cir.1991), cert. denied, — U.S. —, 112 S.Ct. 2942, 119 L.Ed.2d 566 (1992); Corning Glass Works v. Sumitomo Elec. U.S.A., Inc., 868 F.2d 1251, 1260 (Fed.Cir. 1989).
The function of the foot floor is to close up the bottom of the foot, to provide a brace between the sidewalls, and to serve as a contact point between the pallet and the surface on which it is resting. The foot floor of the accused device serves the same function. The patented device performs this function by fusing together the portion of the two sheets between the sidewall. The accused device performs this function in substantially the same way. The ribs in *552the foot floor may aid in this fusion, but they in no way interfere with it.8 With respect to this second element of equivalents, only substantial similarity, not identity, is required. Graver Tank, 339 U.S. at 609, 70 S.Ct. at 856. Certainly, the way in which a foot floor with ribs performs the functions enumerated above is substantially similar to the way in which the patented pallet performs those functions. The result in the patented device is a strong foot floor to support the sidewall and to provide resistance to wear. No evidence offered by defendant refuted Dr. Johnson’s testimony that the foot floor was the strongest section of the pallet. The result in the accused device is the same — a strong foot floor. The foot floor of the accused device may in fact be more resistant to buckling than that of the patented pallet. Any difference, however, is only in degree and not in kind. The court finds persuasive Dr. Johnson’s statement that the foot floor is the strongest portion of the pallet, and that any stress great enough to result in a buckling there would result in the failure of some other portion of the pallet first. From observing the witnesses at trial, it was evident that Mr. Stiller’s and Dr. Singh’s explanation of how the pallet foot floor could buckle considered only the foot in cross-section and failed to acknowledge the contribution of additional structural support from other sidewalls.
B. Sufficiency of the Specification.
In their pretrial filings, defendants contended that the patent was invalid for failure to comply with section 112.9 Because claim nine is the broadest claim and contains all of the features upon which defendants rely in asserting invalidity under section 112, resolution of this issue with respect to claim nine is considered adequate to dispose of the argument with respect to all of the claims. Defendants advanced two hypotheses to support their argument of invalidity: (1) that the claims included subject matter other than that which the applicants regarded as their invention, and (2) that the claims are invalid as indefinite because they are broader in scope than the disclosure portion of the specification. Defendants failed to prove by clear and convincing evidence that the applicants did not regard the subject matter claimed as their invention. Defendants’ second argument is both legally misplaced and factually unsubstantiated.
1. The subject matter the inventors regard as their invention.
Defendants contended in their pretrial materials that the applicants for the patent in suit regarded as their invention the combination of fused and hollow portions in the leg. Because claim nine only requires the presence of fused portions of the leg, defendants contend that it fails to comply with the second paragraph of section 112.
The Court of Customs and Patent Appeals held that in the absence of evidence to the contrary, the subject matter set out in the claims is presumed to be that which the applicant regards as his invention. In re Miller, 441 F.2d 689, 692 (C.C.P.A.1971); In re Moore, 439 F.2d 1232, 1235 (C.C.P.A.1971); see 2 Donald S. Chisum, Patents § 8.03[4] (1993). The “regards” language may only be relied upon to reject a claim where some evidence other than the specification shows that the claim does not correspond in scope to what the applicant regards as his invention. In re *553Conley, 490 F.2d 972, 976 (C.C.P.A.1974); 2 Chisum, supra, § 8.03[4]. Miller, Moore, and Conley, cases in which the court considered the “regards” language, were all appeals from the denials of patent applications. No patent had yet issued. As such, those decisions did not address the statutory presumption of validity. The patent in suit, having been duly issued, is presumptively valid under section 282.10
In closing argument, counsel for third-party defendant stated that the inventors had not testified that they considered a fully fused, solid foot wall to be their invention. This argument mistakes the burden of proof. The inventors, in the declaration submitted to the patent office with their application, swore that the subject matter claimed was what they considered to be their invention. It is up to the party asserting invalidity to prove by clear and convincing evidence that the inventors in fact did not consider the subject matter set forth in the claim as their invention. No such proof was offered.
Moreover, there is ample evidence from which to infer that the inventors regarded the fused portions of the sidewall in the pallet leg as their invention, rather than merely the combination of fused and hollow portions. The inventors designed the pallet foot first. The procedure that the inventors followed, of designing separate molds for testing just the pallet foot, was not their usual practice. Clearly, they were concerned primarily with producing a sufficiently strong foot. It may well be that a foot sidewall with hollow portions as well as solid portions is superior to one having only solidly fused portions. An inventor, however, is not restricted to claiming only the preferred embodiment or best mode of his invention. Inventors may claim that which they regard as their invention as broadly as the prior art will allow; it is in the disclosure portion of the speeifi-cation that the inventor must present the preferred embodiment. A broad claim, supported by a disclosure of the preferred and alternative embodiments, is exactly what the inventors supplied in this patent. The court concludes that the inventors intended claim nine to have the breadth exhibited by the claim language, and that they intended it to cover pallets regardless of whether the sidewalls of the feet have hollow portions.
2. Claim Definiteness.
The court looks to the claims themselves to determine the scope of the invention. Stiftung, 945 F.2d at 1181; Moore, 439 F.2d at 1235. When assessing compliance with section 112, analysis should begin with the second paragraph of that section, Moore, 439 F.2d at 1235, because where the first paragraph of that section uses the term “invention” it could only be referring to the claimed invention as set forth by the applicant. Id. The initial inquiry, therefore, is to determine whether the claims do set forth a particular area with reasonable precision and particularity.
When the meaning of claims is in doubt, especially where there is similar prior art, the claims are properly invalidated. Amgen, Inc. v. Chugai Pharmaceutical Co., 927 F.2d 1200, 1218 (Fed.Cir.), cert. denied, — U.S. —, 112 S.Ct. 169, 116 L.Ed.2d 132 (1991). Only reasonable precision is required, however, in delineating the claimed invention. United States v. Telectronics, Inc., 857 F.2d 778, 786 (Fed.Cir. 1988), cert. denied, 490 U.S. 1046,109 S.Ct. 1954, 104 L.Ed.2d 423 (1989). The exact level of detail required in the claims depends on the particular invention and on the prior art. Shatterproof Glass Corp. v. Libbey-Owens Ford Co., 758 F.2d 613, 624 (Fed.Cir.), cert. dismissed, 474 U.S. 976, 106 S.Ct. 340, 88 L.Ed.2d 326 (1985). The definiteness of claim language is therefore *554analyzed not in a vacuum, but rather in light of the teachings of the prior art and the particular application disclosure, as one of ordinary skill in the art would interpret it. Moore, 439 F.2d at 1235. Just as claims that appear indefinite on their face may prove sufficiently definite when read in light of the prior art, claims that on their face appear definite may prove impermissi-bly indefinite when viewed in light of the prior art. Id.
In their pretrial filings, defendants asserted that the phrase “generally in fixed parallel relationship” is vague. The court does not consider the phrase to be ambiguous on its face, and defendants did not support their position at trial with clear and convincing evidence.
Words of degree often precipitate definiteness arguments that compel the court to determine whether the disclosure portion of the specification provides some standard for defining the outer boundaries of the range permitted by the phrase. Seattle Box Co. v. Industrial Crating & Packing, Inc., 731 F.2d 818, 826 (Fed.Cir.1984) (construing the phrase “substantially equal to”). A phrase is not indefinite merely because the specification does not precisely define it. Rosemount, Inc. v. Beckman Instruments, Inc., 727 F.2d 1540, 1547 (Fed.Cir.1984) (construing the phrase “close proximity”). To accept such a contention would be to turn the construction of a patent into a useless semantic quibble. Rather, a claim is sufficiently definite for purposes of the second paragraph of section 112 if one of ordinary skill in the art would understand what is claimed, when the claim is read in light of the specification. Amgen, 927 F.2d at 1217; Telectronics, 857 F.2d at 786; Seattle Box, 731 F.2d at 826. “If the claims, read in light of the specifications, reasonably apprise those skilled in the art both of the utilization and scope of the invention, and if the language is as precise as the subject matter permits, the courts can demand no more.” Telectronics, 857 F.2d at 786; Hybritech, Inc. v. Monoclonal Antibodies, Inc., 802 F.2d 1367, 1384 (Fed.Cir.1986), cert. denied, 480 U.S. 947, 107 S.Ct. 1606, 94 L.Ed.2d 792 (1987). Even where some experimentation is required to fix the limits of words of degree such as “substantially equal to,” the claim is not indefinite if the experimentation required is not undue. Seattle Box, 731 F.2d at 826; cf. Atlas Powder Co. v. E.I. Du Pont De Nemours & Co., 750 F.2d 1569, 1576 (Fed.Cir.1984) (holding the disclosure portion of a specification enabling where it permitted practice without undue experimentation). The law does not require that the specification, let alone the claims, list every parameter of operation, so long as those of ordinary skill in the art realize that the parameter may be readily obtained. Orthokinetics, Inc. v. Safety Travel Chairs, Inc., 806 F.2d 1565, 1576 (Fed.Cir.1986).
The Federal Circuit listed other factors tending to rebut an argument of claim indefiniteness in Rosemount, 727 F.2d at 1547. The court noted that the party asserting invalidity had no trouble applying the terms of the claim to the prior art references. Id. In that case, one of the prior art references advanced by the party asserting invalidity used the same term that supposedly rendered the claim of the patent there in suit indefinite. The party asserting invalidity used the same term in describing its own products. A witness for the party asserting invalidity asserted that he had no trouble in understanding the claims of that patent. The examiner and the industry had both accepted the language of the claim. Also, the trial court had explicitly found the claim language to be as precise as the subject matter would permit.
As in Rosemount, the defendants in the case at bar introduced numerous prior art references and their witnesses never seemed to have trouble determining whether a particular component of a claim of the patent in suit was disclosed in the reference. The court finds that the language used in claim nine, and in particular the phrase “generally fixed in parallel relationship,” when read in light of the disclosure portion of the specification and the prior art, is sufficiently definite to apprise persons of ordinary skill in the art of both the *555utilization and scope of the invention and is as precise as the subject matter will permit.
In addition to these shortcomings in proof, defendants premise their claim indefiniteness argument on an incorrect legal theory. Defendants assert that because only claims four and ten contain a limitation requiring hollow portions in addition to fused portions in the pallet leg sidewalls, all the other claims are indefinite. Absence of a limitation, however, has a precise meaning, and does not render a claim indefinite. In re Fisher, 427 F.2d 833, 838 (C.C.P.A.1970). An inventor may include subcombination claims that are drawn to just one aspect or combination of elements so long as that combination has a utility separate and apart from other aspects or elements of the invention as a whole. Stiftung, 945 F.2d at 1181. A claim need not recite every element needed for practical utilization of the claimed subject matter, and it is entirely appropriate and consistent with the second paragraph of section 112 to present claims to only one aspect of an invention. Stiftung, 945 F.2d at 1181. Thus, it is error to reason that an inventor may not claim one part of his invention separately from the rest. Stiftung, 945 F.2d at 1181. The scope of a claim omitting a limitation is still definite so long as each recited limitation is definite. Fisher, 427 F.2d at 838.
In its pretrial materials, defendant asserts that claims nine through twelve are invalid “when read in light of the patent and prosecution history.” Certainly it is true, as noted above, that claims are read not in a vacuum, but in light of the application disclosure as it would be interpreted by one of ordinary skill. Moore, 439 F.2d at 1235. Defendant, however, misapplies this rule by looking to the specification for a definition of the invention, then criticizing the claims as being vague because they do not contain limitations directed to specific elements found in the embodiments taught in the disclosure portions of the specification, such as channels in the surface of the pallet’s load-bearing member. It is error, however, to look to the descriptive portion of the specification to determine what the invention is. Stiftung, 945 F.2d at 1181. The claims, not the disclosure, define the scope of the invention. Id.
Defendants’ fundamental mistake lies in asserting that a claim is indefinite because it is broader than the disclosure. Broadening the scope of a claim by omitting a limitation may raise questions of sufficiency of the disclosure under the first paragraph of section 112, but does not render an otherwise definite claim indefinite. Fisher, 427 F.2d at 838. The breadth of the claims, though important in assessing the sufficiency of the disclosure portion of the specification, is irrelevant to definiteness. Fisher, 427 F.2d at 838; see 2 Chi-sum, supra, § 7.03[7][b]. Thus, defendants’ argument concerning matter omitted from claim nine is irrelevant to the asserted defense of claim definiteness.
Defendants’ argument concerning overbreadth of the claims is more properly treated as support for a defense under section 112 paragraph one. That provision sets three requirements for the disclosure portion of the specification: the written description requirement, the best mode requirement,11 and the enablement require*556ment. Moore, 439 F.2d at 1235. Neither the written-description requirement nor the best-mode requirement are implicated by the breadth of the disclosure. The relevant inquiry, therefore, is whether the scope of enablement provided by the disclosure is commensurate with the scope of the claims. In re Bowen, 492 F.2d 859, 861 (C.C.P.A. 1974); Moore, 439 F.2d at 1236. To be sufficiently enabling, the disclosure must teach one of ordinary skill in the art how to make and use the claimed invention without undue experimentation. Telectronics, 857 F.2d at 785; Hybritech, 802 F.2d at 1384. The scope of the claims, therefore, must bear a reasonable relationship to the scope of the enablement that the specification provides to persons of ordinary skill in the art. Bowen, 492 F.2d at 861. The party asserting patent invalidity must present clear and convincing evidence as to why the specification does not reasonably permit one skilled in the art to practice as broadly as the invention is claimed. Telec-tronics, 857 F.2d at 785 (describing the burden as “persuasive” evidence).
The particular embodiments of the pallet illustrated in the disclosures contain features not found in claim nine. This does not mean that the disclosure is not enabling. One apparatus may embody several inventions. Stiftung, 945 F.2d at 1181: The law does not require the applicant to describe every conceivable embodiment of an invention. Telectronics, 857 F.2d at 786 (citing S.R.I. Int’l v. Matsushita Elec. Corp., 775 F.2d 1107, 1121 (Fed. Cir.1985)). Disclosure of one embodiment, along with the general manner of obtaining any information necessary to make and use other embodiments without undue experimentation, is sufficient. Id. 857 F.2d at 786. The patent need not teach, and preferably omits, that which is well known in the art. Hybritech, 802 F.2d at 1384. Thus, in the context of an application pros-
ecution it has been held error to reject the application merely because it did not teach a specific embodiment within the scope of an appealed claim. In re Vickers, 141 F.2d 522, 526 (C.C.P.A.1944). In arts involving predictable factors, such as patents in the mechanical or electrical arts, a single embodiment provides broad enablement because once imagined, other embodiments may be fashioned without undue experimentation. Fisher, 427 F.2d at 839; see Vickers, 141 F.2d at 527 (stating the rule for the mechanical arts); see also U.S. Dept. of Comm., Pat. & Trademark Off., Manual of Patent Examining Procedure § 706.03(z) (rev. 5th ed. 1989) (stating that in mechanical cases broad claims may be supported by disclosure of a single species).12 On the other hand, in arts involving unpredictable factors, such as chemistry and physiology, the requisite scope of enablement varies inversely with the degree of unpredictability of the factors involved. Fisher, 427 F.2d at 839; see also U.S. Dept, of Comm., Pat. & Trademark Off., supra, § 706.03(z). Thus, chemical reactions, which involve a high degree of unpredictability, require a more extensive enabling disclosure. Telectronics, 857 F.2d at 786. The specification, therefore, is deficient only if it does not enable as to some embodiments of the claim. Fisher, 427 F.2d at 839.
The design of pallets is an art involving largely predictable factors. The disclosure portion of the patent in suit amply illustrates why one embodiment may be adequate to enable a broad claim in the mechanical arts. Certainly, there are many conceivable embodiments of claim nine that the disclosure does not teach — pallets having decks without ribs, pallets without hollow portions in the leg sidewalls, etc. Nevertheless, once the inventors have shown a person of ordinary skill in the art one particular pallet design having fused por*557tions in the leg sidewall, other designs having fused portions in the leg sidewalls may be fashioned without undue experimentation. Thus, a fairly narrow disclosure may enable a comparatively broad claim.
The court concludes that the claim language is sufficiently definite reasonably to apprise those of ordinary skill in the art of both the scope and utilization of the claims. Moreover, the court finds that the language is as precise as the subject matter reasonably permits. The omission of limitations from the claim does not implicate concerns of claim definiteness, and because the art involves largely predictable factors, disclosure of a single species is enabling for a broad claim.
C. Nonobviousness.
Defendants admit that claim nine13 of the patent reads on the accused devices, but assert as an affirmative defense that the claim is invalid as obvious under section 103.14 All patents are presumed to be valid under section 282, and each claim is presumed valid individually. Vandenberg v. Dairy Equip. Co., Div. of DEC, 740 F.2d 1560, 1568 (Fed.Cir.1984). Defendants bear the burden of proving invalidity by clear and convincing evidence. Richdel, Inc. v. Sunspool Corp., 714 F.2d 1573, 1579 (Fed.Cir.1983); Lear Siegler, 733 F.2d at 885. This burden of persuasion remains on the defendant throughout the trial, American Hoist & Derrick Co. v. Sowa & Sons, Inc., 725 F.2d 1350, 1360 (Fed.Cir.), cert. denied, 469 U.S. 821, 105 S.Ct. 95, 83 L.Ed.2d 41 (1984), and is all the more onerous where the patent examiner has considered all the pertinent prior art, Kloster Speedsteel A.B. v. Crucible, Inc., 793 F.2d 1565, 1571 (Fed.Cir.1986), cert. denied, 479 U.S. 1034, 107 S.Ct. 882, 93 L.Ed.2d 836 (1987). Furthermore, the party challenging the patent’s validity bears the burden of proving that the examiner failed to consider a pertinent prior art reference. Richdel, 714 F.2d at 1579.
A patent is invalid for obviousness under section 103 “if the differences between the subject matter ... patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains.” Non-obviousness is a question of law. Panduit Corp. v. Dennison Mfg. Co., 810 F.2d 1561 (Fed.Cir.), cert. denied, 481 U.S. 1052, 107 S.Ct. 2187, 95 L.Ed.2d 843 (1987). Analysis of this legal question requires three factual inquiries: the scope and content of the prior art, the differences between the patented item and the prior art, and the level of ordinary skill in the art. Graham v. John Deere Co., 383 U.S. 1, 17, 86 S.Ct. 684, 693, 15 L.Ed.2d 545 (1966); Bausch & Lomb, Inc. v. Barnes-Hind/Hydrocurve, Inc., 796 F.2d 443, 447 (Fed.Cir.1986), cert. denied, 484 U.S. 823, 108 S.Ct. 85, 98 L.Ed.2d 47 (1987). The obviousness inquiry should be addressed to the invention as a whole, not just to the differences between the claimed invention and the prior art. Hybritech, 802 F.2d at 1383 n. 6.
1. Scope and Content of the Prior Art.
The art relevant to a consideration of obviousness is the analogous art. Wang Lab., Inc. v. Toshiba Corp., 993 F.2d 858, 864 (Fed.Cir.1993). Non-analogous art is too remote to constitute prior art. In re Clay, 966 F.2d 656, 658 (Fed.Cir.1992). Two criteria determine whether a particu*558lar reference is analogous art. First, if the reference is within the inventor’s field of endeavor, then it is deemed analogous. Wang, 993 F.2d at 864; Clay, 966 F.2d at 659; Bausch & Lomb, Inc., 796 F.2d at 449. Second, if the reference is reasonably pertinent to the particular problem with which the inventor was involved, it is prior art. Wang, 993 F.2d at 864; Clay, 966 F.2d at 659; Bausch & Lomb, Inc., 796 F.2d at 449; see also, 2 Chisum, supra, § 5.01. As the Federal Circuit recently held in Clay:
A reference is reasonably pertinent if, even though it may be in a different field from that of the inventor’s endeavor, it is one which, because of the matter with which it deals, logically would have commended itself to an inventor’s attention in considering his problem. Thus, the purposes of both the invention and the prior art are important in determining whether the reference is reasonably pertinent to the problem the invention attempts to solve. If a reference disclosure has the same purpose as the claimed invention, the reference relates to the same problem, and that fact supports use of that reference in an obviousness rejection. An inventor may well have been • motivated to consider the reference when making his invention. If it is directed to a different purpose, the inventor would accordingly have had less motivation or occasion to consider it.
966 F.2d at 659.
Thus, to determine whether a particular reference is within the prior art, the court must first determine the inventor’s goals and objectives. The court finds that the inventor’s goal was to design a stronger plastic pallet than previously had been known. This finding is based on all the evidence heard, but with particular reference to the following facts. Mr. Dresen, one of the named inventors, testified that although single sheet plastic pallets worked well in certain applications, the inventors “saw the potential of developing more structurally integrated pallets that would support products that would be heavier.” This perception of a need arose from contact with members of the industry. Mr. Dresen testified that Penda “had received, through our sales people, as well myself, a lot of requests for heavier pallets than we were able to make with the single-sheet methodology that we were using.” The inventors sought to meet these needs with a heavy duty pallet, of strength comparable to wood pallets, that would still offer the well known advantages of plastic over wood. Mr. Dresen testified that in their original meeting he and the other inventors concluded that “there was a real need in the market for a heavy-duty plastic pallet that could potentially compete with wooden pallets.” In addition to being strong enough to support the pallet, the leg had to withstand lateral impacts resulting from misaligned forklifts or jacks. Similarly, Harlon Breezer testified that the inventor’s main goal was to come up with a pallet that would be better than existing pallets by virtue of being stronger. Mr. Price also testified that the inventors came to recognize a commercial need for a stronger plastic pallet.
In particular, the inventors sought to develop a stronger pallet foot as part of a twin-sheet thermoform pallet. The ability of a pallet to bear up under a load is its compressive strength. In distinction, the tendency of a pallet to resist bending and flexing is its rigidity. That the inventors were particularly concerned with compressive strength is indicated by the steps undertaken in developing this pallet. The inventors designed the foot first, and tested it separately to determine whether it was sufficiently strong before proceeding to build a mold for the full pallet. The inventors testified, without contradiction, that it was unusual to test part of a design separately from the whole. If compressive strength had not been a particular concern, the inventors would not have spent so much time on the foot separately. Claim nine is the broadest because it teaches the most essential element of plaintiff’s patent — a stronger pallet foot integrally formed from a load bearing member consisting of two parallel surfaces.
*559Moreover, this finding as to the inventors’ objectives is supported by the language of the specification. “The present invention relates to pallets in general and, in particular, to pallets which can be constructed of thermoplastic materials formed by thermoforming methods.” (the ’306 patent, In. 1.) All the claims’ preambles identify the invention as a “pallet.” The specification contains the following description of the prior art:
The prior art is generally cognizant of the general concept and desirability of manufacturing pallets from moldable materials such as thermoplastic resins. Among the kinds of prior art pallets which were known is one general variety constructed of a single planar member of plastic material which serves as a load-bearing member, with feet formed to elevate the single layer of material. Examples of pallets of this general character can be found in U.S.Pat. Nos. 3,140,672, 3,640,299, 3,680,495, and 3,720,176. Of these patents, U.S.Pat. No. 3,140,672 discloses a molded pallet which includes a planar deck 11 having a plurality of ribs 15 provided in it to stiffen and reinforce it. Another method utilized in the prior art to construct pallets of moldable material has been to construct two substantially planar load-bearing members which are positioned in parallel to each other and spaced by reinforcing or stiffening members spaced so as to allow the projection of the arms of a forklift there-through. Examples of pallets formed of this general character can be seen in U.S.Pat. Nos. 3,404,642, 3,680,496, 3,691,-964, 3,699,902, and 3,757,704. Other examples of patents disclosing pallet configurations which can be constructed of molded materials can be seen in U.S.Pat. Nos. 3,561,375, 3,697,029, 3,717,922, and 3,719,157.
Defendants never challenged these statements of the inventors’ goals. In fact, the inventors’ testimony was consistent with defendants’ statements in their Memorandum of Contentions of Fact and Law. Nevertheless, defendants urge a substantially broader definition of the prior art, as encompassing all of thermoforming. In reliance upon this position defendants proposed many findings regarding developments in the thermoforming industry generally. The court does not consider this to be the inventors’ field of endeavor, however, nor does the court consider every ther-moformed article to be reasonably pertinent to the inventors’ problem of increasing the pallet’s compressive strength. Pri- or art either must fall within the inventors’ field of endeavor or be reasonably pertinent to the particular problem the inventor faced.
The court finds that the inventors’ specific field of endeavor was the creation of a stronger thermoformed plastic pallet.15 All thermoformed plastic pallets are, therefore, prior art. Other specific references, such as other types of pallets or other thermo-formed parts, may be prior art depending upon the specific facts of the reference and whether it is reasonably pertinent to the problem of making a stronger thermo-formed pallet.
Defendants have cited numerous prior art references. The list below, although not including every reference offered at trial, catalogs those prior art references that are most pertinent. Omission of any reference indicates not that the court failed to consider the reference, but merely that the court considered it less pertinent than those included.
a. The Bell Patent.
This reference discloses a plastic pallet made of one or more sheets of material. Bell, U.S. Letters Patent No. 3,640,-229 (1972). The parties agree that the patentees cited this reference to the examiner.16 This is a nestable pallet with ribs in *560the deck and in the legs to impart flex and compressive strength. Bell indicates that the load bearing capacity of the pallet is determined by the thickness of the material used. In an alternative embodiment, Bell suggests that a lighter pallet may be constructed by laminating a plastic exterior to a foam core. Figure 3 shows a cross-section of such a laminated sheet. The evidence tended to establish that the increase in strength-to-weight ratio that the Bell patent professes to provide is explicable in terms of the same underlying phenomena as that which Dr. Johnson used in explaining the strength of the invention’s deck. The court finds that this reference is pertinent prior art, but notes that the examiner duly considered it.
b. The Brown and Howell Patents.
These patents teach a twin-sheet thermoforming machine of the type plaintiff used in making their pallets. One possible application suggested in Howell is plastic pallets. Defendants contend that this machine served as a catalyst, making such pallets economically feasible. The reference, however, contains no teachings as to how such pallets can be made stronger. These patents teach specifically the technology required to make the forming machines themselves. They do not contain any instruction as to what techniques should be employed in designing parts for use on the machine. Although these references suggest that the machine disclosed can be used to make pallets, these references do not teach how to do so. Nothing in this reference would teach one of ordinary skill in the art in 1979 how to overcome the structural problems posed by plastic pallets. The court finds that this reference has a different purpose than the claimed subject matter, is directed toward a different problem, is not within the inventors’ field of endeavor, and is not reasonably pertinent to the problem they confronted.
c. The Toboggan and the Buening Patent.
Defendants offered a plastic toboggan as prior art. The toboggan is fairly conventional in shape and made of two sheets of material. The front consists of a large, enclosed, hollow chamber. The floor contains a number of hat sections extending along the major axis of the toboggan from the front hollow chamber to the rear rail. The sides are rails to which the riders may cling. The normal and intended use of the item was as a toboggan.
Mr. Woodrich, who developed the toboggan, testified that it was intended that the sidewalls be fused. He did not offer any explanation as to why this was desired. As Mr. McConnell pointed out, the purpose of the railing in the toboggan is not to carry heavy loads. Nothing indicates that one of ordinary skill in the art, endeavoring to increase the compressive strength of a pallet, would look to the handrails of a toboggan for guidance.
Buening, shown in Figure 4, teaches a plastic toboggan similar to the one entered in evidence as a physical exhibit. Buening, U.S. Letters Patent No. 3,666,282 (1972). The toboggan has a deck structure with ribs on opposite sheets running perpendicular to one another and fused together at the points of intersection. The toboggan has a raised periphery to serve as a rail, but the only fusing shown in this rail is in the horizontal plane. As with the actual toboggan put into evidence, there is no indication that one of ordinary skill in the art would consult this reference in searching for a solution to the problem of increasing the pallet’s compressive strength. The court finds that with respect to claim nine, *561which addresses primarily the pallet legs, neither the physical toboggan nor the toboggan patent were analogous prior art.
d. The Hammond Patent.
Hammond teaches the use of plastic columns to add compressive strength to a pallet. Hammond, U.S. Letters Patent No. 3,561,375 (1971). The examiner did consider Hammond, inasmuch as it is one of the references cited in the specification of the patent at issue. The columns inserted into the pallet by Hammond may consist of paired frustro-conical sections joined so that the center of the column is more narrow than the ends. The court finds that Hammond is pertinent prior art to the problem of imparting compressive strength to the pallet.
e. The Mcllwraith Patent.
Mcllwraith identifies his field of art as plastic pallets, and in particular, “pallets formed to individual, molded, thermoplastic top and bottom deck members.” Mcllwraith, U.S. Letters Patent No. 3,610,-173 (1969). The Mcllwraith pallet has an upper deck and a lower deck. Each of the two decks has criss-crossed channels “welded” together where they cross. Although the deck structure is susceptible to thermoforming, evidence adduced at trial established, and the court finds, that one of ordinary skill in the art at the date of invention would have understood Mcll-wraith to teach a deck manufactured by blow molding. The two decks are joined by interlocking columns that, in the preferred embodiment, are removable. Figure 5 illustrates this construction. Mcllwraith emphasizes that his pallet is reversible. Mcllwraith teaches that the columns joining the two decks may be cut from a tube of metal, or may be made from high-strength plastic. To increase compressive strength, Mcllwraith suggests using thicker walls. For shipping and storage, Mcll-wraith suggests that the pallet be disassembled.
Mr. Larry Stiller testified that he was generally familiar with the Mcllwraith type construction in twin sheet thermoform products. He described the deck as two series of hat sections, oriented at right angles to one another. He believed the top and bottom surfaces of each deck to be fused together, because the cross-sectional drawing only showed a single sheet of plastic where the two met. Mr. Stiller did not read the patent as suggesting blow molding. The court disagrees with his interpretation. Mr. Stiller conceded that Mcll-wraith did not teach a pallet with legs formed of the same sheet of material as the deck.
Mr. McConnell, plaintiff’s expert, also testified about this pallet. He described the structure as two blow-molded decks separated by columnar spacers. He recog*562nized that the deck taught the use of ribs in the upper and lower surface running criss-cross to each other so as to provide good strength. He saw nothing in Mcll-wraith that would suggest to him that he could make the pallet and the legs out of one sheet of material. According to Mr. McConnell, Mcllwraith requires separate forming operations for the deck and the legs, if the legs are made out of plastic at all. One endeavoring to build a heavy-duty-pallet of unitary construction who looked to Mcllwraith would learn only from the crisscross pattern in the deck, and probably would have known of that technique already.
The court finds that Mcllwraith is pertinent prior art, inasmuch as one of ordinary skill at the time of invention, when working on the problem of inadequate compressive strength, would tend to look to Mcllwraith when seeking a solution. Moreover, this reference was considered by the examiner.
f. The Rowlands Patent.
This reference, which was before the examiner, teaches a blow-molded pallet. Rowlands, U.S. Letters Patent No. 3,467,-032 (1969). The skin forms a top and bottom surface. The molding process deforms pillars in the top and bottom. These pillars are in the shape of frustro-conical sections. These pillars may then be welded together along their tops to provide strength and rigidity. Holes are then cut in the side of the pallet to provide an aperture for forklift tines. Rowlands is pertinent prior art for what it teaches about adding compressive strength through columnar inserts, but like Mcllwraith, it teaches nothing about generating compressive strength through fusing side walls.
g. The Szatkowski Patent.
Szatkowski teaches a layered pallet made of cardboard, pressed board, or the like. Szatkowski, U.S. Letters Patent No. 4,015,544 (1977). The walls for the legs are folded down out of the boards, then intermittently glued together. The pallet, which is nestable, may also have other reinforcements in the legs. After being folded down, the leg portions are fastened together by glue or some other suitable means to form a pallet foot, as shown in Figure 6. The examiner considered this reference when he determined that the inventors’ teaching of fusing the pallet-foot sidewalls together was nonobvious. The court finds that Szatkowski is pertinent prior art.
h. The Vissers Patent.
This pallet has an upper and a lower sheet. Legs depend from each sheet, and the legs of the top sheet fit into those of the bottom sheet. Vissers, U.S. Letters Patent No. 3,267,883 (1965). Mr. Stiller read the phrase “locked against displacement” to mean that the two sheets are fused. Nothing in the patent, however, suggests that these plates are joined at the feet. The only place the patent teaches uniting the plates is at the circumferential rim. Mr. McConnell testified that the specification suggested to him that the sheets were made of a material similar to fiberglass, then snapped together. To perform such a process, the upper sheet would be formed on a female mold and the lower sheet on a male mold, so that the surface of each sheet which was to come in contact with the other sheet would be in contact with the mold and thus more precisely defined. On the whole, the court finds that the Vissers patent is less pertinent a prior art reference than Szatkowski.
i. The Woolworth Patent.
Woolworth teaches a pallet leg made of high-impact plastic. Woolworth, U.S. Let*563ters Patent No. 3,438,342 (1969). Figure 7 is an illustration of this pallet foot, taken from the Wool worth patent. The pallet leg is made of two concentrically situated cylinders connected together by a webbing structure, consisting essentially of radially disposed panels. The bottom of the leg is sealed by another plastic part, labeled in Figure 7. The top of the leg is designed to be bolted to the pallet.
In trying to establish Wool worth as prior art, defendants elicited testimony from Mr. Stiller as to whether he could design a mold to produce this part by thermoforming. Presumably, defendants recognized that in developing a thermoformed pallet, one of ordinary skill in the art would most naturally turn to references that could themselves be thermoformed.
Mr. Stiller’s testimony was more revealing on this subject than defendants might have wished. His point of departure was a confident assertion that he could in fact design a mold to produce such a part by twin sheet thermoforming. As the court observed his verbal and demonstrative description of how he would do so, however, it grew increasingly apparent that the design he was describing simply would not work. Although difficult to articulate, the deficiencies in his design were absolutely clear to the court when hearing his testimony ore tenus and observing his models, his attempts to illustrate his approach, his gestures, and his ultimate frustration. Nor was this an arcane matter, beyond the ken of the lay observer. The court can best relate the deficiencies inherent in Mr. Stiller’s proposal by explaining that to implement his design, the upper mold surface would have to physically pass through the lower mold surface. It requires no specialized knowledge to realize that two objects cannot occupy the same space at the same time. Mr. Stiller tenaciously refused to concede the shortcomings of his proposed design, but, under the pressures of cross-examination, did ultimately abandon this first attempt. The court appreciates that the design process may be iterative in nature. Nevertheless, the inescapable conclusion for anyone observing Mr. Stiller’s testimony was that the conversion of the Woolworth foot from blow molding to twin-sheet molding was far less obvious than he originally thought.
2. Differences Between the Patented Invention and the Prior Art.
Keeping in mind the problem that the inventors confronted — that of enhancing the compressive load-bearing features of plastic pallets — the court now considers the differences between the way in which the inventors accomplished this result and the way the prior art attempted to solve this problem. The prior art must be considered not only for what it may teach toward the patent, but also for the manner in which it may lead one of ordinary skill in the art away from the invention.
Defendants contend that Bell teaches connecting two sheets of plastic in the leg to produce a stronger result. Bell, however, does not teach vertical fusing to add compressive strength. The two laminate sheets in Bell are fused not to each other, but to a central core of foam. The structure of Bell is more similar to the use of inserts in the foot to impart compressive strength. Bell does not teach the use of different designs on the top and bottom. That Bell may be explained by the same principles as explain the patented pallet is irrelevant. In re Van Wanderham, 378 F.2d 981, 988 (C.C.P.A.1967) (observing the difficulties attendant upon not considering *564the subject matter as a whole, but instead focusing on the scientific principles involved.); see also Lindemann Maschinen-fabrik G.M.B.H. v. American Hoist & Derrick Co., 730 F.2d 1452, 1462 (Fed.Cir.1984) (noting that most patentable inventions employ known principles). Dr. Johnson, in analyzing the structure of the patented pallet, explained that tying the upper and lower sheets of material together in the deck provides a shear transfer mechanism, an effect he compared to that produced in an I-beam. The concept of the strength inherent in an I-beam is, however, ubiquitous. The proper inquiry under section 103 is addressed to whether the particular application of this principle taught by the inventors, which is to say the claimed subject matter as a whole, would have been obvious to one of ordinary skill in the art at the time the invention was made. Bell does not teach a fusing of the layers. If the patented pallet is nonobvious over Szatkow-ski, which teaches gluing parts of the leg together, it is nonobvious over Bell.
One method the prior art teaches for achieving greater compressive strength is the insertion of plastic columnar struts between two decks of a reversible pallet. E.g., Hammond, U.S. Letters Patent No. 6,561,375 (1971); Rowlands, U.S. Letters Patent No. 3,467,032 (1969). The struts in Hammond are not integrally formed with the decks, but are the product of a separate forming operation. Hammond teaches the use of struts comprising paired frustro-conical sections for compressive strength; this use of paired frustro-conical sections is very different from plaintiff’s techniques of fusing portions of the sidewalls together in the pallet legs. Hammond is only applicable for a two-deck pallet, which would not have satisfied the inventors’ goal of nestability.
Similarly, Rowlands teaches the use of columns between the decks to provide support. Unlike Hammond, these columns are formed from pockets deformed in each sheet of the two deck pallet. The columns are therefore an integral part of the structure, and in this regard are similar to the patent in suit. This solution is very different, however, from the one taught by the inventors. The columns in Rowlands are of a single thickness throughout. It would not have been obvious to one of ordinary skill in the art upon looking at Rowlands or Hammond that significant increases in compressive strength could be achieved by fusing together portions of the sidewalls in the legs.
Another technique for gaining compressive strength is the use of inserts. Row-lands, for example, teaches that the pockets in either surface which compose the spacer columns may be formed from the insertion of a frustro-conical insert. Mcll-wraith teaches making the columnar supports of a different material, such as steel tubing.
The most common technique found in the prior art for adding compressive strength was to use a thicker sidewall. It was from exactly this confine of thought that the inventors escaped. By fusing the two vertical sidewalls, the inventors achieved the effect of a thicker sidewall without the attendant disadvantages: greater weight, more expense, and formability problems.
No witness testified as to Szatkowski at trial. Plaintiff contends that Szatkowski, which the examiner considered, is a more pertinent reference than Vissers, which defendants advance. The fusing of the sidewalls as shown in the patent was considered by the examiner to be a non-obvious improvement over the adhesion taught by Szatkowski. Vissers teaches locking the legs together against displacement. The two plates are “united”, however, only at the circumferential rim. Plaintiff’s argument in closing, that this reference is less apt than Szatkowski, is well taken. To strengthen the legs, Vissers teaches the use of material of greater thickness. This teaching in no way suggests the solution that the patentees hit upon. The court finds that one of ordinary skill in the art in 1980 would have interpreted the phrase “locked against displacement” to call for some kind of mechanical fit to transfer shear from the upper to the lower sheet. Mr. Stiller’s interpretation, under which Vissers would read on the patent claims, *565results only from impermissibly reading the teachings of the patent back into the prior art.
3. The Level of Ordinary Skill in the Art.
The necessary inquiry when determining obviousness is whether the invention would have been obvious to “one of ordinary skill in the art,” not whether it would have been obvious to the court, to a layman, to those skilled in remote arts, or to geniuses in the art at hand. Environmental Designs, Ltd. v. Union Oil Co., 713 F.2d 693, 697 (Fed.Cir.1983), cert. denied, 464 U.S. 1043, 104 S.Ct. 709, 79 L.Ed.2d 173 (1984). Six factors have been held relevant in determining the level of ordinary skill in the art: (1) the educational level of the inventor, (2) the type of problems encountered in the art, (3) the prior art solutions to those problems, (4) the rapidity with which innovations are made, (5) the sophistication of the technology, and (6) the educational level of active workers in the field. Bausch & Lomb, Inc., 796 F.2d at 449; Environmental Designs, 713 F.2d at 696. Not every case will contain all factors, and in a particular case one or more factors may predominate. Environmental Designs, 713 F.2d at 697.
a. Educational Level of the Inventors.
The court considers this criterion as a direction to consider all aspects of the inventors’ education, including informal education and years of practical experience. In 1980, Mr. Dresen had eleven years experience with thermoforming and was involved in the production of single-sheet pallets at both Penda and Glasfab, its predecessor. He worked for Glasfab starting in 1969, then went to work for Penda when they purchased Glasfab. Mr. Dresen earned a B.S. degree in industrial technology and industrial education. Mr. Breezer began working while in high school and has no formal education beyond high school but has sixteen years thermoforming experience. Mr. Breezer worked for Portage Plastics for eleven years, then worked at a tooling house. He joined Penda in 1978. While he was at Portage, the company produced a single-sheet pallet. Mr. Price, in addition to earning a two-year associate degree at a junior college, had fourteen years thermoforming experience and additional experience in the tool and die trade. He worked first with metal-stamping tools, then started a pattern and model shop called Arborcraft. After working there eleven years, he consulted for three years before joining Penda. He had previously designed single and twin-sheet products, including a single-sheet pallet. The inventor’s educational level, however, is not conclusive of the level of ordinary skill in the art; it is but one factor. Bausch & Lomb, Inc., 796 F.2d at 449; Orthopedic Equipment Co. v. All Orthopedic Appliances, Inc., 707 F.2d 1376, 1382 (Fed.Cir.1983).
h. The Educational Level of Active Workers in the Field,
The court considers the term “education” in the context of this criterion to refer not only to formal education, but also to informal education and practical experience. Mr. Waun, who designed the accused devices, studied mathematics and engineering at four colleges without ever receiving a degree. He went to college about a year before entering the thermoforming business. He completed a five-year tool-and-die apprenticeship during which he served under Mr. Price. He later worked for Mr. Price’s company, Arborcraft, for three years, during which time he was a foreman for Mr. Price. He worked about eight years as head designer for Shuert Industries, a toolmaker that later became a ther-moformer. In 1980, Mr. Waun had approximately ten years experience in thermo-forming and additional experience in the tool-and-die trade. Mr. Constantino, the president of Cadillac, got his start in the thermoforming business straight out of high school. Larry Stiller, defendants’ expert, entered the thermoforming industry directly from high school. Mr. Stiller further testified that others he knew in the business had backgrounds similar to his own. Mr. McConnell, an expert for plaintiff, had about two years of college in aeronautical engineering, but never com*566pleted his degree after serving in World War II. The other experts who testified, Dr. Johnson and Dr. Singh, both have academic careers. The court finds that neither is an active worker in the field in the sense that his background might be useful in assessing the level of ordinary skill in the art.
c. The Types of Problems Encountered in the Art.
The inventors’ precursors in the art of making plastic pallets found it difficult to generate sufficient compressive strength. The plastic pallets that were generally commercially available had problems with sagging and hinging. Additionally, twin-sheet thermoformers encountered problems in making the molds fit together properly to get a good fusing. Makers of pallets in general faced problems in producing durable, reusable pallets cheaply.
Certain articles that defendant offered as prior art17 are instructive as to the types of problems encountered in the art. As early as 1970, industry watchers were suggesting thermoforming plastic pallets. Michael J. Steed, Why Not Thermoform Plastic Pallets?, Plastics Tech., Feb. 1970, at 43. Even at that early date, some recognized that the principal limitation on plastic pallets lay in developing the proper leg design. Steed indicated that research into different leg designs was already underway.
A later article recounts some of the problems encountered with early plastic pallets. See Ronald R. MacBride, Plastic Pallets are Coming on a Lot Stronger Now, Modern Plastics, Oct. 1972, at 60. MacBride recounted how early plastic pallets had been rushed to the market too quickly, and had to be withdrawn because of unanticipated problems. At that point, plastic pallets still could not compete with wood either in terms of rigidity or compressive strength. Returnables, even in 1972, cost over $20 per unit. A later article in a different journal confirms that plastic pallets were having trouble making inroads, and that twin-sheet pallets then in use were not exceptionally strong. Some Straight Talk About Plastic Pallets, Automation, Nov. 1974, at 70.
Two years later, MacBride followed up on his 1972 article. See Ronald R. MacBride, Despite the Odds, Plastic Pallets are Set To Grow, Mod. Plastics, Nov. 1974, at 58. In 1974, plastic pallet designers were still looking for ways to make their products stronger. At least three dozen companies had made major investments in developing a better plastic pallet.
d. The Prior Art Solutions to Those Problems.
Prior art plastic pallets included single-sheet pallets with channels and ribs to add strength, both in the deck and in the legs. Although offering the advantages associated with plastic pallets, these single-sheet pallets were generally too flimsy for heavy use. Prior art twin-sheet plastic pallets contained alternating fused and hollow portions in the deck and patterns of ribs in the upper and lower surfaces intersecting at right angles, but not fused or alternating fused and hollow portions in the leg walls. Others offered injection-molded legs bolted to thermoformed decks. The use of inserts in the legs was a common means of increasing the compressive strength. Some prior art references also suggest increasing the width of the leg wall.
e. The Rapidity with Which Innovations Are Made.
The parties stipulated that “innovation did not occur rapidly in the industry.” Mr. Stiller testified at trial that about the time the inventors were designing their pallet, those in the art were “getting creative” as the effects of the oil embargo eased. That observation, however, does not call into question the stipulation. The problem the inventors were working on — designing a *567stronger foot — had been recognized at least since 1970. Steed, supra, at 43.
f. The Sophistication of the Technology.
The basic technology required to practice this art had been available for almost twenty years. All the equipment needed to practice this technology was commercially available. The molds for making products probably could have been fashioned by a number of toolmakers. This fact, considered in conjunction with the educational level of the inventors and other workers in the field, suggests that the relevant technology was at most modestly sophisticated or relatively unsophisticated.
g. Finding as to Level of Ordinary Skill.
One of ordinary skill in the art of designing and thermoforming plastic parts such as pallets would have had little if any formal education beyond high school. One of ordinary skill would have learned the trade principally from an apprenticeship or similar program of informal study. Although one of ordinary skill would have been familiar with the general principles and basic concepts of thermoforming, including the use of hat sections and furrows in the plastic to increase rigidity and fusing two sheets together in a thermoformed part, the court finds that this general familiarity does not mean that one of ordinary skill in the art would have chosen the proper design to integrate these principles successfully to solve a given problem such as that which confronted the inventors. On the whole, the level of ordinary skill in the art is relatively low.
4. Objective Indicia of Nonobviousness.
Secondary considerations such as commercial success and long-felt but unsolved need may also aid in the nonobviousness inquiry. Graham, 383 U.S. at 17, 86 S.Ct. at 693. Where available, such objective evidence of nonobviousness may be the most probative evidence presented, Bausch & Lomb, Inc., 796 F.2d at 447, and must be considered. Newell Cos. v. Kenney Mfg., 864 F.2d 757, 768 (Fed.Cir.1988), cert. denied, 493 U.S. 814, 110 S.Ct. 62, 107 L.Ed.2d 30 (1989).
a. Commercial Success.
Defendants contend that plaintiff’s failure to put on explicit proof of commercial success is relevant in a determination of obviousness. As a preliminary matter, the court notes that plaintiff has no obligation to offer such proof. The patent is presumed valid under section 282. Defendants, not plaintiff, bear the burden of proof with respect to the question of validity. Furthermore, even if defendants had proven the device a commercial failure, that would not be sufficient to show that it was obvious and unpatentable.
Commercial success suggests patentability because (1) the economic return generated by an item is some indication of its worth to society, (2) others seem to be acknowledging the invention’s validity because they are willing to pay for the invention rather than defend a suit for patent infringement, and (3) if others had seen how to avail themselves of this commercial opportunity, they would have done so. None of these factors, however, warrants application of the rule’s converse. In other words, lack of commercial success does not in and of itself tend to show obviousness.
An invention need not be better, less expensive, or more commercially appealing to be nonobvious. See Demaco Corp. v. F. Von Langsdorff Licensing, Ltd., 851 F.2d 1387, 1390 (Fed.Cir.), cert. denied, 488 U.S. 956, 109 S.Ct. 395, 102 L.Ed.2d 383 (1988); Harmon, supra note 6, at 114. In any event, the record is not devoid of indicia of commercial success. This procurement itself was a substantial commercial coup. Moreover, plaintiff previously had sold as many as 66,000 pallets in one order, to a customer who purchased a total of 132,000 pallets in a year.
Of course, commercial success is only probative of nonobviousness to the extent that a nexus exists between it and the patented aspects of the product sold. *568Cable Elec. Prod,., Inc. v. Genmark, Inc., 770 F.2d 1015, 1026 (Fed.Cir.1985). There is, however, ample evidence of the merit of the Penda pallet. The Postal Service conducted extensive tests of available pallets. It purchased 20,000 pallets of the type covered by this patent for testing in the field. After these tests, the Postal Service concluded that the Penda design was superior to that of its competitors. Penda’s design was strong and durable. As defendants’ expert, Mr. Saunders, observed in regard to damages, this pallet is the type of product in which the commercial embodiment and the invention are largely coextensive. Just as apportionment of damages is not appropriate in such a situation, it is appropriate to impute the commercial success to the structures patented. See Fromson v. Western Litho Plate & Supply Co., 853 F.2d 1568, 1570 (Fed.Cir.1988); Georgia-Pacific Corp. v. United States Plywood Co., 318 F.Supp. 1116, 1132 (S.D.N.Y.1970), modified, 446 F.2d 295 (2d Cir.), cert. denied, 404 U.S. 870, 92 S.Ct. 105, 30 L.Ed.2d 114 (1971).
b. Long-Felt but Unsolved Need— Failure of Others.
Plaintiff did not offer evidence as to any long-felt need for more durable plastic pallets. The inventors did testify that they believed this need existed, but their testimony did not establish how long others may have felt this need. Prior art publications introduced by defendant, however, establish that this need was long-felt and unsolved. The need to develop a better leg design was felt as early as 1970. See Steed, supra, at 43. Those in the industry were working on the problem at that point. Id. A later article recounts the initial failure to solve this problem. MacBride, Plastic Pallets Are Coming on a Lot Stronger Now, supra, at 60. MacBride teaches that plastic pallets were rushed to the market too quickly and had to be withdrawn, and that plastic pallets could not compete with wooden pallets in either compressive strength or rigidity. Id. Two years later, plastic pallets were still having trouble in the market, and twin-sheet pallets were still not very strong physically. Some Straight Talk About Plastic Pallets, supra, at 70. By 1974, at least three dozen companies had made major investments in making pallets stronger. MacBride, Despite the Odds, Plastic Pallets Are Set To Grow, supra, at 58.
Defendants contend that plaintiff’s design would have been suggested by someone else much earlier, except that the oil embargo drove the price of plasties so high that plastic pallets were not a good investment in the mid-seventies. The articles defendants themselves offered for this point, however, show that for at least a four-year span in the early seventies, the plastics industry was working hard on developing a stronger plastic pallet. Even if the Brown machines were what eventually made such pallets commercially viable, they had been in use since the mid-seventies. Yet no one in the industry developed a plastic pallet like the one at issue before the inventors did so. The court finds that the plaintiff’s invention did fulfill a long-felt but unsolved need, and rejects the notion that the particular structure developed by the inventors was an inevitable consequence of the particular machine used.
c. Reaction of Experts.
Mr. McConnell testified as an expert in thermoforming. He has worked for many years in the field, and teaches seminars on thermoforming to various workers. His experience includes some design work and all aspects of the thermoforming industry.18 Mr. McConnell said that he first saw the Penda pallet soon after it was introduced: “[FJrankly I was amazed at what I considered to be an ingenious way [of] increasing load bearing structure with the *569use of those vertical fused sidewalls in the feet.” Later in his testimony he elaborated:
Q: What significance, if any, did you attribute to the Penda nestable twin sheet thermoform pallet when you became aware of it?
A: Well, I felt it opened up another avenue for the pallet manufacturers, ... the next category of weight carrying pallets.
But I also felt that it would help the industry in general, all of a sudden let us know, hey, we’ve been overlooking something for years, and that’s fusing vertical sidewalls for structure, because there are a lot of other items besides pallets [for which] this can be an advantage.
d. Other Secondary Considerations.
Copying may also be evidence of nonobviousness. Avia Group Int’l, Inc. v. L.A. Gear California, Inc., 853 F.2d 1557, 1564 (Fed.Cir.1988); Chisum, supra, § 5.05[5]; Harmon, supra note 6, at 102. Plaintiff urges the court to find that Mr. Waun in fact,copied the patented pallet. Clearly he had access to plaintiff’s pallet. Plaintiff introduced evidence of shipments to WolPac of samples of their pallets. Access and similarity, however, are the elements for proof of copying in a copyright case, not a patent case. These facts standing alone might not indicate that any copying occurred. Regardless, the court considered dubious the testimony of Mr. Waun on this point. It seems at best highly improbable that his employer, a small company, would simultaneously take a rush design job on a twin-sheet thermoform pallet and conduct extensive tests on a large number of such pallets without using that testing knowledge in the designs. Although these two functions may have been located in different buildings, inasmuch as the company owner was involved in both projects, it is difficult to imagine the right hand not knowing what the left hand was doing to the extent indicated by Mr. Waun. The court finds Mr. Waun’s assurances to the contrary not entirely credible.
Acquiescence in licensing the patent may be objective evidence of nonobviousness. 2 Chisum, supra, § 5.05[3]; see Eibel Process Co. v. Minnesota & Ontario Paper Co., 261 U.S. 45, 55-56, 43 S.Ct. 322, 325, 67 L.Ed. 523 (1923). Since this action was commenced, Cadillac has negotiated a license from Penda. Although a clause in the license indicates that the parties to the license did not consider it evidence of an established royalty, that agreement does not mean that the license may not be considered as evidence of some level of acceptance. The license was introduced into evidence by the government, which insisted that the court should allow it to use the license as evidence of a reasonable royalty since it was not a party to the license and did not stipulate that the license should not be used as evidence of an established royalty. Although other factors beside acceptance of patent validity may have contributed toward Cadillac’s entering this license, those same factors would limit the relevance of the agreement with respect to the amount of a reasonable royalty. Although of minimal probative value, the court does consider Cadillac’s acceptance of this agreement to be at least some evidence of the patent’s validity.
Likewise, laudatory statements regarding the patented invention may be some evidence of nonobviousness. See Akzo N.V. v. United, States Int’l Trade Comm’n, 808 F.2d 1471, 1481 (Fed.Cir. 1986), cert. denied, 482 U.S. 909, 107 S.Ct. 2490, 96 L.Ed.2d 382 (1987). But see 2 Chisum, supra, § 5.05[5] (criticizing this use). The Postal Service was apparently quite impressed with the Penda pallets, which embodied the invention. The evidence at trial clearly showed that the Postal Service thought the Penda pallets superi- or to any other they had used, albeit more expensive.
e. Finding as to Secondary Considerations.
The secondary considerations, the so-called objective indicia of nonobviousness, on balance favor patent validity.
*5705. Conclusions as to Nonobviousness.
Defendants have failed to prove by clear and convincing evidence that the Penda patent is obvious within the meaning of section 103. The court concludes that the differences between the subject matter patented and the prior art are such that the subject matter as a whole would not have been obvious to one of ordinary skill in the art to which the subject matter pertains. In reaching this conclusion, the court has first considered and made findings about the scope and content of the prior art, the differences between the prior art and the claimed invention, the level of ordinary skill in the art, and objective indicia of nonobviousness. Much of the prior art discussed at trial had been presented to the examiner during the prosecution of this patent. Most of that which was not was less pertinent than that which was.
Defendants have failed to provide clear and convincing evidence that the examiner erred in his determination that the claimed subject matter was nonobvious over Bell, Hammond, Mcllwraith, Rowlands, Szat-kowski, and Vissers. Although the record does not indicate explicitly that the examiner considered Woolworth,19 the court finds that this prior art reference is no more pertinent than those it is known the examiner did consider.
Defendants have urged the court to declare plaintiff’s patent invalid based on a combination of prior art references. To this end, in their pretrial memorandum defendants cited Orthopedic Equipment Co. v. United States, 702 F.2d 1005, 1013 (Fed. Cir.1983), for the proposition that features may be obvious in view of a combination of references, even if the features of one reference cannot be substituted physically into the structure of another. Defendants further urged the conclusion that the claims are invalid as obvious because they are “mere permutations of known techniques.” Defendants cited no authority for the latter standard.
The law on this point is well settled. There is no special test for inventions that are combinations of old elements, because virtually all inventions are combinations. Intel Corp. v. United States Int’l Trade Comm’n, 946 F.2d 821, 842 (Fed.Cir. 1993); Harmon, supra note 6, at 84. One might as well contend that a book contains nothing new because it coins no new words and is thus merely a combination of old elements. Furthermore, it is permissible to combine prior art references only if the prior art itself suggests the desirability of such a combination. Panduit Corp., 810 F.2d at 1568; 2 Chisum, supra, § 5.04[l][e], Defendants have failed to prove by any measure, let alone prove by clear and convincing evidence, that the prior art does suggest such a combination.
In the preceding sections, the court has determined the scope and content of the prior art, the differences between the claimed subject matter and the prior art, the level of ordinary skill in the art in 1980, and the import of the objective indicia of nonobviousness. The level of ordinary skill in the art is fairly modest in comparison with many technological arts. The differences between the claimed subject matter and the prior art are substantial. The claimed subject matter as a whole would not have been obvious to one of ordinary skill at the time the invention was made.
D. Reasonable and Entire Compensation
The plaintiff has proven infringement, and defendants have failed to prove their affirmative defense of invalidity. We therefore must determine the amount of plaintiff’s “reasonable and entire compensation” for the government’s use of the patented invention. Both sides essentially agree on the methodology for determining *571damages, but sharply disagree on an appropriate amount.
Both sides presented evidence as to damages primarily through expert witnesses. The witnesses who testified for plaintiff on the subject were Mr. James Malikowski and Mr. Howard W. Bremer, Esq., both of whom testified as experts, and Mr. Jack Bauman. Defendant presented the testimony of Mr. Robert Saunders, Esq., as an expert.
Mr. Malikowski is a vice-president and a principal of the I.P.C. Group, a financial consulting firm that focuses in two primary areas: litigation support and business valuation. He has testified on intellectual property matters more than 100 times. Mr. Malikowski holds degrees in both accountancy and philosophy, and is a certified C.P.A. He is a member of the Licensing Manufacturing Association and numerous accounting organizations. He is on the board of trustees of the Licensing Executive Society and the board of the Chicago chapter of the National Organization of Accounting. He worked for Peterson & Company evaluating intellectual property and has previously evaluated licenses for patents. He also serves as a guest lecturer at John Marshall School of Law as an expert in patent damages and reasonable royalties. Mr. Malikowski claims no particular expertise in pallet design or thermoform-ing. He was retained by Penda to determine the value of a reasonable royalty in this case. Dr. Lewis Koppell of I.P.C. Group assisted him in performing this evaluation. Dr. Koppell has an M.B.A. and a chemistry degree. Mr. Malikowski opined, on the basis of a willing-licensor/willing-licensee hypothetical negotiation, that a reasonable royalty would be in the range of $3.00 to $3.50 per pallet.
Since 1988, Mr. Bremer has worked as a consultant in patent and licensing matters. After receiving his undergraduate degree from the University of, Wisconsin-Madison in chemical engineering, in 1944, and after serving in the Navy for approximately two years, he returned to the University of Wisconsin and attended law school, from which he graduated in 1949. He worked first as a patent attorney in the Patent Division of Proctor and Gamble Company, Cincinnati, Ohio. His work there included patent and licensing issues involving royalties. During his twelve years with Proctor & Gamble, Mr. Bremer participated in forty to fifty license negotiations. When he left, Mr. Bremer was responsible for all of the patent work on soaps and detergents, which included patents on packaging machinery, packages, the detergents themselves, and the processes for making them. He also participated in monitoring any preexisting license agreements.
After leaving Proctor & Gamble, Mr. Bremer worked for the Wisconsin Alumni Research Foundation (WARF)20 from December of 1960 until retiring in 1988. For the first fifteen years of his employment with WARF, Mr. Bremer was its sole patent counsel. As such, he evaluated all the faculty’s inventions, patented them, and licensed them out to industry. He participated both in negotiating several hundred new licenses and in monitoring existing ones. The bulk of his work was licensing out, because an entity in WARF’s position would rarely license in. Because the University research function covers almost all disciplines, WARF dealt with any subject matter that arose, ranging in origin from the engineering schools to the biochemistry department. The inventions brought to WARF included very simple mechanical items, complex mechanical contraptions, electronic devices, and much from the chemical art.
Plaintiff called Mr. Jack Bauman as a fact witness. Mr. Bauman’s background is presented because it is relevant in considering certain licenses he negotiated which plaintiff introduced into evidence. He received his undergraduate degree from Colgate University in 1950. He also attended *572Harvard Business School in 1972. Mr. Bauman worked twenty-eight years for Union Camp Corporation, a manufacturer of paper and related converted paper products. He was General Manager of the Bag Division when he resigned on August 1, 1978. Mr. Bauman estimates that he participated in seven or eight negotiations to license products during his last seven or eight years at Union Camp.
After then serving as president of P.C.L. Packaging for three years, Mr. Bauman consulted on several projects and set up United States sales organizations for two Canadian companies, one of which was in the thermoform business and one of which was in the plastic flexible packaging business. His last consulting project, which began in February, 1984, was with Bigelow Sanford (“Bigelow”), with which he remained until 1990 when he took his present position. At Bigelow, Mr. Bauman reported directly to the Chairman of the Board. In January 1985, Bigelow created a separate entity called Bigelow Packaging, which did business under the name Big-Pak. Mr. Bauman was made General Manager of that division, where his principal responsibilities were to commercialize the package covered under a patent Bigelow owned for a returnable, reusable, bulk container system comprising mirror image top and bottom single sheet pallets separated by a triple, uncorrugated sleeve. In developing this package, Mr. Bauman participated in four or five license negotiations.
When Mr. Bauman took over the division in January of 1985, it owned a number of plastic pallet molds scattered among several thermoformers. These thermoformers’ products were not of uniform quality. Also, the pallet was not sufficiently rigid for larger applications. Mr. Bauman testified that, before deciding to deal with Pen-da, he surveyed most of the thermoformers in the United States, first for their ability to get a consistent quality in the existing pallets, and second for solutions to the problem of excessive flexing. Mr. Bauman then negotiated a license with Penda for the patent in suit.
Defendants called Mr. Robert J. Saunders, Jr., as an expert witness. Mr. Saunders completed the basic engineering program at Rose Polytechnic Institute while in the Army during World War II, graduated from Fordham University in physics in 1948, and received his J.D. from Fordham Law School in 1951. He worked first in the legal department of General Electric, principally in the radio and T.V. department and some related product lines. He handled all phases of the patent work, including procurement or evaluation of third-party patents, and negotiated one or two license-in patents and no license-out patents. He then joined the Texaco patent department, where he worked until 1972. As a patent attorney for Texaco, Mr. Saunders handled inventions in electronics and physics, principally nuclear well logging, data processing, or any invention other than primarily chemical inventions. He later worked heavily in licensing technologies in which a petroleum company would be involved, including electronics inventions, seismic data processing, nuclear well logging, and fuel additives.
After leaving Texaco, Mr. Saunders served with Research Corporation. From its formation in 1987 until 1989 he also worked for an affiliate, Research Corporation Technologies, in Tucson, Arizona. At both corporations, he served as general counsel and retired as corporate secretary. As general counsel of both corporations, Mr. Saunders spent about ninety-five percent of his time on the practice of patent law and technology transfer, developing programs to license or otherwise commercialize technology. The corporations evaluated inventions to assess potential for generating revenue. Ordinarily, patent protection would be sought, followed by licensing out. After seeking patents through outside counsel, the corporations negotiated the license agreements themselves. Mr. Saunders testified that he participated extensively in the licensing aspect of the business. The practice encompassed diverse arts. Since retiring, he has worked as a consultant in the private practice of patent law, mostly in the field of technology transfer and the transfer of patents and know *573how. He also does some consulting work on the prosecution and preparation of patents.
1. General Principles.
The government may take a license in any United States patent. Decca, Ltd. v. United States, 640 F.2d 1156, 1166 (Ct.Cl.1980), cert. denied, 454 U.S. 819, 102 S.Ct. 99, 70 L.Ed.2d 89 (1981); Calhoun v. United States, 453 F.2d 1385, 1391 (Ct.Cl. 1972). The patentee’s sole remedy is a suit in this court for its “reasonable and entire compensation” under section 1498. Decca, 640 F.2d at 1166. This remedy is premised on a theory of eminent domain. Motorola, Inc. v. United States, 729 F.2d 765, 768 (Fed. Cir.1984); Leesona Corp. v. United States, 599 F.2d 958, 966 (Ct.Cl.), cert. denied, 444 U.S. 991, 100 S.Ct. 522, 62 L.Ed.2d 420 (1979); Tektronix, Inc. v. United States, 552 F.2d 343, 346 (Ct.Cl.), modified, 557 F.2d 265 (Ct.C1.1977), cert. denied, 439 U.S. 1048, 99 S.Ct. 724, 58 L.Ed.2d 707 (1978). The property interest taken by the government is a compulsory, compensable, nonexclusive license. Motorola, 729 F.2d at 768; Leesona, 599 F.2d at 968.
The “reasonable and entire compensation” owed the patent holder is equivalent to just compensation under the fifth amendment. Calhoun, 453 F.2d at 1395. Thus, the court has traditionally turned to the law of eminent domain for legal precedents to apply in determining “reasonable and entire compensation.” Leesona, 599 F.2d at 967. Basic principles of fairness are the governing consideration in determining just compensation for an eminent domain taking. Tektronix, 552 F.2d at 347. These same principles guide in assessing reasonable and entire compensation. Id.
This court and its predecessors have employed several different methods in fixing the amount of reasonable compensation due one whose patent the government has infringed. Three of the measures the court has used are (1) a reasonable royalty,21 (2) the patent owners’ lost profits, and (3) the savings to the government. Decca, 640 F.2d at 1167. One way to monitor the reasonableness and fairness of an award is to compare the results derived from one measure, such as a reasonable royalty, to the results derived from another available method, such as lost profits or savings to the government. Leesona, 599 F.2d at 973.
Courts most frequently determine reasonable and entire compensation by calculating a reasonable royalty. Decca, 640 F.2d at 1172 (stating that “the reasonable royalty method is the preferred method of ascertaining the value of patent rights taken by the government”); Leesona, 599 F.2d at 968. They use the other two measures, lost profits and government savings, less often. A court may award lost profits only after the strictest proof that the plaintiff would in fact have earned and retained those sums on its sales to the government. Tektronix, 552 F.2d at 349; see Leesona, 599 F.2d at 971 (lost profits may only be awarded as an alternative to, not in addition to, a reasonable royalty). In a hypothetical negotiation, however, it is possible to consider the losses the licensor *574might suffer as one factor in determining the rate it would have sought. Leesona, 599 F.2d at 972 (“Some elements of business injury might be weighed in determining what the owner might have sold for.”). Similarly, savings to the government are not used as a measure of compensation in preference to a reasonable royalty, Decca, 640 F.2d at 1172, but these savings may be employed in estimating the amount a willing buyer would offer a willing seller. Leesona, 599 F.2d at 971.
Because Penda did not establish a commercial licensing rate,22 the court must turn to other methods, as did the court in Tektronix, 552 F.2d at 348. The plaintiff in that action was the owner of a patent on an oscilloscope and some related electronics that the government had used without permission. Id. at 345. Rejecting both the “nominal sliding-scale royalty” proposed by defendant and the plaintiffs theory of lost profits, the court concluded that the best method of computing compensation would be to determine a reasonable royalty by applying the willing-buyer/willing-seller method, id. at 348-349, observing that the method is exemplified by Georgia-Pacific, 318 F.Supp. at 1116.
Tektronix approved the use of a hypothetical negotiation to determine a reasonable royalty in a section-1498 action and noted that Georgia-Pacific exemplifies that method. 552 F.2d at 348-49. The Federal Circuit subsequently has approved the use of the factors enumerated in Georgia-Pacific to determine a reasonable royalty in the context of a private suit for infringement. Smithkline Diagnostics, Inc. v. Helena Lab. Corp., 926 F.2d 1161, 1168 (Fed.Cir.1991).
Just as in Tektronix, the parties in the case at bar agree that there is no established royalty rate and differ substantially in their evaluation of what a reasonable royalty rate should be, drawing on the fifteen factors listed in Georgia-Pacific. The ultimate purpose of applying the Georgia-Pacific factors is to determine what a willing licensee would pay a willing licensor for the rights taken by the government. Most of the evidence on these factors came from expert witnesses.
2. The Georgia-Pacific Factors.
The Georgia-Pacific factors “state the practical considerations in royal*575ty negotiations between a hypothetical willing licensor and licensee.” Sun Studs, Inc. v. A.T.A. Equip. Leasing, Inc., 872 F.2d 978, 994 (Fed.Cir.1989), overruled in part on other grounds, A.C. Aukerman Co. v. R.I. Chaides Const. Co., 960 F.2d 1020, 1039 (1992). Whether to treat the government or Cadillac as the licensee for purposes of this analysis depends on the approach used and the factor involved. The contractor operates under the aegis of the sovereign’s power to take a license under the second paragraph of section 1498(a). Decca, 640 F.2d at 1166, 1167. Hence, at some points in the analysis, it is more useful to treat the government as the hypothetical licensee. See Leesona, 599 F.2d at 971 (may use savings to the government in determining the amount a willing buyer would pay a willing seller). In the context of a hypothetical negotiation, however, it is sometimes more appropriate to consider the hypothetical licensee to be the manufacturer of the product, rather than the government. See Tektronix, 552 F.2d at 349 (willing buyer of the license is considered to be the party which will manufacture the item).
The Georgia-Pacific factors were postulated in the context of a private infringement action. As such, they do not always fit squarely in an analysis under section 1498. Mr. Saunders, defendants’ expert witness, recognized as much and referred to the test as setting up “artificial criteria.” He stated that because Georgia-Pacific was a commercial case, applying the factors in the context required by the case at bar was questionable. It should be remembered that the entire hypothetical negotiation is a legal fiction, and that the “whole notion of a reasonable royalty is a device in aid of justice, by which that which is really incalculable shall be approximated.” Cincinnati Car. Co. v..New York Rapid Transit Corp., 66 F.2d 592, 595 (2d Cir.1933); accord Tektronix, 552 F.2d at 351.
1. “The royalties received by the pat-entee for licensing of the patent in suit, proving or tending to prove an established royalty."
Penda licensed Bigelow to use the ’306 patent at $2.00 per pallet. Mr. Malikowski was of the view that the amount in the Bigelow license establishes “a floor under the royalty.” Because Bigelow was a retailer of plastic pallets and a customer of Penda, Mr. Malikowski believed that Penda would seek a higher royalty rate from Cadillac, a thermoformer that competes with Penda. Also, Mr. Malikowski thought that a very large, fixed price contract would justify a larger royalty than would the mere opportunity for the licensee to solicit business. Mr. Bremer concurred with Mr. Malikowski that the $2.00 royalty was a floor, in the sense that he would ask for more than that in a negotiation to license this patent and would not be willing to settle for less. Mr. Bremer testified that he considered this and other royalty agreements, but did not believe that any of them, alone or in conjunction, served to establish a royalty. The only opinion he could draw from the agreements was that within this industry it was quite common to grant licenses on a per pallet basis. The court finds that although this license does not establish a royalty rate, it is the most relevant of the various license arrangements discussed at trial.
Two other license figures are of less value. While this litigation was ongoing, Penda and Cadillac entered into a license to govern any future use of the patent. Cadillac agreed to pay Penda three percent of sales or $1.00 per pallet, whichever proved less. The license includes a specific provision that forbids its consideration as an “established royalty” for purpose of this litigation. Nevertheless, inasmuch as the government was not a party to that arrangement, it has introduced the license as evidence of a reasonable royalty rate. Because of the disclaimer, Mr. Malikowski discounted the significance of the license. He also considered it less persuasive because the parties entered into it after the time when the hypothetical negotiation would have occurred. Mr. Saunders testified for defendant that the license entered into with Cadillac and a substantially similar license offered to Shuert-Oakland indicate what Penda would be willing to accept *576on a license negotiation. The probative value of these licenses is somewhat limited, however, because they show only what Penda would be willing to accept when mired in litigation and after at least one competitor has already paid the start up costs necessary to compete. What would have been reasonable compensation for Penda in 1988 might have been a different, and probably higher, figure.
On the question of establishing a floor under the royalty, it has been held that developmental expenses provide such a floor. Leesona, 599 F.2d at 978. On the facts of the case at bar, however, it is not clear that Penda necessarily should expect to recoup its entire developmental cost on this one use by the government. In any event, Mr. Dresen testified that the developmental cost was about $250,000. That amount is less than the amount determined to be a reasonable royalty in this ease. Whether or nor it is considered a “floor,” therefore, is immaterial.
2. “The rates paid by the licensee for the use of other patents comparable to the patent in suit.”
As discussed supra, it is the government, in this case the Postal Service, which actually takes a license. That license authorizes it to contract with a third party to manufacture the infringing item. At trial, there was no evidence that the Postal Service ever paid for a license in a comparable patent. Nevertheless, in the context of applying the tool of a hypothetical negotiation to establish a reasonable royalty, it would not be unreasonable to look to what Cadillac has paid for rights under other patents comparable, to the ’306 patent as indicative of the reasonable and entire compensation due Penda. The only agreement under which Cadillac is the licensee is the Cadillac-Penda agreement discussed above.
3. “The nature and scope of the license, as exclusive or non-exclusive; or as restricted or non-restricted in terms of territory or with respect to whom the manufactured product may be sold.”
Plaintiff and defendants disagree sharply on this point. Plaintiff’s experts contended that the license taken in this situation was exclusive, while defendants’ expert contended that it was not. Mr. Malikowski testified that he believed that the defendants owed Penda compensation for an exclusive license.
An exclusive license would require greater compensation. As Mr. Malikowski pointed out, if a licensee has an exclusive license, “it will be permitted to [practice] the technology and no one else will.” Thus, exclusivity in a license is best seen as an additional burden on the licensor and a benefit to the licensee, for which the licensee should be willing to pay more. The licensor thereby undertakes not to license any others, a thing he would otherwise have the right to do.
In rationalizing his position on exclusivity, Mr. Malikowski stated:
It is my opinion that if Penda were obligated to license Cadillac for this proposal to the post office, they would have desired to maintain license to that one proposal only. They would not have wanted Cadillac out there competing with them generally in the marketplace. And so this license would be exclusive for the contract that we’re talking about.
This position mistakenly discusses the exclusivity of a license as a benefit to Penda, the licensor. The fact that Penda would want to license no others is a valid consideration in determining a reasonable royalty, but it enters into the mix under the fourth Georgia-Pacific factor, not the third.
Mr. Malikowski then employed his conclusion that the license was exclusive to impute a royalty of $3.30. Mr. Malikowski reasoned that because this is an exclusive license from Penda to Cadillac, Cadillac would have offered as its bid the lowest noninfringing bid and would have paid Pen-da the difference as a royalty. In connection with this factor, it becomes apparent that it is significant that the United States, not Cadillac, is the entity that has the authority under section 1498 to take a compulsory license. The United States was free to award this contract to a party other *577than Cadillac, subject to the limitations of the bidding process.
When counsel for defendant pointed out that Penda had already granted a nonexclusive license under the ’306 patent to another entity, Bigelow, Mr. Malikowski sought to preserve his conclusions by suggesting that Penda could have granted Cadillac an exclusive license within the specific market of bidding on this Postal Service contract. Penda could not have done so. It had no power to prevent others from bidding on or receiving the Postal Service contract with articles that would infringe the ’306 patent.
Mr. Malikowski explained during his testimony that he did not consider his assignment to be to determine a reasonable royalty from the Postal Service to Penda. He believed “reasonable and entire compensation” to be a royalty rate which Cadillac would be willing to pay to Penda. It is true that at least one previous case determined reasonable compensation from “a willing-buyer/willing-seller concept, in which a suppositious meeting between the patent owner and the prospective manufacturer of the infringing article is held to negotiate a license agreement.” Tektro-nix, 552 F.2d at 349. The court in that case went on to apply the test by considering a negotiation between the plaintiff and the third-party defendants. Although this approach is useful and does justify considering the third-party defendant as at least participating in such hypothetical negotiations, Mr. Malikowski’s conclusion illustrates the fallacy of applying this reasoning too rigidly. The patent owner and the manufacturer could not, in the hypothetical negotiation, agree on an amount to charge the government because the government takes the compulsory license by eminent domain.
Mr. Bremer stated on direct that as a rule of thumb in the licensing community, an exclusive license is worth two times a nonexclusive license because the grantor of an exclusive license assumes the obligation to exclude all others, to protect the licensee, and to preserve the licensee’s right to practice that invention. The court finds no indication that Penda assumed any such obligation here with respect to either the Postal Service or Cadillac. It may well be that Penda would choose to enforce its patent right to exclude others, but that choice to exercise a right would differ substantially from incurring an obligation so to act. Nevertheless, Mr. Bremer agreed with Mr. Malikowski’s statement that this is an exclusive license. Although he acknowledged that the Postal Service must also be considered a licensee, he referred to the Postal Service and Cadillac as being in consort. He did not explain, however, why he believed that Penda assumed, as a result of the government’s actions, an obligation to exclude others. Mr. Bremer did point out that in a real negotiation the prospective licensor would probably have to stop anyone who was infringing the patent in order to maintain a viable negotiating stance. That observation, however, is not relevant to whether the license taken by the government was exclusive in nature. As with Mr. Malikowski’s statements, it pertains more to what Penda might choose to do than to what Penda was obligated to do. Mr. Bremer’s acknowledgment that the Postal Service would not care whether it received an exclusive license is, on this point, telling. It is the government’s use that determines the scope of the license for which the parties will bargain in the hypothetical negotiations. That Cadillac might have wished for exclusive rights is irrelevant.
Mr. Saunders, defendants’ expert, agreed with Mr. Bremer that a promise of exclusivity may double the value of a license, but maintained that the license here being valued was not exclusive. Under a nonexclusive license, the licensor grants only a bare right to work the patent, but gives the licensee no market control.
The Federal Circuit has said that under section 1498, the government is in the position of “a compulsory, nonexclusive licensee.” Motorola, 729 F.2d at 768. Moreover, in an earlier Court of Claims case it was said, “It is true that the patent right taken by the government was not exclusive in a legal sense.” Decca, 640 F.2d at 1181. Decca did proceed to hold that “the patent *578rights taken by the Government were, practically speaking, exclusive.” Id. That holding is best limited, however, to the unusual facts in Decca. The patent taken by the government pertained to satellite communication. Inasmuch as the government was implementing a world-wide system, the Court of Claims decided that no one would incur the cost of establishing a duplicative system.23 Unlike that situation, plaintiff in the ease at bar may continue to derive benefit from the patent. In sum, the court rejects plaintiff’s contention that the reasonable royalty should be set higher to cover an exclusive license.
4. “The licensor’s established policy and marketing program to maintain his patent monopoly by not licensing others to use the invention or by granting licenses under special conditions designed to preserve that monopoly.”
This court’s predecessor answered in the affirmative the question whether a paten-tee’s reluctance to license could be considered in determining reasonable and entire compensation for a government use. See Leesona, 599 F.2d at 976. The plaintiff in that case owned a patent on a type of battery. Id. at 962. Until the government’s infringement deprived it of its exclusive right to manufacture the batteries in the United States it had never licensed them domestically. Id. at 976. The Court of Claims there found that the plaintiff did not want to relinquish its exclusive manufacturing right and would in fact have charged a high fee for a license surrendering that right. Id. The court held that although the license agreement was deemed to involve a willing licensor, that licensor is still the plaintiff and embodies all of the plaintiff’s particular interests. Id. Moreover, the court justified its decision by reference to the general law of eminent domain, in which the property is regarded from the point of view of the owner. Id. at 977.
Before the events giving rise to this suit, Penda had licensed the ’306 patent only once, for $2.00 per pallet. The licensee was Bigelow, one of Penda’s customers. Penda did the actual thermoforming work for Bigelow. Furthermore, Penda had also received a license from Bigelow on other pallet technology. Mr. Malikowski argued that this factor mitigated strongly in favor of a higher royalty here. Based on his conversations with management at Penda, he determined that Penda’s marketing policy did not include the grant of licenses. The company intended to exploit its patent rights by being the sole manufacturer of this article. Mr. Saunders found this factor to be of neutral import. He discussed Penda’s licensing program, however, only in terms of the licenses it had entered into or offered. He did not indicate whether he considered the absence of extensive licensing indicative of a desire not to license.
The court is wary of adopting Mr. Mali-kowski’s rationale in using the Bigelow licensing circumstances to leverage upward the calculations here. The Bigelow arrangement was informal, confirmed after the fact by a letter. The evidence plainly suggests that there was an amicable and ongoing business relationship between the two companies, which might prompt Penda to ask for a lower license in order to get access to Bigelow’s own technology. It can also be argued, however, that the very fact of licensing suggests Penda’s willingness to sell the right to utilize the patent. The most that the court derives with a degree of certainty is that the Bigelow arrangement was not contrived after the fact and is one of the most probative items adduced by either party as to a reasonable royalty. Neither it nor the Cadillac arrangement prompt a finding that Penda had an established licensing program.
5. “The commercial relationship between the licensor and licensee, such as, whether they are competitors in the same territory in the same line of business; or *579whether they are inventor and pro-motor. ”
This is one of the factors in which it is appropriate to consider Cadillac as the licensee. See Tektronix, 552 F.2d at 349. The commercial relationship between the patent owner and the manufacturer of the infringing article is likely to be more instructive as to the amount of a reasonable royalty than is the relationship between the patent owner and the government. In determining reasonable and entire compensation, it should not be disregarded that the government has essentially created a potential competitor for plaintiff. Even if, as defendants maintain, Cadillac made no profit on this job, it did cover the cost of obtaining tools to make this article. Moreover, as Mr. Malikowski observed, the work provided Cadillac with valuable manufacturing experience that it could subsequently use in competition against Penda. Even without the benefit of the subsequently executed license, this places Cadillac in a more favorable position from which to compete with Penda for business when the patent expires. This change in position is part of what plaintiff has lost, and should be considered in determining its compensation. This harm is dissimilar to consequential damages for lost business opportunity, for which plaintiff would not be entitled to recover.
Mr. Malikowski, treating Cadillac as the licensee, contrasted the commercial relationship between Penda and Bigelow on the one hand and between Penda and Cadillac on the other. Penda and Cadillac are competing thermoformers, both located in the industrial upper midwest. Bigelow, in contrast, is a regular customer of Penda’s. The license with Bigelow authorized it to have pallets covered by the Penda patent made elsewhere, but contemplated that Bigelow would still get at least some of its pallets from Penda. Mr. Bremer generally agreed with Mr. Malikowski’s assessment of this factor. On cross-examination, Mr. Malikowski acknowledged that the Postal Service had in the past been a customer of Penda’s and that at the time of the taking was at least a potential future customer. This relationship contrasts to the competitive manufacturer-manufacturer relationship between Penda and Cadillac, but is analogous to the manufacturer-customer relationship between Penda and Bigelow. Mr. Saunders, in contrast, focused on the government as licensee. He contended that because the Postal Service was potentially a big customer, Penda would have been less likely to press for a high royalty.
Both Penda’s relationship with the government and Penda’s relationship with Cadillac are valid considerations, but they do not entirely cancel out one another. On balance, the harm Penda would suffer from Cadillac (or whoever won the pallet contract) being able to pay for the cost of tooling up, eventually to compete with Pen-da, militates in favor of a higher royalty rate. Any similarities between the government’s and Bigelow’s relationships with Penda avails defendants little, because the amount the court ultimately sets as a reasonable royalty infra is somewhat less than the amount set in the license between Bigelow and Penda.
6. “The effect of selling the patented specialty in promoting sales of other products of the licensee; the existing value of the invention to the licensor as a generator of sales of his non-patented items; and the extent of such derivative or convoyed sales. ”
A convoyed sale occurs when the sale of the patented product naturally brings with it some additional components. Both Mr. Malikowski and Mr. Bremer dealt with this factor summarily on the basis that no convoyed sales were involved. Mr. Saunders agreed. He also suggested, however, that this large procurement might encourage others to buy this pallet and, therefore, work to plaintiff’s benefit. This factor, as expressed in Georgia-Pacific, seems directed only to the extent to which convoyed sales might benefit the licensee. It does not concern the more general benefits of good publicity from sales. Insofar as this type of spillover effect is relevant, the court notes that, as Mr. Saunders pointed out on cross-examination, a company is not always associated with a product. Here, *580Penda received no publicity as a result of the large government procurement because there was no acknowledgment of the use of Penda’s patent.24 This factor does not suggest a diminished royalty.
7. “The duration of the patent and the term of the license. ”
With respect to the duration of the patent, there is little room for dispute. All the experts agree that at the time of the hypothetical negotiation the patent would have had twelve to thirteen years of remaining viability. This period is about two-thirds of the patent’s entire term. If the patent were close to expiration, it would generally tend to depress associated royalties. Neither Mr. Malikowski nor Mr. Bremer addressed the question of the duration of the license. According to Mr. Saunders, the duration of the license would have been about two years. In any event, Mr. Saunders did not believe that this factor had any substantial impact on the assessment of a reasonable royalty in this case.
8. “The established profitability of the product made under the patent; its commercial success; and its current popularity.”
Mr. Malikowski asserted that Penda has been successful in selling the patented pallet, sometimes at prices up to fifty percent greater than that charged by Cadillac to the Postal Service. He also testified that it had sold in large numbers. The source of Mr. Malikowski’s information and the basis of his opinion with regard to commercial success were a meeting he had with Penda personnel and sales figures they supplied to him. On cross-examination, he was unable to recount any of the specific figures on which he based his opinion of commercial success, because it had been over a year and a half since he had actually conducted the analysis.
Mr. Bremer agreed with Mr. Malikowski that the pallet was profitable, commercially successful, and popular. He indicated that this belief was based on discussions in which counsel for Penda indicated that these pallets sold for as much as $26 apiece at the time of the taking.
Mr. Saunders subdivided the eighth Georgia-Pacific factor into three components: the established profitability of the product, its commercial success and current popularity, and the licensee’s profit expectations. As to the first two components he stated that he had insufficient information on which to form an opinion.25 As to the third component, he merely stated that the Postal Service considered this an expense and not a profit item. Additionally, Mr. Saunders stated that he did' not think a nonexclusive license to the government would affect the commercial market. This particular factor, however, deals not with any effects of the licensing on the market, but on the market conditions as they existed at the time of the license. In the context of a normal infringement suit, this factor would be relevant because a willing licensee intending to exploit the commercial market should be willing to pay more if the product is already in demand.
In the context of a section-1498 action, exploitation of the existing commercial market is usually not an objective. The Postal Service had no intention of selling pallets. Even from the perspective of the manufacturer of the infringing item, the conditions in the commercial market are of little import. Absent a license from the patentee, the manufacturer of the infringing article may not exploit the patent except for government use until the patent term expires. The factor would be significant, however, in determining what value the reasonable licensor would place on the patent. The license is valued from the perspective of the licensor. Leesona, 599 F.2d at 977. In this case, plaintiff invested significant quantities of time and money in developing a product that, based on the need perceived to exist by the inventors, it expected to be a commercial success. *581There is no question that in some respects that expectation was fulfilled.
9. “The utility and advantages of the patent property over the old modes or devices, if any, that had been used for working out similar results.”
Mr. Malikowski opined that the patented pallet was stronger than other, similar devices. He said the basis of his opinion was conversations he had with Penda personnel. He recounted no independent research. Mr. Bremer concurred in this assessment, based on conversations he had with plaintiffs counsel. Mr. Bauman, a fact witness, testified that the patented invention was very useful commercially to him and to Big-Pak, his employer. He said that, in addition to being light, the Penda pallets provided a structural integrity unavailable from single sheet pallets. These features allowed him to exploit markets previously closed to him.
Mr. Saunders compared the patented pallet not to other available plastic pallets, but to other pallets in general. Although he conceded that the patented pallets were more durable, he pointed out that they were priced at a higher level in an inelastic market. He said that it was difficult to quantify the mixed effects of these two factors, and refused to draw a conclusion on the issue. Mr. Saunders’ analysis fails clearly to focus on the patented invention. Plaintiff’s patent did not cover all plastic pallets; it covered a particular structure of twin-sheet thermoform plastic pallets. Thus, there were other pallets of comparable durability that were substantially less rugged. There is no indication, however, of how the patented pallets would compare in price to these analogous pallets.
The most convincing evidence of the pallet’s worth, however, comes from the Postal Service itself. This material is discussed in the context of the next factor.
10. “The nature of the patented invention; the character of the commercial embodiment of it as owned and produced by the licensor; and the benefits to those who have used the invention.”
In addressing this factor, Mr. Malikowski formed his opinion as to the benefits of the invention based on Postal Service employee statements about the extent to which these pallets were superior to other products. Mr. Malikowski used these statements to impute a numerical value to the Postal Service’s use of the product. His goal in calculating these figures was to ensure that the opinion he formed as to a reasonable royalty was not so prohibitively high that the Postal Service would not accept it. It served as a test factor, to make sure the assessment would in fact, be reasonable to the Postal Service.
Mr. Malikowski computed what he understood to be the Postal Service’s willingness to pay based on an internal assessment done by the Postal Service. He then used that calculation to place an upper boundary on the royalty. Exhibit 200 reflects a summary of Postal Service documents discussing the cost per trip of the thermoform pallet in comparison to composition pallets. The documents summarized are primarily memos from Marshall Weiss, the Director of acquisition or agent responsible for this contract. Mr. Malikowski used the estimates in these Postal Service documents as the bases of his calculation of cost savings to the government.
Mr. Malikowski relied on a memo of May 15, 1987, from Mr. Weiss to Prank Hanson to show that the Postal Service expected a thermoform pallet to cost about 45 cents a trip, whereas the composition pallet would cost anywhere between $1.00 and $1.50 per trip. Mr. Malikowski considered the differences between 45 cents and $1.00 and $1.50 respectively to be what the Postal Service would have considered the minimum and the maximum benefit per trip by using the thermoform pallets.
The second Postal Service document to which Mr. Malikowski referred was a memo dated November 20, 1987, from Mr. Weiss to Mr. Gary Litwinowicz. Mr. Mali-kowski had to make an adjustment on pricing to use this document. In 1987, the Postal Service hoped to buy the pallets at $16.00 each, but there were no responsive bids at $16.00. Mr. Malikowski therefore *582adjusted the price from $16.00 per pallet to Penda’s bid price. His third note in the column on Exhibit 200 labeled “Documents” reflects this adjustment, which resulted in a cost per trip estimate of 43 cents. In this second memorandum, the Postal Service estimated that composition pallets would cost $1.50 to $2.00 apiece. Mr. Malikowski then derived the figures 1.069 and 1.569 by subtracting the 43 cents from the $1.50 and the $2.00, respectively, as shown in the exhibit.
After calculating these several possible values for the benefit to the Postal Service, Mr. Malikowski adopted the most conservative estimate he had calculated: fifty-five cents. He multiplied this amount by the number of trips the Postal Service expected to get out of each new pallet. According to Mr. Malikowski all the documents anticipated 50 trips. Mr. Malikowski opined that the number 50 was in fact very conservative. He suggested that the pallets shipped are still being used and have been used many more than 50 times. He was principally interested, however, in conservatively estimating the value the Postal Service would place upon the pallets. Multiplying 55 by 50 gives a value to the Postal Service of $27.50 per pallet. Mr. Malikow-ski therefore believed that the Postal Service should have been willing to pay up to $27.50 per pallet because it would still be economically advantageous over composition pallets.
Mr. Malikowski also used these cost calculations to determine what he thought the Postal Service should have been willing to bring to the bargaining table in the form of a reasonable royalty. Although in general Mr. Malikowski considered Cadillac to be the licensee, for the purposes of this ealeu-lation he treated the Postal Service as the licensee. To derive this figure he started with the savings as shown in the first block of exhibit 197 and multiplied it by the number of trips. This operation results in the $27.50 seen before. He then applied a rule of 25 percent of the benefit going to the licensee26 and multiplied $27.50 by 25 percent to get $6.87. Mr. Malikowski felt that this calculation indicated that the Postal Service, if this negotiation were between them and Penda directly, should have been willing to pay up to $6.87 in the form of a reasonable royalty. Exhibit 197 also reflects Mr. Malikowski’s calculations based on the possibility of a different number of trips.
Mr. Bremer adopted Mr. Malikowski’s statement on this factor. He studied the same Postal Service documents as Mr. Mal-ikowski pertaining to the value of the pallet in issue.
The court is reluctant to place great weight on these calculations. The very size of the figure Mr. Malikowski derived by this methodology suggests that it is questionable. It is far out of line with any royalty offered into evidence. It has certainty only in the mathematical sense, but it is based in large measure on subjectively determined figures. Collectively, there are simply too many points at which the hard figures introduced into Mr. Malikowski’s calculations are in fact only estimates. The cost savings per trip, the number of trips, the percentage of savings to allocate to a royalty, are all inherently uncertain. While it is appropriate to look to Postal Service estimates as to the fact of an anticipated cost saving, however, the court does not believe the figures relied upon were *583intended to bear the weight imposed by this analysis.
Mr. Saunders’ analysis of the tenth factor differed substantially. He observed that the patented invention is a plastic pallet for transporting mail and other products. It is a commercial industrial product that in a license situation would be made in rather large quantities and sold directly to the user. It is not a consumer product. He described the benefit to those who have used the invention to be the increased lifetime, but at additional cost. In his view, factor ten differs from nine in adding another element, the nature of the patent invention. It is a commodity-type product, in the sense that it’s made in volume and sold to the customer. He explained that usually, the mark-up is less than some other types of product. This is not an ironclad rule, but ordinarily industrial commodities, especially where there is competition, do not have the same mark-ups as, for example, pharmaceuticals.
If there really is no competition because an item is unique, the mark-up probably would be greater. Its value would be reflected in the mark-up and the margin of profit. In Mr. Saunders’ view, however, there is competition. The analysis is complex because the patented pallets have different characteristics from the competing items, but they serve the same purpose. Different pallets have different lifetimes and different costs. The patented pallet is not so distinctive as to have no competing products. Mr. Saunders identified a longer lifetime as the principal benefit. Mr. Saunders’ analysis of factor ten led him to decide that because this was the kind of product that bears a lower mark-up, a prospective licensor could not push for an unreasonably high royalty, or the kind of royalty one would get in the case of pharmaceuticals, where the licensee has tremendous mark-ups and market exclusivity because of factors in addition to patents, such as government approvals from the FDA. Because he believed that there were nonin-fringing alternatives to which the government could have resorted, Mr. Saunders did not believe this factor militated in favor of a higher royalty.
Once again, the court is not entirely persuaded. Mr. Saunders did not give full credit to the uniqueness of the invention. It is true that there are competing products, but the advantages of the patented invention created a performance gap that had monetary significance. In some sense a hand spade can do the same work as a shovel, but the shovel is more cost effective for larger projects. Mr. Saunders’ assumption that the patented invention would receive a lower mark-up was not persuasively established.
11. “The extent to which the infringer has made use of the invention; and any evidence probative of the value of that use. ”
The experts’ opinions again diverged as to factor number eleven. Mr. Malikowski, maintaining his position that Cadillac was the infringer, returned to the argument that this license essentially put Cadillac in the business of making twin-sheet thermo-form pallets. He asserted that “the invention” allowed Cadillac to submit quotes to the Postal Service and to other customers. What he meant by this statement is not entirely clear. On its face, it does not seem consistent with his earlier testimony that he considered this to be an exclusive license because Penda would want it to be exclusive to this job.
Mr. Bremer’s analysis of this factor was more cogent. He agreed with Mr. Mali-kowski that the compulsory license essentially set Cadillac up in the business of making twin-sheet thermoform pallets. He explained that Cadillac, once having set up the operation and having resolved any problems encountered in implementation, would be in a position to bid to the Postal Service or other postal customers. The extent of the use under this compulsory license, however, would be limited to this contract. In particular, Cadillac would not be able to bid on contracts to private industry without securing an appropriate license from Penda.
*584Mr. Saunders addressed this factor from the perspective of the extent of the government’s use. First, Mr. Saunders did not believe that the size of the order, 640,000 pallets, would be likely to have any effect on the amount to be negotiated as a reasonable royalty in hypothetical negotiations. Certainly, he did not believe that an order for a large volume would raise the royalty. Second, Mr. Saunders, in the context of this factor, concluded that there was not enough reliable information upon which he could decide whether there were any actual government savings.
Mr. Saunders’ analysis on this point illustrates a troubling aspect of his testimony. He showed virtually total reluctance to concede that any factor militated in favor of the plaintiff. Mr. Malikowski, in contrast, though generally arguing on the plaintiff’s behalf, seemed at times willing to give the government the benefit of the doubt in his analysis. In addition, some of Mr. Saunders’ explanations of why a factor should be interpreted to favor the defendants and not the plaintiff were often facile. In any event, his argument with respect to the savings to the government, whether considered under factor eleven or factor ten, was unpersuasive.
Although Mr. Malikowski’s analysis of the benefit to the government did not reduce all amounts to present value, and though the imprecision inherent in his calculations gives the court pause, the court finds ample evidence that the Postal Service expected to save a significant amount of money by using these thermoform pallets. A hypothetical licensee in 1988, likewise expecting to save money, should have been willing to pay a royalty based on that expectation..
12. “The portion of the profit or of the selling price that may be customary in the particular business or in comparable businesses to allow for the use of the invention or analogous inventions. ”
Mr. Malikowski and Mr. Bremer both considered the agreements discussed under factors one and two to be the only pertinent ones in this case. Neither added any substantively new analysis under factor twelve. Mr. Saunders, however, employed this factor to discuss other licenses. The plaintiff presumably put these licenses in evidence as indicative of a reasonable royalty.
Mr. Saunders first discussed a license from Ford Motor Company to Penda with respect to the use of a certain tool. Under this license, Penda must pay Ford a royalty of $2.00 per part manufactured with the molds. It is not clear that this represents a royalty on a patent. It appears to be merely a usage fee for the molds. Mr. Saunders, therefore, discounted this agreement’s significance under factor twelve. Mr. Saunders also was unable to extract any information he considered useful from any of the licenses granted by Bigelow.27 Mr. Bauman negotiated these licenses, and they were introduced through him. Bige-low licensed a certain company to use one "of Bigelow’s patents, but Mr. Saunders did not find the $2.50 royalty specified in that agreement instructive because he knew neither the value of the pallet involved nor the quantum of that market, and because the minimum amounts involved were rather small. Mr. Saunders also did not consider significant another license to a different company, also for $2.50, because it was an exclusive license for a foreign country and included trademark rights and obligations and know-how transfer. Mr. Saunders did not consider comparable another license for pallets to yet a third company, set at five percent of sales, because he did not know the volume involved under that license or whether the products were comparable.
Although Mr. Saunders is correct in noting that none of these licenses is directly on point with the hypothetical license for the patent in suit, that is not the standard called for under factor twelve. These licenses were in the same or a comparable business. Both Bigelow and Penda are in the business of selling pallets, and all of these licenses pertain to that business. *585The court duly notes the differences between these licenses and the hypothetical license for the patent in suit. The court also notes that while each of these licenses includes unique terms, they all are for comparable amounts. Although the court does not consider this sufficient evidence to show that a license of about $2.50 was the prevailing rate on pallet technology, the court does consider it some evidence of a reasonable rate in similar circumstances.
13. “The portion of the realizable profit that should be credited to the invention as distinguished from non-patented elements, the manufacturing process, business risks, or significant features or improvements added by the infringer.”
Neither Mr. Malikowski nor Mr. Bremer considered this factor as a major component of their analysis. Mr. Malikowski opined that at least twenty-five percent of the realizable profit should be attributed to the invention. It was plain from his testimony, however, that he meant instead to use twenty-five percent of incremental profit. Mr. Bremer generally agreed with Mr. Malikowski.
This factor was the focal point of Mr. Saunders’ analysis. He first calculated Penda’s expected profit as a percentage of its sales cost. Next, he postulated that a reasonable royalty rate would be twenty-five percent of Penda’s expected percent profit. Mr. Saunders, by applying this technique, decided that 2.1% of Cadillac’s sales price of $18.93, or approximately $.40, would be a reasonable royalty.
The Court of Claims relied on this factor in assessing a reasonable royalty in at least one case. See Tektronix, 552 F.2d at 349. To determine the amount an infringer would have been willing to pay, the court took the selling price of the article, deducted both fixed and variable manufacturing costs to find gross profit, then allocated to the infringer its “normal” profit and assigned the rest to the patentee as a royalty. Id. This reasoning would seem to assume implicitly that the selling price of the article in some way incorporates the value of the property taken, so that the manufacturer would be making more on this sale than it otherwise would have.
The court does not accept Mr. Saunders’ analysis. His use of this factor differed sharply from the use approved of in Tek-tronix. He made his calculations based on Penda’s expected profit on this job, rather than on the basis of the profit of the accused infringer. See id. at 349-351. He testified that he did so because, though Cadillac claimed to have made no profit, it would have been unfair to deny Penda a royalty on that basis. The court agrees that it would be unfair to deny Penda a royalty. That Cadillac apparently bid this job without the expectation of making any profit runs counter to the implicit assumption of the court in Tektronix and indicates that the approach used there is not appropriate on the facts of the case at bar. The court does not agree, however, that the alternative method Mr. Saunders employed produced a reasonable royalty. Mr. Saunders’ method produces a royalty value of less than forty cents per pallet. The next lowest royalty value shown in any license offered in evidence was a minimum of $1.00 or 3% of sales. The evidence, viewed as a whole, does not support a conclusion that Penda’s royalty should be that low.
14- “The opinion testimony of qualified experts.”
Mr. Malikowski ultimately concluded that a reasonable royalty would be between $3.00 and $3.50. Mr. Bremer also stated that he believed a reasonable royalty would be about $3.00 per pallet or, on the high end, as much as $3.50. Mr. Saunders believed that 2.1 percent of the sales price of $18.93, or thirty-seven cents per pallet, would be a reasonable royalty for the first 320,000 pallets. Although instructive, the court found none of these opinions completely persuasive.
15. “The amount that a licensor (such as the patentee) and a licensee (such as the infringer) would have agreed upon (at the time the infringement began) if both had been reasonably and voluntarily trying to reach an agreement; that is, the amount which a prudent licensee — who desired, as a business proposition, to ob*586tain a license to manufacture and sell a particular article embodying the patented invention — would have been willing to pay as a royalty and yet be able to make a reasonable profit and which amount would have been acceptable by a prudent patentee who was willing to grant a license. ”
Mr. Malikowski testified extensively with respect to this factor. He constructed an elaborate negotiation model based on his calculations of both Penda’s and Cadillac’s incremental incomes. Although his technique is intriguing, the court does not discern clear support for this methodology in the case law that governs our decision on this matter. The approach may be appropriate in some cases, but it is not necessary to devise a novel method to calculate reasonable and entire compensation. Mr. Mal-ikowski’s testimony on the government savings from using this pallet is, however, reflective of the fact that the government should have been willing to pay something less than its anticipated savings in such a negotiation.
3. Finding as to a Reasonable Royalty.
After reviewing the factors pertinent to assessing a reasonable royalty and considering the opinions of qualified experts, the court does not consider any single factor or opinion determinative or fully convincing. After considering all of the above mentioned evidence, the court finds that a reasonable royalty would have been $1.85 per pallet. Most relevant in the court’s determination is the clustering of analogous royalties in the $1.00 to $2.50 range. The royalty paid by Bigelow on the Penda pallet ($2.00) is the most closely on point. It is the most material arrangement and does not suffer from the effects of litigation. The court does not share the defendants’ concern that this royalty was a reflection of the present litigation. If $2.00 is the fulcrum of those royalty amounts, it is also relevant to the court that the size of the sale and the potential for future business with the government suggests a somewhat lower royalty. The $1.85 figure, slightly less than ten percent of the sales price, lies well within the range of royalties discussed in this case for generally similar products, and is within the range of royalties previously awarded under section 1498 to successful plaintiffs.
Conclusion
Defendants concede that the accused pallets infringe claims nine and ten of the patent. The court finds that the accused pallets literally infringe claim eleven because they contain every limitation of that claim. In addition, the court finds that the accused pallets perform substantially the same function in substantially the same way to achieve the same result as the pallet of claim eleven.
Defendants have failed to prove by clear and convincing evidence that the patent in suit does not meet the requirements of the second paragraph of section 112. The court finds that the inventors did consider a pallet having a foot floor with fused portions in the sidewall to be their invention, and so were entitled to claim it separately. The court finds that the language of claim nine reasonably apprises those of skill in the art of both the utilization and scope of the invention, and is as precise as the subject matter permits.
Defendants have failed to prove by clear and convincing evidence that the Penda patent is obvious within the meaning of section 103. The court concludes that the differences between the subject matter patented and the prior art are such that the subject matter as a whole would not have been obvious to one of ordinary skill in the art to which the subject matter pertains.
The clerk is directed to enter judgment in favor of plaintiff and against defendant the United States in the amount given by the following formula,28 where n is the number *587of days in 1993 preceding the entry of judgment:
Costs to plaintiff.