The opinion of the court was delivered by: OBERDORFER
Plaintiff Ferag AG is a family-owned Swiss corporation, with its principal place of business near Zurich; and plaintiff Ferag, Inc. is Ferag AG's U.S. subsidiary. The two companies are referred to here collectively as "Ferag". Defendant Grapha-Holding AG is also a family-owned Swiss corporation, with its principal place of business near Zurich -- a few kilometers from Ferag. The parties are "two of only a very few companies that manufacture bindery equipment used to assemble or join together pages of brochures and magazines." Ferag AG v. Grapha-Holding AG, 905 F. Supp. 1, 2 (D.D.C. 1995). They compete furiously worldwide. One arena of competition is patent litigation. Each holds numerous patents on its products and elements of them issued by several governments, including the United States, and are vigorously litigating issues about the validity of the patents and infringement.
At issue in this case is Grapha's U.S. Patent No. 4,735,406, referred to by the name of "Weber", the Grapha employee who disclosed and claimed as his invention a rotary drum "gatherer-stitching" machine "for making brochures and the like." A copy of the Weber patent, issued on April 5, 1988, is attached here as Appendix A. The issue is drawn by Ferag's prayer for a declaratory judgment that the Weber patent is invalid and that no Ferag machine infringes it. Grapha counterclaims that Ferag's rotary drum "IPEX" machine infringes the Weber patent, although Grapha has never manufactured or sold the rotary drum machine disclosed by the Weber patent. During a 13-day bench trial concluded in May 1996, the parties adduced elaborate proof, illuminated by models and prolific expert testimony. The evidence yields the following:
In contrast, Grapha's Weber patent discloses a gatherer-stitcher that would not only move signatures linearly down a chain, but also circularly around a rotary drum. That is, the Weber machine would deposit individual sheets spine-up on carriers which resemble the fins of a paddle-wheel-type rotary drum. Sheet feeders arranged axially along the drum, one downstream from another, would deposit individual sheets on the fins one after the other as the fins rotate past the feeders. At the same time, the individual sheets would move axially along the gathering fins. Each sheet would follow a helical path while moving along the length of the drum. The carriers would advance the sheets to the stitching station, where they would be stapled or stitched together. The stitchers would include two or more staple applicators mounted on a yoke that would be concentric to the drum and would move in a pendulum manner with the drum.
The machine disclosed by Orapha's Weber patent would also differ from Grapha's own conventional in-line gatherer-stitcher in that the patent claims that the Weber machine would be capable of producing 40,000 copies per hour, whereas the top in-line stitchers, including Grapha's, are capable of producing only 20,000 copies per hour. Although Grapha commands a substantial U.S. share position in the gatherer-stitching market, it has never made or sold a machine covered by any claim of the Weber patent, and in fact, has never constructed a prototype or working model of the machine shown in figure 1 of the Weber patent. See Stip. Facts PP 38, 40. Grapha's top product is an in-line gatherer-stitcher which produces a maximum of 20,000 copies per hour.
Unlike Grapha, in 1991, Ferag briefly offered for sale a rotary drum machine known as "Print '91"; and in 1993, Ferag constructed and exhibited a demonstration machine of a rotary drum gatherer-stitcher at its facility in Bristol, Pennsylvania. More important, Ferag presently manufactures and sells an unpatented, high speed, rotary drum gatherer-stitcher, the "IPEX" machine: Ferag's IPEX machine is capable of binding and stitching at 40,000 copies per hour. There are six IPEX machines operated by Ferag customers in the United States.
The IPEX machine consists of a rotary drum, the circumference of which has forty radially protruding fins. A rotary drum's "fin" is one in-line chain's "saddle". These fins extend along the length of the drum. A gripper conveyor delivers folded signatures to the drum. As each signature, held in a gripper at the fold line, moves toward the drum, a star wheel opener separates the two panels of the folded signature. A gripper conveyor and the feeder screw position each signature over an outer edge of one of the fins, guide the signature onto the fin with the signature's panels straddling the fin, whereupon the gripper conveyor releases the signature. As the drum rotates, the first signatures move circularly for approximately 270 degrees. Only as the signatures reach the last quarter of their rotation around the drum do they move axially by a control slide to the next feeding station, where the process repeats itself. When the signatures reach the stitching station near the end of the drum, a rotatable stapling apparatus, comprised of a carrier rotating in a direction opposite to that of the gatherer-stitcher drum and containing ten stapling heads, staples serially, one signature at a time. The stapled signatures then move axially into the exit station.
In order to protect its rotary drum gatherer-stitchers from claims of infringement, Ferag seeks a declaratory judgment that Grapha's Weber patent is invalid and unenforceable, and even if valid, is not infringed by Ferag's IPEX, Print '91, and Bristol demonstration machines.
Primarily, Ferag contends that the Weber patent is invalid because it claims a gatherer-stitcher machine that was known before Weber described his invention and that the machine Weber claimed to have invented was obvious to those of ordinary skill in the art as it developed prior to June 1985, the date the Weber patent application was first filed in Switzerland. Grapha defends Weber's patent and counterclaims for a declaratory judgment that the Weber patent is valid and enforceable, and that Ferag infringed the Weber patent by manufacturing, using, offering for sale, and/or selling the unpatented IPEX, Print '91, and Bristol demonstration machines.
In explanation of the competing claims, the parties generally agreed about the role of gatherer-stitching machines in the print-bindery industry: After the printing press finishes printing the pages of a newspaper or magazine, the bindery machine assembles and stitches them into the finished product. For the last 40 years, the in-line gatherer-stitcher has been the industry standard for assembling magazines. This machine gathers or accumulates individual folded sheets or "signatures" on a linear gathering chain, one over the other along their folded edge or spine with the cover being the outermost sheet. This is known in the trade as accumulating sheets from the "inside-to-outside." The in-line gatherer-stitcher places the sheets over a saddle, and stitches or staples them together along the fold line. Well-known examples of gatherer-stitched magazines are Time and Newsweek. Currently the top speed or rate of production for an in-line gatherer-stitcher is 20,000 copies per hour. Because the output of a printing press typically outpaces the output of bindery machines, it is often necessary to store unbound pages, a costly and time-consuming step. To save this time and expense, the bindery industry is constantly in search of binding equipment which more nearly matches the speed of the printing press.
The industry's demand for greater bindery speed has been frustrated by the fact that on a conventional in-line gatherer-stitcher, each sheet must be individually transported through the machine; therefore, higher output can be achieved only by increasing the speed of the sheets as they travel along the gathering chain. But speeding the movement of paper linearly can create many difficulties such as air currents, which flutter and tear the sheets. To minimize such difficulties, the speed at which the sheets can be moved linearly is limited, with consequent finite limits on the production rate achievable by an in-line machine. See Weber patent, col. 3, ll. 8-9. Moreover, on in-line chains, the sheets change the direction of their movement roughly ninety degrees when they drop from the feeder to the gathering chain. This abrupt change in direction, if it occurs at high speeds, often causes the pusher on the gathering chain to drive through the sheets and damage them. Id. at col. 1, ll. 34-37.
Weber's patent addressed these problems by designing a machine that would change the conventional in-line gatherer-stitcher processing of sheets by putting several straight in-line chains onto a rotary drum, thereby moving paper circularly as well as linearly. Weber claimed that such a rotary drum machine would increase the time to deposit sheets; reduce the axial speed of the sheets thereby eliminating the problems associated with paper traveling at high speeds; reduce the abrupt changes in the direction of the paper; and increase the time for stitching. The evidence is clear and convincing, however, that the idea of putting straight in-line carriers on a rotary drum did not originate in 1985 with Weber or with Grapha; Ferag earlier designed and produced a rotary' drum binding machine.
In 1976, Ferag received a U.S. patent for a rotary drum inserter for handling sheets of paper in newspaper assembly. A copy of this patent, the "Reist" patent, U.S. Patent No. 3,951,399, which issued on April 20, 1976, is attached as Appendix B. In contrast to magazines, newspapers are normally assembled from the "outside-to-inside" in a pocket in the shape of an open "V", that is, the cover or "outside" goes into the pocket first, followed by the inside pages. This process is known as "inserting"; there is no stapling or stitching in newspaper assembly. The Reist patent's rotary drum allowed the signatures to move closely together, reducing the speed at which they must move through the machine. The Reist rotary, drum concept also permitted reduction of the distance between the signatures below the length of the signature--impossible on an in-line machine -- and also enabled the machine to work with the shortest possible dimension or depth of paper. This design resulted in an extraordinarily high output capacity even though paper did not move through the rotary drum machine at the high speed which creates disturbing air flow in on-line machines.
Specifically, the Reist machine included a horizontally elongated drum. In the embodiment shown in figures 1 to 6 of the Reist patent, the outer sheet or cover of a publication is fed into a drum between the dividing walls of a section of the drum as it rotates. As the drum rotates, the sheets move axially along the drum in a "generally helical" path until the sheets reach the second section of the drum. Pl.'s Findings of Fact P 24. The path is only "generally helical" because, as illustrated in figure 2 of the Reist patent, the path actually becomes circular (i.e.; no axial movement takes place) over the top of the drum. As the sheets rotate under the drum, an endless belt forms a kind of trough under the drum, preventing the sheets from falling out at the bottom of the drum's cycle. After the sheets have rotated completely around the drum and have moved under the second feeding station, more sheets are inserted into the first section of the drum. The sheets continue to rotate with the drum and to move axially, following a "generally helical" path.
Although the Reist patent disclosed a machine primarily for inserting or newspaper assembly, it also proposed some modifications: "the above described apparatus is not limited to the insertion of inserts into articles.... It is possible for numerous other operations to be performed . . . ." Reist patent, col. 8, ll. 7-11. The relevant modification here is that the Reist patent claimed that: "Assembly may also take place from the inside to the outside, the innermost section being supplied to one end of the wheel and the successive further outwardly disposed sections being supplied subsequently." Id. at col. 8, ll. 36-37, 42-46. This statement strongly suggested that the Reist rotating drum machine could assemble sheets from the "outside to inside" in a pocket and could also assemble them from the "inside to outside" on a saddle. Although the Reist patent did not specifically state that "inside to outside" assembly occurs on a saddle, it is a necessary inference that "inside to outside" assembly is feasible only on a saddle such as traditionally used in an in-line machine. May 22, 1996 Trial Tr. at 15-16 (Boss cross).
As early as 1980, Ferag produced and sold the Reist machine and quickly became identified with the high-capacity rotary drum concept in the bindery industry. Its commercial newspaper inserters were the state of the art and well-known in the industry, including Grapha. Pl.'s Ex. 395 (Grunder dep. at 152-53). Ferag's Reist rotary drum inserting machines were also a tremendous commercial success. To date, it has sold over sixty of these machines in the United States, including one at a newspaper bindery in Springfield, Illinois and another at Bergen, New Jersey. See Pl.'s Exs. 372, 373, 379 (Springfield); Pl.'s Exs. 381, 382 (Bergen).
In 1980, Ferag's head of research and development at that time, Jacques Mejer, used the Ferag rotary drum concept to design a rotating drum gatherer-stitcher. The U.S. patent for this design, the "Meier '755" patent, U.S. Patent No. 4,408,755, issued on October 11, 1983 based on a Swiss application filed March 11, 1980. The Meier '755 patent shows a rotary drum with 24 saddles spaced evenly from each other around the drum's circumference. On these saddles, webs of paper (as distinguished from individual sheets of paper) move axially along the length of the drum until they reach the drum's end, where a rotating cylinder stitching apparatus at the end of the drum inserts staples serially on one set of sheets after another.
In 1982, Ferag assigned a team of research and development engineers, namely, Werner Honegger, Egon Hansch, and Erwin Muller, the task of designing a high speed, rotating drum gatherer-stitcher capable of operating on individual sheets of paper using the Reist patent's rotary drum inserter technology. Muller's initial responsibility was to perfect a stitching system for a rotating drum gathering machine. In the process, Muller conceived of an entire system for opening, gathering, and stitching signatures using the Reist rotating drum inserter. Although the Reist inserter used pockets for inserting, and not fins or saddles, it was well-known by 1980, as evidenced by the 1976 Macke patent, that the equipment used to accomplish different ways of assembling products, that is, inserting for newspapers and gatherer-stitching for magazines, could be broadly and flexibly combined to do either. See Pl.'s Ex. 71 (Macke patent); Pl.'s Ex. 389 (Higgins stmt. at 39). In his engineering drawings completed in September and October 1982, Muller proposed a gathering drum for single sheets with a stitching unit for each gathering fin or saddle; Muller also proposed that each sheet be opened and placed in a straddling relationship on the saddles. See Pl.'s Ex. 391 (tabs 3-15).
After this initial design work in 1982, Ferag did not immediately pursue a rotary drum gatherer-stitcher nor did it seek a patent on Muller's concept; it believed that the system was covered by the Reist patent. May 15, 1996 Trial Tr. at 124-25, 148-50 (Muller direct, cross). Two years later, in April 1984, Ferag turned once again to the task of designing a high speed, rotary drum gatherer-stitcher and assigned it to Egon Hansch, one of the original engineers on the project in 1982. From April 1984 to July 1986, Hansch worked on the project using, as the nucleus for his design, Muller's unpatented concept and the patented Reist rotary drum inserter technology, as well as the patented innovations of Meier and Macke. Hansch considered two variations on the Reist rotary drum technology: a "ferris wheel" drum, in which the sheets always remain vertical, and a "star-shaped" drum. Hansch designed, among other things, the schematic drawings of an inside-to-outside-gathering, rotary drum saddle machine which could open and feed individual sheets onto the fins of a rotating drum using moving belts to prevent the product from falling off the fins when rotated to the bottom of the drum, drawn directly from the Reist invention. See Pl.'s Ex. 392 at tab 13. In so doing, Hansch borrowed considerably from existing Reist inserter technology. For example, in designing the use of moving belts to prevent the product from falling off the fins when rotated to the bottom portion of the drum, he proposed using the same belts used on Ferag's inserter drums. The feeding mechanism he designed is identical to the feeding mechanism employed on the Reist inserter drum where the product is gripped at the fold.
Ferag built and tested a prototype of the ferris wheel, rotary drum gatherer-stitcher in 1984. Between 1985 and mid-1986, Ferag continued to develop the ferris wheel rotary drum gatherer-stitcher and shipped a prototype of this design to a German client. In 1987, after studying the advantages and disadvantages of a number of variations of the rotary drum concept, Ferag replaced the ferris wheel approach with the star-shaped drum design. From 1987 to 1990, Ferag worked on the engineering and construction of a star-shaped rotary drum gatherer-stitcher. In 1990, after the issuance of Grapha's Weber patent, Ferag built and exhibited a prototype of the rotary drum gatherer-stitcher at the DRUPA exhibition in Dusseldorf, Germany. Ferag modified this prototype and since 1993, has manufactured and sold it as the IPEX machine, which Grapha claims infringes the Weber patent.
In August 1984, independent of the earlier design work of Heinz Boss and without any knowledge of it, Weber, also a Grapha engineer, designed the rotary drum gatherer-stitcher, which he patented and assigned to Grapha as the Weber patent. Like Boss, Weber was aware of Ferag's Reist patent and inserter drum. Weber recorded his work in, among other documents, a morphological box in which he divided tasks to be performed by a machine into partial functions and sources where one could look for solutions to each part of the problem. Pl.'s Ex. 5. From this morphological box, he could determine the best solution to particular problems. Column 5 of this box addressed how to put a signature over or into another signature. In Box D-5, Weber suggested the solution as the inserting process of the Reist patent. Boxes F-10 and F-11 in the morphological box also illustrate the Reist rotating-drum concept.
Weber also wrote a "Study Report," in which he detailed the results of his experimentation and work and discussed the pertinent prior art and possible patent protection. In it, he noted that "there will have to be a comprehensive and clear delineation of features in terms of patent protection since a similar design approach already exists in the realm of newspaper processing." Pl.'s Ex. 21 at 6. This "design approach" was Ferag's Reist inserter drum. And the same advantages claimed by Weber in his patent -- that a rotary drum machine would reduce the axial speed of the sheets thereby eliminating the problems associated with paper traveling at high speeds and reduce the abrupt changes in the direction of the paper -- were those achieved by Ferag's Reist inserter machine.
Although it has always been Grapha's goal to increase the speed of its gatherer-stitcher to equal that of the printing presses and although Grapha considered Weber's design to be revolutionary, Grapha never made either of the rotary drum gatherer-stitchers described in Weber nor made a machine covered by any claim of the Weber patent. Stip. Facts PP 38, 39, 40. Grapha did not attempt to commercialize the Weber patent because it believed, among other things, that in 1985, the market was not ready for rotary drum gatherer-stitcher. May 22, 1996 Trial Tr. at 38 (Boss cross). Furthermore, Grapha realized that by 1985, because of the Reist inserter drum, the rotating drum principle had become Ferag's marketing image, so that Grapha's manufacture of a rotary drum could be perceived as a "me-too" to Ferag's Reist rotating drum inserter. May 22, 1996 Trial Tr. at 38-42 (Boss cross); Pl.'s Ex. 395 (Grunder dep. at 80-81).
The legal principles applicable to the foregoing facts are not in serious dispute. As a matter of statute, Grapha's Weber patent-in-suit "shall be presumed valid." 35 U.S.C. § 282. It remains valid until the party asserting invalidity proves by "clear and convincing evidence" that the patent "is no longer viable as an enforceable right." Roper Corp. v. Litton Sys., Inc., 757 F.2d 1266, 1270 (Fed. Cir. 1985). Nevertheless, it was long ago established that "it was never the object of [the patent] laws to grant a monopoly for every trifling device, every shadow of a shade of an idea, which would naturally and spontaneously occur to any skilled mechanic or operator in the ordinary progress of manufactures. Such an indiscriminate creation of exclusive privileges tends rather to obstruct than to stimulate invention." Atlantic Works v. Brady, 107 U.S. 192, 200, 27 L. Ed. 438, 2 S. Ct. 225 (1883); see also Goodyear Tire & Rubber Co. v. Ray-O-Vac Co., 321 U.S. 275 279, 88 L. Ed. 721, 64 S. Ct. 593 (1944) (Black, J., dissenting). Indeed to justify the issuance of a patent, a "genuine 'invention' or 'discovery' must be demonstrated 'lest in the constant demand for new appliances the heavy hand of tribute be laid on each slight technological advance in the art.'" Sears, Roebuck & Co. v. Stiffel Co., 376 U.S. 225, 230, 11 L. Ed. 2d 661, 84 S. Ct. 784 (1964) (Black, J.) (quoting Cuno Engineering Corp. v. Automatic Devices Corp., 314 U.S. 84, 92, 86 L. Ed. 58, 62 S. Ct. 37 (1941)). Therefore, a patent may not be obtained for the product of mere mechanical skill or for simply carrying forward the application of a prior device with a change only in degree. See Phillips v. Detroit, 111 U.S. 604, 608, 28 L. Ed. 532, 4 S. Ct. 580 (1884); Smith v. Nichols, 88 U.S. 112, 22 L. Ed. 566 (1874).
Ferag identifies five independent grounds for its challenge to the validity of the Weber patent, any one of which would invalidate the patent. Specifically, Ferag argues that the Weber patent is invalid because: (1) it defines a gatherer-stitcher machine that was known before Weber designed his machine and that the concept of the Weber machine was obvious to those of ordinary skill in the art prior to June 1985, the date the Weber patent application was first filed in Switzerland, in violation of 35 U.S.C. § 103; (2) it fails to disclose the best mode of the invention as required by 35 U.S.C. § 112, P 1; (3) it does not contain an enabling disclosure as required by 35 U.S.C. § 112, P 1; (4) it lacks utility as required by 35 U.S.C. § 101; and (5) certain of its claims are not definite as required by 35 U.S.C. § 112, P 2. In the circumstances here, it is unnecessary to resolve the issues of best mode, utility, and indefiniteness. Ferag has shown by clear and convincing evidence that, at ...