112 F. 3d 1163 - Robotic Vision Systems Inc v. View Engineering Inc

Robotic Vision Systems, Inc. appeals from the summary judgment of the United States District Court for the Central District of California holding that U.S. Patent 5,463,227 is invalid on the grounds of failure to disclose the best mode of carrying out the invention and violation of the on-sale bar. Robotic Vision Sys., Inc. v. View Eng'g, Inc., No. 95-CV-7441, 1996 WL 383900 (C.D.Cal. Mar. 5, 1996). Because the district court erred in holding that the patent is invalid on the grounds of failure to comply with the best mode requirement and the on-sale bar, we reverse-in-part, vacate-in-part, and remand.

The '227 patent discloses a method of using a three-dimensional sensor in order to scan and inspect the leads of integrated circuit chips. The method requires a device that contains motors for moving the sensor along an x-axis and a y-axis. The device also has an x-axis linear encoder and a y-axis linear encoder, which produce information relating to a position of the sensor along the corresponding axis. The patent contains a figure illustrating an embodiment of the machine which lacks means connected to the motors for controlling them, and the figure likewise lacks a connection to the linear encoders for receiving positional information. In addition, the specification does not indicate how the motors that move the sensor are controlled.

A prior art system scanned all four sides of one chip before inspecting the next chip. The method of the '227 invention scans rows and columns of chips, reducing the number of changes in the direction of the scanner when compared with the prior art method. This results in improved scanning time, because the reduced number of changes in direction translates into reduced numbers of scanner movements of acceleration and deceleration. Claim 1 of the patent is the only claim in question on appeal. It recites the row and column scanning feature, referred to as full-tray scanning, and reads as follows:

1. A method for obtaining three-dimensional data from devices having corresponding sides, comprising the steps of: providing a multi-pocketed tray with tray pockets arranged in rows and columns; scanning sequentially with at least one three-dimensional sensor corresponding sides of said devices in a row or column; and repeating said scanning step for all rows and columns containing sides of said devices from which data is to be obtained.

The application for the '227 patent was filed on June 24, 1992, and it thus has a critical date of June 24, 1991 for purposes of 35 U.S.C. § 102(b). In March 1991, before the critical date, Robotic employees agreed to a request from a representative of Intel to supply Intel with a full-tray scanning system. According to Robotic, the invention at that time was still in a developmental stage, involving rewriting the software program contained in a prior art system, adding new sections of software code, and testing the software in order to demonstrate that the invention would work effectively. The patent issued on October 31, 1995, and on that day Robotic sued View Engineering, Inc., alleging that certain of View's three-dimensional scanning machines infringed the patent.

View moved for summary judgment, arguing that the patent was invalid on the grounds of failure to disclose the best mode of carrying out the invention and violation of the on-sale bar. The district court granted View's motion, concluding that, at the time the application for patent was filed, the inventors knew that using software was the only mode of practicing the invention, that it was undisclosed, and that the patent was accordingly invalid. Our prior holding in In re Hayes Microcomputer Prods., Inc. Patent Litigation, 982 F.2d 1527, 1537-38, 25 USPQ2d 1241, 1248-49 (Fed.Cir.1992), that disclosure of software code is not necessary if the functions of the software are disclosed was distinguished on the ground that, contrary to Hayes, there was no disclosure at all in the patent that would lead one skilled in the art to use software. With respect to the on-sale bar, the court concluded that the full-tray scanning method was sufficiently developed before the critical date to trigger the bar. While the court noted that the software program for implementing the invention was completed in draft form in May 1991, before the critical date, it also noted that the invention does not depend upon software and thus that completion of the software program was not necessary to trigger the bar. Accordingly, the court held that there were no genuine issues of material fact and that View was entitled to judgment as a matter of law. Robotic now appeals to this court.

Summary judgment is appropriate when there is no genuine issue as to any material fact and the moving party is entitled to judgment as a matter of law. Fed.R.Civ.P. 56(c); Johnston v. IVAC Corp., 885 F.2d 1574, 1576-77, 12 USPQ2d 1382, 1383 (Fed.Cir.1989). Thus, summary judgment may be granted when no "reasonable jury could return a verdict for the nonmoving party." Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248, 106 S.Ct. 2505, 2510, 91 L.Ed.2d 202 (1986). In determining whether there is a genuine issue of material fact, the evidence must be viewed in the light most favorable to the party opposing the motion, with doubts resolved in favor of the nonmovant. Transmatic, Inc. v. Gulton Indus., Inc., 53 F.3d 1270, 1274, 35 USPQ2d 1035, 1038 (Fed.Cir.1995). We review de novo a district court's grant of summary judgment. Conroy v. Reebok Int'l, Ltd., 14 F.3d 1570, 1575, 29 USPQ2d 1373, 1377 (Fed.Cir.1994).

Robotic argues that summary judgment was inappropriate because there were genuine issues of material fact concerning whether the specific software program used by Robotic was necessary for successfully carrying out the invention and therefore had to have been disclosed. Robotic argues that the court erred in concluding that software was the best mode known to the inventors of carrying out the invention. View responds that the inventors had a best mode, the nature of which they failed to disclose, and that that best mode included software. Robotic also argues that, since software is not part of the claims, it need not have disclosed software to satisfy the best mode requirement. In any event, Robotic argues that the specification is adequate to lead one of ordinary skill in the art to use software in the invention. View responds that no amount of skill in the art can overcome the lack of disclosure.

The patent statute requires that a patent specification "shall set forth the best mode contemplated by the inventor of carrying out his invention." 35 U.S.C. § 112 (1994). Determining whether a patent satisfies the best mode requirement involves two factual inquiries. First, a fact-finder must determine whether at the time an applicant filed an application for a patent, he or she had a best mode of practicing the invention, which is a subjective determination. Second, if the inventor had a best mode of practicing the invention, the fact-finder must determine whether the best mode was disclosed in sufficient detail to allow one skilled in the art to practice it, which is an objective determination. United States Gypsum Co. v. National Gypsum Co., 74 F.3d 1209, 1212, 37 USPQ2d 1388, 1390 (Fed.Cir.1996); Chemcast Corp. v. Arco Indus. Corp., 913 F.2d 923, 927-28, 16 USPQ2d 1033, 1036 (Fed.Cir.1990).

We conclude that there is no genuine issue of material fact regarding the best mode and that the specification is not deficient concerning the disclosure of the best mode. As noted, the requirement to disclose the best mode of carrying out one's invention is statutory and one does not usually meet a statutory requirement for disclosure with silence. While a disclosure necessary to meet a statutory requirement is to be understood from the standpoint of one skilled in the relevant art, a certain basic disclosure is needed of the best mode. Cf. Genentech, Inc. v. Novo Nordisk, A/S, 108 F.3d 1361, 1366-67 (Fed.Cir.1997) (stating that the knowledge of one skilled in the art may supplement a disclosure for purposes of satisfying the enablement requirement, but it is "not a substitute for a basic enabling disclosure"). Even an only mode must be disclosed, expressly or implicitly. Moreover, the fact that the use of software or a computer is not mentioned in the claims of the Robotic patent does not, contrary to Robotic's argument, exempt such use from the requirements of a best mode disclosure, since carrying out the invention usually involves more than what is expressly claimed.

On the other hand, the inventors in this case disclosed a device for carrying out their method, and it is plainly apparent that a computer, operating under software control, is to be interfaced to the device for controlling the movement of the sensor. Something must be connected to the device for providing control signals to the motors and for receiving information from the linear encoders concerning a position of the sensor, and there is no dispute that that something is a computer.

The facts support the conclusion that software is to be used. William Yonescu, one of the named inventors, averred in an affidavit that, as of the filing date of the patent, he and the other inventors knew of no mode other than what was described in the specification. View cites a different portion of the same affidavit, in which Yonescu averred that, other than the software code developed to implement the invention, he and the other inventors "did not know of any other software program, code, or other mode which existed as of June 24, 1992, that was useful to carry out the mode disclosed in the specification of the '227 patent." View interprets this statement as an admission that at least two modes existed, the mode disclosed in the patent and the mode consisting of software, but we do not consider that this statement raises a genuine issue of fact. It is not an admission. It simply states that a software program was the only means contemplated of carrying out the invention. From the record before us, it is clear that a software program was involved in the carrying out of the invention and that no other mode existed.

Moreover, as asserted by Robotic, it would have been apparent to one skilled in the art, knowing that software was used in the prior art system, to use software for implementing the improved scanning method claimed in the patent. Yonescu averred that: "A person of ordinary skill in the art to which the '227 patent pertains would know and understand that software is needed to perform the patented method. The details of such software would also be within the skill of a person of ordinary skill in the art to which the '227 patent pertains." View has not provided a basis for concluding that Yonescu's assertions are not correct. Thus, one cannot conclude that a person skilled in the art would not have known that software was the best mode of carrying out the invention and how to implement it. The patent cannot be held to fail to comply with the best mode requirement for lack of the word "software," the use of which was plainly apparent to one skilled in the art. Such a disclosure was implicit in the specification.

Finally, it has not been shown that there is a genuine issue as to whether one skilled in the art would have known how to create specific source code for this purpose. We have previously held in Hayes and in Fonar (after the district court decided this case) that when disclosure of software is required, it is generally sufficient if the functions of the software are disclosed, it usually being the case that creation of the specific source code is within the skill of the art. Fonar Corp. v. General Electric Co., 107 F.3d 1543, 1549-50 (Fed.Cir.1997); Hayes, 982 F.2d at 1537-38, 25 USPQ2d at 1248-49. The functions that a software program would instruct the computer to perform for controlling the machine are readily apparent from the specification of the patent at issue here, which describes the scan paths and parameters for full-tray scanning. View has not presented any evidence to controvert Robotic's assertion that one skilled in the art could generate the necessary software program to implement the disclosed functions. We therefore must conclude that the district court erred in granting summary judgment to View that the '227 patent is invalid for failure to disclose the best mode.

The district court also held the patent invalid on the ground that, by virtue of Robotic's acceptance of Intel's request in March 1991, the invention was on sale before the critical date. Robotic argues that, at the time of that acceptance, the invention was not sufficiently completed to have triggered the on-sale bar. View responds that the remaining development and testing related solely to software, which was not a part of the claimed invention. Thus, according to View, because Robotic was not working on a part of the invention that is recited in the claims, its alleged experimentation cannot prevent the imposition of the on-sale bar.

Application of the on-sale bar under section 102(b) is a question of law based upon underlying issues of fact. See KeyStone Retaining Wall Sys., Inc. v. Westrock, Inc., 997 F.2d 1444, 1451, 27 USPQ2d 1297, 1303 (Fed.Cir.1993). A determination that an invention was on sale within the meaning of the statute requires that "the claimed invention asserted to be on sale was operable, the complete invention claimed was embodied in or obvious in view of the device offered for sale, and the sale or offer was primarily for profit rather than for experimental purposes." Id. In explaining the application of the on-sale bar, we recently stated that:

even though the technical requirements of a reduction to practice have not been met, a sale or a definite offer to sell a substantially completed invention, with reason to expect that it would work for its intended purpose upon completion, suffices to generate a statutory bar.

Micro Chem. Inc. v. Great Plains Chem. Co., 103 F.3d 1538, 1545, 41 USPQ2d 1238, 1244 (Fed.Cir.1997); see Seal-Flex, Inc. v. Athletic Track and Court Constr., 98 F.3d 1318, 1322, 40 USPQ2d 1450, 1452 (Fed.Cir.1996) (stating that application of the on-sale bar requires consideration of whether "the invention was in fact complete and was known to work for its intended purpose").

The claimed invention in this case is a method. What Robotic is accused of putting on sale in March 1991 is a device for carrying out that method. Thus, an offer to sell the device (or otherwise provide it in a commercial context), if it met the requirements for an on-sale bar, would constitute a bar to the patentability of the claimed method. However, an alleged offer to sell the device is not an on-sale bar to the patentability of the method unless the activities involving the device meet the criteria for the on-sale bar. In this case the software program, even though not recited in the claims, was necessary for the operation of the invention. Without workable software, there was no device that could have been on sale, and the method of using such a device therefore could not have been on sale. Since it is not clear from the record before us when the software was completed, we conclude that there is a genuine issue as to whether the making of the device was substantially completed before the critical date.

There is no question that a completed device necessary to carry out the claimed method did not exist in March 1991. Daniel Briceno, who wrote the software program for the prior art system, averred that when he subsequently rewrote the software program for the prior art system in order to implement the full-tray scanning concept, it was not until May 22, 1991, well after the alleged offer to sell, that it existed in a form that could be tested. He also stated that upon initial testing of the software program, the system did not work satisfactorily and that problems with the code were discovered. Thus, it is undisputed that the software program for implementing the invention was not even close to completion at the time of the alleged offer. The record in fact indicates that the invention was apparently completed in the latter part of June 1991, perhaps after the critical date.

The precise date when this occurred makes all the difference to the resolution of this case. That is because subsequent completion of an invention after the critical date does not relate back to the date of an earlier alleged offer of sale. An offer of sale, to be a bar within the meaning of section 102(b), must be of an invention that is substantially complete at the time of the offer. See Micro Chem., 103 F.3d at 1545, 41 USPQ2d at 1243. If mere discussions prior to the critical date, or even an agreement to develop and provide a device that had not yet been invented, developed, or completed were to be held to be a bar to patentability, then collaboration between inventors and customers would be greatly impeded. Patent applications would be required to be filed prematurely, before an invention was completed. The on-sale bar was not intended to prevent discussions between potential inventor-suppliers and customers concerning inventions not yet completed. Thus, the later completion of an invention concerning which an alleged offer to sell had been made earlier does not relate back to the date of that offer.

On the other hand, the record shows that the development of the software for the device in question was completed no later than soon after the critical date. We do not know whether in fact it might have occurred just prior to that date. If it did occur prior to that date, it clearly cannot be considered in isolation from the alleged offer since the development of the software was pursuant to the offer. Completion of the invention prior to the critical date, pursuant to an offer to sell that invention, would validate what had been theretofore an inchoate, but not yet established, bar. It would validate it, however, as of the date of that completion, not the date of the original offer. Completion after the critical date, even if pursuant to the offer, would not create a bar.

View argues that finalization of the software program was unnecessary for completion of the invention as claimed and that the invention as claimed could have been on sale without the completion of the program. It notes that the claims of the patent do not expressly require software. As indicated above, however, the development of the software was essential to the substantial completion of the device to be provided to Intel for use in practicing the claimed method, even if it was not part of the claims. Without a substantially completed invention, there could be no on-sale bar.

We therefore conclude that there is a genuine issue of material fact relating to whether a statutory bar was created by the discussion with Intel concerning possible supply of a product for which the software program had not yet been developed. That issue is: when was the software completed? If it was completed before the critical date, then it validates the offer to sell the machine to be used in the method of the patent, and the patent is invalid because of the on-sale bar. If it occurred after the critical date, then the earlier offer was not an on-sale bar. The alleged offer by itself before the critical date could not have triggered the on-sale bar, as the invention was not substantially complete at that time. See Micro Chem., 103 F.3d at 1545, 41 USPQ2d at 1244. Accordingly, the district court erred in granting summary judgment that the '227 patent is invalid on the ground of the on-sale bar. We therefore vacate the grant of summary judgment on that ground and remand for further fact-finding.

This case presents the interesting situation in which the use of software was sufficiently within the skill of the art so as to not run afoul of the best mode requirement, but it may not have been sufficiently developed by the critical date to cause an offer of the product to constitute an on-sale bar. The reason for this apparent inconsistency is that the relevant statutory provisions have different purposes, one to promote prompt filing of patent applications after commercialization has begun, and the other to compel full disclosure of an invention. Knowledge of one skilled in the art is relevant to satisfaction of the best mode requirement, but such knowledge does not create a completed invention in an on-sale context when the invention at the time of the offer is nothing more than a concept for improving a method of using a prior art machine. The software program is an issue in both contexts, but in different ways.

The district court erred in holding that View was entitled to judgment as a matter of law that the patent was invalid for failure to disclose the best mode. With respect to the on-sale bar, the court erred in concluding that there was no genuine issue of material fact concerning whether the invention was on sale within the meaning of section 102(b). Accordingly, we reverse the summary judgment that the patent is invalid for failure to disclose the best mode, vacate the judgment that it is invalid based on the ground of the on-sale bar, and remand for further proceedings consistent with this opinion.

REVERSED-IN-PART, VACATED-IN-PART, AND REMANDED.