Center for Advanced Study & Research on Innovation Policy


CASRIP Newsletter - Spring/Summer 1997, Volume 4, Issue 2

Best Mode for Software-Related Inventions

Since the Federal Circuit's In re Alappat en banc decision and the changeof the USPTO's examination guidelines on software patentability, long-debated Section 101issues regarding the patentability of computer software-related inventions appear to besettled. Discussions now focus on issues more common to other technology fields, such asnovelty, nonobviousness and Section 112 requirements. Two recent Federal Circuit decisionsinvolving the best mode requirement offer practical guidelines on preparing patentapplications and in managing the pre-filing process of software-related inventions.

The first case is Fonar Corp. v. General Electric Co.1,in which a patentee of a technique for scanning a patient's body sued a competitor forinfringement and for inducement of infringement. The patentee requested an award ofdamages in the form of lost profits and reasonable royalties.

The patentee's technique related to a magnetic resonance imaging ("MRI")machine which used the principle of nuclear magnetic resonance ("NMR") toproduce cross-sectional images of body tissues. Although prior art machines could providemultiple image views of a patient's body at different angles, the prior art machinesobtained images along the same axis in a single scan, and thus required multiple scans toobtain the images at varying angles. The patentee developed a method for obtaining suchimages in a single scan, a method known as "multi-angle oblique (MAO)," whichsignificantly shortened imaging time.

Claims in the disputed patent recited steps for positioning and scanning an object. Thepatentee also held another patent which relates to an application of the MAO technique todetect cancer.

The accused infringer raised a defense of invalidity against both patents, allegingthat the patentee did not meet the best mode requirement, because it did not disclose thesource codes which accomplished the MAO imaging technique. A jury found the two patentsvalid and infringed, and awarded both lost profits and reasonable royalties; the districtcourt agreed with the jury, and denied the accused infringer's motion for JMOL regardingthe alleged best mode violation. Accordingly, the accused infringer appealed to the U.S.Court of Appeals for the Federal Circuit.

In reviewing the grant of motion which denied the alleged best mode violation, theFederal Circuit agreed with the district court that the patentee met the best moderequirement. After reciting the two-part test in determining whether a patent satisfiesthe best mode requirement,2 the court gave a general rulethat "where software constitutes part of a best mode of carrying out an invention,description of such a best mode is satisfied by a disclosure of the functions of thesoftware."

The court explained the reason of this rule by stating that "normally, writingcode for such software is within the skill of the art, not requiring undueexperimentation, once its functions have been disclosed." As long as the functions ofthe best mode software were adequately disclosed, flow charts or source code listingswould not be required under Section 112. Since substantial evidence supported thesufficient disclosure of the functions of the patentee's basic patent to satisfy the bestmode requirement, the court affirmed the district court's ruling, and denied the allegedviolation.

Unrelated to software, the accused infringer charged another best mode violation withrespect to the disclosure of an element called a "gradient multiplier board(GMB)". The accused infringer insisted that GMB was a critical element to accomplishMAO imaging, but a comparator used as part of the GMB was not disclosed in thespecification. The Federal Circuit disagreed. Important testimony offered by one of theinventors was that an ordinary skilled engineer would know about the use of a comparator.The court also indicated that the GMB used in the patentee's machine was not the onlymeans to accomplish MAO imaging, and an ordinary skilled engineer would have known fromthe disclosure of functions in the patent what parts were required, depending on theselected type of GMB.

Relying on this testimony, the Federal Circuit found that substantial evidencesupported the jury's finding of no best mode violation with respect to the disclosure ofGMB. The court also denied the need to specifically identify a particular manufacturer'sproduct to conduct functions disclosed in the patent.

The Fonar Court's general rule on the best mode of software-related inventionswas further reaffirmed in Robotic Vision Systems3.Like the patentee in Fonar, Robotic's patent related to a scanning technique andused a three-dimensional sensor to inspect the leads of integrated circuit chips.

A prior art system scanned all four sides of one chip before inspecting the next chip.The patented system scanned rows and columns of multiple chips and thus shortened scanningtime by reducing the number of direction changes of the scanner. The disputed claimrecited the movement of the scanner with respect to a multi-pocketed tray on whichpositioned chips were arranged in rows and columns.

The accused infringer moved for summary judgment, alleging that the patent violated thebest mode requirement because Robotic knew that software was the only mode to accomplishthe scanning process, but did not disclose it. Agreeing with the accused infringer, thedistrict court granted the motion, ruling that the patent was invalid for a best modeviolation. The accused infringer also argued that the patent was invalid for a violationof the on-sale bar, because Robotic agreed with Intel to supply the scanning system beforethe critical date for purpose of an on-sale bar, i.e., one year prior to the applicationdate. Robotic argued that the on-sale bar did not apply because the software to operatethe system was incomplete when the agreement was made.

However, the district court held that the system was sufficiently completed before thecritical date to trigger the on-sale bar, finding the completion of software inessentialto the completion of the invention and irrelevant to the application of the on-sale bar.Ruling that there were no genuine issues of material fact, the district court granted asummary judgment for the accused infringer.

On appeal, the Federal Circuit reversed the district court's ruling on the best modeviolation and vacated the issue of an on-sale bar. The court disagreed that Robotic's lackof discussion on the use of software gave rise to a best mode violation, because it wasapparent to one skilled in the art, particularly from the disclosure of the prior artcited in the specification, that a computer must be interfaced with the device forcontrolling the movement of the scanner and software should be used to accomplish thecontrol.

Further, the testimony from an inventor revealed that Robotic did not know of anycomputer software which could properly control the movement on the application date, sincethey were still in the development stage of such software. Because the accused infringerdid not produce any evidence to show otherwise, the court found that a disclosure of theuse of software was implicit in the specification. Regarding the lack of disclosure on thesoftware, the court repeated the general rule in Fonar and applied it to this casesince no evidence to negate the enabling quality was introduced by the accused infringer.

The discussions on the on-sale bar are worth mentioning because the Federal Circuitfound that the software may not have been sufficiently developed by the critical date soas to trigger the on-sale bar, although it found that the use of software and thedisclosure of its functions were sufficiently within the skill of the art so as to meetthe best mode requirement. The court explained this apparent inconsistency as resultingfrom different policy considerations relating to the best mode requirement, i.e., topromote full disclosure of an invention and the on-sale bar, i.e., to encourage promptfiling of patent applications after the commercialization of the invention.

To apply the on-sale bar, an invention must be ready for sale; this requires theclaimed invention asserted to be on sale must be operable and the completed inventionbeing embodied in or obvious in view of the device offered for sale. The degree ofcompletion of an invention to trigger the bar does not necessarily correspond to thedegree of completion necessary to establish a reduction to practice under the priorityrule. The Federal Circuit recently reaffirmed this different standard, noting 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 toexpect that it would work for its intended purpose upon completion, suffices to generate astatutory bar." 4

Applying this rule, the Federal Circuit found that the software program, even thoughnot recited in the claim, was essential for the operation of the invention. Thus, thesubstantial completion of the claimed invention depended on the degree of softwaredevelopment. Robotic's offer to sell the system did not give rise to a bar until thesoftware program was sufficient to provide the system with an intended operation. Tosupport the view requiring a substantial completion to trigger the bar, the courtexplained that holding otherwise would discourage communication between inventors andprospective customers, and encourage applications of premature inventions. No evidence toindicate whether the software was sufficiently developed to meet the criteria before thecritical date was available on record. Accordingly, the Federal Circuit concluded thatthere was a genuine issue as to the date of substantial completion of the invention, andthe court vacated the grant of summary judgment on the ground of an on-sale bar, andremanded the case for further fact-finding.

-- Toshiko Takenaka

1107 F.3d 1543, 41 USPQ 2d(Fed. Cir. 1997)
2The test consists of two factualinquiries (1) whether an applicant had a best mode of practicing the invention at the timewhen the application was filed (subjective determination); and (2) whether the best modecontemplated by the applicant was sufficiently disclosed to enable one skilled in the artto practice it (an objective determination).
3Robotic Vision Systems Inc. v.View Engineering Inc. -- F.3d --, 42 USPQ2d 1619 (Fed. Cir. 1997).
4Micro Chem. Co. v. Great PlainsChem. Co., 103 F.3d 1538, 1545, 41 USPQ2d 1238, 1244 (Fed. Cir. 1997)

Connect with us:

© Copyright 2015, All Rights Reserved University of Washington School of Law

4293 Memorial Way Northeast, Seattle, WA 98195