Case No. 2012-1254 (Prost, Mayer, Moore) (nonprecedential)
Seeing a trend here? In re Chevalier is yet another case involving an obviousness rejection affirmed by facts under substantial evidence review on appeal from the USPTO. In addition to the difficult review standards facing applicants at the Federal Circuit, here the facts on appeal underpinning the Board’s determination of obviousness are stacked up fairly high against the patentee. The case reminds that combining two prior art embodiments to reject a claim under § 103 does not require actual, physical combinability of the embodiments.
Chevalier, through American Air Liquide, applied for a patent, in application 11/407,778, on a liquid stirrer that formed a gas-liquid dispersion in a reactor. Specifically, the ‘778 invention included a rotor sandwiched by upper and lower deflectors on a single shaft. A gas injector could introduce gas to the mixture above the rotor, and the upper deflector managed the froth level while the lower deflector ensured axial fluid flow was redirected radially. The claim at issue for this invention looked like this:
A device for stirring a liquid in a reactor and for injecting a gas into the said liquid in order to form a gas-liquid dispersion, comprising:
a) a drive device positioned above the reactor, provided with a vertical output shaft fitted at its lower end with at least one axial flow rotor immersed in the liquid;
b) means for introducing gas above the axial flow rotor; and
c) a deflector placed above the axial flow rotor preventing the gas-liquid dispersion from rising, wherein said device comprises a deflector secured to the shaft placed below the axial flow rotor converting the axial flow of the said rotor into a radial flow.
The Examiner rejected this claim as obvious over two US Patents to Kwak and Howk (no relation?), and the Board affirmed the rejections. Specifically, Kwak taught everything in the claim except for putting the deflector on the same shaft as the rotor; Kwak put its rotor on a separate shaft extending up from the bottom of the reactor. Howk disclosed a same shaft rotor-and-deflector combo for mixing gasses and liquids. The Examiner and Board reasoned that modifying Kwak to use only the same-shaft arrangement from Howk would be obvious, because it would be easy to do and it would result in a more reliably-deflected flow.
Chevalier appealed, arguing that Kwak and Howk taught away from their combination because such combination would be inoperable or not achieve a desired functionality of Kwak. Specifically, Chevalier illustrated that adding a deflector right below the rotor in Kwak would drive fluid into a tube wall, complicating mixing and aeration, while not achieving a desired bottom-scouring provided by the bottom-mounted deflector in Kwak.
The Federal Circuit instead credited the Board’s rationale that the Kwak-Howk combination would improve radial flow and that a person skilled in the art would have merely adjusted the length of the shaft in Kwak as modified by Howk with the lower deflector to avoid loss of function. As the opinion noted, from In re Gartside, “presence or absence of a motivation to combine references in an obviousness determination is a pure question of fact” reviewable only for substantial evidence. Here, the motivation to combine and avoid inoperability were quite within that standard.
Moreover, the panel found Chevalier’s arguments of a strict melding of Kwak and Howk resulting in loss of function to be legally misguided. Obviousness is not defeated by physical incompatibility resulting from wholly melding two prior art embodiments or from strictly combining prior art structures without any adjustment:
The obviousness inquiry does not ask “whether the references could be physically combined but whether the claimed inventions are rendered obvious by the teachings of the prior art as a whole.” In re Etter, 756 F.2d 852, 859 (Fed. Cir. 1985) (en banc); see also In re Keller, 642 F.2d 413, 425 (CCPA 1981) (stating “[t]he test for obviousness is not whether the features of a secondary reference may be bodily incorporated into the structure of the primary reference”). Rather, in a case such as this where each of the elements of the claim are known to the art, the obviousness inquiry requires a finding that the combination of known elements was obvious to a person with ordinary skill in the art. KSR, 550 U.S. at 420. Here, each of the elements of the claim is taught by Kwak with the exception that the Kwak deflector is not attached to the drive shaft. This deficiency in Kwak is supplemented by Howk, which teaches a deflector attached to the drive shaft.
The Applicants also did not help themselves by admitting on appeal that the deflectors in Kwak and Howk were equivalents that both converted axial flow to radial flow. The court seized on this admission as further evidence of obviousness because, under KSR, “when a patent claims a structure already known in the prior art that is altered by the mere substitution of one element for another known in the field, the combination must do more than yield a predictable result.”
Lessons: Those recalling my previous posts know that I’m no fan of the cited Keller opinion, because it sometimes short-circuits consideration of impermissible or inoperable modification insofar as it allows such arguments to be dismissed as premised on strict physical incorporation. This effectively sidesteps more subtle points dealing with loss of an element’s prior art function in the claimed combination. Keller does, however, have viability and proper applications. Chevalier is clearly one, where an applicant relied on an exact placement of one prior art structure into another to show inoperability. The take-away is, if arguing that swapping an element from a secondary reference into a primary reference is not permitted under § 103 because previous functionality is changed or lost, always be critical of whether such functionality can be preserved through further simple or routine accommodations in the modification. If one specific way of joining two prior art references, or importing separate and unapplied structures or teachings from the secondary reference, is the only way to achieve an argued loss of functionality, you’re probably doing it wrong. As seen in Chevalier, simple movement/repositioning of combined elements within the skill of the ordinary practitioner is certainly fair game to undo any argument of loss of function. Arguments of non-combinability should instead focus on how any reasonable combination of the prior art destroys or changes operation of prior art elements.
Also consider some other lines of attacking obviousness rejections by substitution that are more insulated from Keller: arguing that a prior art element performs a wholly new, unrecognized function only in the claimed combination; or arguing that functionality-preserving modification was beyond the skill of the ordinary artisan (the latter likely requiring evidence, like a non-enabling argument). Finally, if a substitution seems too narrowly or arbitrarily selected out from one piece of prior art or strangely spliced into another, avoid general statements that a skilled person would not have pursued such a strange physical incorporation; these sometimes raise Keller-based rebuttals that the combined teachings of the prior art, not their bizarro physical incorporation, is being relied upon to reject the claims. Instead, take a look at the rationale provided for combination. Gerrymandered excisions and junctions between the prior art are often hindsight-based and not fully explained by simple rationales for combining. Clearly attacking the rationale as not supporting such picking and choosing can result in a more thoughtful, and Keller-free, response from the Examiner.
This horse has been beaten dead in the last few posts, but Chevalier also reminds that motivation to combine/modify distinct prior art embodiments is a question of fact that will be affirmed on appeal unless unreasonable or supported by only a scintilla of evidence in the record. Consider the previous lessons from In re Hyon, suggesting that legal arguments be advanced on appeal over factual ones. Of course, whatever your arguments against an obviousness rejection, it is probably a good idea to avoid any admission that prior art structures are known equivalents providing a same structure, unlike the appellants in Chevalier.
Comment: Sometimes it is a relief to see an opinion like Chevalier designated nonprecedential, because it involves lopsided facts and poor arguments (of course, with facts this poor, there are few, if any, “good” arguments to be had). The label makes it clear that the court is not treading or glossing doctrinal edges, while preventing what could be a harsh and/or too-easy conclusion from ever finding inappropriate application in other contexts.
Also, what is up with the claim structure here? Why bury the lower, critical deflector – a claim element – in a wherein clause while explicitly reciting each other element, including the upper deflector?