Two PTAB decisions recently designated as informative show that failure to provide detailed evidence of motivation to combine references for an obviousness challenge, can sink a Petition before or after institution of trial.  In Johns Manville Corp. v. Knauf Insulation, Inc. (IPR2018-00827, Paper 9), the PTAB denied institution of an IPR for lack of reasons to substitute compositions in one reference with those of another.  In Hulu, LLC v. Sound View Innovations, LLC (IPR2018-00582, Paper 34), the PTAB reached a final decision that the claims of a telecommunications patent were not obvious where Petitioner’s motivation to combine rationale was undermined by contrary evidence.  These decisions illustrate the demanding level of proof required of Petitioners to articulate the why and the how a person of ordinary skill in the art (POSA) would combine or modify the teachings of the references to arrive at the claimed invention. 

In the first case, Johns Manville petitioned for review of numerous claims of Knauf’s U.S. Patent No. 9,828,287 (“the ’287 patent”), asserting unpatentability for obviousness based on several combinations of four prior art references.  The ’287 patent is directed to “a thermal or acoustical fiberglass insulation material [that comprises] a collection of glass fibers and . . . binder[s] with various characteristics.”  A key feature of the binders is that they included a reaction product between “a reducing sugar reactant and an amine reactant.” Johns Manville, Paper 9 at 3-4.  Patent Owner, Knauf, successfully argued that Petitioner “picks and chooses binder components without explanation,” and did not provide a valid rationale for combining references.  Id. at 9.

In agreeing that the proposed obviousness grounds failed to adequately show a motivation to combine or modify the references, the Board focused primarily on Petitioner’s first ground based on the Srinivasan and Worthington references.¹ The PTAB described Srinivasan as disclosing an aqueous binder composition for making glass fiber products, such as fiberglass insulation.  Id. at 6.  Its binder is not the same as the claimed product of a reducing sugar and amine.  Petitioner relied on Worthington for teaching the claimed binder which is disclosed as useful for “the making of shell molds or cores and for other purposes.”  Id. at 7.  Petitioner argued that a POSA would have substituted Worthington’s binder for Srinivasan’s because Worthington is analogous art and both references disclose thermosetting binder compositions with overlapping constituents. Id. at 10.  Relying on Securus Techs.² and Personal Web Techs.,³the PTAB held that neither “mere compatibility of the references” nor a showing “that the prior art references [were] analogous art or in the same field of endeavor as the challenged patent” were sufficient “to establish that the skilled artisan would have had reason to combine the teachings of the prior art as claimed in the challenged patent.” Id.

The Board found Petitioner’s other alleged motivations to combine similarly defective.  Worthington’s statement that its thermosetting compositions may be used for “other purposes for which thermosetting compositions are used” was found insufficient to support Petitioner’s assertion that the challenged claims represent the predictable use of prior art elements according to their established functions. Id. at 11.  Noting that the “cited references teach considerably different binding systems,” the Board faulted Petitioner for failing to explain a specific reason to substitute Worthington’s binder for Srinivasan’s other than the unsupported assumption that all thermoset binders are useful in the Srinivasan’s type of fiberglass insulation product.  Id.  More generally, the Board wanted to see evidence concerning “which of the prior art elements allegedly has a known function, what that function is, and why that function is allegedly predictable.  Id. at 12.  As a corollary, the Board found that in view of the extensive modifications proposed, Petitioner also failed to adequately support existence of a reasonable expectation of success in making the modifications. Id. at 13.

The PTAB’s ruling in Hulu, LLC v. Sound View Innovations, LLC shows that even where an IPR trial is instituted, detailed consideration of the facts may ultimately demonstrate a lack of motivation to combine.  In Hulu, Petitioner challenged claims of US Patent No. 6,502,133, which relate to an apparatus and method for processing real-time events in, e.g., a telecommunications system.  The ’133 patent discloses that conventional database management systems are often too slow to meet real-time performance requirements relating to “switching services” in telecommunication systems. The ’133 Patent at 1:25-42.  At the time of the invention, custom databases were effective, but expensive and inflexible solutions to such problems.  Id. at 1:43-56. The ’133 patent claims include a real-time analysis engine associated with a main-memory database where recovery information is stored. The Board instituted trial on two grounds of obviousness, but noted that one ground was flawed from the start.⁴  Here, we focus on the second ground, the combination of O’Neil, Kao and De Witt, which was the primary basis for institution.  Hulu, Paper 9 at 2.

O’Neil discloses a telecommunications system for providing prepaid and credit limited cell phone service and includes a real-time telephone call monitoring, rating, and response system.  This system includes at least two databases, but does not disclose whether those are in main memory.  Petitioner therefore relied on Kao for teaching that in real-time database systems, main memory databases provide fast and predictable access times.  Id. at 14.  Kao also noted that telephone switching is an application with “stringent timing requirements” for which a real-time database would be useful.  Petitioner contended that a POSA would have been motivated “‘to use Kao’s main-memory database in O’Neil’s real-time telephone call monitoring system . . . because (i) it was known that telephone routing and billing systems required tight timing requirements, and (ii) it was known that main memory databases are particularly suitable for these types of systems.’” Id.  De Witt, which is cited by Kao, was relied on for details concerning storing recovering information on memory. Id.

Patent Owner, Sound View Innovations, was able to rebut this specific and seemingly credible motivation to combine, by adducing evidence that undermined the alleged “stringent timing requirements” of O’Neil’s databases.  First, Patent Owner pointed to evidence that O’Neil’s system obtains information from its databases during a call within a few seconds rather than milliseconds.  Petitioner’s expert was then forced to admit that “once a second” updates from O’Neil’s databases were acceptable in O’Neil’s telecommunications system.  Id. at 16-17. Second, Patent Owner’s expert provided uncontested testimony that the “vast majority of database systems” at the time of the ’133 patent were disk-based systems and that these were fast enough to meet O’Neil’s requirements. Id. at 17-18. Third, Patent Owner also put forth evidence that using a main memory database for O’Neil’s databases would not have had any impact on O’Neil’s call routing speeds because call routing was complete before the system requires information from the databases. Id. at 18.  Finally, the Board also found that main memory disadvantages of high cost, low capacity and volatility would have weighed against a POSA combining O’Neil and Kao/De Witt.⁵Id. at 19-20. 

In summary, the Johns Manville and Hulu decisions provide a cautionary tale to Petitioners to thoroughly vet the motivation to combine evidence and arguments supporting an obviousness challenge in an IPR Petition.  These decisions are also a cogent reminder to Patent Owners that even if all elements of the claims are met by combined references, a detailed attack on the factual bases for the alleged motivation to combine is a worthwhile strategy to defeat an obviousness challenge.

¹ One or both of Srinivasan and Worthington appear in the other three proposed obviousness grounds and the Board found that those grounds failed for similar reasons as ground 1. Id. at 14.

²  Securus Techs., Inc. v. Global Tel*Link Corp., 701 F. App’x 971, 977 (Fed. Cir. 2017).

³  Personal Web Techs., LLC v. Apple, Inc., 848 F.3d 987, 993 (Fed. Cir. 2017).

⁴ In the Institution Decision and in the Final Written Decision, the Board indicated that the first ground of obviousness failed to teach or suggest at least one of the claimed elements.  

⁵ Other motivation arguments raised by Petitioner in its Reply were deemed either unsupported or untimely.  Id. at 21