This disposition is nonprecedential

TARANTO, Circuit Judge.

Liberty Mutual Insurance Company initiated seven overlapping proceedings against Progressive Casualty Insurance Company in the Patent and Trademark Office, challenging many claims of several of Progressive's insurance-related patents under the statutory program for review of "covered business method" patents. See GTNX, Inc. v. INTTRA, Inc., 789 F.3d 1309, 1310 (Fed. Cir. 2015) (describing transitional program for review of such patents under 35 U.S.C. §§ 321-329, pursuant to the Leahy-Smith America Invents Act, Pub. L. No. 112-29, § 18(a)(1), 125 Stat. 284, 329-31 (2011)). We now have before us Progressive's appeals from the Patent Trial and Appeal Board's final written decisions that many claims are invalid over prior art. See 35 U.S.C. §§ 328, 329. We affirm. We need not address Liberty's cross-appeal in one proceeding, because that cross-appeal concerns claims held invalid in another proceeding whose result we affirm.

The claims at issue come from five patents: (1) claims 1-20 of U.S. Patent No. 8,140,358; (2) claims 1-78 of U.S. Patent No. 8,090,598; (3) claims 1 and 3-18 of U.S. Patent No. 6,064,970; (4) claims 1-46 of U.S. Patent No. 7,124,088; and (5) claims 1-9 and 13-59 of U.S. Patent No. 7,877,269. The challenged claims of the '358, '598, and '970 patents relate to pricing automobile insurance based on vehicle use, such as the number of sudden stops over a given period. The challenged claims of the '088 and '269 patents relate to adjusting insurance policies online.

For oral argument, we organized the appeals from the Board's seven decisions into four (single-case or multicase) groups. We follow that grouping here and address a subset of the many issues raised by the parties. We have jurisdiction under 28 U.S.C. § 1295(a)(4)(A).

A

The first group of appeals consists of Progressive's 14-1636 and 14-1639 appeals and Liberty's 14-1637 crossappeal. The appeals are from two final written decisions of the Board that address the '358 patent. In CBM 2012-3, 2014 WL 651401 (Feb. 11, 2014), the Board invalidated all claims but 1, 19, and 20 for obviousness. In CBM 2013-9, 2014 WL 651402 (Feb. 11, 2014), the Board invalidated all claims of the '358 patent for obviousness. As Progressive and Liberty agreed at oral argument, Oral Arg. (14-1336, -1337, -1339) at 13:30-14:40, the decision in CBM 2012-3 need not be addressed if we affirm the decision in CBM 2012-9—which we do.

As a preliminary matter, we reject Progressive's argument that, under 35 U.S.C. § 325(e)(1), the Board was estopped from entering its CBM 2013-9 decision. Section 325(e)(1) provides:

Progressive argues that § 325(e)(1) bars the Board's entry of its CBM 2013-9 decision because the Board posted that decision to its electronic docketing system just over an hour after, but the same day as, it posted the CBM 2012-3 decision.

There are two problems with Progressive's contention. First: § 325(e)(1) by its terms does not prohibit the Board from reaching decisions. It limits only certain (requesting or maintaining) actions by a petitioner. Nothing in the provision, or chapter 32 more generally, equates that limitation on a petitioner with Board authority to enter a decision. Cf. § 327(a) (Board may enter decision even after petitioner settles and drops out of the proceeding).

Second: § 325(e)(1) does not say when a final decision begins to have estoppel effect (on the petitioner). Here, the Board stated in its two decisions that they were being entered "concurrently." CBM 2012-3, 2014 WL 651401, at *2; CBM 2013-9, 2014 WL 651402, at *1. The Board previously agreed to the parties' joint request to "synchronize the timelines in both trials" because the reviews involved the same parties, the same patent, and overlapping prior-art references. A8387 (14-1639 appeal). We see nothing in the statute (or any regulation or other source) that forecloses the Board's treatment of the two same-day decisions as simultaneous and therefore outside § 325(e)(1)'s scope, regardless of the precise times of posting on an electronic docketing system. See also §§ 325(d), 326(a)(4) (conferring authority on the PTO to decide how to deal with multiple related proceedings).

On the merits, we conclude that the Board did not err in invalidating all of the '358 patent's claims in CBM 2013-9. We address two of Progressive's challenges to that ruling—one procedural, the other substantive.

Progressive argues that the Board committed a procedural error by relying, in its final written decision, on a portion of Kosaka, a published Japanese patent application (Jap. Pub. App. H4-182868), that neither Liberty's petition nor the Board's institution decision had relied on. Kosaka describes a device that determines insurance premiums by evaluating risk in moving bodies. It describes an embodiment that uses fuzzy logic, and the Board, in its institution decision, pointed to Kosaka's fuzzy-logic processor and memory as disclosing certain limitations of Progressive's patent claims. Progressive responded in its patent-owner response that a relevant skilled artisan would neither have experience with nor understand fuzzy logic. The Board rejected the argument in its final written decision, relying in part on Kosaka's statement that "`determination [of risk evaluation values] may be carried out without using fuzzy logic'" by using a "`common insurance table'" instead. A5036 (quoting A5162) (14-1639 appeal). Neither Liberty's petition nor the Board's institution decision had pointed to the insurance-table passage of Kosaka; it was raised for the first time, as teaching claim limitations, in Liberty's postinstitution reply brief—after Progressive no longer had a full opportunity to respond to it with evidence.

Liberty, along with the Director of the PTO, argues that the Board's reliance on the previously unmentioned passage in Kosaka was not a "new ground of rejection" because that label applies only in examination and reexamination. See 37 C.F.R. § 41.50(b); 37 C.F.R. § 41.77(b). But the Administrative Procedure Act imposes its own similar obligations on Board actions, including in covered business method reviews. For example, 5 U.S.C. § 554(b)(3) requires that "[p]ersons entitled to notice of an agency hearing shall be timely informed of . . . the matters of fact and law asserted"; § 554(c) requires that agencies give "all interested parties opportunity for . . . the submission and consideration of facts [and] arguments. . . [and] hearing and decision on notice"; and § 556(d) "entitle[s]" a party "to submit rebuttal evidence." Indeed, § 554(b)(3) has been applied to mean that "an agency may not change theories in midstream without giving respondents reasonable notice of the change" and "the opportunity to present argument under the new theory." Rodale Press, Inc. v. FTC, 407 F.2d 1252, 1256-57 (D.C. Cir. 1968). The APA's requirements may well embody, in substance, the inquiry at the heart of the newgrounds-of-rejection analysis: "whether applicants have had fair opportunity to react to the thrust of the rejec-tion." In re Biedermann, 733 F.3d 329, 337 (Fed. Cir. 2013) (internal quotation marks and citation omitted).

We need not pursue the legal analysis to a conclusion, even if Progressive has a substantial argument that the Board—by relying on a factual assertion about the teaching of a passage of Kosaka that Progressive had no opportunity to respond to—committed a notice error. If there was error, it was harmless. The Board did not rely solely on the insurance-table passage of Kosaka. It independently rejected Progressive's lack-of-knowledge argument, making a determination supported by the evidence that a relevant skilled artisan would have understood fuzzy logic. In light of the independent ground, any error regarding the insurance-table passage in Kosaka did not affect the outcome.

Progressive also attacks the substance of the Board's obviousness conclusion, which was based on Kosaka and an informational booklet by the Geostar Corporation called Understanding Radio Determination Satellite Service (RDSS). Specifically, Progressive argues that the Board lacked evidence to find, and did not adequately articulate, a motivation for a relevant skilled artisan to combine the two references. Motivation to combine is a question of fact that we review for substantial evidence. In re Gartside, 203 F.3d 1305, 1316 (Fed. Cir. 2000).

The Board's affirmative explanation for its finding of a motivation to combine is cursory. It consists of a singleparagraph recitation of three points from Liberty's petition, some not even addressing motivation to combine, followed by the Board's declaration of agreement. The Board's affirmative treatment would have benefited from elaboration to display the needed effort to place itself into the mindset of the hypothetical person of ordinary skill in the art considering prior-art knowledge without hindsight use of the patent claims as a guide.

Nevertheless, the Board's findings make its rationale clear, and the brief citations to the record accompanying the enumeration of Liberty's points provide substantial evidence to support that rationale in the present setting. In particular, the Board found that a relevant skilled artisan would have understood the benefits of the RDSS model (using a central terminal for resource-intensive computing) for use in an insurance system. CBM 2013-9, 2014 WL 651402, at *13. The Board adequately supported this proposition with a declaration from Liberty's expert Scott Andrews, which stated that a relevant skilled artisan would realize that "various components of Kosaka. . . could be advantageously implemented at a remote central computer system/server of the insurer" because "processing at a central computing system" would allow one to "tak[e] advantage of, e.g., larger storage capacity, centralized collections of data involving multiple customers, and faster and more sophisticated computing capabilities." A7846-47 at ¶¶ 37-38 (14-1639 appeal). We have been presented no persuasive reason to conclude that the Andrews evidence is insufficient to support the Board's finding here.

B

Progressive's appeal in 14-1586 challenges the Board's determination in CBM 2013-4, 2014 WL 1252839 (Mar. 13, 2014), that all claims of the '598 patent are anticipated by at least one piece of prior art (with some claims also held to claim matter that is unpatentable because of obviousness). Progressive's chief argument seeks to antedate the prior art based on the contention that the '598 patent is entitled to the benefit of the earlier priority date of its immediate parent application, U.S. Patent Application No. 09/571,650. The Board rejected that contention, finding that the "interface module" required by all claims of the '598 patent lacks written-description support in the '650 application.

The correctness of the written-description finding is the only issue raised by Progressive that warrants discussion—which can be brief. The adequacy of a written description is a question of fact, and we review the Board's finding for substantial evidence. Noelle v. Lederman, 355 F.3d 1343, 1348 (Fed. Cir. 2004). Substantial evidence is present here.

All claims require that the interface module produce either a "driver safety score" or "driver safety data." See, e.g., '598 patent, col. 28, lines 44-49. The Board construed "driver safety score" as "a calculated insurance risk value associated with driver safety"; it construed "driver safety data" as "encompassing `driver safety score' and other data associated with driver safety." CBM 2013-4, 2014 WL 1252839, at *7. In either case, Progressive shows only that the '650 application discloses an interface module that calculates insurance premiums and that this calculation is based on "rating factors," which might be of a variety of types, not merely factors based on safety. Progressive fails to show error in the Board's finding that the '650 application does not disclose an interface module producing a "driver safety score" or "driver safety data," which are particular examples of possible "rating factors"—examples not described in the '650 application.

Progressive's overarching argument is that a "rating factor" is a type of "driver safety score" or "driver safety data." But as the Board found, that gets matters backwards: a "rating factor" is broader than the two claim terms. Id. at *7, *15. The '598 patent's own written description refers to "a total discount [that] is based upon a calculation including . . . a rating factor, such as a safety score, "'598 patent, col. 22, lines 19-21, with other examples of "rating factors" including "a daytime mileage adjustment, a nighttime mileage adjustment and a high risk mileage adjustment," id., col. 23, lines 11-13. Substantial evidence supports the notion that these latter examples are not associated with driver safety, as "mile-What age" suggests that they track risk levels based on a vehicle's amount of use independent of how safely one drives. In short, the '650 application's disclosure of a component producing a rating factor does not, without more, imply disclosure of the particular types of rating factor required by the '589 patent's claims—namely, a driver safety score or driver safety data. And Progressive has pointed to no other disclosures that might suffice; in particular, Progressive acknowledged at oral argument that it was not contending that the "estimated cost" disclosed in the '650 application is a driver safety score or driver safety data. Oral Arg. (14-1586 appeal) at 41:14-42:08; see also CBM 2013-4, 2014 WL 1252839, at *15 ("There is no dispute that an insurance cost is not a driver safety score.").

We conclude that substantial evidence supports the Board's written-description-based determination that the '598 patent is not entitled to the '650 application's priority date. Progressive has raised no other significant challenge to the CBM 2013-4 decision, which we affirm.

C

Progressive's appeals in 14-1466 and 14-1656 challenge the Board's invalidations, for obviousness, of all claims of the '970 patent. In CBM 2012-2, 109 U.S.P.Q.2d 1833 (Jan. 23, 2014), the Board invalidated all claims but 7 and 8. In CBM 2012-4, 2014 WL 2213411 (Jan. 23, 2014), the Board invalidated all claims without exception.

Progressive raises only two issues that warrant discussion. Our discussion of one, a notice-error challenge involving the insurance-table portion of Kosaka, can be limited to the observation that the notice-error challenge here is not meaningfully different from the notice-error challenge raised in Progressive's 14-1639 appeal. In Part A, supra, we have rejected the challenge in that appeal. We do so in the present appeals too.

What remains is Progressive's challenge to the Board finding of a motivation to combine certain references— some focused on vehicle-use monitoring but not premium determination, others focused on premium determination but not vehicle-use monitoring. The Board relied on the finding that the prior-art Florida Guide, a 1988 shoppers' guide for automobile insurance in Florida, taught that insurers were required to generate insured profiles in issuing policies. That rationale is incomplete in isolation, and the Board did not set forth in one place the totality of its rationale for finding a motivation to combine. Nevertheless, although the Board's full rationale must be pieced together through a reading of the Board's full decisions at issue, the Board's findings make its rationale clear, and substantial evidence supports that rationale. In particular, the Board made two key findings that, when put together, provide sufficient reason to combine the references at issue.

First, the Board reasonably found that a relevant skilled artisan would have had knowledge of basic actuarial principles, standards, and practices, which, crucially, aim to determine insurance premiums based on an accurate assessment of risk. See, e.g., CBM 2012-2, 109 U.S.P.Q.2d 1833, at *11, *15-16, *22-23. For example, the Board relied on portions of the Florida Guide and declarations from Liberty's expert Mary O'Neil. The passages relied on show that insurers set insurance premiums at prices commensurate with projected risk (in part to comply with nondiscrimination laws). E.g., id. at *10, *13-14, *21-23; see also A164-68 (Florida Guide); A3564-65 at ¶¶ 20-22, A3567-68 at ¶ 26 (Ms. O'Neil's original declaration); A4538-41 at ¶¶ 8-13 (Ms. O'Neil's rebuttal declaration). Thus, as the Board recognized, a relevant skilled artisan would have been motivated to develop more reliable methods for creating and applying actuarial classes, see CBM 2012-2, 109 U.S.P.Q.2d 1833, at *27-28—something consistent with both Kosaka and the '970 patent itself. A121 (-1466 appeal) (Kosaka identifying, as a problem to be solved, that insurance systems "do not consider the role of the insurance agreement customer's environment and movements in governing risk levels"); '970 patent, col. 2, lines 47-51 ("[T]he limited amount of accumulated relevant data and its minimal evidential value towards computation of a fair cost of insurance has generated a long-felt need for an improved system.").

Second, the Board reasonably found that a relevant skilled artisan would have had familiarity with the idea of use-based insurance, as, among other things, the '970 patent itself acknowledges that it had "`been suggested to detect and record seatbelt usage to assist in determination of the vehicle insurance costs.'" '970 patent, col. 2, line 66 through col. 3, line 2; CBM 2012-2, 109 U.S.P.Q.2d 1833, at *11 (citing this part of the '970 patent); id. at *20 (citing the same). References directly and indirectly related to insurance confirm such knowledge. See, e.g., A120 (14-1466 appeal) (Kosaka); A146 (14-1466 appeal) (Black Magic); A149 (14-1466 appeal) (Herrod); A7135 (14-1656 appeal) (Pettersen).

Therefore, because a relevant skilled artisan would have been motivated to make insurance determinations more accurate and would have known of at least some use-based actuarial methods, he or she would have been motivated to look at use-monitoring methods in other safety-related contexts to see what else might yield helpful actuarial data. The Board did not clearly state this rationale in any single portion of its decisions. But see CBM 2012-2, 109 U.S.P.Q.2d 1833, at *10-11, *27-28; CBM 2012-4, 2014 WL 2213411, at *10-11, *20. Nevertheless, the rationale is apparent when the decisions are read as a whole. Because substantial evidence supports this motivation-to-combine rationale, we affirm the Board's CBM 2012-2 and CBM 2012-4 decisions.

D

Progressive's 14-1538 and 14-1549 appeals challenge the Board's invalidations, for obviousness, of all claims of the '088 and '269 patents in CBM 2012-10, 2014 WL 869413 (Feb. 24, 2014), and CBM 2013-2, 2014 WL 824373 (Feb. 24, 2014), respectively. We have fully considered Progressive's arguments and find them unpersuasive. We affirm the Board's decisions without further discussion.

CONCLUSION

For the foregoing reasons, the judgments of the Board are affirmed.

No costs.

AFFIRMED