US 8674809

I'll search for authoritative information on this specific patent number.

Let me refine the searches with more specific queries.

Let me check specifically for any CAFC 2026 activity and confirm the claim set.

US Patent 8,674,809 — Summary

Important caveat up front: I could not locate any Court of Appeals for the Federal Circuit docket (2026 or otherwise) naming this patent. See "Litigation / CAFC check" below. Also, the claim text itself was not included in the authoritative material I was given, so my claim discussion below is derived from the patent's "Summary of the Invention" and specification rather than from verbatim claim language — treat claim numbering as approximate.

Bibliographic data (confirmed via Google Patents, FreePatentsOnline, and the patent's own front matter)

Source: https://patents.google.com/patent/US8674809/en and https://www.freepatentsonline.com/8674809.html

Abstract (as issued): Systems and methods for activating one or more devices. The device listens for an activate code having a length field and a mask field, the mask field including a mask value, the length field specifying a length of the mask field to a final bit of the mask value. Upon receiving the activate code, the length field is compared to a stored length value to determine whether the length field meets a predefined criterion. If so, an address of the activate value is loaded (if an address field is present) and the appropriate bits (mask value) of the mask field are compared to a stored activate value. An activate signal is generated if the mask value matches the stored activate value, and can activate additional circuitry including the entire device.

Plain-language overview of the disclosed inventions / independent claim concepts

The patent is about selectively waking up a subset of RF devices (RFID tags) instead of all of them, to save battery. The activate command has (in order) a Length field, an optional Address field, and a Mask field. The Length field acts as a cheap early filter: a tag compares the received length against a locally stored minimum-mask-length (MML) register; if the length doesn't qualify, the tag bails out immediately without processing the rest of the code. Only qualifying tags compare the mask bits (using length as the number of bits and address as the bit offset into a 96-bit internal mask register) and, on a match, generate an activate/wake-up signal.

The independent-claim concepts described in the Summary of the Invention are:

1. Device activation method (length + mask): Listen for an activate code with a length field and a mask field; compare the length field to a stored length value to see if a predefined criterion is met; if it is, load an address of the activate value (if an address field is present) and compare the relevant mask bits to a stored activate value; generate an activate signal on a match.
2. Position-of-final-bit method: Length field specifies the position of the final bit of the mask value in the mask field; compare mask bits to a stored activate value and terminate the comparison at that final bit; generate an activate signal on a match.
3. Address-offset method: Receive an activate code having an address field and a mask field, where the address field indicates a start position of the mask value; compare that mask value to a stored activate value on the device.
4. Method of analyzing an activate code (length + mask): Receive the length field, compare it to a stored length value, determine whether a predefined criterion is met; receive the mask field; compare the mask value to a stored activate value only if the criterion is met; generate an activate signal on a match.
5. Method of analyzing an activate code (address + mask): Receive address and mask fields where the address indicates the location of the mask value; compare the mask value to a stored value; generate an activate signal on a match.
6. System claim: An interrogator (reader) plus a plurality of devices in RF communication, where a first subset responds to an activate command of a first length and a second subset responds to an activate command of a second length — i.e., code length itself partitions the tag population.
7. Circuit for selectively generating an activate signal: An interrupt circuit that checks whether an interrupt period falls within a predetermined range (or matches predetermined values) and outputs an interrupt signal; plus a data comparison circuit that compares the received activate code (length field + mask field) to a stored activate value and generates the activate signal on a match.
8. Circuit for analyzing an activate code: A data register storing an activate value, a length counter receiving the length field, and a data comparison circuit comparing at least a portion of the mask field to the activate value and generating an activate signal on a match.
9. Method of selectively activating a subset of devices: Transmit an activate code with a length field and a mask field, where the length field dictates which of the plurality of devices process the mask field.
10. Alternative selective-activation method: Transmit several activate codes to a plurality of remote devices, where only a subset of the devices analyze a particular one of the codes based upon the variable size of the code.

Supporting hardware described (not necessarily claimed): a class-3 tag chip with a battery activation circuit, an ultra-low-power narrow-bandwidth self-biasing preamplifier, a self-clocking interrupt detector (FIGS. 10 and 11B), current-mirror-based low-power inverters, a 96-bit internal mask register, and a comparison stage using a 96×1 MUX driven by an address counter (FIG. 12), with an optional multi-code register variant (FIG. 13). Notable practical details: length 0 (with MML=0) can act as a "wake everyone" broadcast; lengths greater than the maximum mask length cause the packet to be ignored; rollover can either be treated as an error or handled circularly; trailing "dummy" bits after the mask value are ignored so different tags' codes can share a common length.

Certainty / uncertainty notes

• High confidence in title, inventors, application number, filing date, issue date, priority date, abstract, and assignee history — these are corroborated across multiple sources and the patent front matter.
• Moderate confidence in the precise number and wording of independent claims. I identified ten independent-claim concepts from the Summary of the Invention, but without the verbatim claim set I cannot state with authority how many independent claims actually issue (e.g., whether some of these were merged or presented as dependent claims, or whether "address field" variants appear as separate independents). Verify against the issued claims at the USPTO/Google Patents "Claims" tab.
• Assignee: Google Patents lists "Intelleflex Corp" as original assignee and Zest Labs Inc as current assignee via a 2017-02-10 "change of name" record. FreePatentsOnline's page shows only the older "Intelleflex Corporation (San Jose, CA, US)". The two are consistent as a name-change chain, but note Zest Labs, Inc. is the same entity that was formerly Intelleflex Corporation.

Litigation / CAFC check — no 2026 Federal Circuit docket found for this patent

• My searches did not surface any CAFC 2026 docket (or any Federal Circuit appeal) naming US 8,674,809. I cannot confirm one exists, and I'm flagging that explicitly rather than implying a search hit.
• The well-documented litigation involving the former Intelleflex/Zest Labs entity is Zest Labs, Inc. (f/k/a Intelleflex Corporation) and Ecoark Holdings, Inc. v. [Walmart Inc.](/litigations/by-defendant/Walmart%20Inc.), No. 4:18-cv-00500-JM (E.D. Ark.) — a trade-secret misappropriation / breach of contract / fraud case (Defend Trade Secrets Act and Arkansas Trade Secrets Act), not a patent-infringement suit asserting the '809 patent. Reported outcomes include a 2021 verdict, a 2024 order for a new trial, a roughly $223M new jury award, and a July 2025 settlement. An Eighth Circuit matter, In re Zest Labs, Inc., No. 23-1411 (8th Cir., filed 2023-03-01), was a petition for writ of mandamus, again not a Federal Circuit patent appeal.
• I found no evidence that US 8,674,809 was ever asserted in litigation, and no IPR/PGR or Federal Circuit activity tied to it in my searches. If you need a definitive litigation/PTAB check, the authoritative sources would be USPTO PatentCenter/PTAB (https://patents.google.com/patent/US8674809/en links to PatentCenter) and the Federal Circuit's PACER/CM-ECF docket search — I do not have direct access to those and could not verify beyond web search.

Related family note (not the same number, provided only for context): US 8,248,211 B2, also titled "Selective RF device activation," shares the same three inventors and appears to be a sibling in the same family (both trace to Ser. No. 11/186,666). It is a different patent and its data should not be substituted for the '809.

I'll search for litigation involving US 8,674,809 across patent litigation databases.

Let me dig deeper into specific case databases.

Let me check the Zest Labs complaint and look for any assertion of the '809 patent specifically.

Bottom line

I found no litigation that names U.S. Patent No. 8,674,809. My searches did not return a single complaint, docket entry, verdict, PTAB filing, or litigation-database record identifying the '809 patent (or its application/continuation lineage) as an asserted, challenged, or adjudicated patent. On the record available to me, the answer is: no known litigation involving US 8,674,809.

I want to be explicit that this is a negative finding from incomplete sources, not a certified "never litigated" conclusion (see caveats below).

What I searched

• General web searches for "8,674,809" and "8674809" combined with litigation terms (lawsuit, infringement, asserted patents).
• Targeted searches tying the '809's owner chain — Intelleflex Corporation → Zest Labs, Inc. (per the Google Patents record for the '809: filed 2012‑07‑18 as US 13/552,514, continuation of Ser. No. 11/186,666 filed 2005‑07‑20; assignee reassigned Intelleflex → Zest Labs, Inc. on 2017‑02‑10) — to litigation.
• Court documents from the E.D. Ark. docket (via CourtListener/archive.org RECAP and courthousenews) and Law360 case pages.
• The Unified Patents patent page for an Intelleflex/Zest patent surfaced in results; no '809-specific litigation record appeared.

The only hits on "8674809" were unrelated generic patent-litigation statistics documents (e.g., on NPE studies and fee awards) that do not mention this patent. No result for a similar number was substituted.

The closest related litigation (and why it is not '809 litigation)

Zest Labs, Inc. f/k/a Intelleflex Corporation, et al. v. [Walmart Inc.](/litigations/by-defendant/Walmart%20Inc.) f/k/a Wal-Mart Stores, Inc.

• Plaintiff(s): Zest Labs, Inc. (f/k/a Intelleflex Corporation); later also Zest Labs Holdings LLC, RisKon International, Inc., and Ecoark Holdings, Inc.
• Defendant(s): Walmart Inc. (f/k/a Wal-Mart Stores, Inc.)
• Jurisdiction: U.S. District Court, Eastern District of Arkansas (Central Division), Little Rock — Judge James M. Moody Jr.
• Case number: 4:18-cv-00500-JM
• Filing date: August 1, 2018 (complaint)
• Outcome/status: Jury verdict for Zest, 2021; new trial ordered December 2023; retrial verdict May 13, 2025 — $72.7M compensatory + $150M exemplary ($222.7M total); settled July 28, 2025 per Law360 reporting. A related malpractice action by Zest against its former counsel was active into 2026.

Critically, this was not a patent infringement case, and the '809 was not asserted in it. Walmart's own filing describes the complaint as asserting ten counts: trade secret misappropriation under the DTSA and Arkansas Trade Secrets Act (Counts I–II), contract claims (Counts III & VII), tort claims (Counts IV–VI), unjust enrichment (Count VIII), and remedies (Counts IX–X). Court orders state the patent-adjacent "various other claims... have since been dismissed by the Court." The patents at issue in the dispute were Walmart's Bohling applications/publications (the '396 Bohling Application, and the '020/'325 applications), used as evidence of misappropriation — not Zest-assigned patents asserted against Walmart.

⚠️ Caution on a common mischaracterization: some Law360 headlines in the results describe a "$115 million patent infringement verdict" against Walmart. The underlying verdict was on trade secret misappropriation (plus contract/fraud findings); treat "patent infringement verdict" wording about this case as imprecise.

Other Zest-adjacent proceedings surfaced (e.g., In re Zest Labs, Inc., No. 22-2306 (8th Cir. 2022) mandamus on discovery; Zest's 2026 malpractice/confidentiality dispute with Walmart) — none involve the '809.

Caveats on confidence

• No patent was asserted against anyone on the '809 in any record I could retrieve, including in the Zest/Intelleflex enforcement activity.
• I did not obtain a docket-level PACER/Unified Patents litigation report keyed specifically to patent number 8,674,809. Unified's litigation caselist and PTAB records are paywalled/registration-gated in significant part, so a licensed search of those systems is the appropriate next step before treating this as conclusive.
• I found no IPR, PGR, CBM, reexam, or ITC Section 337 record naming the '809; again, this reflects my search coverage, not a verified negative.
• Note a date inconsistency in my sources: the instruction states today is April 26, 2026, but search results include items dated September 2026. Per the operating rules I am preferring the search results; flagging it so you can reconcile the timeline.

Recommendation: run the '809 (and family members US 8,248,211 and the 11/186,666 continuation line) against a Docket Navigator / Lex Machina / Unified Patents patent-number query to convert this "not found" into a verified negative.

I'll verify the structured source against live search results before drafting.

Proceedings overview

Total AIA trial proceedings on file for US 8,674,809: 0. The canonical structured source (USPTO ODP ingest) returns no IPR, PGR, or CBM proceedings, and my independent web checks surfaced no PTAB case naming this patent — so there is no active/claims-invalidated/claims-sustained/settled/institution-denied breakdown to report, and the defensive posture is: no PTAB record exists at all — no claim of the '809 has ever been tested at the Board, no petitioner is estopped under § 315(e)(2), and the patent is already expired (status "Expired - Lifetime"; anticipated expiration 2025-07-20 per Google Patents), so the only remaining exposure is backward-looking damages inside the 35 U.S.C. § 286 six-year window.

No proceedings to itemize

There is no proceeding number, petition, institution decision, FWD, settlement, or appeal to describe. I am not going to construct one — the instructions in the structured block are explicit that "no PTAB activity on file" is the default, and nothing I found displaces it.

False positive to flag explicitly (do not misattribute): web search on this patent returns PTAB records citing IPR2017-00577 (FWD of 2018-07-05 invalidating the parent of the '907 patent) and IPR2018-00362, both styled as "Inter Partes Review of U.S. Patent No. 9,606,907." Those are different patents, and the exhibit record in those files (Dr. R. Jacob Baker declarations, a certified Chinese SIPO examination decision, § 102/§ 103 art on memory circuitry) is inconsistent with the Intelleflex/Zest Labs RFID-activation family of the '809. They are artifacts of same-numbered-art collisions in the search index, not proceedings on US 8,674,809. Source example: USPTO PTAB PTACTS petition record.

Context worth knowing (not a PTAB event): the current assignee, Zest Labs, Inc. (f/k/a Intelleflex Corporation), has been in heavy litigation — but over trade secrets, not over this patent: Zest Labs, Inc. v. [Walmart Inc.](/litigations/by-defendant/Walmart%20Inc.), No. 4:18-cv-00500-JM (E.D. Ark.), with jury verdicts in 2021 and again in May 2025 ($72.7M compensatory + $150M punitive). See E.D. Ark. Dkt. 811 (2025-06-25) and Dkt. 813. I found no infringement assertion of the '809 in that docket; that is a scope limitation on my search, not a certification.

Family details relevant to possible future challenge

• The '809 (app. 13/552,514, filed 2012-07-18, granted 2014-03-18) is a continuation of Ser. No. 11/186,666, filed 2005-07-20, which issued as US 8,248,211 ("Selective RF device activation," same inventors Bares / Paley / Schlicht). See US 8,248,211 PDF and FPO record for 8,674,809.
• Substantive disclosure is a Length/Address/Mask activate-command structure with an MML (minimum mask length) register, an optional Address offset into a 96-bit mask, and rollover-vs-circular mask comparison. That is the claim-relevant subject matter a future petitioner would have to map.
• I did not verify the exact claim count or claim text of the '809 from the authoritative record provided (the full text supplied omits the claims), so I am deliberately not quoting claim numbers. Do not rely on any claim-number statement about the '809 from this memo without pulling the issued claims.

Strategic summary

Claim status. Every claim of the '809 is UNTESTED — none canceled, none sustained, none construed by the Board. There is no narrowing to describe and no surviving-claim list to give you, because no FWD exists. If you receive a demand letter citing the '809, there is no PTAB disposition (e.g., a canceled independent claim) you can wave at the sender; the letter's claim citations are not contradicted by any administrative record.

Estoppel landscape. This is the rare clean slate. With no prior IPR/PGR petitioner, no one is barred by § 315(e)(2) — no petitioner, privy, or real party in interest has estoppel attaching to any ground, so a defendant today enjoys the full universe of § 102/§ 103 art and § 112 theories. Conversely, nothing has been pre-litigated or pre-rejected by the Board, so you get no free roadmap either; you would be building the invalidity case from zero. Two practical gates: (1) if you have been served with a complaint alleging infringement of the '809, the § 315(b) one-year clock to petition starts on service; (2) the patent expired 2025-07-20, which eliminates prospective injunctive exposure and confines recovery to the § 286 six-year lookback for pre-expiration conduct — but the Board has historically permitted IPRs to proceed against expired patents (with claim amendment effectively unavailable to the patent owner), so expiration does not close the PTAB door if you need a cancellation for res judicata/collateral-estoppel reasons in co-pending litigation. Treat that last point as practice guidance to confirm against current Board law rather than as a citation-backed holding.

Pattern signals. None of the usual tells are present: no repeat petitioner on this patent, no patent-owner appeal history to the Federal Circuit concerning the '809, and no evidence of a defensive aggregator (Unified Patents, RPX, etc.) in the chain — the Unified/RPX hits in search results are generic FAQ and PTAB-success exhibits from unrelated cases, not filings against this patent. The absence is itself informative: the '809 sits in an RFID/semi-active-tag activation space, and the patent's asserted-relevance window appears to have passed without ever drawing a post-grant challenge. That pattern is typical of patents whose commercial assertion value was consumed by a trade-secret dispute rather than a patent case.

Recommended next steps

• If you are a defendant and the demand cites the '809: there is no FWD to link to and no canceled claim to quote — do not represent otherwise. Your best record-based pressure points are (a) the 2025-07-20 expiration and the resulting § 286 damages cutoff; (b) the 2012-07-18 filing as a continuation of the 2005-07-20 application, which fixes the priority date and therefore the art window; and (c) the applicant's own cited art in the '809/'211 file history (e.g., Mickle US 7,375,637; Corbett US 2005/0057368; Collins US 2005/0052283; Eber US 6,831,548; Juels US 2004/0223481; O'Toole US 6,774,685), per the FPO record.
• If active assertion is threatened with a live complaint: calendar the § 315(b) one-year bar from service and pull the expiration calculus — an IPR filed now would be against an expired patent, so petition strategy should be aimed at cancellation for estoppel/offensive-collateral purposes rather than at prospective relief.
• Verify the negatives directly before you rely on them, since the ODP ingest can lag: (1) USPTO PTAB E2E / PTAB Center case search on patent number 8674809 and on the parent 8,248,211; (2) USPTO PatentCenter assignment/adverse-proceeding history; (3) CourtListener and the CAFC docket for any appeal or co-pending § 271 action in which the '809 was asserted. If any of those returns an IPR/PGR/CBM number for this patent, this memo's central conclusion changes and should be rewritten around that proceeding's FWD.
• Bottom line for a defendant today: the PTAB record gives the patent owner nothing to hide behind and gives you no estoppel to manage — but it also gives you no free win. The binding constraints on this patent are expiration and the damages lookback, not claim cancellation.

I'll research this patent's ownership chain using the Assignment Center data, Google Patents legal events, and company/litigation sources.

Let me dig into the specific assignment records, the litigation's patent counts, and the corporate status.

US 8674809 — Ownership & Assignment Reconstruction

Patent: US 8674809 B2 — "Selective RF device activation"
App. No.: 13/552,514 (continuation of Ser. No. 11/186,666, filed 2005-07-20)
Issued: 2014-03-18 · Legal status: Expired – Lifetime (anticipated expiration 2025-07-20)

Data-availability caveat (read first). Google Patents' legal-events feed for this patent (embedded verbatim in the authoritative patent text I was given) confirms two recorded post-filing assignment entries, but that feed does not carry reel/frame numbers, execution dates, or the correspondent-of-record. My attempts to pull the individual Assignment Center abstracts (reel/frame + correspondent) for this patent did not return those fields, and I will not invent them. Everything below that is labeled "confirmed" comes from the patent's own legal-events record or from SEC/court filings; everything dependent on the Assignment Center abstract is marked NOT RETRIEVED. You can verify the abstracts directly at https://assignmentcenter.uspto.gov/ (search "8674809") or https://assignment.uspto.gov/patent/index.html.

Inventors

All three are named on the recorded inventors' assignment to Intelleflex Corporation (legal-event entry dated 2013-01-22, "ASSIGNMENT OF ASSIGNORS' INTEREST," assignors listed as BARES, WILLIAM HENRY; PALEY, DANIEL NOAH; SCHLICHT, LAUREN MARIE). Their Bay Area residence cities are consistent with the assignee's San Jose headquarters; the underlying invention was filed in 2005, so employment at filing = Intelleflex for all three (confirmable only from the 2005-era assignment abstract, which I could not retrieve).

• Unusual-pattern check — NOT RETRIEVED. I found no evidence one way or the other about whether any inventor left Intelleflex within 12 months. Note that the 2005→2012 delay before this continuation was filed is a prosecution/continuation artifact, not an inventor-departure signal. A secondary aggregator (patentleaderboard.com) attributes Bares's patents to Intel; that conflicts with the Intelleflex assignment of record and is likely a current/last-known-employer attribution artifact, not evidence of an assignment transfer. Flagging it as an unverified data point only.

Original assignee

Intelleflex Corporation, San Jose, California — a battery-assisted passive (Class-3 / ISO 18000-6C, EPCglobal C1G2) RFID company, founded 2003.

• Product ship status: YES. Intelleflex commercialized hardware and services: XC3 Technology temperature-sensor tags (e.g., TMT-8500), handheld/cellular/fixed readers, antennas, starter kits, and the "ZEST Data Services" cloud platform. Marketing collateral (Zest/Intelleflex solution briefs, zestlabs.com) documents these offerings.
• Primary line of business: semi-active RFID tags/readers for cold-chain, asset, and yard management; later repositioned as "Zest Fresh" / "Zest Delivery" shelf-life and freshness management.
• Current status: Acquired and renamed, then re-sold. Ecoark Holdings, Inc. (OTC: EARK, later RiskOn International, Inc., Nasdaq: ROII) acquired Intelleflex in September 2013. Effective 2016-10-28, Intelleflex Corporation changed its name to Zest Labs, Inc. On 2023-08-25/28, RiskOn sold 100% of Zest Labs stock to Zest Labs Holding, LLC (an entity owned by Gary Metzger, a Zest Labs manager). No bankruptcy of Intelleflex/Zest Labs was found. (RiskOn's other unit, Agora Digital Holdings, filed bankruptcy on 2023-11-01 — unrelated to this patent.)

Assignment timeline

Confirmed USPTO legal-event entries for this patent (recording/entry dates as published in the Google Patents legal-events feed; execution dates, reel/frame, and correspondent NOT RETRIEVED):

• exec. date NOT RETRIEVED / recorded 2013-01-22 — Reel/Frame NOT RETRIEVED

  • Conveyance: Assignment (of assignors' interest)
  • Assignor: Bares, William Henry; Paley, Daniel Noah; Schlicht, Lauren Marie
  • Assignee: Intelleflex Corporation
  • Correspondent: NOT RETRIEVED — the Assignment Center abstract (and any recurring attorney/firm) could not be pulled from available sources; do not treat this as a negative finding.
  • Context: Ordinary inventors' assignment to the original operating assignee (the only inventor-side recording in the chain).

• exec. date NOT RETRIEVED / recorded 2017-02-10 — Reel/Frame NOT RETRIEVED

  • Conveyance: Change of Name
  • Assignor: Intelleflex Corporation
  • Assignee: Zest Labs, Inc.
  • Correspondent: NOT RETRIEVED
  • Context: Change of name only — no change in beneficial ownership; confirms the 2016-10-28 corporate rename recorded at USPTO (per Ecoark's 10-Q: "Effective on October 28, 2016, Intelleflex Corporation changed its name to Zest Labs, Inc.").

No further USPTO-recorded assignments appear in the legal-events feed. Critically, the ownership-changing events below are corporate/stock-level transactions that were never recorded as patent assignments, which is why a two-entry USPTO record sits on top of a four-company reality:

• 2013-09 — Ecoark Holdings, Inc. acquires Intelleflex (stock purchase; not recorded against this patent). Source: Ecoark 10-Qs / PitchBook.
• 2022-10-28 — Ecoark and Ecoark, Inc. "assigned all of its residual intellectual property rights and rights in the Zest Labs lawsuits to Zest Labs." Not shown as a recorded patent assignment in the legal-events feed.
• 2023-08-25 / 2023-08-28 — RiskOn International sells 100% of Zest Labs, Inc. stock to Zest Labs Holding, LLC (Gary Metzger). Stock purchase; not recorded against this patent.

Timeline diagram

timeline
    title Ownership of US 8674809
    2005 : Priority application filed
    2012 : Continuation filed by Intelleflex
    2013 : Inventors assign to Intelleflex
         : Ecoark acquires Intelleflex
    2014 : Patent issues as US 8674809
    2016 : Intelleflex renamed Zest Labs
    2017 : Change of name recorded
    2018 : Zest Labs sues Walmart
    2021 : First verdict 115M
    2022 : Ecoark assigns IP rights to Zest Labs
    2023 : Zest Labs sold to Zest Labs Holding
    2025 : Second verdict 222M
         : Case settles
         : Patent expires

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. The only recorded assignee change is Intelleflex → Zest Labs, a change of name of the same operating company (recorded 2017-02-10), and the assignee had a real product line (XC3 tags, readers, Zest Data Services). The 2023 stock sale to Zest Labs Holding, LLC superficially resembles a holdings-LLC pattern, but (a) it was a stock purchase of a company with employees and products, not an IP-only assignment, (b) it was not recorded at USPTO against this patent, and (c) the LLC was owned by Zest Labs' own manager (Gary Metzger), i.e., a related-party reorg rather than a transfer to a third-party licensing vehicle. No registered-agent address, single-member-DE/TX formation, or "IP/Patents/Licensing" naming evidence was found for the recorded assignees.

2. Known asserter in the chain — NOT PRESENT. None of the listed NPEs (Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities) appears as an assignor or assignee on this patent. Zest Labs, Inc. is indexed in Unified Patents' public patent database (assignee listing), but that is a defensive-tech patent tracker, not an NPE-designation directory — do not read it as an NPE listing.

3. Repeat correspondent across the chain — INSUFFICIENT DATA. This signal requires the correspondent-of-record fields, which are NOT RETRIEVED for either entry. No attorney or recording firm can be named without fabricating it. This is the single most important gap in this analysis; the two recorded entries should be re-checked at the Assignment Center for a recurring correspondent.

4. Cascading transfers — NOT PRESENT (at USPTO). Only one patent-level transfer of record exists (2017-02-10), and it is a name change. There is a corporate cascade (Intelleflex → Zest Labs → Ecoark subsidiary → Zest Labs Holding) spanning 2013–2023, but those are parent/stock deals, not the sub-24-month chained-LLC patent assignments the signal describes.

5. Pre-litigation transfer — NOT PRESENT. The suit Zest Labs Inc. et al. v. Wal-Mart Inc., No. 4:18-cv-00500-JM (E.D. Ark.) was filed 2018-08-01. The only inventor-side recording (2013-01-22, inventors → Intelleflex) predates it by ~5 years, and the recorded name change (2017-02-10) by ~18 months. The 2022 Ecoark→Zest Labs IP assignment postdates the suit. No transfer within 6 months of filing was found.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11 for Intelleflex or Zest Labs was found. Zest Labs exited Ecoark/RiskOn via a stock purchase agreement (2023), not a bankruptcy sale. (The unrelated Agora Digital bankruptcy of 2023-11-01 involves a different RiskOn subsidiary.)

7. Privateering — NOT PRESENT. The pattern here is the inverse: the operating company (Zest Labs) asserted its own patents and trade secrets against a customer-turned-competitor (Walmart), rather than transferring to an NPE to assert on its behalf. No SEC/EFF/Patent Progress coverage describing an NPE front for Intelleflex/Zest was found.

8. Defensive aggregator (anti-NPE) — NOT PRESENT. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN.

Verdict

Operating-company assertion.

The chain is a single operating company's corporate history, not an NPE formation: inventors → Intelleflex Corporation (recorded 2013-01-22) → Zest Labs, Inc. via a 2017-02-10 change-of-name only, where the assignee shipped real products (XC3 temperature tags, readers, Zest Fresh/Zest Delivery) and filed Zest Labs Inc. v. Wal-Mart Inc., 4:18-cv-00500-JM, on 2018-08-01. None of the eight NPE tells is present on the evidence retrieved; the only superficially "shell-like" event — the 2023 stock sale to Zest Labs Holding, LLC — was a manager-buyout of a going concern, unrecorded at USPTO, not a transfer to a licensing vehicle. Residual uncertainty is material on signal 3 (repeat correspondent) and on the reel/frame + execution dates, because the Assignment Center abstracts were not retrievable here — those two entries should be pulled directly to confirm there is no recurring recording attorney hiding beneath the name change.

Verify at: https://assignmentcenter.uspto.gov/ (search patent 8674809) — cross-check the two entries (2013-01-22 and 2017-02-10) for reel/frame and correspondent. Google Patents legal events: https://patents.google.com/patent/US8674809/en

Note: related sibling patent US 8248211 B2 ("Selective RF device activation," Bares et al.) shares this owner chain and should be analyzed together if you are mapping the Intelleflex/Zest portfolio.

I'll search for the authoritative record of US 8,674,809 and its cited prior art references.

Let me pull the full cited-reference list from the primary sources.

Let me retrieve more of the citation list from both the Google Patents and Justia/FPO records.

Prior-Art Analysis — US 8,674,809 B2 ("Selective RF device activation")

0. Scope, sources, and an important limitation (read first)

What I could verify: The identity and bibliographic data of US 8,674,809 B2 are confirmed and match the earlier-generated summary. Application 13/552,514; filed 2012-07-18; continuation of Ser. No. 11/186,666 (filed 2005-07-20); issued 2014-03-18; published as US 2012/0319824 A1 on 2012-12-20; Intelleflex Corp. → Zest Labs, Inc.; expired-lifetime, anticipated expiration 2025-07-20. Sources: https://patents.google.com/patent/US8674809/en, https://www.freepatentsonline.com/[8674809](/patent/8674809).html, https://patents.justia.com/patent/8674809.

What I could NOT fully retrieve — flagging this explicitly:
I was unable to obtain the complete, verbatim "References Cited" / "U.S. Patent Documents" list from the '809 front matter. I have direct access only through web-search snippets, and both Google Patents and FreePatentsOnline return that section truncated. I therefore have a partial list (roughly the older end via Justia and the newer end via FPO), with a gap in the middle. Do not treat the tables below as the complete cited-reference set. The authoritative source is the "Citations" tab on Google Patents or the front page of the granted patent PDF at USPTO PatentCenter.

Second limitation: The issued claim text was not in the authoritative material provided (flagged in the prior section as well — this is a agreement, not a contradiction). Claim mapping below is therefore keyed to the claim concepts enumerated in the earlier summary (I'll call them C1–C10), not to verbatim claim language. Treat "potentially anticipates C__" as provisional pending verbatim claim review.

Critical date: Because 13/552,514 is a continuation claiming benefit of 11/186,666 (filed 2005-07-20), the effective filing/critical date for § 102 analysis is 2005-07-20. References dated on or after that date are prior art only if they qualify under § 102(e) (pre-AIA) via an earlier effective filing date, or via another subsection. (Note: the America Invents Act first-inventor-to-file provisions do not apply to an application with this priority chain.)

1. Confirmed cited U.S. patent references (partial list, ascending – Justia record)

Retrieved from https://patents.justia.com/patent/8674809. Titles shown are as they appear in the retrieved record or from corroborating sibling records; where I could not confirm a title, I have not invented one.

2. Confirmed cited U.S. patent references (partial list, descending – FPO record)

Retrieved from https://www.freepatentsonline.com/8674809.html.

The FPO list continues below this point (older references down to the Sharpe '412 at the other end); I could not retrieve the middle of the list. Note the overlap: Justia's ascending list and FPO's descending list are the same set from opposite ends, and both are incomplete in my retrieval.

3. Which references are the strongest § 102 candidates (preliminary)

Ranked by conceptual proximity to the '809 claim concepts (C1–C10):

1. US 6,831,548 (Eber et al.), 2004-12-14 — "Data carrier with means for reducing the power consumption upon reception of data." Directly addresses not consuming power to process received data addressed to other devices. Best § 102 candidate against C1, C2, C4, C7, C8 (the "compare a qualifier field before comparing the payload" architecture). Verify claim-by-claim.
2. US 5,621,412 (Sharpe et al.), 1997-04-15 — "Multi-stage transponder wake-up, method and structure." Multi-stage wake-up maps onto the '809 staged listen→interrupt→length-qualify→mask-compare flow. Best candidate against C1/C2/C7/C8.
3. US 2004/0046642 A1 (Becker et al.), 2004-03-11 — "Protocol for addressing groups of RFID tags." Best candidate against the population-partitioning concepts C6, C9, C10.
4. US 2004/0223481 A1 (Juels et al.), 2004-11-11 — selective blocking/selection of RFID devices; C1/C6/C9.
5. US 6,239,690 (Burbidge et al.), 2001-05-29 — selective wake-up of tags; C1/C2/C9.
6. US 6,693,511 (Seal), 2004-02-17 — communication with dormant tags; C7/C8.
7. US 2005/0057368 A1 (Corbett et al.), 2005-03-17 and US 2005/0052283 A1 (Collins et al.), 2005-03-10 — multi-frequency trigger/architecture; relevant to the two-tone (2 kHz/8 kHz) interrupt-and-code signaling and to C7.
8. US 7,375,637 (Mickle et al.) — power reduction in active transponders; C1/C2/C5/C7/C8 only if its effective filing date precedes 2005-07-20 (qualify under § 102(e)); its issue date (2008-05-20) is after the critical date.
9. US 6,690,264 (Dalglish), 2004-02-10 — selective cloaking; C6/C9.

No retrieved reference appears to disclose the specific length-field-qualifies-the-tag-before-mask-processing mechanism as literally recited (length field compared to a stored minimum-mask-length register, with early bail-out). On the record I have, the novelty/§ 102 exposure for '809 rests on (a) whether Eber, Becker, Juels, or Sharpe individually teach the length-as-early-filter step, or (b) obviousness combinations of these references. I did not find a single reference I can state with high confidence anticipates the independent claims on an element-by-element basis — that determination requires the verbatim claim text, which I do not have.

4. References that are NOT § 102 prior art (important)

Two entries on the FPO cited list are the applicants' own family members and, as a matter of law, are not prior art against '809:

• US 8,248,211 B2 (Bares et al.), 2012-08-21 — same inventors, same family/specification, different patent number. This is the sibling the earlier summary correctly identified.
• US 2007/0018794 A1 (Bares et al.), 2007-01-25 — the publication of the parent application Ser. No. 11/186,666 from which '809 claims benefit.

Their appearance in the "References Cited" field of the FPO/Google record is a family/continuation citation artifact, not an admission that they are prior art. Flagging this because an unwary reader could mistakenly cite 2007/0018794 as § 102 art against '809.

Also note two applications referenced in the body of '809 (not in the "References Cited" list, but incorporated by reference):

• Ser. No. 11/007,973, filed 2004-12-08, "BATTERY ACTIVATION CIRCUIT" (cited in the FIG. 3C discussion).
• Ser. No. 10/902,683, filed 2004-07-28, "SECURITY SYSTEM AND METHOD" (cited at the security encryption circuit 222).
  These are incorporated-by-reference material and could bear on enablement/§ 112 and on what the specification teaches, but they are not "cited references" against the claims.

5. Explicit uncertainty and verification steps

Because of the retrieval limits above, I am stating the following with the indicated confidence:

• High confidence: the bibliographic data and priority chain, and that US 8,248,211 and US 2007/0018794 are family members (not prior art), and that the specific citation list is larger than what I retrieved.
• Moderate confidence: the dated citations in § 1 and § 2 are accurate as to number, date, and inventor (they come from the Justia and FPO records for '809), but I could not confirm every title.
• Low confidence / do not rely on: my § 102 claim-level mapping, for two reasons — (i) I lack verbatim claim text (flagged in the earlier section too), and (ii) I only have a partial reference list. Any statement that a specific reference "anticipates claim X" is a lead to investigate, not a conclusion.
• I could not verify examiner-cited (as opposed to applicant-cited / IDS-listed) references, since the front-page "References Cited" annotations distinguishing applicant vs. examiner citations were not retrievable in my search snippets.

Recommended authoritative checks (I do not have direct access to these):

• USPTO PatentCenter for application 13/552,514 — full "References Cited" and the IDS/PTO-1449 record: https://patents.google.com/patent/US8674809/en links to PatentCenter.
• USPTO PTAB (https://ptacts.uspto.gov/) for any IPR/PGR against '809 — none surfaced in my searches.
• Federal Circuit / PACER docket search — as the earlier section noted, no 2026 (or other) CAFC docket naming US 8,674,809 was found, and the well-documented Zest Labs v. Walmart litigation (E.D. Ark. No. 4:18-cv-00500-JM) was a trade-secret case, not a '809 infringement suit. That remains consistent.

If you can supply the verbatim claim set (or the full front-page citation list from the granted PDF), I can produce a proper element-by-element § 102 chart against the references in § 3.

I'll ground this analysis in the actual prior-art record for the '809 before drawing § 103 conclusions.

Now let me pull the full cited-art list and the key references' substance.

Let me verify the substance of the remaining cited references most likely to matter for § 103.

Obviousness Analysis — U.S. Patent No. 8,674,809 under 35 U.S.C. § 103

0. Evidentiary basis, and one limitation you must know about

A disclosure problem I will not paper over. The task asks me to "use the results from the Prior Art section of this page." The authoritative full text furnished to me for US 8,674,809 was the description/definitions extract — it does not contain the tabulated Citations / Cited By / Similar Documents lists, and it does not contain the issued claims. My live fetch of the Google Patents page likewise returned the Definitions section rather than the prior-art tabulation. I therefore reconstructed the citation record from the patent's own front-page "US Patent References" list mirrored at FreePatentsOnline and Justia Patents. No reference number has been auto-corrected or substituted, per the strict rule. Where I could not confirm a reference's substance or date from a primary document in this session, I say so.

Consequence for this analysis: I cannot map limitations to claims verbatim. The element-by-element mapping below is built on the independent-claim concepts recited in the patent's own Summary of the Invention and BEST MODE sections. Treat claim numbering and any quoted claim language as unverified pending a pull of the issued claims from PatentCenter or the Google Patents Claims tab. Flagging a related contradiction from the earlier sections: the previously generated summary states the claim text "was not included in the authoritative material," and that remains true here — I am not going to manufacture claim language to make the § 103 case look tighter than the record supports.

Governing law. The '809 issued from App. 13/552,514, filed 2012-07-18, as a continuation of Ser. No. 11/186,666, filed 2005-07-20, and it never contained a claim with an effective filing date on or after 2013-03-16. Pre-AIA § 102/§ 103 therefore governs, and the critical date for § 102(a)/(b) art is 2005-07-20 (with § 102(e) art available as of a reference's earliest supporting U.S. filing date). This matters: it is why a reference granted in 2008 (Mickle) can still be prior art, via its 2005 provisional.

Prior-art window established by that date:

The '809's own Background contains an admission that is fatal to any argument that selective wake-up was novel as a concept: "Wake up codes have been used in RFID systems to selectively 'wake up' individual tags and not others, thereby conserving the battery life of the tags that are not needed… Typically, the reader broadcasts a wake up code, and each tag activates just long enough to determine whether the broadcast code matches a code stored in the tag's memory. If the codes match, the tag fully activates." The stated problem the invention set out to solve is narrower: "it would be desirable to eliminate the need for all tags receiving a broadcasted wake up code to analyze the wake up code." That framing hands the § 103 analysis its motivation, from the patentee's mouth.

1. Decomposition of the claimed subject matter into elements

Independent-claim concepts (C1–C10) are those identified in the earlier-generated summary; I add element labels for mapping.

Everything else in the patent — the current-mirror inverters (FIG. 5–6), the narrow-band self-biasing preamplifier (FIG. 8), the 250 µs/1 ms/120 µs/2 ms mirrored-inverter delay chains (FIGS. 10, 11B), the 96-bit register and 96×1 MUX (FIG. 12), the circular-mask rollover option — is disclosed but, on the Summary's own account, not recited in the independent claims. If the issued claims do recite any of it (particularly E9 implemented with the specific delay-based self-clocking architecture), the § 103 analysis changes materially and the strongest nonobviousness argument moves to that architecture. That is the single most important thing to verify against the real claim set.

2. What each principal reference teaches

2.1 Becker (US 2004/0046642 / WO 2004/023387) — the closest art, by the patentee's own admission

Becker discloses a tagging system in which a reader addresses a group that is "a subset of a plurality of RFID tags," each tag having a unique identifier. The reader inserts a group address into a message, where the group address comprises:

• a first set of data elements substantially equal in value to corresponding ones of the data elements in the identifiers of the group of RFID tags (→ the mask); and
• a second set of data elements representing any value for corresponding ones of the data elements in the identifiers (→ don't-care / wildcard bits).

The tag-side method: "comparing the group address to an identifier, wherein the identifier comprises K data elements; and determining a match even if only the first set of data elements favorably compares to corresponding data elements in the identifier," where L + M = K, L ≥ 0, M ≥ 0, K > 0. Source: US 2004/0046642 A1.

Four things follow, and they are the spine of the § 103 case:

1. Becker selects a subpopulation of tags by comparing a transmitted value against a stored identifier (E6), and expressly to avoid requiring all tags to respond — "the tag reader can address all of the tags simultaneously… Alternatively… it is also useful to interrogate some but not all of the available RFID tags."
2. Becker's L is a variable number of compared elements (E3) and M is the complementary wildcard count — the germ of a length field.
3. Becker responds on a partial match — the tag need not match its full identifier.
4. Becker's K/L/M framework is a length-parameterized comparison against a stored value — E4/E5/E6 in substance.

Critical admission. The European family member of the '809's own parent, EP 1904949 B1, states at [0016]: "US 2004/0046642, against which the present claims are delimited, describes a tagging system including a tag reader and a plurality of RFID tags, each tag having a unique identifier. The tag reader addresses one or a plurality of the tags by transmitting an individual or group address." Source: EP1904949B1. The applicant conceded that Becker is the closest prior art and delimited the claim over it. That concession is admissible evidence of the level of ordinary skill and of what the art already disclosed — and it means the only remaining § 103 question is whether the characterizing feature (the length/MML early-exit mechanism) was itself nonobvious.

2.2 Mickle (US 7,375,637) — the low-power "smart buffer" that decides before waking

Mickle teaches an active transponder whose receiver feeds a buffer device / "smart buffer" that is "structured to: (i) receive an information signal based on the RF signal from the receiver, (ii) determine whether the information signal includes the identifier, and (iii) cause the processing unit to move from the inactive state to the active state and transmit at least a portion of the information signal to the processing unit only if it is determined that the information signal includes the identifier." The stated purpose is that "all non-addressed transponders remain in a sleep (standby) state, thereby reducing total system or collection energy." Mickle's FIG. 4 filtering circuit is "a state machine that will generate a wake-up signal only if a particular [combination]…" and is "any low power electronic device that can be turned on for a short period of time, increment a counter, measure a burst length, and then go back to sleep." Sources: US 7,375,637 B2, US 2007/0205873 A1, WO 2007/101080 A3.

Mickle supplies E1 (partial), E5 (conditional-wake gating), E8 (wake-up signal), E9 (a low-power detection/comparison circuit that gates the wake signal), E11 (counter register) — and, crucially, the motivation ("reduce total system or collection energy"; "prevent spurious RF energy (noise) from inadvertently causing the processing unit to move from an inactive state to an active state"). Mickle also expressly contemplates subpopulation selection: "a number of transponders 60 may be deployed with different burst switch frequencies and/or different passive logic combinations 65 such that an RFID reader can generate appropriate RF signals to selectively wake-up certain ones of the RF transponders."

2.3 Eber (US 6,831,548) — the "don't spend power on data you don't need" teaching

Titled "Data carrier with means for reducing the power consumption upon reception of data" — a data carrier (tag) that reduces power consumption upon reception of data. This is the § 103 motivation for the early-exit: a device that cannot yet tell whether a message is for it should not have to fully receive and process that message. (I was rate-limited before I could pull the Eber specification's exact text in this session; the title and claim-level subject matter are confirmed from the '809's own front-page citation list, but I have not quoted its disclosure and will not.)

2.4 Corbett (US 2005/0057368) and Collins (US 2005/0052283) — code-based selective activation

Corbett: a continuous-wave, fixed-multiple-frequency-triggered RFID tag where the interrogator "cycles a [code] that represents a… tag 100,200. This code is then transmitted by the RFID interrogator 300 in the form of selected CW, un[modulated] frequency signals during a tag interrogation stage," and only the tag whose stored code matches responds. Collins: a multiple-frequency RFID tag architecture for the same purpose. These supply E6 (compare received code to stored code) and E12 (a transmitted code selects which tags respond), and they establish that using the form of the code itself as the population-selection mechanism was squarely in the art. The '809's own FIG. 3C discussion confirms the lineage, noting its activate code is "based on a F2F modulation protocol which will allow the transmitter (reader) to select which populations of receivers (tags) it wishes to activate in a Class-3 mode."

2.5 O'Toole (US 6,774,685 / 6,731,548 / 6,721,289) — the RF data-communications device with command-matched wake-up

The O'Toole family (MICRON) discloses an RF data communications device with command decoding, stored comparison values, and transition between low-power and active states upon a command match. These are § 102(b) art and supply generic but complete disclosure of E6/E8/E11 (registers, comparators, and command-matched state change). They are weak as primary references but strong as secondary references confirming that "compare a received command to a stored register, then wake" was routine by 2004.

2.6 Gallagher (US 2004/0140884) and the EPC/Auto-ID Class-1 "filter string" — the pointer-plus-bitstream idiom

Gallagher discloses an anticollision protocol with fast read request and schemes for reading multiple transponders — i.e., partial-identifier matching to reach subsets. And the Auto-ID/EPC Class-1 air interface — a pre-2005 standard — uses a filter string consisting of a pointer location and a bit stream, where "the pointer location indicates where the bit stream starts in the EPC. Each tag tests the relevant portion of its EPC." That is, functionally, address (pointer) + mask (bit stream), with the compared length set by the length of the bit stream. (I verified this description from a vendor manual excerpt in this session; I have not independently confirmed the 2002–2003 publication date of the Class-1 candidate specification from a primary source, so treat this as a lead to confirm rather than a proven § 102(b) reference.)

2.7 The 1997–2002 patent backbone

US 5,856,788 (Walter), US 5,621,412 (Sharpe), US 6,104,333/6,265,963 (Wood), US 6,362,738 (Vega), US 6,249,212 (Beigel), US 5,673,037/6,172,596 (Cesar), US 6,130,603 (Briechle), US 6,310,558 (Minami), US 6,469,618 (Gaultier) all pre-date the critical date by years and collectively show: stored wake-up/select codes, selective activation of a subset, low-power listening states, and code-length-dependent behavior. They are the "background art" that makes any claim to the general idea of selective activation by stored code untenable.

3. Combination theories rendering the claims obvious

I give the primary reference, the secondary reference(s), the elements covered, and the articulated motivation. All motivations are drawn from the references' own stated purposes or from the patentee's admitted problem, which is the strongest form of motivation because it is the patent's own framing.

Ground 1 — Becker + Eber (+ EPC Class-1 filter string / Gallagher): the core method claims (C1, C2, C4) and circuits (C7, C8)

Motivation to combine. Becker's own object is "to interrogate some but not all of the available RFID tags" by group addressing. Eber's own object is to cut power consumed on reception of data by a data carrier. The '809's own Background supplies the bridge verbatim: it wishes "to eliminate the need for all tags receiving a broadcasted wake up code to analyze the wake up code." A POSITA confronting Becker's group address — where a tag must receive the entire K-element group address before it can compare — and Eber's reception-power-reduction teaching, would have had every reason to make the transmitted comparison length itself a parameter, so that a tag that fails the cheap length test never expends power on the mask. That is precisely E4/E5, and it is a predictable, mechanism-level improvement of Becker's architecture, not a new architecture.

Ground 2 — Becker + Mickle: the "smart buffer" claims (C1, C4, C7, C8, C9)

Mickle supplies the conditional-wake gating architecture (a low-power front-end circuit that determines whether the incoming signal includes the identifier and only then powers the processing unit) and its express motivation — "all non-addressed transponders remain in a sleep (standby) state, thereby reducing total system or collection energy" and "prevents spurious RF energy (noise) from inadvertently causing the processing unit to move from an inactive state to an active state." Becker supplies the comparison content (group address vs. identifier, partial match, variable compared-element count). Combining them yields: a low-power comparison circuit that first checks a cheap length parameter against a stored register (Mickle's counter/state-machine) and only if it passes compares the mask to a stored value (Becker), then emits a wake signal (Mickle).

Motivation. Both references are in active/semi-active RFID tag power management; both are directed at the identical problem (battery drain from waking tags that are not needed); both describe the same physical locus (a low-power circuit ahead of the main controller). The Mickle "smart buffer" and the '809 "battery activation circuit" are the same functional block — "an entity that is cheaper to run than the tag's full decoder and that decides whether the tag is addressed." The "Active RFID Tag Power Optimization Architecture" summary of Mickle's own work makes this explicit: "The energy saving mechanism on the active RFID tags is termed a smart buffer which is a circuit fabricated in silicon that looks at the destination of each packet… and produces a wake-up signal only if the base station inquiry is for the specific 1 of m or k of m sensor equipped tags." Under KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398, 417 (2007), where a technique has been used to improve one device and a POSITA would recognize that it would improve similar devices in the same way, the combination is obvious; and under In re Keller, 642 F.2d 413 (CCPA 1981), the test is not whether features of one reference may be bodily incorporated into another, but whether the combined teachings would have suggested the claimed subject matter.

Ground 3 — Corbett / Collins + Becker + Eber: the "length dictates which devices process the mask" claims (C9, C10, C12)

Corbett and Collins establish code-form-as-selection-mechanism (only tags whose stored code matches the transmitted multi-frequency code respond). Becker adds variable compared length and wildcard/any-value elements. Eber adds the power rationale for stopping early. Combined: a transmitter sends a code of a given length; tags whose stored criterion the length satisfies proceed to compare the value; others do not. That is E12 and E10.

Motivation. Corbett's express objective is enabling the reader "to select which populations of receivers (tags) it wishes to activate." Once code content selects a population (Corbett/Collins) and code length is a parameter (Becker), using length as an additional selection dimension — the very thing claim concept C9 recites ("the length field dictates which of the plurality of devices process the mask field") — is an obvious design choice among a finite number of identified, predictable options, and the '809 itself concedes there is no unexpected result: it explains length shortening as "less chance of a bit being unrecognized due to noise" and lengthening as "greater security." The patent frames the length/value trade-off in purely expected terms.

Ground 4 — For the system claim (C6 / E10): Becker + Eber, or Corbett/Collins + O'Toole

The claim concept is that a first subset responds to an activate command of a first length and a second subset responds to an activate command of a second length. This is the '809's own "Group A MML = 16, Group B MML = 32" worked example. Mechanically it is: (a) Becker's group addressing with different stored comparison parameters per tag; plus (b) Mickle's expressly disclosed deployment of "a number of transponders 60 … with different burst switch frequencies and/or different passive logic combinations 65 such that an RFID reader can generate appropriate RF signals to selectively wake-up certain ones of the RF transponders 60," including "another RF transponder 60 … wherein a particular two of the burst switch frequencies are required to wake it up, another … wherein a different particular two … are required." Mickle literally discloses a population partitioned into subsets by the required complexity of the wake criterion. Substituting "number of compared bits" for "number of required frequencies" is an obvious substitution of a known equivalent parameter — In re Ethicon, Inc., 844 F.2d 1578 (Fed. Cir. 1988) (substitution of a known equivalent yields predictable results and is obvious).

Ground 5 — For the interrupt-circuit claim (C7 / E9): O'Toole / Eber / Mickle + the admitted ISO 18000-7 wake-up tone

The concept recites an interrupt circuit that determines whether an interrupt period matches predetermined values or falls within a predetermined range and outputs an interrupt signal. The '809's Background admits that "in RFID systems implemented according to the ISO 18000 Part 7 standard, when one or more tags are to be queried, the reader will send out a 30 KHz tone lasting for a period of approximately 2.5 seconds. Upon receipt of the tone, the RF receiver in each tag will wake-up the controller in the tag." That is period/range detection against a reference — E9 in substance — and it is admitted prior art in the specification (and likewise described in Mickle). O'Toole and Eber supply the receiver/detector circuitry. The concept, divorced from the specific self-clocking mirrored-inverter delay architecture, is squarely met.

Ground 6 — For the circuit claims' register/counter/comparator (C8 / E11): O'Toole + Eber + Becker

Data register storing an activate value (O'Toole's command registers; Becker's stored identifier), length counter receiving the length field (Mickle's "increment a counter, measure a burst length"; Eber's reception control), comparison circuit (Becker's comparator). This is the routine-implementation limb of the case: In re Aller, 220 F.2d 454 (CCPA 1955) — a change in a known process is obvious where the result is merely a "more expeditious" working and no new or unexpected result is obtained.

4. Where the § 103 case is weakest — and I am not going to hide it

1. The specific self-clocking interrupt architecture and the current-mirror delay chains (FIGS. 5–11B, 12–13). The 120 µs / 2 ms / 256 µs ±135 µs/+1.74 ms mirrored-inverter delay matching, the weak feedback latch, the XOR-and-pass-gate shift-register synchronization with the extra delay in the low path, the P-side transistor at minimum width/long channel to stretch the delay, the MUX-bit-at-a-time on-the-fly comparison eliminating a 96-bit input shift register — none of this reads on Mickle, Becker, Eber, Corbett, Collins, O'Toole, or the admitted ISO tone. If any issued claim recites an "interrupt circuit" with the structure of FIG. 11B or the on-the-fly MUX comparison of FIG. 12, the obviousness case is materially weaker than the case for the method/length-mask concepts. This is the hinge. The patentee's best nonobviousness argument is not "length then mask," it is "no clock, therefore no oscillator, therefore sub-microamp standby, achieved by exploiting matched-inverter propagation delay as the timing reference" — and that argument is only available if the claims tie the interrupt circuit to that structure.
2. Becker's group address is a fixed-length structure with variable compared length; the '809 makes the transmitted message itself variable-length. The '809 also allows a length field of zero to mean "wake everything," and permits trailing dummy bits so that codes for different tags share a common length. A patentee may argue that this is a different protocol architecture, and that the "length field-as-field" (an explicit, separately-transmitted, explicitly-interpreted length) is not the same as Becker's implicit L/M partition. The counter is that Becker's own claim language recites a message "contain[ing] a group address comprising K data elements... L + M = K" in which the compared count is itself a message parameter — so the difference, if any, is one of field encoding (explicit length octet vs. a parameterized structure), which is a design choice among a finite number of predictable encodings, supported by the Class-1 filter-string idiom.
3. Claim breadth is unverified. I cannot state the actual independent claim count, the exact preambles, or whether claims recite "address field" variants separately. If the issued claims include limitations on the MML register, the specific bit-ordering, the rollover-vs-circular selection, or the password-locked reprogrammability, some of those may survive Grounds 1–3, and Grounds 4–6 may need a different secondary reference. Do not brief this case without the claim text.
4. Secondary considerations are unevidenced either way. I found no evidence of unexpected results, long-felt need, industry praise, copying, or licensing attributable to the '809. The patent's own prosecutorial record does contain one datapoint in the other direction on the § 103 side: the EPO’s treatment of the counterpart EP 1904949 B1 was to allow the claim with a characterizing feature over Becker (US 2004/0046642), i.e., a foreign patent office found Becker alone insufficient. But allowance over a single reference is not a finding of nonobviousness over the combination, and the EPO's novelty-based wording ("against which the present claims are delimited") is not a § 103 holding. It is, however, the best single statement of the patentee's position and should be read in full.
5. One more honest caveat. The Juels reference (US 2004/0223481, selective blocking of RFID devices) is cited on the '809 but does not help the obviousness case on the claimed subject matter; it is a privacy/blocking mechanism. Citing it as support would be padding. I list it only because it appears on the face of the '809 and bears on the motivation for longer/secret codes, which the '809 itself invokes ("in a pharmacy or financial situation where privacy and security concerns are paramount, a longer length provides greater security") — an argument that cuts toward predictability, not away from it.

5. Bottom line

• The strongest § 103 theory is Ground 1 (Becker + Eber), reinforced by Mickle. US 2004/0046642 establishes group addressing of a subset of RFID tags by comparing a transmitted group address, containing fixed elements and "any value" wildcard elements, against a stored identifier, with a variable number of compared elements (L, M) and a match declared on a partial comparison. US 6,831,548 establishes that data carriers exist specifically to reduce power consumption upon reception of data. US 7,375,637 establishes the low-power "smart buffer" that decides whether a received signal contains the tag's identifier before waking the main processor, expressly to keep non-addressed transponders asleep and to prevent noise-triggered wake-ups. The '809's own admitted problem — let tags bail out early rather than analyze the whole wake code — supplies the motivation to combine, and the combination is a predictable, mechanism-level optimization of a known architecture.
• The strongest § 103 theory for the population-partition claims (C6/C9/C10) is Mickle + Becker, because Mickle literally discloses distinct transponder groups woken by distinct criteria-complexities (different burst-switch frequency combinations), which makes "first subset responds to a length-1 command, second subset to a length-2 command" an obvious parameter substitution.
• The concept-level method claims are, in my assessment, more likely than not obvious over Becker in view of Eber and/or Mickle, with the '809's own Background admission doing much of the work.
• The circuit claims are the real battleground, and the outcome turns entirely on whether they recite the self-clocking, clockless, current-mirror delay-based interrupt detection and on-the-fly MUX comparison of FIGS. 10–13 or merely a generic "interrupt period within a range" plus register/counter/comparator. The generic version is met by admitted ISO 18000-7 practice plus O'Toole/Eber; the specific version is not disclosed or suggested by anything I located in the '809's citation record.

6. What to do before relying on any of this

1. Pull the issued claims of US 8,674,809 from USPTO PatentCenter or the Google Patents Claims tab, and the claims of the parent US 8,248,211 — my element chart in § 1 is concept-derived and must be replaced with literal claim text before any invalidity contention is written.
2. Read the full file histories of 11/186,666 → US 8,248,211 → 13/552,514 → US 8,674,809, and of the EP counterpart EP 1904949 B1, to capture every examiner citation, every applicant remark about US 2004/0046642, and every amendment. The EP prosecution's D1 designation is the most probative document in the record and I have only its one-sentence characterization.
3. Verify Eber (US 6,831,548) and Collins (US 2005/0052283) at the specification level — I confirmed their existence, dates, and titles from the '809's own front-page citation list but was rate-limited before pulling their disclosure text, so I have not quoted them.
4. Confirm the Auto-ID/EPC Class-1 filter-string standard's publication date against a primary source before relying on it as § 102(b) art; the pointer-plus-bitstream description is currently sourced to a vendor manual, not the standard itself.
5. Check for secondary-consideration evidence attributable to the '809 specifically (the trade-secret litigation previously summarized concerns Zest Labs/Intelleflex's commercial system, not this patent, and no litigation nexus is established — a patentee asserting nexus to that commercial success would need to prove the patented feature drove it, which the trade-secret framing makes difficult).

Explicit uncertainty statement: I do not have the claim text, and I could not retrieve the Google Patents "Prior Art" tabulation in this session; my reference list is reconstructed from the patent's own front-page citation list and Justia's listing. Any statement in this memo about a specific claim number, claim preamble, or claim ordering is unverified. I found no reference in the '809's citation record that discloses or suggests the specific clockless, current-mirror-delay interrupt architecture of FIGS. 10–11B, and I would not represent otherwise.