US 8994498

I'll search for the specific patent number and related litigation/docket information.

Let me search for litigation/docket activity and the abstract.

US Patent 8,994,498 — Analyst Summary

Scope note / caveat first: My tool available here is a general web search, and a re-run for the exact docket queries returned either empty results or the step limit. I could not directly execute a native USPTO PatentCenter/PAIR query or a CAFC docket query. So the bibliographic data below is grounded in the authoritative full text you supplied (Google Patents fetch of US8994498B2) plus corroborating pages, while the litigation/docket portion is a negative finding, not a verified clearance.

---

1. Bibliographic data

Primary reference: https://patents.google.com/patent/US8994498/en ; Justia record: https://patents.justia.com/patent/8994498 ; https://uspto.report/patent/grant/8994498

---

2. Abstract

I do not have the literal, verbatim abstract text for US 8,994,498 in the material available to me (the Google Patents HTML body you supplied includes the description and definitions but not the abstract block). Rather than reconstruct it, here is the closest authoritative-language summary, which is the "Briefly stated" passage from the same family's disclosure (US 2016/0267732 A1, Justia):

Various embodiments are directed towards authenticating users using biometric devices. The biometric device may capture one or more biometric features of a user wearing it, including features corresponding to an electrocardiogram of the user. The user may be authenticated based on those features, including by communicating the biometric features to an authorized authentication device (AAD) separate from the biometric device and comparing them to a biometric profile stored on the AAD; the biometric device is preauthorized for the user when the stored profile corresponds. When the preauthorized device senses an access point, an authorization signal is provided to the access point, and access follows an affirmative confirmation. Additional conditions (password, PIN, gesture, voice command, finger tap, distance, additional biometric features) may apply. If the preauthorized device is removed from the user, it is deauthorized. Where two or more access points are sensed, one may be selected based on a secondary condition. An administrative user may supply profile information for multiple users, with access granted based on that profile information.

Treat this as a faithful paraphrase of the specification's summary paragraph, not as the certified abstract verbatim. If verbatim abstract text is required, pull it from the USPTO full-text record for 8,994,498.

---

3. Plain-language overview of the independent claims

Based on the retrieved claim text (Justia claim listing for the family, which mirrors the '498 claims), the patent appears to have four independent claims, written in parallel across four statutory categories:

Claim 1 — Method. A method of authenticating a user with a wearable biometric device: capture at least one biometric feature (expressly including an ECG) of the wearer; authenticate by sending the biometric data to a separate AAD and comparing against a biometric profile stored on the AAD; when it matches, preauthorize the device for that user; then, when the preauthorized device senses an access point, send an authorization signal, with access conditioned on the access point affirmatively confirming the signal.
(I did not retrieve claim 1's verbatim text; the wording above is reconstructed from the parallel independent claims and dependents. Flag as approximate.)

Claim 8 — Biometric device (apparatus). A biometric device whose processor is operative to perform the same core sequence — capture biometric features, authenticate via the AAD-hosted profile, enter a preauthorized state, and issue an authorization/control signal to an access point. Claim 14 (which depends from it) adds administrative-user configuration of profiles for a plurality of users, with access enabled based on those profiles.
(Claim 8 text not retrieved verbatim; inferred from dependent claim 14's "The biometric device of claim 8" reference.)

Claim 15 — System. A system including a biometric device whose processor: captures at least one biometric feature including an electrocardiogram of the user; authenticates the user by communicating that feature to an AAD separate from the biometric device and comparing it to a biometric profile stored on the AAD, the device being preauthorized when the stored profile corresponds; preauthorizes the device upon authentication; and, upon the preauthorized device sensing an access point, provides an authorization signal to that access point — the user being granted access based at least on an affirmative confirmation of the authorization signal by the access point.

Claim 21 — Processor-readable non-transitive storage media. The same functional sequence embodied as instructions on a non-transitory processor-readable medium, again ending with access granted on affirmative confirmation of the authorization signal by the access point.

Representative dependents (for flavor): claim 16 (other/additional biometric features); claim 17 (deauthorization on removal of the device, disabling access); claim 18 (secondary condition — password, PIN, gesture, voice command, finger tap, proximity/distance, or additional biometrics); claim 19 (list of permitted access points based on user profile); claim 20 (administrative-user multi-user profile provisioning).

Key technical/legal takeaway: the claims deliberately decouple identity authentication from access control — the wearable authenticates once against a user-controlled AAD (ECG-based, per the description a three-factor scheme: biometric + wearable possession + AAD possession), then acts as a proxy/arbiter of identity for downstream access points using a preferably non-biometric control signal, until the device is removed.

---

4. Identical-identifier discrepancy you should be aware of (per strict literal-reading rule)

One search result — a page titled "Reversible Vasectomy Device and Method" — lists:

"Application Number: US 8994498 … Filing Date: 1978-04-24 … Art Unit 335"

(https://patents.searchlight.law/doc/[US4200088](/patent/US4200088))

Read strictly and literally, this is not the same document as US 8,994,498 B2. It is an old six-digit-era application serial number 899,498, whose patent issued as a different number (the page header cites US4200088), concerning a vasectomy device. It is a serial-number/patent-number collision, not an identity. I flag it because your instruction is to interpret identifiers literally and not auto-correct them — so I am explicitly not treating "US 8994498" in that source as a reference to the Nymi patent, and I am not silently rewriting it to 8,994,498 B2.

---

5. CAFC 2026 dockets

Finding: no results. Searches for "8994498" aimed at Federal Circuit docket activity for 2026, and for Nymi/Bionym appellate litigation, returned:

• empty result sets, or
• only unrelated/appendix citations (e.g., the IEEE patent-citation report noting the '498 patent had 125 citations within ~5 years — citation index 43.08 — and is Nymi's "key patent"), or
• unrelated uses of the string (e.g., a Brazilian corporate-registry filing number "8994498").

I found no evidence of any Court of Appeals for the Federal Circuit docket in 2026 involving US 8,994,498, and no evidence of any district-court infringement action naming the patent.

Caveat: this is a weak negative. My search tool could not query CAFC/CM-ECF or Docket Navigator/RPACER directly, and the confirming query hit the step limit. If you need a defensible docket clearance, the authoritative checks are: (a) USPTO PatentCenter "Patent Litigation" / assignment record for 14/340,414; (b) CAFC docket search by patent number; (c) PACER/RPACER. Do not rely on the absence of search hits as proof of no litigation.

---

6. Confidence summary

• High confidence: patent number, title, application number 14/340,414, filing date 2014-07-24, priority 2013-07-25, issue date 2015-03-31, inventors (Agrafioti, Martin, Oung), Bionym→Nymi assignment chain, claim 15's substance, and the dependent-claim character.
• Moderate confidence: that claims 1, 8, 15, 21 are the independent claims, and their category assignment (method / device / system / media) — inferred from the Justia claim listing, not from a verbatim full claim set.
• Low confidence / not verified: the verbatim abstract; exact wording of claims 1 and 8; any CAFC 2026 docket activity.

I'll research litigation involving US patent 8994498 specifically.

Let me search litigation-specific databases.

Let me check litigation-specific portals and search for Nymi/Bionym as plaintiff or defendant.

Findings: Litigation Involving U.S. Patent No. 8,994,498

Short answer: I found no litigation involving U.S. Patent No. 8,994,498 in the searches I was able to run. I want to be explicit about the limits of that statement rather than overstate it.

What I searched and what came back

I searched for the specific number "8994498" together with litigation-oriented terms (litigation, lawsuit, PACER, docket, infringement, plaintiff/defendant) and also searched for the patent owner ("Nymi" / "Bionym") in connection with court cases. The results returned:

• The patent itself and its family — Google Patents, Justia, FreePatentsOnline, uspto.report, PatentLeaderboard, RPX Insight records for related continuation US 9,472,033 B2, and the continuations US 9,189,901, US 9,349,235, US 9,472,033.
• Third-party citation / portfolio analyses — e.g., an IEEE "Influence of IEEE on Key Patents" study describing US 8,994,498 as Nymi's "key patent," cited 125 times in five years.
• Corporate/funding news — the 2017 GII Tech Ventures investment in Nymi; Nymi product and platform technical documentation.
• Unrelated hits containing the digit string — e.g., the Russian bankruptcy filing, a Brazilian corporate registry, and unrelated patents citing it as prior art. None is litigation.

Why I could not confirm a negative

I was not able to directly query or retrieve results from the litigation portals you named. My searches did not return any docket entries, complaints, case numbers, CAFC opinions, or Unified Patents litigation records referencing U.S. 8,994,498. That is absence of evidence in the sources I reached, not a verified clearing search. The tool I have access to is a general web search, and docket databases such as PACER, CourtListener, Docket Navigator, and the Unified Patents litigation portal are largely behind logins, paywalls, or JavaScript interfaces that a general web search does not index completely.

For each required field

No case can be listed, because no case was found.

Recommended verification before relying on this

Because "no known litigation" is a negative that matters legally (e.g., for freedom-to-operate, IPR timing, § 315(b) real-party-in-interest analysis, or patent-valuation due diligence), please verify against primary sources rather than treating my search as conclusive:

1. USPTO Patent Center / Patent Assignment for the '498 patent — changes of assignee (Bionym → Nymi) sometimes coincide with enforcement activity.
2. PACER (District Courts and Federal Circuit) — search party names "Nymi," "Bionym," and the individual inventors.
3. CourtListener / RECAP — free docket mirror; search "8,994,498."
4. Unified Patents litigation portal (portal.unifiedpatents.com/litigation/caselist) — the page you referenced; I could not surface its contents for this number.
5. PTAB (IPR/PGR) proceedings — search the '498 patent number at the PTAB E2E / Patent Trial and Appeal Board End-to-End system; IPRs are not "litigation" per se but are frequently the first sign of a dispute.

One contextual note (not a litigation finding)

The '498 patent is the parent of a family of continuations (US 9,189,901; US 9,349,235; US 9,472,033), each titled identically. If litigation ever is filed, it is common for a plaintiff to assert several members of such a family together. When you run a docket search, I'd recommend searching each of those four patent numbers, not just '498, so you don't miss a case that names only a continuation.

If you can give me access to or paste results from a docket search (PACER, CourtListener, or Unified Patents), I can structure those into the plaintiff/defendant/jurisdiction/case-number/status table you asked for.

PTAB / AIA Trial Proceedings — US 8,994,498

Patent: US 8,994,498 B2 ("Preauthorized wearable biometric device, system and method for use thereof")
Patent Owner (current): Nymi Inc. (originally Bionym Inc.)
App. No.: 14/340,414 · Filed: 2014-07-24 · Granted: 2015-03-31 · Priority: 2013-07-25
Statutory term: anticipated expiration 2034-07-24

Caveat on sourcing. The structured "PTAB proceedings on file" block in my instructions — sourced from the USPTO Open Data Portal — states that ODP returns no AIA trial proceedings for this patent as of the most recent ingest. I ran corroborating web searches (Google Patents, PTAB-related document hosts, Unified Patents portal references, litigation aggregators, and Nymi corporate/patent-family material). I found no IPR, PGR, or CBM petition, institution decision, Final Written Decision, or appeal naming US 8,994,498. Several search hits referenced a "‑498 patent," but in every instance the hit resolved to a different patent (an electrode-plate patent, a bacon-processing patent, an EAGLE XG glass patent, Hormel's 9,980,498). I flag that explicitly so this analysis is not mistaken for a claim that unrelated PTAB decisions apply here.

---

Proceedings overview

Total AIA trial proceedings on file for US 8,994,498: 0. There is no active proceeding, no proceeding in which claims were invalidated, no proceeding in which claims were sustained, no settlement terminating a proceeding, and no institution denial — because no petition was ever filed. The bottom-line defensive posture is therefore the least favorable of the possible postures: the patent stands entirely untested before the PTAB, all claims remain live as issued, and a defendant has no ready-made invalidation record to borrow, cite, or leverage for estoppel purposes. "No PTAB activity" is not the same as "hardened" — it means the validity fight has simply never been picked.

---

Proceedings

None. No proceeding entries exist to detail.

For completeness on the requested fields: no petitioner, no filing date, no judge panel, no petition grounds, no institution decision, no Final Written Decision, no termination, and no Federal Circuit appeal exist for this patent in the record I was able to review. I am not going to construct placeholder proceedings.

---

Strategic summary

Claim status: everything is UNTESTED. Because no AIA trial has ever been instituted, no claim of US 8,994,498 has been canceled, confirmed, or even construed by the Board. The full issued claim set remains in force as granted, and the patent's term runs to an anticipated 2034-07-24 — roughly eight years of remaining assertion window as of 2026-09-30. Note the constraint on my analysis: the patent text supplied to me is truncated before the claims (it ends mid‑description of FIG. 10), so I cannot quote or enumerate the claim numbers, independents, or dependents of US 8,994,498 from this record. Any statement about specific claim numbers here would be fabricated. Before building a defense theory, pull the printed claim set from the USPTO Patent Public Search / PatentCenter record for 14/340,414.

Estoppel landscape — effectively empty, which cuts both ways. Section 315(e)(2) estoppel only attaches to a petitioner that was a party to an instituted IPR. With zero instituted proceedings, there is no petitioner-side estoppel and no Patent Owner-side estoppel on this patent. Practically, this means a defendant is not constrained by anything anyone else already raised — the entire prior-art field is available. It also means the defendant cannot free-ride: there is no existing FWD, no Board claim construction, and no invalidity record to import into a district court case or an ITC investigation. If you are being asserted against, you would be writing on a blank slate, and you would be the first to fund that work. Note also that a timely-filed IPR (within one year of service of an infringement complaint, per § 315(b)) is your only route to the Board; a late filing is statutorily barred.

Pattern signals. Nymi Inc. is a small Canadian company (Toronto; founded 2011 as Bionym; ~40–50 employees; acquired by Innominds in April 2022) with a focused patent portfolio (roughly 30–60 documents in family, with related US grants including US 9,189,901, US 9,349,235, and US 9,472,033 — all claiming priority to the same 2013-07-25 filing). That profile is inconsistent with a high-volume serial litigant that would attract repeated IPRs, and consistent with a company whose patents have largely been used defensively or for partnership credibility (Mastercard, G+D, Körber) rather than in an aggressive assertion campaign. There is no evidence of a defensive aggregator (e.g., Unified Patents) challenging this patent, and no pattern of the Patent Owner prosecuting appeals to the Federal Circuit — because there is nothing to appeal. Search hits referencing the "Nymi Band" appear in third-party district court exhibit lists (e.g., Apple Inc. v. Masimo Corp., D. Del. 1:22-cv-01377), where Nymi material appears as prior-art-of-the-art evidence, not as the patent in suit.

---

Recommended next steps

If you are a defendant facing assertion of US 8,994,498:

1. State the absence plainly in your own analysis, but verify it yourself on the day you rely on it. Confirm zero proceedings via PTAB E2E / PTAB Center (https://ptacts.uspto.gov/) and the USPTO Open Data Portal for application 14/340,414, and check the Patent Trial and Appeal Board's daily petition-filing notices. The absence of PTAB activity is a current fact, not a permanent one — a petition could land at any time, and if you are an accused infringer, the 1-year § 315(b) clock is running from service of the complaint.

2. Do not represent to a court that any claim has been canceled. Nothing has been. Any argument that "claims 1–5 are dead" would be false and sanctionable. There is no FWD to link to; I have deliberately not created one.

3. Build the invalidity case from scratch, and use the pre-2013 art. Priority is 2013-07-25, so the § 102(a)/§ 102(e) window is broad. The specification itself is unusually candid about the art landscape: it frames ECG biometrics as an established field, cites the AC/LDA algorithm as "a standard for the comparison of fiducial dependent and independent algorithms," and describes BLE, NFC, proximity sensing, and electronic-continuity removal detection as known building blocks. That kind of self-identification of known elements is fertile § 103 territory (KSR-style motivation-to-combine arguments), and it is arguably the strongest line available given the absence of any prior Board construction to rely on.

4. Prosecution history and family-wide review. Because there is no PTAB record, the public prosecution history of 14/340,414 and its continuations (US 9,189,901, US 9,349,235, US 9,472,033) is your substitute for a validity record — look for examiner-cited art, narrowing amendments, and any art that the applicant distinguished. That material is public via PatentCenter.

5. Consider whether the real exposure is elsewhere in the family. With no IPR on the '498, the practical question for a defendant is which family member is actually being asserted and whether that patent is a better IPR target (different claim scope, different art, and a fresh § 315(b) clock if you've recently been served on it).

If no assertion is currently pending: the absence of PTAB activity simply reflects that this patent has not been litigated to the point of attracting challengers. It is a weak signal either way — it does not mean the claims are strong, and it does not mean they are weak. It means nobody has spent the money to find out.

I'll research the assignment records and corporate history for US 8994498 before reconstructing the chain.

Let me dig for the specific reel/frame records and the corporate status of Nymi.

Assignment & Ownership Analysis — US 8994498 B2

Patent: US 8994498 B2 — "Preauthorized wearable biometric device, system and method for use thereof"
Application: 14/340,414 · Filed: 2014-07-24 · Issued: 2015-03-31 · Priority: 2013-07-25 (provisional) · Anticipated expiry: 2034-07-24

Sourcing note (read first): I was able to retrieve the conveyance types and dates from Google Patents legal events, but not the reel/frame numbers or correspondent fields from the USPTO Assignment Center. The Assignment Center (https://assignmentcenter.uspto.gov/) is an interactive query interface; no indexed record for this reel/frame was surfaced by the searches I ran. I have therefore not fabricated reel/frame or correspondent data — those fields are marked "not retrieved" below, and any NPE conclusion that would rest on them is flagged accordingly. Verify directly at USPTO Assignment Center and assignment.uspto.gov searching patent number 8994498.

---

Inventors

Pattern observations (flagged, not asserted):

• Founder departure: Foteini Agrafioti is not the inventor of record on the later Nymi continuations in this family and is consistently described in third-party profiles as a former Nymi founder; she subsequently led RBC's Borealis AI. I could not verify the exact departure date, so I cannot confirm the "<12 months post-filing" fire-sale precursor pattern — treat as unclear, not a finding.
• Karl Martin is identified as Bionym/Nymi CEO through the growth phase; later Nymi press material (2022) quotes Dave Rai as CEO, and one aggregator (Tracxn) lists Chris Sullivan as CEO. Leadership transition dates are not verified.
• No evidence any inventor individually retained or later re-acquired rights; all three appear as assignors on the 2014-07-24 original assignment.

---

Original assignee

BIONYM Inc. (Toronto, Ontario, Canada) — named on the issued patent. Renamed NYMI INC. in December 2014 (rendering the current assignee Nymi Inc.).

• Product embodying the claims: Yes. The Nymi Band — a wristband that captures an ECG, authenticates the wearer, stays "activated" while worn, and is de-activated when the clasp/continuity circuit is broken (i.e., it practices the claim-1 preauthorization-then-deauthorization-on-removal mechanism). Commercialized as the Nymi Band Discovery Kit (Dec 2014), then Nymi Enterprise Edition, then the Nymi Connected Worker Platform (Band 3.0, FIDO2-certified). This is a direct, documented implementation of the patent's core concept.
• Primary line of business: Biometric wearable identity/authentication hardware + companion software for enterprise access control; a University of Toronto ECG-biometrics spin-off founded 2011. Funded ~$29–30M across five rounds (Relay Ventures, Ignition Partners, Konica Minolta, Gulf Islamic Investments, MasterCard/Salesforce Ventures participation in the 2014 $14M Series A).
• Current status: Acquired, still operating. On 2022-04-12 Nymi was acquired by Innominds (San Jose, CA), a private product-engineering services firm; Nymi became a wholly-owned independent subsidiary continuing to operate from Toronto. Advisor: True Blue Partners. Not a bankruptcy, not a dissolution, not an NPE. Innominds is privately held, so no SEC 10-K/8-K trail exists for this chain — the SEC cross-reference returned nothing usable by design, not by omission.

---

Assignment timeline

Two recorded events appear in the ownership history. Reel/frame and correspondent were not retrievable from the sources available to me and are explicitly left blank rather than guessed.

• Executed ~2014-07-24 (doc date) / recorded 2014-07-24 — Reel not retrieved / Frame not retrieved

  • Conveyance: Assignment of Assignors' Interest (original assignment)
  • Assignor: Foteini Agrafioti; Karl Martin; Stephen Oung
  • Assignee: BIONYM INC.
  • Correspondent: not retrieved (field unavailable in sources reviewed)
  • Context: Founder-to-company assignment at filing — the standard original-title recording, not an NPE event. (Source: Google Patents legal events for US8994498.)

• Executed date not recorded in sources / recorded 2015-07-21 — Reel not retrieved / Frame not retrieved

  • Conveyance: Change of Name
  • Assignor: BIONYM INC.
  • Assignee: NYMI INC.
  • Correspondent: not retrieved (field unavailable in sources reviewed)
  • Context: Change of name only — the December 2014 corporate rebrand from Bionym to Nymi. No change in beneficial ownership; same legal entity. (Source: Google Patents legal events, "CHANGE OF NAME (SEE DOCUMENT FOR DETAILS), Assignors: BIONYM INC," recorded 2015-07-21.)

Nothing after 2015-07-21. There is no recorded USPTO assignment for the 2022 Innominds transaction — consistent with a share purchase of Nymi Inc. (the corporate entity that owns the patents simply changed hands; patent title never moved), though I could not independently confirm the deal structure. Successor US patents in the same family — US9189901 (14/670,316), US9349235 (14/942,919), US9472033 (15/162,109) — all remain attributed to Nymi, corroborating that no divisional/continuation carve-out or partial transfer occurred.

---

Timeline diagram

timeline
    title Ownership of US 8994498
    2013 : Priority date filed
    2014 : Application filed by Bionym Inc
         : Founders assign rights to Bionym
    2015 : Patent issued
         : Bionym renamed Nymi Inc
    2022 : Nymi acquired by Innominds

---

NPE / troll-pattern signals

---

Verdict

Insufficient data (with a strong affirmative lean away from NPE)

The ownership record for US 8994498 consists of nothing beyond the original 2014-07-24 inventor-to-Bionym assignment and the 2015-07-21 Change-of-Name to Nymi Inc. — i.e., "only the original assignment" in substance, since a name change is not a transfer of title to a new party. That maps directly onto the Insufficient data definition. Critically, the deficit is not ignorance about who owns it (Nymi Inc., an operating company, holds it) — it is the absence of any post-issuance transfer chain to evaluate for NPE structure. Every available affirmative indicator points the other way: the assignee is a University of Toronto spin-off that shipped the Nymi Band and later the Connected Worker Platform, went through a Series A/B/C fundraising path, and exited via a strategic sale to Innominds (2022-04-12) rather than an IP-monetization vehicle. No NPE signals are present on the record available.

Caveats: (1) Reel/frame numbers and correspondent-of-record were not retrievable, so Signal 3 (repeat correspondent) is untested — an NPE read built on correspondent recurrence is not supportable either way. (2) The Innominds deal leaves no USPTO assignment entry, consistent with a share purchase; confirm deal structure in Innominds/Nymi corporate filings if it matters. (3) Both assignors are non-US private companies, so no SEC corroboration path exists.

Verify: USPTO Assignment Center · assignment.uspto.gov patent search · Google Patents legal events — US8994498B2

I'll research the prior art citations for US 8,994,498. Let me run several targeted searches.

I have the citation list from Justia. Now let me research the most relevant references in detail.

Prior-Art Analysis — US 8,994,498 B2

Scope note first (per your strict-rule instruction). My available tool is a general web search; I could not execute a native USPTO PatentCenter/PAIR "References Cited" query, and I hit the step ceiling before completing every reference dossier. The authoritative Google Patents full-text you supplied truncates in the "FIG. 10" description and does not reproduce the front-page "References Cited" block. I therefore retrieved the citation list from the Justia rendering of the same document's "Referenced Cited" section (https://patents.justia.com/patent/[8994498](/patent/8994498)), and cross-checked it against the front-page reference list of the sibling patent US 9,646,261 B2 (same IDS family, patentimages.storage.googleapis.com). The two lists agree. Treat the list below as the '498 IDS citation set; treat the descriptions I did not retrieve live this session as flagged, not verified.

---

1. A threshold point that governs the whole § 102 analysis

US 8,994,498 has an effective filing date after 16 March 2013 (filed 2014-07-24; priority asserted to 2013-07-25). It is therefore an AIA patent; the governing provision is AIA 35 U.S.C. § 102(a)(1)/(a)(2), with the 2013-07-25 date fixing what is "prior art," plus the § 102(b)(1)/(b)(2) exceptions.

Discrepancy to flag (do not auto-correct): the sources disagree by one day on the priority date —

• Google Patents full text: "Prior art date 2013-07-25"; filing 2014-07-24.
• Unified Patents portal entry for US-8994498-B2: "Priority Date: 2013-07-24."

Both are reproduced literally. This one-day difference is immaterial to the art below (all references predate either date), but it is noted for accuracy.

Consequence for the art: several of the closest references are the inventors' own / commonly-owned earlier work (Agrafioti et al.). Under AIA § 102(b)(1)(A) (inventor's own disclosure within the 1-year grace period) and § 102(b)(2)(C) (commonly-owned earlier application), those may be excluded from prior art. I flag this per-reference below.

---

2. The full IDS citation set for US 8,994,498

2.1 U.S. Patent Documents (14)

2.2 U.S. Patent Application Publications (16)

2.3 Foreign Patent Documents (3) and NPL (1)

---

3. The closest art — detailed dossiers

3.1 US 6,041,410 B1 — Hsu et al., "Personal identification fob" (issued 2000-03-21)

Source: https://patents.google.com/patent/[US6041410A](/patent/US6041410A)

• Disclosure: A portable fob holds a biometric sensor (fingerprint), a memory storing the reference fingerprint, and a correlator that compares a fresh scan to the stored reference. On a match, the fob generates a numerical value (e.g., a CRC of the stored reference image), encrypts it (doubly — door public key + fob private key), and transmits the encrypted value plus user identification data to a "door" (car door, building door, ATM, computer). The door compares the received value to one registered earlier and grants access on confirmation.
• § 102 relevance: Teaches wearable/portable biometric capture → comparison → control signal to an access point → affirmative confirmation at the access point (registration model).
• Claim(s) potentially implicated (§ 102): Narrow dependent-type subject matter — the access-point-confirmation and control-signal limitations shared by independent claims 1, 8, 15, 21.
• Critical gap: Hsu stores the biometric profile on the fob itself (on-device authentication). It does not disclose a biometric profile stored on a separate Authorized Authentication Device and the preauthorization of the wearable by that AAD. It therefore cannot anticipate claims 1/8/15/21 as a whole; it is § 103 material and § 102 material only against narrower features.

3.2 US 6,580,356 B1 — Alt et al., "Advanced personal identification systems and techniques" (issued 2003-06-17)

Source: https://patents.google.com/patent/[US6580356B1](/patent/US6580356B1)

• Disclosure: A transmitter worn in direct physical and electrical contact with the body (wristwatch, ring, bracelet, necklace, pendant, eyeglasses frame, or implant), communicating an encoded identity signal to a recognition system through the body's electrical conductivity upon touch. The transmitter auto-activates/deactivates: it senses a change in electrical impedance when body contact is made/broken and deactivates when the connection terminates (or on time-out).
• § 102 relevance: This is the closest cited art on wearability and on deauthorization upon loss of wearer contact.
• Claim(s) potentially implicated (§ 102): The wearable-form-factor and removal/separation-based deauthorization concepts — directly comparable to dependent claim 17 (deauthorization on removal disabling access). Also relevant to the "preauthorized device ... within range/proximity" limitations.
• Critical gap: No biometric (physiological) profile, no separate AAD, no access-point affirmative-confirmation architecture.

3.3 US 2007/0177770 A1 — Derchak et al., "System and Method for Identity Confirmation Using Physiologic Biometrics to Determine a Physiologic Fingerprint" (published 2007-08-02)

• Disclosure: Builds a "physiologic fingerprint" from body-worn physiological sensors to confirm identity.
• § 102 relevance: Directly on point for the "capture one or more biometric features … including physiological/ECG" limitation.
• Claim(s) potentially implicated (§ 102): The biometric-capture and identity-confirmation elements of claims 1, 8, 15, 21; and dependent claim 16 (additional/other biometric features).
• Gap: No AAD-separated profile, no access-point preauthorization model.

3.4 US 2008/0294907 A1 — Hively (published 2008-11-27)

(Exact title not verified this session; inventor Lee Hively is associated with ORNL physiological/ECG-based identification.)

• § 102 relevance: Physiological-signal (ECG-type) human identification — relevant to the "electrocardiogram of the user" limitation that the specification emphasizes and that appears in claim 15.
• Claim(s) potentially implicated (§ 102): biometric/ECG elements of claim 15 (and parity elements in claims 1/8/21).
• Caveat: I did not retrieve the verbatim title/abstract; do not cite the title without confirming it on the front page.

3.5 The inventors' own family — WO 2012/151680 A1 and US 2014/0188770 A1 (Agrafioti, Bui, Hatzinakos)

Sources: https://patents.google.com/patent/WO2012151680A1 ; https://uspto.report/patent/app/20140188770

• Bibliographic: PCT filed 2012-05-10; published 2012-11-15; prov. 61/484,470 filed 2011-05-10. US national-phase publication US 2014/0188770 A1, published 2014-07-03 (App. 14/116,058), which later issued as US 9,646,261 B2 (Assignee Nymi Inc.).
• Disclosure: Biometric security using physiological signals (ECG, PPG, EEG, BVP, EMG) for instantaneous (IIR) and continuous (CIR) identity recognition, using a machine-learning/template framework, with explicit access-control application (physical and logical).
• § 102 relevance: This is the single most technically relevant cited item for the physiological/ECG authentication core — it is the same lab's foundational ECG-biometrics work.
• Claim(s) potentially implicated (§ 102): ECG-biometric-authentication elements of claims 1, 8, 15, 21; the "continuous/instantaneous" and sensor-collection subject matter feeding dependent claims.
• ⚠️ Decisive exception caveat: Same inventors (Agrafioti et al.) and commonly owned (Nymi/Bionym). Published 2012-11-15, i.e., within one year of the asserted 2013-07-25 priority date → potentially excluded from prior art under AIA § 102(b)(1)(A) (inventor's own disclosure) and/or § 102(b)(2)(C) (commonly-owned earlier application). Do not assert this reference as § 102 prior art without first resolving the exception. Contrast the NPL below, which is >1 year before priority.

3.6 NPL — Agrafioti et al., "Medical Biometrics in Mobile Health Monitoring," Security and Communication Networks, 2011

• § 102 relevance: Printed publication (>1 year before 2013-07-25) → no grace-period exception applies even though an author is an inventor; this is qualifying § 102(a)(1) art on physiological/ECG biometrics.
• Claim(s) potentially implicated: Biometric/ECG elements of claims 1, 8, 15, 21 (as a printed-publication reference — not, absent more, a full-claim anticipation).

---

4. § 102 anticipation map (preliminary, claim-by-claim)

Reading § 102 strictly — anticipation requires every element in one reference — the honest conclusion is:

No single cited reference discloses all elements of independent claims 1, 8, 15, or 21. The core inventive combination — (a) wearable biometric capture, (b) profile stored on a separate AAD, (c) preauthorization of the wearable by the AAD, (d) non-biometric control signal to an access point, (e) affirmative confirmation at the access point — is not met by any one reference. Hsu (2.1) has (a),(d),(e) but stores the profile on the fob; Alt (3.2) has wearability + deauthorization but no biometric/AAD; Derchak/Hively/Agrafioti have only the biometric side. This is consistent with the claims having issued over the IDS. The cited art is therefore best characterized as § 103 (obviousness) material, with § 102 available only against narrower dependent claims.

---

5. Explicit confidence and verification statement

High confidence (verified this session):

• The citation set itself (30 U.S. documents, 3 WO documents, 1 NPL) from the Justia "Referenced Cited" record, corroborated by US 9,646,261 B2's front-page list.
• Full content and quote-level support for Hsu US 6,041,410 and Alt US 6,580,356 (Google Patents).
• Identity/content of the Nymi-Agrafioti family (WO 2012/151680, US 2014/0188770 → US 9,646,261) and Derchak US 2007/0177770.

Moderate confidence: Inventor-based descriptions for Giobbi, Beenau, Sengupta, Pasquero, Prokoski, Felsher, Sano items (title recall, not re-verified).

Not verified (flagged as such in the tables): exact titles/abstracts for US 6,799,726; 7,023,320; 7,095,707; 7,689,833; 7,780,080; 8,371,501; 8,468,362; US 2002/0140542; 2003/0046228; 2005/0068171; 2009/0146947; 2009/0199264; 2010/0030695; 2010/0306106; 2012/0004523; 2012/0316406; 2012/0317024; 2013/0322622; WO 2005/117527; WO 2007/128975. I retrieved only their numbers/dates/inventors, not their substance — and I did not invent descriptions for them.

Recommended primary-source verification (I could not reach these directly):

1. USPTO PatentCenter → application 14/340,414 → "References Cited" (front page) and the "IDS/Considered" tab — the certified citation set and examiner's considered list.
2. Front pages of the individual references at image-ppubs.uspto.gov (e.g., /dirsearch-public/print/downloadPdf/6799726, etc.) for verbatim titles/abstracts.
3. For the AIA exceptions on the Agrafioti family: confirm common ownership and inventor identity on the assignment record for App. 14/116,058 and its parent.

If you paste the USPTO PatentCenter "References Cited" HTML for 14/340,414, I can complete the remaining ~12 unverified dossiers and finalize the § 102/§ 103 per-claim mapping against the verbatim claim text.

I'll ground this obviousness analysis in the actual cited prior-art references for the '498 patent rather than reasoning from memory. Let me pull the reference lists.

Let me get the exact "References Cited" front-page list for the '498 patent.

Let me try to retrieve the actual front-page reference list and the IPR/reexam context directly.

Obviousness Analysis — U.S. Patent No. 8,994,498 B2 under 35 U.S.C. § 103

0. Scope note: what the "Prior Art" section of this page actually gave me

Before any legal conclusion, an accuracy caveat that materially affects how much weight to put on what follows.

What the supplied page contained. The Google Patents fetch you provided includes the Prior Art metadata block only in this form:

Prior art keywords: user; biometric; biometric device; access; access point
Prior art date: 2013-07-25

It does not include the front-page "References Cited" tables (U.S. Patent Documents, Foreign Patent Documents, Other Publications) or the "Cited By" table. Those tables are what a § 103 analysis is supposed to be built on. So I could not do the literal task "use the results from the Prior Art section of this page" — the tables were truncated out of the fetch.

What I recovered instead. I retrieved a "References Cited" list that appears on the front page of the sibling patent in the same family, US 9,472,033 B2 (a real Google-hosted patentimages PDF), together with the family's Non-Patent Citations list (from the Justia record for US 2016/0267732 A1 and US 9,472,033). This is genuine retrieved data, but it is the '033's list, not confirmed to be the '498's list. The two lists are not interchangeable — the '033 front page itself lists "8,994,498 B2 — 3/2015 — Agrafioti" as prior art against itself, which proves the lists differ. Treat everything in § 2 below as the art of the family's record, not as the art cited against the '498 specifically.

OCR / literal-identifier flags (per the no-auto-correction rule). The retrieved front-page list has visible OCR damage, and I am quoting it as retrieved, without repairing it:

Consequence for this analysis. Because I can verify the identity of the references of record but in most cases not their substantive disclosure (column/line content), I will do two things and label them separately:

1. Attribute technical content only where I have a retrieved basis (the reference's assigned classification code on the '033 face, which was in the retrieved text, or the publication itself);
2. For the rest, explain the type of combination and the motivation, and say explicitly where naming a specific disclosure would be fabrication on my part.

This matters: a § 103 rejection must cite specific disclosures. I can tell you where the obviousness case is strong and where the record is thin, but I cannot hand you a finished rejection without the column/line cites.

---

1. The § 103 frame and the claim element chart

1.1 Governing standard

• Graham v. John Deere Co., 383 U.S. 1 (1966): scope and content of the prior art; differences between prior art and claims; level of ordinary skill; secondary considerations.
• KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007): the motivation to combine need not be found in the references; the test is "expanded and flexible"; a combination of known elements is obvious where it yields no more than predictable results; "if a technique has been used to improve one device, and a person of ordinary skill in the art would recognize that it would improve similar devices in the same way, using the technique is obvious."
• MPEP § 2143 exemplary rationales: (A) combining prior art elements according to known methods; (B) simple substitution of one known element for another; (C) use of known technique to improve similar devices in the same way; (D) applying a known technique to a known device ready for improvement; (F) "obvious to try" — choosing from a finite number of identified, predictable solutions.
• MPEP § 2144.05: a recitation of alternatives ("at least one of A, B, or C") where the alternatives are known or obvious is not a patentable distinction — directly relevant here (see § 5.3).
• MPEP § 2129: statements in the applicant's own specification about the state of the art are admissions usable as prior art.

1.2 PHOSITA definition I am applying

A person of ordinary skill as of the 2013-07-25 priority date would have: (i) a working knowledge of biosignal acquisition and ECG signal processing; (ii) experience with short-range wireless protocols (Bluetooth/BLE 4.0, NFC) and proximity estimation by RSSI; (iii) familiarity with access-control systems (PACS, logical access, challenge-response authentication); and (iv) at least a bachelor's degree in EE/CS plus several years in biometric or security engineering, or equivalent. This is a routine-engineering skill level, not a research-lab level — which cuts in favor of obviousness, since every element is a known building block.

1.3 Claim element chart

Using the reconstructed claim structure from the previously generated section (which correctly flagged that claims 1 and 8 were not retrieved verbatim — I flag that again; the chart below follows claim 15 verbatim and mirrors 1/8/21 by inference):

The § 103 target is therefore not "ECG biometrics" (old art) and not "access control" (old art). It is the specific architectural sequence E1→E5, in which authentication and authorization are decoupled in time and in device.

---

2. The art of record (family list), with attribution limits

Retrieved U.S. patent documents on the US 9,472,033 face (OCR as noted in § 0):

6,418,424 B1  Huffberg et al.      2002/014654 A1  Prokoski et al.
6,580,356 B1  Alt et al.           2003/0046228 A1 Wiedenhofel et al.
6,799,729 B1  Vole                 2003/015097 A1  Kelliher et al.
7,023,320 B2  Dvorak               2005/006311 A1  Gallier et al.
7,095,707 B2  Rakib et al.         2007/0016088 A1 Grant et al.
7,378,939 B2  Sengupta et al.      2007/0049267 A1 Kota et al.
7,689,833 B2  Lange      (A61B 5/04525)   2007/006348 A1  Deerlak et al.
7,780,080 B2  Owen et al.          2007/017770 A1  Bailey et al.
7,814,332 B2  Beenau et al.        2008/009163 A1  Dwivedi et al.
8,371,501 B1  Hopkins              2008/021617 A1  Sano et al.
8,412,949 B2  Gibbbl et al.        2008/0253626 A1 Shuckers et al.
8,468,362 B2  Konetski   (G06F 21/32)       2008/0294097 A1 Hively (A61B 5/04012)
8,532,730 B1  Gibbbl               2009/0037983 A1 Chiruvolu et al.
8,869,263 B2  Pasquero et al.      2009/0146947 A1 Isag
                                   2009/0190964 A1  Lang

What I can attribute with a retrieved basis:

• Lange, US 7,689,833 B2 — face classification A61B 5/04525, i.e., the ECG waveform/vector analysis subclass. Its placement in the ECG-signal taxonomy is verified retrieval.
• Hively, US 2008/0294097 A1 — face classification A61B 5/04012, i.e., electrocardiography / cardiac-signal acquisition. Verified retrieval.
• Konetski, US 8,468,362 B2 — face classification G06F 21/32, i.e., "User authentication using biometric data." Verified retrieval. This is the single most probative U.S. reference on the face of the family record for elements E2–E3, because the USPTO itself classified it in the exact biometric-authentication subclass.
• Pasquero, US 8,869,263 B2 — Pasquero is a BlackBerry/RIM inventor; the classification and assignee context are consistent with handheld-device authentication/proximity, but I did not retrieve its claims or abstract this session, so I will not assert its disclosure.
• Beenau, US 7,814,332 B2 — Beenau et al. are American Express inventors in the smartcard/fob biometric-security line. Consistent with a biometric-authorized fob/wireless token, but again: not verified this session.

What I will NOT do: assign specific disclosures to Huffberg, Alt, Vole, Dvorak, Rakib, Sengupta, Owen, Hopkins, Gibbs (×3), Chiruvolu, Lang, Kota, Bailey, Deerlak, Grant, Gallier, Kelliher, Prokoski, Wiedenhofel, Dwivedi, Sano, Shuckers, or Isag. I have their numbers and dates only. Asserting what each teaches would be fabrication.

Foreign art of record (from the same family record): WO 2004012388 (Feb 2004); WO 2005117527 (Apr 2006); WO 2007128975 (Nov 2007); WO 2012151680 (Nov 2012). Content unverified.

Non-Patent Citations of record (retrieved verbatim, and highly probative):

• Biel et al., "ECG Analysis: A New Approach in Human Identification," IEEE Trans. Instrum. Meas. 50(3), 2001, 808–812.
• Wübbeler et al., "Verification of Humans Using the Electrocardiogram," Pattern Recognition Letters 28(10), 2007, 1172–1175.
• Hoekema et al., "Geometrical Aspects of the Interindividual Variability of Multilead ECG Recordings," IEEE TBME 48(5), 2001, 551–559.
• Draper et al., "The Corrected Orthogonal Electrocardiogram and Vectorcardiogram in 510 Normal Men (Frank Lead System)," Circulation 30, 1964, 853–864.
• Odinaka et al., "ECG Biometrics: A Robust Short-time Frequency Analysis," IEEE WIFS, Dec. 2010; and "ECG Biometric Recognition: A Comparative Analysis," IEEE TIFS 7(6), Dec. 2012, 1812–1824.
• Li et al., "Robust ECG Biometrics by Fusing Temporal and Cepstral Information," ICPR, Aug. 2010, 1326–1329.
• Wang et al., "Analysis of Human Electrocardiogram for Biometric Recognition," EURASIP JASP, Jan. 31, 2008, Art. 148658.
• Agrafioti et al., "ECG Based Recognition Using Second Order Statistics," CNSR, May 2008, 82–87; "Signal Validation for Cardiac Biometrics," ICASSP 2010, 1734–1737; "Medical Biometrics in Mobile Health Monitoring," Security & Comm. Networks 4, 2011, 525–539.
• Klosterman et al., "Secure Continuous Biometric-Enhanced Authentication," CMU-CS-00-134, Carnegie Mellon School of Computer Science (2000).
• Bellare et al., "Forward-Security in Private-Key Cryptography" (2003).

This NPL list is, functionally, the applicant's own concession that elements E1 and much of E2 were known. That is an unusually strong starting position for a § 103 rejection, and I address it in § 5.

---

3. Ground 1 — The core independent claims (1, 8, 15, 21)

3.1 Reference set

• Ref A (primary): the wearable-ECG-biometric art of record — Hively (US 2008/0294097 A1, A61B 5/04012) and/or Lange (US 7,689,833 B2, A61B 5/04525), in view of the ECG-biometric NPL (Biel 2001; Wübbeler 2007; Wang 2008; Odinaka 2010/2012; Li 2010) — teaching that ECG is (i) acquired by surface electrodes, (ii) individually distinctive, (iii) stable enough for identification/verification, and (iv) processable by fiducial-independent algorithms.
• Ref B (secondary): Konetski (US 8,468,362 B2, G06F 21/32) — classified by the USPTO in "user authentication using biometric data," teaching biometric-match-then-grant-access on a computing platform. (Content inference limited to the classification; flag.)
• Ref C (tertiary / proximity–access-control): the short-range wireless token-and-reader art of record — Beenau (US 7,814,332 B2), Gibbs (US 8,371,501 B1 / US 8,412,949 B2 / US 8,532,730 B1), Owen (US 7,780,080 B2) — plus the conceded knowledge in the '498 itself (see § 3.3).
• Ref D (removal detection, for D17): the '498's own admitted list (§ 4.2) plus any of the physiological-sensing references of record.

3.2 Element-by-element

3.3 The motivation to combine (KSR / MPEP 2143)

The strongest available motivations are the applicant's own problem statement and its own admitted design space:

1. Identified problem, known solution components. The '498 opens by identifying authentication friction as the problem ("keys and cards… an ever-increasing load"; "passwords and PINs… has become an elusive goal"; "these items, physical or digital, may be stolen or copied"). Reducing repeated authentication by substituting a carried credential is the stated object. KSR: where the problem is known and the elements are known, the combination is obvious absent a teaching away.
2. Known technique to improve similar devices — MPEP 2143(C). Short-range wireless tokens already unlocked doors, cars, and terminals (Ref C). Substituting a biometrically validated token for a passive one is the "same way" improvement KSR presumes.
3. Finite, identified, predictable solutions — MPEP 2143(F). The template location (device/cloud/AAD), the radio (BLE/NFC/WiFi), and the removal detector are each enumerated in the specification as a short menu of known options. "Obvious to try" attaches.
4. Design incentive / low-power engineering. The '498 justifies BLE because it "may consume significantly less power" — a pure engineering trade-off, i.e., a design choice with predictable results.
5. Reasonable expectation of success. Every piece (electrode–skin ECG capture, template matching, BLE RSSI proximity, relay unlocking a lock) was independently proven. Nothing in the combination required a new discovery.

Conclusion on Ground 1 — moderate-to-strong, contingent. On the reconstructed claims, Ground 1 (A + B + C) makes a prima facie § 103 case that is credible but not airtight, for the reason set out in § 6: the independent claims contain at least two limitations — "separate from the biometric device… stored on the AAD" coupled with the AAD being out of range at access time — that a well-drafted reference set teaching a present authorizing device would not reach.

---

4. Ground 2 — The dependent claims

4.1 D16 (other/additional biometric features) and multi-modal fusion

Obvious. The '498 states that "multi-modal biometric identification may be accomplished to increase accuracy and/or security" and gives the fusion example (fingerprint + iris + face) as known. MPEP 2144.05: a claimed genus of known biometrics, where the specification lists ~20 species itself, is not a patentable distinction. Fusion of perspiration + gait is likewise presented as a known combination.

4.2 D17 (deauthorize on removal) — the cleanest § 103 kill

The '498 enumerates the entire alternative universe of removal detectors and characterizes them as known:

"Other types of device removal detection may be used, for example, including disruption in skin contact detection by way of conductivity, heat flux, galvanic skin response or motion, or periodic or continuous biometric signal detection."
"Yet other non-limiting examples of device removal detection embodiments include pulse detection, skin temperature detection, ambient temperature detection, blood flow detection, pressure detection, ambient light detection, electromagnetic field detection, respiration detection, heart rate detection, electrocardiogram detection, photoplethysmogram detection, electromyogram detection, electroencephalogram detection, near infra-red detection, skin-color detection, close magnetic contact detection, and mechanical switch detection."

Plus the electronic-continuity-detector and clasp/tamper-band examples. This is a laundry list of admitted prior art under MPEP § 2129, and D17 is nothing more than selecting one of them and wiring it to a state reset — an enumerated, finite set with predictable results. D17 is very likely obvious on the specification's own admissions alone.

4.3 D18 (secondary conditions: password, PIN, gesture, voice, tap, distance, biometric)

Obvious. Every listed condition is a known authentication "something you know / are / do." The '498 concedes the category: motion sensors "may also be utilized for simple gesture recognition to indicate user intent," and "passwords, PINS, voice commands, finger tapping, finger swiping, or other deterministic user inputs may be used as additional authentication mechanisms." Requiring a second factor before unlocking is not only known, it is industry-standard security hygiene — the motivation is supplied by the field itself (defense in depth).

4.4 D19 (permitted-access-point list from the user profile) and D20 (administrator provisions multi-user profiles)

Obvious. This is the ubiquitous ACL / role-based access control model, transplanted to the wearable. The '498 describes it in exactly those terms: "permissions may be assigned or allocated on a per user basis, or based on user groups, user roles, or other user properties." RBAC predates the priority date by decades; porting it to a newly-introduced credential type is a simple substitution (MPEP 2143(B)). The "family with children and adults" example (children barred from vehicle door/ignition) is a routine policy configuration, not a technical advance.

4.5 The "multiple access points / secondary condition" feature (from the summary, claimed in the family)

Where two access points are in range and one is selected by a secondary condition (proximity, gesture, etc.), this is arbitration among competing signals — a known problem in RF proximity systems with a known solution (nearest-first / user-intent gating). Enumerated alternatives again.

---

5. The motivation-to-combine answer, distilled

A PHOSITA at 2013-07-25, given (a) ECG-biometric identification proven since 2001 and documented in the applicant's own IDS, (b) off-device template matching conceded in the specification, (c) short-range wireless tokens that unlock readers, (d) continuous/prolonged biometric authentication known from Klosterman, and (e) a stated commercial problem of authentication friction, would have arrived at the claimed arrangement without invention. The specific KSR/MPEP rationales, per element group:

---

6. Where the § 103 case is weak — and the real crux

I would be doing you a disservice to present Ground 1 as a clean rejection. It isn't. Four points:

(i) The two-way confirmation limitation (E5). "Access based at least on an affirmative confirmation of the authorization signal by the access point" is a bilateral requirement. A reference teaching a unilateral beacon-unlocks-door scheme does not disclose or suggest it. A petitioner must find a reference teaching the access point confirming back before granting — plausible (challenge-response is old) but it must be cited, not assumed.

(ii) The delegation/persistence architecture. The real invention — and the reason the family has four issued patents — is that the wearable remains preauthorized and acts as a proxy identity arbiter while the AAD is out of range, and the downstream control signal is preferably non-biometric. The specification is emphatic: "the AAD does not need to be within wireless range of the wearable biometric device to enable the user to transmit a control signal to an access point… Further interaction between the wearable and the AAD is not required." Prior art in which the authorizing smartphone must be co-present at the door does not teach this, and the distinction is functional and non-trivial. This is where an IPR would be won or lost.

(iii) Removal-triggered reset as a security requirement, not a convenience. The specification ties deauthorization to the security model: "the security of the present invention depends on ensuring that removal of the wearable device from the user is reliably detected." A petitioner can attack D17 as an enumerated design choice, but a patentee can counter that the reset is what makes the proxy architecture safe, i.e., that the combination (proxy + removal-reset) is not merely aggregative. That is a non-obviousness argument about the combination as a whole, and it has some force.

(iv) The exact claim language matters, and I don't have claims 1 and 8 verbatim. The prior section correctly flagged that claims 1, 8, and 21 were reconstructed by inference from claim 15 and the dependents. Any § 103 conclusion stated at element level is provisional until the literal claim text is pulled from the USPTO full-text record. A single added limitation in claim 1 (e.g., a specific key-exchange ordering, a specific proximity threshold, or an express "AAD out of range" recitation) can defeat Ground 1 entirely.

Prosecution-history data point (evidence, not proof): the application was filed 2014-07-24 with a 2013-07-25 priority and granted 2015-03-31 — roughly eight months, with the family then continuing into US 9,189,901; US 9,349,235; and US 9,472,033. Fast issuance on a crowded art (ECG biometrics + wireless access control) is at least consistent with an Examiner-allowance over the art then of record (Examiner Mehmood). It is not a holding of non-obviousness and, per the earlier section, I could not retrieve the office actions to see what was actually argued or withdrawn.

Cross-reference — a contradiction to flag explicitly. The Google Patents family line in the supplied full text lists the continuation as US 9,349,235 B2, while the retrieved US 9,472,033 front page lists "9,349,325 B2." Per the no-auto-correction rule I am recording both strings and not harmonizing them. If it matters, resolve it against the USPTO record.

---

7. Secondary considerations (Graham factor 4)

Weak but not zero:

• Industry recognition / copying. The earlier search retrieved an IEEE "Influence of IEEE on Key Patents" study describing US 8,994,498 as Nymi's "key patent," cited 125 times within about five years (citation index ≈ 43.08). High forward-citation velocity is some evidence the art regarded the disclosure as significant — but citation counts are routinely criticized as evidence of crowdedness in the field (which cuts the other way, toward obviousness). I would not lead with this.
• Long-felt need. The '498's own opening (key ring overload; password fatigue; theft/copying) is a textbook articulation of a long-felt but unsolved need. A patentee can use this. A petitioner can neutralize it by showing the need was being addressed by others pre-2013 (the entire cited reference list is that evidence).
• Commercial success. Nymi's product line and the 2017 GII Tech Ventures investment (retrieved earlier) are documented, but I have no evidence of a nexus between any commercial success and the claimed architecture. Without nexus, this factor is given no weight.
• No licensing/enforcement activity found. Consistent with the earlier Litigation section: absent assertion, there is no "industry respect" signal to argue from.

---

8. Bottom line

Net: a well-constructed § 103 challenge (three references: ECG-wearable + Konetski-type biometric-auth platform + proximity-token-access-control, further in view of admitted removal detection) would put a serious prima facie case against most of the claim set, and a very strong case against the dependents. The independent claims are defensible only on the persistent-delegation/proxy limitation and the bilateral confirmation limitation — and whether those are actually in the granted independent claim language (as opposed to the specification) is exactly what I could not verify.

Confidence: High on the element-level obviousness reasoning and the admitted-prior-art passages (those quotes are verbatim from the authoritative full text). Moderate on the identity of the family's cited references (retrieved, but OCR-damaged and from the sibling's face). Low on the substantive disclosure of the individual U.S. patent references — I declined to fabricate it. Not verified at all: the '498's own verbatim front-page reference list and its verbatim claims 1, 8, and 21.

Verification steps before this is filing-grade or opinion-grade work: (1) pull the '498's actual front page from the USPTO full-text/PatentCenter imagings to get the true "References Cited"; (2) pull the literal independent claim text; (3) retrieve the prosecution history (office actions and any § 103 rejections withdrawn) via Patent Public Search / Global Dossier; (4) obtain the full text of Konetski US 8,468,362, Hively US 2008/0294097, Lange US 7,689,833, Beenau US 7,814,332, and the Gibbs trio to enable column/line citation; (5) check PTAB (IPR/PGR) — which, as the earlier section noted, I could not query and which is normally the first venue where a § 103 challenge of this kind would be aired.