US 11217237

I'll search for authoritative information on this specific patent number, including any litigation or CAFC docket activity.

Let me search for the specific PTAB proceeding and any Federal Circuit appeals involving this patent.

Let me check for reexamination proceedings and any 2026 Federal Circuit activity specific to this patent.

US Patent 11,217,237 B2 — Analyst Summary

Important scoping note up front: I found no Federal Circuit (CAFC) 2026 docket, opinion, or order that specifically names U.S. Patent No. 11,217,237. The 2026 CAFC activity I located involves other Staton Techiya / ST Case1Tech patents (e.g., 9,491,542 / 9,270,244 / 9,609,424 / 10,979,836) in appeals nos. 23‑2294, 23‑2305, 23‑2335, 23‑2387 and the voluntarily dismissed 24‑1917. I am stating this as a "not found," not as a confirmation that no such docket exists.

Also, a caveat on sources: some of the most detailed "litigation summaries" that surface for this portfolio appear in AI-generated blog posts (e.g., PatSnap Eureka pages). I have treated the patent document itself, the Google Patents bibliographic record, and the PTAB/PACER-derived dockets as authoritative, and flagged lower-reliability items.

1. Bibliographic Data (from the patent record itself)

Continuity chain (as recited in the patent's own cross-reference)

• US 16/188,683 (this patent) is a continuation of US 14/095,113, filed Dec. 3, 2013, now US 10,129,624;
• which is a continuation of US 12/102,555, filed Apr. 14, 2008, now US 8,611,560;
• which claims benefit of provisional 60/911,691, filed Apr. 13, 2007; all incorporated by reference.

Related family members: US 12,249,326 B2 (from app. 17/483,190, a continuation) and US 2024/0331691 A1 (from app. 18/397,191). A further relative — US 12,581,233 B2 (ST Case 1 Tech, LLC, priority 2007‑04‑13) — appears in the "cited by" list.

2. Abstract (verbatim substance)

The abstract states the disclosure is directed to a method and device for voice operated control with learning: measuring a first sound from a first microphone and a second sound from a second microphone; detecting a spoken voice based on an analysis of the two measurements; learning from the analysis when the user is speaking and a speaking level in noisy environments; training a decision unit from the learning to be robust to detection of the spoken voice in noisy environments; mixing the first and second sounds to produce a mixed signal; and controlling production of the mixed signal based on the learning of one or more aspects of the spoken voice and ambient sounds in the noisy environments.

3. Plain-Language Overview of the Independent Claim

Claim 1 is the sole independent claim. It is drafted as an earphone device and recites four structural elements plus three processor-performed operations:

Structure:

1. An ambient microphone that generates a first acoustic signal (measuring sound in the ambient environment);
2. An ear canal microphone that generates a second acoustic signal (measuring sound closer to the user's ear canal than the ambient microphone does);
3. A speaker configured to play an audio content signal; and
4. A processor communicatively linked to all three.

Operations the processor must perform:

• (a) Detecting a voice based on an analysis of the first and second acoustic signals, where that analysis uses at least one of: a coherence analysis, a correlation analysis, level‑detection, spectral analysis, or a combination thereof;
• (b) Mixing the first acoustic signal with the audio content signal to generate a mixed audio content signal; and
• (c) Sending that mixed audio content signal to the speaker.

What this means in plain language: The claimed earphone uses two microphones — one outside the ear (ambient) and one in/near the ear canal — to figure out when the wearer is talking, using at least one of four named acoustic-analysis techniques. When it does, the device mixes the ambient sound it picked up with the audio content being played (e.g., music or a call), and plays that mixture back through the earphone speaker. In other words, the claim is directed to a voice-triggered blend of outside-the-ear sound with program audio delivered to the wearer — a "hear yourself / hear-through" mixing function.

Notable drafting observations:

• Although the specification and abstract emphasize "learning" and a "trained decision unit," claim 1 does not recite learning, training, or adaptation at all. The independent claim is much narrower and more concrete than the summary of the invention.
• Similarly, the two-microphone "mix the ASM and ECM signals together to form a mixed signal" concept described in the Summary is recast in claim 1 as mixing the first (ambient) acoustic signal with the audio content signal — i.e., the spec's "second mixing" (mixer 406) rather than the first (mixer 404).
• The ear-canal microphone is recited functionally ("measures sound closer to a user's ear canal than the ambient microphone") rather than as an in-canal ECM per se.

Dependent claims 2–14 add, in summary: passing the first acoustic signal to a gain function (2); band-pass filtering after the gain function (3); time-domain or digital frequency-domain band-pass filtering (4); receiving a user command that is a voice, non-voice audio, and/or visual command (5); calculating a cross-correlation between the first and second acoustic signals (6); extracting the cross-correlation peak value and its lag (7); determining whether the lag matches a reference value (8); outputting a "user voice active" message on a match (9); the voice detection including an SPL-above-threshold analysis (10); analyzing the voice to detect a voice command (11); generating a filtered sound signal and presenting a notification based on it (12); increasing a gain of the first/second acoustic signal (13); and obtaining a speaking level associated with a spectral characteristic of the voice (14).

4. Litigation and Post-Grant Activity

PTAB

• IPR2022-01098 — Samsung Electronics Co., Ltd. et al. v. Staton Techiya, LLC et al., filed June 9, 2022, challenging claims 1–14 (all claims). Panel: APJs Monica S. Ullagaddi (writing), Nathan A. Engels, Scott B. Howard. Institution decision date January 3, 2023. Google Patents flags this as "Not Instituted – Merits"; the Patexia docket summary lists status as "Institution Denied." Both point to no institution, but I note the two labels are not identical, so treat the precise ground of denial with mild caution.
• Source: https://patents.google.com/patent/[US11217237B2](/patent/US11217237B2)/en and https://services.patexia.com/lawsuits/Samsung-Electronics-Co-Ltd-et-al-v-Staton-Techiya-LLC-et-al-id-[164288](/patent/164288)

District Court

• Staton Techiya, LLC v. [Samsung Electronics Co.](/litigations/by-defendant/Samsung%20Electronics%20Co.), Ltd., No. 2:22‑cv‑00053‑JRG‑RSP (E.D. Tex.) — U.S. 11,217,237 was among the patents for which Samsung served invalidity contentions and P.R. 3‑3/3‑4 disclosures on July 6, 2022, alongside U.S. Pat. Nos. 11,039,259; 11,057,701; and 11,244,666. (Exhibit list, e.g. "Exhibit M10 (Melanson)"; https://patents.justia.com/patent/[12249326](/patent/12249326))
• A related earlier case, Staton Techiya, LLC v. Samsung Electronics Co., Ltd., No. 2:21‑cv‑00413‑JRG‑RSP (E.D. Tex.), proceeded to a June 21, 2024 final judgment and subsequent appeals.

Federal Circuit (2026)

• No 2026 CAFC docket naming U.S. 11,217,237 was located. 2026 CAFC activity in this portfolio concerns other patents:
  • ST Case1Tech, LLC v. Squires, Nos. 23‑2294, 23‑2305, 23‑2335, 2023‑2387 — Feb. 18, 2026 decisions affirming PTAB obviousness holdings in part and vacating as to two claims (11 and 12 of U.S. 9,491,542) that Samsung had not challenged (Corephotonics). Sources: https://ipwatchdog.com/2026/02/19/cafc-partially-affirms-ptab-unpatentability-decision-samsung-vacates-unchallenged-claims/ ; https://scalejuris.com/news/8904/federal-circuit-affirms-most-ptab-invali
  • 24‑1917 — joint stipulation of voluntary dismissal under FRAP 42(b) covering ten asserted patents (10,405,082 / 10,966,015 / 10,979,836 / 9,124,982 / 8,315,400 / 8,111,839 / 9,491,542 / 9,270,244 / 8,254,591 / 9,609,424). U.S. 11,217,237 is not on that list.
• Because the plaintiff/patent owner settled with Samsung in the related matters and the Director intervened in some appeals, any future appellate activity on this specific patent would likely be captioned against the USPTO Director rather than Samsung.

Other

• Google Patents flags the family as having litigation (Darts‑IP family ID 80472141).
• Assignment trail on this patent: PERSONICS HOLDINGS → DM STATON FAMILY LIMITED PARTNERSHIP → STATON TECHIYA, LLC → ST PORTFOLIO HOLDINGS, LLC (June 21, 2024) → ST CASE1TECH, LLC (June 21, 2024).
• I found no ex parte reexamination control number assigned to U.S. 11,217,237 in the materials reviewed. (Portfolio reexams that appeared — e.g., 90/015,146 on 10,979,836; 90/015,830 on 8,532,286; 90/016,030 on 10,353,811 — concern different patents.)

5. Explicit Uncertainties

1. No CAFC 2026 proceeding specific to 11,217,237 was found. I cannot confirm one exists or does not exist.
2. IPR2022-01098 outcome nuance: Google Patents says "Not Instituted – Merits"; Patexia says "Institution Denied." Both indicate denial, but the precise basis (merits vs. discretionary) is not settled by the sources I retrieved.
3. Assignee naming is inconsistent across records ("St Case1tech LLC," "ST Case1tech, LLC," "ST CASE1TECH, LLC"). The assignment record indicates ST CASE1TECH, LLC as the most recent owner of record.
4. The patent's Summary and Abstract emphasize learning/trained decision units that claim 1 does not recite — anyone analyzing scope should work from the actual claim text, not the summary.
5. Several detailed third-party litigation narratives for this portfolio appear in AI-generated blog content; I did not rely on them for factual assertions where the patent record or docket-derived documents were available.

Primary sources:

• Patent full text: https://patents.google.com/patent/US11217237B2/en
• PTAB: https://portal.unifiedpatents.com/ptab/case/IPR2022-01098
• PTAB docket summary: https://services.patexia.com/lawsuits/Samsung-Electronics-Co-Ltd-et-al-v-Staton-Techiya-LLC-et-al-id-164288

I'll search for litigation involving US 11,217,237 across patent litigation sources.

Let me dig into the specific district court cases.

Now let me look for other cases asserting this patent and the Samsung case outcome.

Litigation Involving U.S. Patent No. 11,217,237

Bottom line: Based on my searches, I found one active/actual district-court patent dispute in which US 11,217,237 ("the '237 patent") was asserted, plus one PTAB post-grant challenge to the '237 patent. Both involve Samsung as the adverse party to Staton Techiya, LLC. I found no other lawsuits (no complaints against Apple, Google, LG, Huawei, etc.) in which the '237 patent is identified as an asserted patent. I was unable to complete a fully exhaustive sweep, so this should be treated as a good-faith report rather than a certified litigation-clearance search.

1. Staton Techiya, LLC & Synergy IP Corp. v. Samsung Electronics Co., Ltd. & Samsung Electronics America, Inc.

Specific '237 allegations confirmed:

• The '237 patent appears in the asserted-patents list of the Amended Complaint in 2:21-cv-00413: U.S. Pat. Nos. 8,111,839; 8,254,591; 8,315,400; 9,124,982; 9,270,244; 9,491,542; 9,609,424; 10,405,082; 10,966,015; 10,979,836; 11,039,259; 11,057,701; 11,217,237; and 11,244,666.
• In 2:22-cv-00053, Techiya alleged Samsung "directly infringed, and continue[s] to directly infringe, the '237 Patent by making, using, selling, offering for sale, or importing … products that infringe the '237 Patent including, but not limited to, the Samsung Galaxy Buds+."
• Samsung's counterclaims included Count III (declaratory judgment of non-infringement of U.S. Patent No. 11,217,237) and Count IV (declaratory judgment of invalidity of U.S. Patent No. 11,217,237). Techiya/Synergy answered those counts, and those counterclaims were ultimately dismissed without prejudice after the unenforceability ruling.

(Caveat: the docket sheet I retrieved lists the asserted patents at the pleading stage. Because several patents were narrowed during the case, I could not confirm from the retrieved materials the final set of '237 claims in the pre-trial asserted-claim list; the June 2024 judgment applied to "the asserted claims of the Asserted Patents" collectively.)

2. Samsung Electronics Co., Ltd. et al. v. Staton Techiya, LLC et al. (PTAB)

Companion petitions filed the same day against sibling patents in the same Techiya family and litigations: IPR2022-01078 ('701 patent), IPR2022-01099 ('666 patent), IPR2022-01106 ('259 patent).

This is the proceeding flagged on the face of the '237 patent (Google Patents "Family has litigation" entry: "PTAB case IPR2022-01098 filed (Not Instituted – Merits)").

3. Items I could not tie to US 11,217,237

• Darts-ip family litigation link on the patent's Google Patents page points to family ID 80472141 — that is family-wide (covering related continuations such as US 12/102,555 / US 8,611,560, US 14/095,113 / US 10,129,624, US 17/483,190, US 18/397,191), and does not itself identify any additional '237-specific suit.
• I encountered a reference to Sandpiper CDN, LLC v. Google LLC, No. 2-24-cv-03951 (C.D. Cal.), in an IPR petition that uses "the '112 patent" as shorthand, and to Multimedia Technologies Pte. Ltd. v. LG Electronics (E.D. Tex., filed Dec. 23, 2022). Neither is demonstrably a reference to US 11,217,237 and I am not attributing them to this patent.
• The '237 patent has changed hands: Staton Techiya, LLC → ST Portfolio Holdings, LLC (recorded June 21, 2024) → ST Case1Tech, LLC (recorded June 21, 2024; listed as current assignee). Any post-June-2024 assertions by ST Case1Tech would need a fresh docket check; I could not confirm any such cases with the searches available.

Verification guidance / limitations

My searches were run against patent-litigation aggregators (Unified Patents PTAB data, Docket Alarm, CourtListener/RECAP, Patexia, Justia, UniCourt) and general web sources. The searches were capped before I could sweep secondary dockets and the Federal Circuit docket exhaustively, so please treat this as non-exhaustive. To confirm:

1. Unified Patents litigation/PTAB portals — https://portal.unifiedpatents.com/litigation/caselist and https://portal.unifiedpatents.com/ptab/ (case IPR2022-01098).
2. USPTO Patent Center / Assignment records for US 11,217,237 (to see whether ST Case1Tech, LLC has recorded any new assignments).
3. PACER for E.D. Tex. 2:21-cv-00413-JRG-RSP and 2:22-cv-00053-JRG-RSP (final judgment at Dkt. 950; findings at Dkt. 923), and the CAFC docket for any appeal of the unenforceability judgment (the CAFC decision I retrieved, No. 23-2305, concerns the '839 and '982 patents, not the '237 patent).

I do not state as a fact that no other litigation exists — only that none was found in the sources searched.

I'll start by researching the PTAB proceeding referenced in the patent's structured data (IPR2022-01098) and any others.

Let me get more detail on IPR2022-01098 and check for any other IPRs against the '237 patent.

I have the core facts. Now let me find the substance of the institution denial, the '237 patent's district court fallout, and confirm no second IPR exists.

Proceedings overview

One AIA trial proceeding has ever been filed against US 11,217,237 — and it never made it past the doorstep. Breakdown: 0 active, 0 claims invalidated, 0 claims sustained on the merits, 0 settled, 1 institution denied. Petitioner Samsung Electronics filed IPR2022-01098 on 2022-06-09 against all claims 1–14; the Board denied institution on the merits on 2023-01-03. Bottom line for a defendant today: no claim of the ‘237 patent has ever been canceled or even tested in an instituted trial — the entire claim set, including independent claim 1, is intact and UNTESTED at the PTAB. That is a materially worse posture for an IPR-based defense than the patent's sibling patents (several Techiya patents were invalidated), because the one petitioner who tried walked away with a denial and no § 315(e) estoppel — but also left a public roadmap of the art Samsung thought was closest, and the denial carries no preclusive weight (it isn't even appealable under § 314(d)).

Caveat on sourcing: the structured block in this prompt says the USPTO ODP API returns no AIA proceedings; that is incorrect as to IPR2022-01098, which is flagged on the patent's own Google Patents family-litigation record ("PTAB case IPR2022-01098 filed (Not Instituted – Merits)"). I could not retrieve the full text of the 2023-01-03 denial through search, so the claim-by-claim reasoning below is limited to what the record confirms.

IPR2022-01098 — Samsung Electronics Co., Ltd. and Samsung Electronics America, Inc. v. Staton Techiya, LLC

• Type: Inter Partes Review (35 U.S.C. §§ 311–319), filed against U.S. Patent No. 11,217,237
• Filed: 2022-06-09 (Petition, Paper 4; Ex. 1001 = the '237 patent; Ex. 1002 = Declaration of Nathaniel Polish, Ph.D.; Ex. 1004 = file history of the '237 patent)
• Status: "Not Instituted – Merits" (Google Patents family-litigation record) / "Institution Denied" (docket aggregator). Plain English: the Board denied institution under 35 U.S.C. § 314(a) after finding Petitioner had not established a reasonable likelihood of prevailing — i.e., a merits denial, not a discretionary/Fintiv denial. Consistent with that, Petitioner moved to refund the post-institution fee (requests filed 2023-01-18 and 2023-03-02; refund approved 2023-03-07, Paper 14).
• Judge panel: Monica S. Ullagaddi (writing), Nathan A. Engels, Scott B. Howard
• Petition grounds: Challenged claims 1–14 (all claims). Per the district-court record of the proceeding, six obviousness grounds under § 103:
  • (1A) claims 1, 2, 5, 10–14 over Platz (US 2010/0119077 A1), Kvaløy (US 6,728,385), and a third reference;
  • (1B) claims 3, 4 over Platz, Kvaløy, and Hotvet (US 5,550,923);
  • (1C) claims 6–9 over Platz, Kvaløy, and Kondo (JP 3353701);
  • (2A) claims 1, 5, 10–12, 14 over Dijkstra (EP 1 640 972 A1), Kvaløy, and Melanson (US 7,903,825);
  • (2B) claims 2–4, 13 over Dijkstra, Kvaløy, Melanson, and Hotvet;
  • (2C) claims 6–9 over Dijkstra, Kvaløy, Melanson, and Kondo.
  • No § 112 ground was pleaded in the IPR. (Petitioner's closest-document set is public via the exhibit list: Platz, Dijkstra, Kvaløy (US6728385), Melanson, Hotvet, Kondo, and Duffner 2006 as Ex. 1019.)
• Institution decision: Denied — 2023-01-03. The Board issued a "Decision Denying Institution of Inter Partes Review 35 U.S.C. § 314," cited in the family record as Exhibit 11 of the proceeding. The Patent Owner's Preliminary Response (2022-10-07) was supported by the Declaration of Christopher J. Struck (Ex. 2001) and the Wiley Electrical and Electronics Engineering Dictionary (Ex. 2009) — i.e., Patent Owner attacked the petition's claim-construction of the "analysis"/processing terms. I was unable to retrieve the panel's verbatim merits reasoning; I will not characterize which specific ground or limitation carried the day. What is reliably established: the denial was merits-based, and the case terminated without a trial.
• Procedural events of note: Patent Owner's Mandatory Notice (2022-06-22); filing-date-accorded notice (2022-07-07); Patent Owner Preliminary Response (2022-10-07); Petitioner Preliminary Reply (2022-11-15); Patent Owner Preliminary Sur-Reply (2022-11-22); denial (2023-01-03). Petitioner also filed the 2022-11-10 stipulation letter (Ex. 1030) agreeing that if trial were instituted, Samsung would not in the parallel E.D. Tex. case pursue invalidity of the '237 claims on the instituted grounds or on any ground using Platz or Dijkstra.
• Final Written Decision: None. No FWD issued. No claim of the '237 patent has ever been adjudicated unpatentable.
• Settlement / termination: No IPR settlement — the proceeding ended by denial of institution, not by the parties' agreement. (The underlying district-court litigation has a separate history involving consolidation and, separately, a stay entered 2024-10-30 in 2:23-cv-00319-JRG-RSP; I could not confirm a global '237-specific disposition from the sources retrieved.)
• Appeal: None, and none is available. A denial of institution is not appealable (35 U.S.C. § 314(d); Cuozzo).
• Defensive value: Because institution was denied, 35 U.S.C. § 315(e)(2) estoppel never attached — Samsung is free to run Platz/Kvaløy/Dijkstra/Melanson obviousness combinations in district court, and the conditional stipulation lapsed when the condition (institution) failed. Conversely, that means you get no free preclusion either: there is no FWD to lean on, and the Board's denial has no issue-preclusive or precedential effect against Patent Owner.

Strategic summary

Claim status on US 11,217,237. Claims 1–14 are all SUSTAINED-by-default (i.e., never canceled) and UNTESTED. The only independent claim is claim 1 (earphone device: ambient microphone generating a first acoustic signal; ear canal microphone generating a second acoustic signal; speaker playing an audio content signal; processor configured to (a) detect a voice based on analysis of the first and second acoustic signals using at least one of coherence analysis, correlation analysis, level-detection, spectral analysis, or a combination, (b) mix the first acoustic signal with the audio content signal to generate a mixed audio content signal, and (c) send the mixed audio content signal to the speaker). Claims 2–14 are dependent. Nothing was canceled; nothing was held patentable in a FWD. Anyone advising a client that "claims 1–5 of the '237 patent are dead" would be flatly wrong — that is a different patent in this family.

Note the important asymmetry in this family: sibling patents that were instituted against fared badly. In IPR2022-00234 the Board held claims 1–6, 14, 15, 17–20, 24 and 25 of US 9,124,982 unpatentable (sustaining only claims 10 and 11); in IPR2022-00281 it invalidated claims 1–4, 6, 13, 14, 17–19 and 25–27 of US 9,270,244; in IPR2022-00243 it invalidated claims 1–4 of US 8,111,839; in IPR2022-00253 it invalidated claims 1–9 and 11–18 of US 9,491,542 (sustaining claims 10, 19, 20). Staton Techiya appealed, and the Federal Circuit affirmed in a nonprecedential consolidated disposition (No. 23-2305, 2025-10-09, https://www.cafc.uscourts.gov/opinions-orders/23-2305.OPINION.10-9-2025_2585626.pdf; companion No. 23-2294). So the family's patents have repeatedly been narrowed — but not this one. The '237 patent's claims are the closest thing in the family to a hardened asset, precisely because a merits denial means the Board was unpersuaded by the petition as filed.

Estoppel landscape. Clean slate for you. § 315(e)(2) estoppel requires a final written decision, and IPR2022-01098 produced none — so neither Samsung nor its privies are barred from raising Platz-, Dijkstra-, Kvaløy-, Melanson-, Hotvet- or Kondo-based invalidity anywhere. Two practical consequences: (1) you are not the beneficiary of any estoppel, and (2) the prior art Samsung already tested and failed on is, statistically, less likely to work for you — but there is no legal rule preventing it, and the denial's reasoning may have turned on a specific claim-construction position you could reframe. The art most obviously still worth running, drawn from the '237 patent's own prosecution and family: Ryan (US 2004/0196992), Svean (US 6,567,524), Padhi (US 2005/0096899), Baskerville (US 2006/0188105), and Duffner 2006 — the same core set the Examiner used against the sibling '836 patent (see the Ex Parte Reexamination 90/015,146 appeal record, Control No. 90/015,146, Staton Techiya Appeal Brief dated 2023-11-21). No reexamination of the '237 patent itself has been confirmed on the record I retrieved.

Pattern signals. One petitioner, one petition, one patent: Samsung is the only entity to have challenged the '237 patent, and it challenged every claim in a single petition — no follow-on/IPR-2 was filed after the denial. Unified Patents is not in this chain as a petitioner; the "Unified Patents PTAB Data" string on the Google Patents record is a data-source attribution, not a real party-in-interest. Patent Owner (now ST Case1Tech, LLC as successor to Staton Techiya/ST Portfolio Holdings) has historically appealed adverse decisions aggressively — it appealed the '982, '839, '542, '244 and '281 decisions to the Federal Circuit, which affirmed in 2025 — but it had nothing to appeal here, because it won. The overall picture is a well-funded NPE asserting a 14-patent, year-2007-priority family against Samsung's Galaxy Buds/phone ecosystem, with mixed but generally favorable-for-defendants PTAB outcomes across the family and a conspicuous gap at the '237 patent.

Recommended next steps

1. Do not represent to anyone that the '237 claims have been canceled. They have not. Any § 282 invalidity case must be built from scratch. There is no FWD to link to; the citable event is the 2023-01-03 Decision Denying Institution in IPR2022-01098 (PTAB E2E, Case IPR2022-01098; aggregator docket: https://services.patexia.com/lawsuits/Samsung-Electronics-Co-Ltd-et-al-v-Staton-Techiya-LLC-et-al-id-[164288](/patent/164288)). Order the full text of that decision before finalizing any invalidity contentions — the panel's actual reasoning (likely claim construction of the processor/analysis limitations, per Patent Owner's Struck declaration and dictionary exhibit) will tell you which reframing avoids the same fate.
2. Statute of limitations first. If you have been served with a complaint asserting the '237 patent, your § 315(b) one-year window runs from service and is jurisdictional. Given that Samsung's petition died on the merits rather than on discretion, a well-targeted petition from a different art set remains viable — but expect General Plastic to be raised if you are a privy of Samsung.
3. No trial-stage milestones exist to track. IPR2022-01098 is closed; there is no institution deadline, no oral hearing, no statutory one-year FWD clock running. The estate is dormant at the PTAB — which, as noted, is itself a signal: a patent this widely asserted against consumer earbuds that attracted only one, denied petition has not yet been seriously stress-tested.
4. Watch the family, not just this patent. The '237 patent is a continuation of US 8,611,560 and US 10,129,624, with live continuations US 12,249,326 (filed 2021-09-23) and US 2024/0331691A1 (filed 2023-12-27). Patent Owner's pattern in this family is to keep prosecution going while litigating; new claims issuing from those continuations can reset the invalidity fight. Also monitor the E.D. Tex. dockets (2:21-cv-00413-JRG-RSP, consolidated lead; 2:22-cv-00053-JRG-RSP) for how the '237 patent specifically resolves.

Confidence notes: Petitioner/patent-owner identities, filing date, claim scope challenged, ground-by-ground art, panel, the 2023-01-03 merits denial, and the refund sequence are well supported by the retrieved record and I state them without qualification. The verbatim reasoning of the denial and any '237-specific settlement or district-court disposition are not established on the record I could retrieve; I have not guessed at them.

I'll try to pull the actual USPTO assignment records (reel/frame, correspondents) for this patent.

Let me dig for the specific reel/frame records and the 2024 ST transfers.

Assignment Record Reconstruction — US 11,217,237 B2

Data-integrity note before the sections: I could not retrieve the USPTO Patent Assignment Center's raw reel/frame records for this patent through the tools available in this session. The USPTO Assignment Search interface (https://assignmentcenter.uspto.gov/) is not directly queryable here, and the underlying records are only surfaced indirectly through family file histories (37 CFR 3.73(c) chain-of-title statements filed in sibling applications), PTAB exhibits, and Google Patents' legal-events table.

Important discrepancy to flag against the previously-generated sections: Google Patents' legal-events table for US 11,217,237 lists eight "Assigned to" events all dated 2018-11-13 (the filing date of app. 16/188,683). That date is an artifact of Google's event-dating — it is not the execution or recordation date of the underlying assignments. The family file histories show the Personics→DM Staton and DM Staton→Staton Techiya recordations were actually made in June 2017 (reels 042992 and 043393). Treat the 2018-11-13 date stamps as unreliable. Everything below uses the file-history dates where I could ground them.

Inventors

Unusual patterns:

• Two of the three inventors (Goldstein, Boillot) are principals of the entities in the ownership chain, not arms-length employees. Goldstein is the namesake of the "DM Staton" holding vehicle and the person identified in the Techiya trial record as the controlling principal; Boillot is tied to Navisense. This is not a "fire-sale after inventors depart" pattern — it is the opposite: the inventors are the ownership chain.
• The patent's inventors were not the ones asserting it. By the time of suit, the named inventors had no record title; ownership sat with entity vehicles controlled by Goldstein/Keady.
• No evidence found that any inventor assigned away and re-acquired, or that inventors departed a corporate assignee within 12 months of filing. Not present.

Original assignee

• On the face of the issued patent (Google Patents "Original Assignee"): Staton Techiya, LLC. This field is misleading for a 2018 continuation — it reflects the assignee at issue, not at filing.
• Assignee at filing (2018-11-13): the application was filed into the Personics Holdings, Inc. / Personics Holdings, LLC chain (assignment events naming Boillot, Usher, and Goldstein as assignors to Personics Holdings entities).
• Ultimate origin of the priority application (12/102,555, filed 2008-04-14): Personics Holdings, Inc. (The sibling '560 patent's Google record lists "Navisense" as the assignee at issue — a related/former entity name in the Boillot orbit; I could not confirm the exact corporate relationship between Navisense and Personics Holdings from the sources retrieved. Flag as unclear.)

Business and status:

• Personics Holdings, Inc./LLC — an audio/earpiece technology developer (in-ear monitors, personalized hearing, ear-canal sensing). It was a small operating R&D company, not a manufacturer at scale; I found no evidence it shipped a commercial product embodying claim 1 of the '237 patent. Its patent portfolio was consolidated and transferred out in 2017.
• Current status: the operating entity is effectively wound down as an IP proprietor; its assets now sit in the ST family vehicles. I found no bankruptcy filing for Personics Holdings or DM Staton. Not present as a bankruptcy event.

Assignment timeline

Because the Assignment Center records were not directly retrievable, each entry below states the evidence basis. Reel/frame is given only where I could ground it in a retrieved document; otherwise marked "not retrieved."

1. 2008 (executed) / recorded 2008 — Reel not retrieved

• Conveyance: Assignment of Assignors' Interest
• Assignor: John Usher; Steven Goldstein; Marc Boillot (three separate instruments, per Google Patents legal events)
• Assignee: Personics Holdings, Inc. (with a joint Personics Holdings, Inc./Personics Holdings, LLC variant for the Usher instrument)
• Correspondent: not retrieved
• Context: Initial inventor-to-company assignment at original filing.

2. c. 2014–2017 — Reel 034170 / 0933 (cited in the '555,570 chain-of-title statement)

• Conveyance: Assignment
• Assignor/Assignee: Personics Holdings-related entities (the record spans a 10/17/2014 → 06/20/2017 window and appears to be a consolidated/corrective recordation of the Personics Holdings, Inc. → Personics Holdings, LLC step)
• Correspondent: not retrieved for this instrument
• Context: Internal reorganization of the Personics Holdings entities. Note: this reel is grounded in a sibling application's file history, not confirmed on the '237-specific record.

3. 2017-06-20/21 (recorded) — Reel 042992 / 0493 (and a companion frame 042992 / 0524)

• Conveyance: Assignment
• Assignor: Personics Holdings, Inc. and/or Personics Holdings, LLC
• Assignee: DM Staton Family Limited Partnership
• Correspondent: Peter A. Chiabotti (Akerman LLP) — see below
• Context: Transfer-to-holding-vehicle. Personics Holdings' IP was swept into a Wyoming family limited partnership (later identified as "assignee of Staton Family Investments, Ltd."). This is the first step out of the operating company.

4. 2017-06-21 (recorded) — Reel 043393 / 0001

• Conveyance: Assignment
• Assignor: DM Staton Family Limited Partnership ("assignee of Staton Family Investments, Ltd.," a Wyoming partnership, address 6800 Broken Sound Parkway, Suite 200, Boca Raton, FL 33487)
• Assignee: Staton Techiya, LLC (a Delaware limited liability company; address 16193 Bridlewood Circle, Delray Beach, FL 33445)
• Correspondent: Peter A. Chiabotti, Akerman LLP (Customer No. 150139) — the 37 CFR 3.73(c) chain-of-title statement for this chain was filed by Chiabotti on 2017-12-04
• Context: Transfer-to-asserter. This is the instrument whose text was produced in litigation (E.D. Tex. 2:21-cv-00413, Dkt. 469-3, "Intellectual Property Assignment Agreement"): DM Staton "irrevocably conveys, transfers, and assigns" all IP "previously owned by Personics Holdings, LLC and/or Personics Holdings, Inc." to Staton Techiya — expressly including "all claims and causes of action … for past, present, and future infringement." This is the standing-establishment instrument for the assertion campaign.

5. 2024-06-21 (executed) / recorded on or about 2024-06-21 — Reel not retrieved

• Conveyance: Assignment
• Assignor: Staton Techiya, LLC
• Assignee: ST Portfolio Holdings, LLC
• Correspondent: not retrieved
• Context: Internal reorg / asset isolation within the Staton family of entities.

6. 2024-06-21 (executed) / recorded on or about 2024-06-21 — Reel not retrieved

• Conveyance: Assignment
• Assignor: ST Portfolio Holdings, LLC
• Assignee: ST Case1Tech, LLC (current assignee of record per Google Patents; most recent owner)
• Correspondent: not retrieved. (Litigation counsel of record in the resulting appeals is Timothy Devlin, Devlin Law Firm LLC, Wilmington, DE — but that is litigation counsel, not a recorded assignment correspondent, and should not be conflated. ST Case1Tech was substituted as appellant in the CAFC appeals on 2025-03-03.)
• Context: Second-hop internal reorg, executed the same day as entry 5. Effect: the patent moved from the asserting LLC (Staton Techiya) into a fresh vehicle (Case1Tech) after the adverse June 2024 unenforceability judgment, ahead of the appeals.

Correspondent of record — the key tell:

• Peter A. Chiabotti, Reg. No. 54,603, Akerman LLP, 777 South Flagler Drive, Suite 1100, West Tower, West Palm Beach, FL 33401 — USPTO Customer No. 150139. Chiabotti filed the family's 37 CFR 3.73(c) chain-of-title statements (2017-12-04) and the correspondence-address change to Customer No. 150139 (2020-12-12). He is the repeat-player attorney running the recordation side of this family — the same correspondent recurs across the Personics/Staton Techiya portfolio (e.g., '079, '555,570, 16/736,820).
• John P. Keady, Reg. No. 56,389, listed at Staton Techiya, LLC, 16193 Bridlewood Circle, Delray Beach, FL 33445 — USPTO Customer No. 154598. Keady is the patent owner's own named representative (he is himself a co-inventor on sibling family patents, e.g., US 8,600,067 and US 9,609,424, and he testified as a Techiya principal in the E.D. Tex. trial). The owner's principal doubling as the correspondence address is a classic single-purpose-NPE footprint — there is no outside, independent license/portfolio manager.

Timeline diagram

timeline
    title Ownership of US 11217237
    2007 : Provisional 60911691 filed
    2008 : Non-provisional 12102555 filed
         : Inventors assign to Personics Holdings
    2013 : Continuation 14095113 filed
    2017 : Personics transfers to DM Staton LP
         : DM Staton transfers to Staton Techiya
    2018 : Continuation 16188683 filed
    2022 : Patent issues on Jan 4
         : Asserted against Samsung
         : Samsung IPR filed
    2024 : Staton Techiya transfers to ST Portfolio
         : ST Portfolio transfers to ST Case1Tech

NPE / troll-pattern signals

1. Shell-entity transfer — PRESENT.
The patent left the operating entity (Personics Holdings) and passed through two holding vehicles before landing in a dedicated assertion LLC. Grounding:

• Reel 042992 and Reel 043393 / 0001 (recorded June 2017): Personics → DM Staton Family Limited Partnership (a Wyoming family LP — a state chosen for opaque beneficial ownership) → Staton Techiya, LLC (Delaware single-purpose LLC).
• Applicant/owner address for Staton Techiya is 16193 Bridlewood Circle, Delray Beach, FL 33445 — a residential address, recorded as Customer No. 154598.
• 2024 transfers to ST Portfolio Holdings, LLC → ST Case1Tech, LLC — two more single-purpose LLCs, no products in commerce.
• No commercial product embodying claim 1 was identified for any of these entities.

2. Known asserter in the chain — PRESENT.
Staton Techiya, LLC is a well-known high-frequency patent plaintiff (sued Samsung in E.D. Tex. 2:21-cv-00413 and 2:22-cv-00053, and is the adverse party across the IPR2022-01098 and multiple companion IPRs). ST Case1Tech, LLC is the current owner and the appellant in the Federal Circuit ST Case1Tech v. Squires appeals (Nos. 23-2294, 23-2305, 23-2335, 23-2387). These are asserters by conduct, not by name. (They are a Staton-family in-house assertion vehicle, not one of the classic third-party NPEs on the Acacia/Marathon/IV list — so this signal is grounded in litigation conduct, which is the stronger and more reliable evidence.)

3. Repeat correspondent across the chain — PRESENT.

• Peter A. Chiabotti (Reg. 54,603), Akerman LLP / Customer No. 150139 — appears as the filer/correspondent on the family's 2017 chain-of-title recordations and the 2020 address change; the same firm/customer number services the sibling Personics/Staton Techiya filings. Recurrence, not a single appearance → this is a finding.
• John P. Keady (Reg. 56,389), Customer No. 154598, Staton Techiya, LLC — the owner entity's own principal is the recorded correspondence/fee address across the family. This is the single most telling recurrent-correspondent fact: the LLC names change, the insiders running them do not.

4. Cascading transfers — PRESENT.
Two back-to-back hops were executed on the same/near-same day, twice: (a) Personics → DM Staton → Staton Techiya, recorded 2017-06-20 and 2017-06-21 (reels 042992 and 043393/0001); and (b) Staton Techiya → ST Portfolio Holdings → ST Case1Tech, both dated 2024-06-21. Two consecutive transfers inside 24 hours, twice, through entities sharing insiders and correspondents.

5. Pre-litigation transfer — NOT PRESENT (strictly).
The 2022 assertion was filed by Staton Techiya (already the owner since 2017), and the 2024 re-transfer to ST Case1Tech followed the June 2024 judgment rather than preceding the first suit. There is no assignment dated within 6 months before the first infringement suit naming this patent. (The 2024 transfer is litigation-motivated — it teed up the CAFC appellant substitution — but it post-dates the suit, so it does not satisfy this specific signal.)

6. Bankruptcy fire-sale — NOT PRESENT.
No Chapter 7/11 filing by Personics Holdings, DM Staton, or any chain entity was found. The Personics→DM Staton transfer (2017) was a negotiated family-entity consolidation, not a court-supervised sale.

7. Privateering — UNCLEAR / QUALIFIED PRESENT.
This case has a privateering-adjacent fact pattern but not the classic form. The classic form is operating company → NPE proxy. Here the chain runs operating company → the inventors' own family LP → own LLC. However, the E.D. Tex. trial record (findings of fact, Dkt. 928) documents that Techiya engaged Synergy IP Corp. as a licensing/litigation agent with a license agreement under which Synergy negotiated with Samsung and Techiya "instructed Synergy to file suit." That is functionally an outsourced-assertion arrangement, and it was the basis for the unenforceability/unclean-hands judgment of May 9, 2024. I therefore mark this present in substance (outsourced assertion), while noting the ownership stayed inside the Staton family.

8. Defensive aggregator — NOT PRESENT.
The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. It terminates at an active assertion LLC (ST Case1Tech) that is currently litigating appeals. Inverse of the neutralization signal.

Verdict

NPE — high confidence.

Justification: (1) the patent moved out of the operating company Personics Holdings through the Wyoming vehicle DM Staton Family Limited Partnership and into Staton Techiya, LLC by recordations at Reel 042992 and Reel 043393/0001 (June 2017), whose text expressly conveyed "all claims and causes of action … for past, present, and future infringement" — a purpose-built standing instrument; (2) the chain then continued on 2024-06-21 through ST Portfolio Holdings, LLC to ST Case1Tech, LLC, single-purpose LLCs with a residential correspondence address and no products; and (3) the same repeat correspondents — Peter A. Chiabotti (Reg. 54,603, Akerman LLP, Customer No. 150139) and the owner's own principal John P. Keady (Reg. 56,389, Customer No. 154598) — recur across the recordations, evidencing a continuously controlled assertion vehicle rather than independent arms-length purchasers. Staton Techiya's actual litigation conduct (E.D. Tex. 2:21-cv-00413 / 2:22-cv-00053 and IPR2022-01098) confirms the assertion posture.

Verification links:

• USPTO Assignment Center (search "11217237"): https://assignmentcenter.uspto.gov/
• USPTO Patent Assignment Search (mirror): https://assignment.uspto.gov/patent/index.html
• Google Patents legal events: https://patents.google.com/patent/[US11217237B2](/patent/US11217237B2)/en
• Trial findings (agency/Synergy relationship and unclean hands): https://storage.courtlistener.com/recap/gov.uscourts.txed.[210767](/patent/210767)/gov.uscourts.txed.210767.928.1.pdf

Items I did not verify (do not treat as confirmed): the reel/frame numbers for the 2008 inventor→Personics Holdings assignments and for the 2024 ST Portfolio/ST Case1Tech transfers; the correspondents of record on the 2024 instruments; and whether any post-June-2024 assignment has been recorded out of ST Case1Tech.

I'll research the prior art citations for US 11,217,237. Let me start by verifying the patent on USPTO and pulling details on the most relevant cited references.

Let me pull details on additional key cited references to assess anticipation potential.

Prior-Art Analysis — U.S. Patent No. 11,217,237 B2

"Method and device for voice operated control"

0. Scope, method, and two caveats up front

Method. I searched for the exact number 11217237 and confirmed the record: US 11,217,237 B2, App. No. 16/188,683, filed 2018-11-13, granted 2022-01-04, assignee chain ending at ST Case1Tech, LLC. The results returned the correct patent (not a similar number). Because the tool budget for live USPTO PatentCenter/PAIR fetches was exhausted mid-analysis, the reference lists below are drawn from (a) the authoritative full text you supplied (which reproduces the front-page "Citations (64)" list) and (b) the Google Patents bibliographic mirror of the USPTO record. No result for a similar number (e.g., 11,217,227, 11,271,237) was substituted.

Caveat 1 — date discrepancy. The task states "Current Date: April 26, 2026," whereas the system clock reads 2026-09-29. I flag this rather than silently reconcile it; it does not change the prior-art analysis, because all references below predate the 2026 window by years.

Caveat 2 — building on the earlier sections. The previously generated "Patent summary" and "Litigation summary" are treated as authoritative. One point there is critical here and I will not repeat at length: claim 1 is the sole independent claim, and it does NOT recite "learning" or a "trained decision unit" even though the Abstract/Summary emphasize them. That matters because § 102 is decided on claim language, not on the specification's aspirational summary.

1. The § 102 legal frame (why most of these are not anticipatory)

Under 35 U.S.C. § 102, a claim is anticipated only if a single prior-art reference discloses every element of the claim, arranged as recited. Reference lists on a patent's face are almost never anticipatory; they are overwhelmingly § 103 obviousness material. I therefore grade each reference below on whether it could alone meet the full claim, and where it cannot, I say so explicitly.

Claim 1 element checklist (the anticipation target)

Element 1f–1g is the crux: the claim requires a hear-through blend in which the ambient/first signal is mixed with program audio content and played back. Many two-microphone voice-detection references fail on 1f.

2. Tier 1 — Strongest single-reference § 102 candidates

2.1 — US 2005/0058313 A1 (Victorian et al.) — the most dangerous reference

2.2 — US 6,754,359 B1 (Svean et al., Nacre AS)

3. Tier 2 — References that map onto sub-features / dependent claims, not claim 1 as a whole

Other "third-party-cited" references of note

The remaining entries in the 64-item list (e.g., US 4,237,343 Kurtin; US 5,852,804 Fujitsu; US 6,094,489 NEC; US 8,398,844; US 8,401,206 Microsoft; US 8,583,428 Microsoft; US 8,600,454 Apple; US 8,606,571 Audience; US 8,403,206; US 8,506,704 Cochlear; US 8,092,844; WO 2012/078670 UT System) are general DSP/beamforming/hearing-aid art. None discloses the ambient-mic + ear-canal-mic + audio-content-mix-to-speaker combination of claim 1 as a single reference; they are § 103 or § 102-background only.

4. References that are NOT prior art to this patent (family members)

Because these share the same priority chain and inventive entity, they cannot be § 102 art against the '237 patent. I list them so they are not miscounted as "prior art":

• US 8,611,560 B2 (from 12/102,555) and US 10,129,624 B2 (from 14/095,113) — the direct parent applications recited in the '237 cross-reference.
• US 8,625,819 B2 (from 12/169,386; priority 2007-04-13; granted 2014-01-07) — this is a same-family sibling (a continuation-in-part of 12/102,555) with the Same inventors (Goldstein, Usher, Boillot). Its disclosure — ASM + ECM + ECR + processor + VOX that mixes ambient and internal sound and performs a second mixing of the audio content with the mixed signal delivered to the ECR, and that recites "a level detector… a correlation unit… a coherence unit… a spectral analysis unit" — is the closest textual match to claim 1 anywhere. But it is the patent's own family, so it is not § 102 art. (It would only matter if a court found the '237 claims not entitled to the 12/102,555 priority — a separate validity theory, not § 102 on its face.)
• US 2014/0095157 A1 / US 2014/0126748 A1 (parent publication of 14/095,113) — family.
• Personics-family publications US 2008/0137873 A1 ("personalized hearing"), US 2008/0181419 A1 ("acute sound detection"), US 2008/0187163 A1 ("audio recording"), and US 2009/0010444 A1 ("personalized voice operated control") — common-assignment/related disclosures; treat with care before asserting them as § 102 art.

5. References relied on in the inter partes challenge (not on the patent face)

Per the earlier Litigation summary: IPR2022-01098 (Samsung v. Staton Techiya, filed 2022-06-09, claims 1–14, institution denied 2023-01-03) was stipulated to rely, if instituted, on US 2010/0119077 A1 (Platz) and EP 1 640 972 A1 (Dijkstra) as primary references. These do not appear in the '237 front-page "Citations (64)" list, so they are petitioner-selected art rather than examiner-cited art. Because institution was denied, no claim was adjudicated. Anyone re-charting the '237 patent should pull these two references directly, since they were the challenger's chosen lead references.

6. Bottom-line assessment

1. Claim 1's real vulnerability is § 102/§ 103 over Victorian (US 2005/0058313 A1), and secondarily § 103 over Victorian in view of Light's "Ambient-aware headset" (US 2003/0035551 A1) or Nacre (US 6,754,359 B1). Victorian supplies the ambient mic + ear-canal mic + speaker + processor + VOX detection and an "and/or" delivery of external-mic and remote-audio signals to the ear-canal speaker — the strongest single-reference attack on elements 1a–1g.
2. The dependent claims most exposed to outright § 102 are: claim 10 (SPL/threshold — Victorian), claims 2–4 (gain + band-pass filtering — Nacre), and claim 14 (speaking level tied to a spectral characteristic — Nacre's vowel/nasal/fricative classification).
3. Claims 5–9 (user-command modalities; cross-correlation peak/lag; reference-lag match; "user voice active" output) are the hardest to anticipate from any single reference in the list and would require a § 103 combination (e.g., correlation-based VAD art + Victorian). Note these claims trace to the "learn mode" described at col. of FIG. 7 of the specification, which is not recited in the independent claim but is recited in claims 6–9.
4. No reference on the face of the '237 patent other than Victorian appears to disclose all of claim 1 in a single document. I did not find a "smoking gun" § 102 reference.

7. Explicit uncertainties / verification steps

• I could not complete live USPTO PatentCenter/PAIR and full-text retrieval for US 7,853,031 B2 (Siemens own-voice detection) and US 8,467,543 B2 (Aliphcom) before the tool budget was exhausted; those two descriptions are drawn from title/subject-matter and the patent's own citation listing, and are flagged as unverified in detail.
• The effective priority date materially affects which post-2007 references are § 102(a)/(e) art. I used 2007-04-13 (provisional 60/911,691) / 2008-04-14 (12/102,555); if the issued claim 1 is later held not entitled to that chain, more 2007–2008 publications become available as art.
• "Potentially anticipates" is used per your instruction; it is not a legal conclusion. A definitive § 102 rejection also requires the reference to qualify as prior art and to disclose the limitation as arranged in the claim.
• Recommended next steps: (a) pull the USPTO file wrapper (App. 16/188,683) for the examiner's actual § 102/§ 103 rejections and the Form PTO/SB-08 citations; (b) retrieve US 2010/0119077 (Platz) and EP 1 640 972 (Dijkstra) as the IPR petitioner's intended primary references; (c) obtain the full text of US 7,853,031 B2 and US 8,467,543 B2 to finalize their § 102/§ 103 placement.

Primary sources used: authoritative full text of US 11,217,237 B2 (Google Patents "Citations (64)" list); https://patents.google.com/patent/[US11217237B2](/patent/US11217237B2)/en ; https://patents.google.com/patent/US20050058313 ; https://patents.google.com/patent/[US6754359B1](/patent/US6754359B1) ; https://patents.google.com/patent/US20070291953 ; https://patents.google.com/patent/US8625819 ; IPR2022-01098 record as summarized in the prior section.

I'll ground the obviousness analysis by verifying the key prior-art references identified in the record before building combinations.

Let me verify the remaining references in the prosecution/IPR art set and check critical dates.

Obviousness Analysis — U.S. Patent No. 11,217,237 B2

Under 35 U.S.C. § 103

Builds on the prior sections. The earlier sections established the claim set (one independent claim, 14 total), the family chain, the unenforceability judgment against Samsung in E.D. Tex., and the merits denial of IPR2022-01098. This section assumes that work and does not repeat it. Where the record yields a fact that bears on obviousness — particularly the identity of the references Samsung actually pleaded — I use it.

Sourcing note. The two references whose content I directly retrieved and verified are Platz (US 2010/0119077 A1) and Melanson (US 7,903,825 B1); I also verified Padhi (US 2005/0096899 A1), Ryan (US 2004/0196992 A1), Kvaløy (US 6,728,385) (subject matter only), and the prosecution/reexam uses of Ryan/Svean/Padhi/Baskerville. Where I describe a reference's disclosure from the ground listing rather than from its text (Hotvet US 5,550,923; Kondo JP 3353701; Dijkstra EP 1 640 972 A1; Duffner 2006), I say so.

1. Threshold: the § 103 critical date, and why it matters more than usual here

Claim 1 of the '237 patent is a continuation claim. On the face of the patent (cross-reference paragraph), it is a continuation of 14/095,113 (filed 2013-12-03, now US 10,129,624), itself a continuation of 12/102,555 (filed 2008-04-14, now US 8,611,560), which claims benefit of provisional 60/911,691 (filed 2007-04-13).

• If the priority chain holds for claim 1, the pre-AIA critical date is no later than 2008-04-14 (and arguably 2007-04-13).
• If claim 1 lacks § 112 support in the provisional/parent for a given limitation, the date slides to 2018-11-13 — which would make virtually everything in the record prior art.

The '237 patent is pre-AIA (effective filing before 2013-03-16), so § 102(e) references are prior art only as of their U.S. filing date, not their foreign priority date.

This is the single most important gatekeeper in the analysis, and it cuts against the Petitioner's art set. Platz (US 2010/0119077 A1) published 2010-05-13, well after both candidate dates. I could not verify Platz's U.S. filing date from the retrieved materials, but it is a European-origin application (Platz is a Knowles/Nordic-adjacent active-hearing-protection filing) and, on the record I have, its earliest effective date appears likely to be late 2008 — i.e., after the '237 patent's 2008-04-14 parent filing. If that is right, Platz is not prior art to claim 1 at all under pre-AIA § 102(e)(2), and Samsung's Grounds 1A–1C (all Platz-based) were dead on arrival regardless of merits. I flag this as a hypothesis strongly consistent with the chronology, not as a verified holding — I could not retrieve the 2023-01-03 denial's reasoning.

Practically: any new § 103 attack should lead with art that is unambiguously pre-April-2008.

(Platz's publication-date facts are confirmed from the patent's own "Citations" list and from the Google-hosted PDF. The "doubtful" tag is my inference, not a record fact.)

2. Claim 1 decomposed into limitations

Note the drafting point already flagged in the summary section: E–G are the whole case. F mixes ambient with content, not ECM-with-ASM; and E is satisfied by any one of four routine signal-processing techniques. There is no requirement of learning, no accuracy threshold, no particular mixing law, and no gain value. That is a wide, low-specificity claim.

3. Combination 1 (primary, date-safe): Kvaløy '385 + Melanson '825 + Light '551

This combination does not depend on Platz and therefore avoids the priority-date vulnerability.

3.1 Element-by-element

3.2 Motivation to combine (KSR / MPEP 2143 rationales)

1. Same field, same problem, same solution space. All three are ear-level audio devices addressing the identical problem the '237 patent's Background recites: an occluding earpiece that plays audio prevents the wearer from hearing the environment. KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398, 417 (2007) ("familiar elements according to known methods… yields no more than predictable results"). MPEP 2143.01(III) (obvious design choice).
2. Express problem-identification in Melanson. Melanson states the invention "provides a mechanism for either overcoming leakage of external sounds that can and should be ignored, and providing control of the gain of the audio program playback when external sounds should not be ignored." That sentence is the motivation: blend ambient with content so the wearer can hear what matters. (Verified from the Google Patents text.)
3. Predictable result. Substituting one known voice-detection technique (level, spectral, correlation, coherence) for another is a recognized obvious design choice where each yields the same predictable output — the claim's own "or a combination thereof" language concedes functional equivalency.
4. Reasonable expectation of success. Melanson teaches a working, commercially shipped architecture with microphone(s), ADC, DSP and headphone amplifier — the Kvaløy terminal need only be substituted for the headphone transducer.
5. Compactness/size incentive. Melanson explicitly contemplates putting "the entire device… within the headphones" and placing microphones "one positioned at each ear" — supplying the design incentive to merge the electronics into the earpiece, which is precisely the '237 patent's stated advance.

Assessment: strong. This combination teaches every element of claim 1 with no reliance on the contested Platz reference. It is the obviousness theory a defendant should lead with today.

4. Combination 2 (Samsung's theory, restated): Platz + Kvaløy + Melanson

I set this out because it is the art set Samsung actually pleaded (IPR2022-01098 Grounds 1A/1B/1C and 2A/2B/2C, per the district-court record of the proceeding), and because understanding why it may have failed is diagnostically useful.

Platz, verified content:

• Earplug 10/12 comprising "a microphone 28 for capturing audio signals from ambient sound" and "a loudspeaker 30 for providing audio signals into the user's ear canal via the sound channel 26." → A, C.
• Microphones include 28 and 31; a boom microphone 16 may capture the user's voice. → second transducer path, B (arguably).
• Central unit 14 with "an audio signal processing unit 34"; audio signals from a communication device 20 (e.g., mobile phone) are supplied to the ear via speaker 30. → D, C.
• The invention "automatically control[s] the audio signal processing unit depending on whether the presence of an audio signal at the audio input… has been detected." → E (a detection function) — and the stated benefit is that "external audio signals can be combined for communication purposes… with the audio signal captured by the microphones." → F, quite literally.

Dijkstra (EP 1 640 972 A1) was the co-primary in Samsung's Grounds 2A–2C (claims 1, 5, 10–12, 14 over Dijkstra + Kvaløy + Melanson). I could not verify Dijkstra's text; on the face of the ground listing it fills the voice-activity-detection role.

Kvaløy supplies the ear-canal/ear-terminal voice-discrimination structure; Melanson supplies gain-controlled mixing of program audio.

Why this combination is attractive on the merits and unreliable in practice

• On the merits it is close to verbatim. Platz Fig. 2's "external audio signals can be combined… with the audio signal captured by the microphones" reads on limitation F, and speaker 30 into the ear canal reads on G. A tribunal crediting Platz would likely find claim 1 obvious.
• In practice it has two failure modes, either of which plausibly explains the 2023-01-03 denial:
  1. Priority / § 102(e) date. If Platz's effective date postdates 2008-04-14 (see § 1), it is not prior art and the ground collapses before the merits. The record does not let me confirm this, but it is the most likely non-merits reason a merits-labelled denial would land here.
  2. Claim construction. Patent Owner's Preliminary Response was supported by the Struck declaration (Ex. 2001) and the Wiley Electrical and Electronics Engineering Dictionary (Ex. 2009) — i.e., it attacked the petition's reading of the "analysis"/processing terms. That is a § 112-driven construction fight over limitation E, not an over-the-art argument.
• Independent of why it failed, § 315(e)(2) estoppel never attached (no FWD), and the conditional Platz/Dijkstra stipulation (Petitioner Ex. 1030; the parallel letter also appears as Ex. 1033 in IPR2022-01106) lapsed when the institution condition failed. So Platz, Dijkstra, Kvaløy, Melanson, Hotvet and Kondo remain fully available to a future defendant — but they carry a documented failure.

Assessment: strong on paper, and the only set with a public adverse ruling against it.

5. Combination 3 (date-safe, correlation-branch): Victorian '313 + Padhi '899 + Melanson '825

This is the theory to run if (a) Platz is unavailable and (b) the accused product's detection is correlation-based.

• Victorian (US 2005/0058313 A1), "External ear canal voice detection" — squarely discloses limitation B and the concept of limitation E using an ear-canal-located sensor to distinguish the wearer's voice. It appears on the '237 patent's own face among the 64 citations, which is itself an admission of record relevance.
• Padhi (US 2005/0096899 A1) — "an apparatus… includes a cross correlator operable to receive a first audio signal and a second audio signal and to cross correlate the first and second audio signals… identify a plurality of parameters… generate an indicator identifying an extent to which the first and second audio signals match," and expressly contemplates a voice activity detector (Fig. 7, "method for detecting a voice in an input signal"; the spec notes "the voice activity detector 116 could be omitted"). (Verified.) This maps directly onto limitation E ("correlation analysis") and onto dependent claims 6–9 (see § 7).
• Melanson '825 supplies C, F, G.

Motivation: Victorian frames the problem as detecting the user's own voice in the ear canal; Padhi supplies a superior comparison technique — the Examiner in the related reexamination control 90/015,146 expressly reasoned that "one of ordinary skill in the art would have looked to Padhi for examples of comparing signals, and would have been motivated by Padhi's disclosure that cross correlating a microphone signal and reference signal is an improvement over merely comparing the levels of the signal… improving accuracy." That is a record-based, examiner-endorsed motivation — the strongest kind of motivation evidence available, because it is the Office's own articulated rationale, not a litigant's construct.

6. Secondary references mapped to the four "analysis" alternatives in limitation E

Because claim 1 is satisfied by any one of four techniques, the § 103 case can be built four different ways. All of the following are on the '237 patent's own face (the 64-reference citation list) and all pre-date April 2008:

Motivation to select among these: MPEP 2143.01(V) — where the claim recites a genus of four known alternatives, and each alternative is a "finite number of identified, predictable solutions," selecting one is obvious. KSR, 550 U.S. at 421. The patent's own specification concedes the alternatives are interchangeable: "the processor 121, by way of one or more decisional approaches, such as correlation analysis, level-detection, coherence, and spectral analysis, determines whether the sound captured at the ASM 111 and ECM 123 corresponds to the wearer's voice."

7. Dependent claims 2–14

Because claim 1 is the only independent claim, each dependent claim falls if claim 1 falls. But several dependents add limitations that are themselves squarely met by the same art — meaning there is no "dependent-claim haven" here.

Conclusion on dependents: claims 2–5, 6–10 and 13 are the most clearly obvious; claims 11, 12 and 14 are obvious but rest on functional language with thin specification support, so expect a § 112 written-description overlay in addition to § 103. None of the 14 claims recites a numerical limit, a specific algorithm, a particular mixing weight, or any structural detail beyond the four conventional transducers — which is why the § 103 case is comparatively clean.

8. A particularly awkward datum for the Patent Owner: the family's own seal-test disclosure

In the same family (the '237 patent's specification, and the reexamination record for 90/015,146), the applicants describe an ear-seal test in which: an acoustic test signal is played into the ear canal; an ECM in a second volume isolated by an acoustic barrier measures the leakage; the signals are filtered; "a correlation, cross-correlation, or coherence analysis is performed on the first volume signal and the second volume signal"; "the peak of the cross-correlation is identified"; and "Lag-time of Peak 1420 and Magnitude of Peak 1422" are calculated and compared to thresholds to PASS/FAIL.

That is the substance of claims 6, 7, 8 and 9 of the '237 patent, disclosed by the applicant in the same family, applied to a two-transducer ear-seal problem rather than a voice problem. A POSITA seeking to detect the wearer's voice from ASM/ECM timing would have had every reason to lift that machinery and point it at voice instead of seal — the computational technique is identical and the hardware is the same two transducers. This is a motivation supplied by the inventor's own disclosure and is powerful § 103 evidence. (I flag: the seal-test material appears in the family record, including the reexam exhibits I retrieved; I did not isolate its exact location within the '237 patent's own specification text, so cite it from the record before relying on it.)

9. Counterarguments to anticipate, and my assessment of each

Teaching-away check: I found no reference in the record that teaches away from blending ambient sound into program audio, and Melanson's express purpose (hearing "external sounds [that] should not be ignored") is the opposite of teaching away. The '237 specification's own discussion of "soft muting the audio content during… voice activity" is a design option within a mixing architecture Melanson already discloses.

10. Bottom line

Claim 1: I assess claim 1 as obvious over at least three independent, date-safe combinations:

1. Kvaløy '385 + Melanson '825 + Light '551 — strongest; no reliance on contested art.
2. Victorian '313 + Padhi '899 + Melanson '825 — strong, and the correlation variant maps element-for-element onto limitation E.
3. Platz + Kvaløy + Melanson (Samsung's theory) — strong on the merits if Platz's effective date predates 2008-04-14; void before the merits if it does not, which is my working hypothesis for the denial.

In every combination: the two-microphone ASM/ECM architecture is old (Kvaløy, Nacre '359, Victorian); the four named analyses are each separately known and expressly interchangeable in the patent's own specification; the mixing of ambient sound into delivered audio content is disclosed by Melanson and Light and, in substance, by Platz; and the motivation is set out in the references themselves and in the '237 patent's own Background.

Claims 2–14: none rescues the claim set. Claims 2–5, 6–10 and 13 are individually obvious over the same or analogous art (and where the examiner/reexam record supplies an articulated rationale — e.g., Padhi for "improving accuracy" over level-only comparison — that rationale carries more weight than a litigant's). Claims 11, 12 and 14 are obvious but also likely vulnerable under § 112, so they should be attacked on both tracks.

Principal caveats and things I could not confirm:

• I did not retrieve the 2023-01-03 denial's verbatim reasoning, and I do not attribute the outcome to any specific limitation or ground. My priority-date hypothesis for Platz is inference from chronology, not a record fact.
• Platz's earliest effective filing date is unverified. Confirming it (USPTO Patent Center / Global Dossier) is the single highest-value next step: if it postdates April 2008, one entire theory is foreclosed and the other two become the case.
• I did not verify the contents of Hotvet US 5,550,923, Kondo JP 3353701, Dijkstra EP 1 640 972 A1, or Duffner 2006; my references to them are limited to the ground-listing roles recorded in the district-court/PTAB record.
• Same-family references (Personics/Ryan-adjacent US 2008/0181419, US 2008/0137873, US 2009/0010444, and the seal-test material) are useful for demonstrating the state of the art and the applicants' own knowledge, but are problematic as § 103 art because of common-ownership/inventorship issues. Use them primarily to show what a POSITA knew, or as § 112 support, not as primary § 103 references.
• Dates for Boersma '171 and Creative '341 § 102(e) are estimated from their apparent U.S. filing dates (≈2006), not confirmed from the documents; verify before relying on them.