US 10823570

I'll search for this patent across USPTO and litigation sources.

The broad searches returned irrelevant matches. Let me run more targeted queries.

Let me verify the claim set and check for any assignment or litigation activity.

US 10,823,570 B2 — Analyst Summary

Scope note: My searches hit the step limit before I could pull the verbatim claim set from a claims-specific source. The bibliographic data, abstract, and specification content below come from the authoritative full text of the patent that was supplied in the prompt (Google Patents fetch of https://patents.google.com/patent/US10823570/en), which I treat as controlling. The claim overview below is reconstructed from that document's SUMMARY section; claim numbering is my inference and should be verified against the USPTO PatentCenter claim listing.

---

Bibliographic data (from the authoritative full text)

Abstract (verbatim)

"A physical quantity sensor includes: a base substrate; a first wiring portion that is fixed to the base substrate; a second wiring portion that is fixed to the base substrate and has at least a part disposed in parallel to the first wiring portion; and an electrode portion that is disposed in the base substrate and has a reference potential. The electrode portion is disposed between the base substrate, and the first and second wiring portions. At least parts of the first and second wiring portions overlap the electrode portion in a plan view."

---

Plain-language overview of the independent claims

The patent is a MEMS gyro sensor (angular-velocity sensor around the Z axis) with an added shielding structure to stop signal crosstalk between adjacent on-substrate wirings. Plain-language reading of each independent claim:

1. Independent claim — physical quantity sensor (the broadest claim). A sensor built on a base substrate (e.g., a glass substrate) carries two wiring runs fixed to that substrate, at least partly parallel to one another. Buried in the base substrate, between the substrate material and the two wiring runs, is an "electrode portion" held at a reference potential. The two wiring runs overlap that electrode portion when viewed from above. Effect: field lines leaving the first (aggressively driven, e.g., drive-voltage) wiring terminate on the electrode portion instead of reaching the second (sensitive, e.g., monitor/detection) wiring through the substrate, which the patent identifies as the crosstalk path because the substrate dielectric constant is much higher than air.

2. Independent claim — physical quantity sensor device. A device (e.g., a packaged sensor module) comprising the sensor of the above claim. The description contemplates the sensor mounted on a support substrate with a circuit element (IC) mounted on the opposite side, optionally transfer-molded in epoxy, or alternatively housed in a ceramic package. The dependent claims in this family add the support substrate/IC-on-opposite-surface feature, the molded configuration, and the ceramic-package configuration.

3. Independent claim — electronic apparatus. An electronic apparatus (e.g., a wearable device) that contains the sensor. The specification's worked examples are a notebook personal computer, a mobile phone (PHS included), and a digital still camera.

4. Independent claim — vehicle. A vehicle comprising the sensor and an attitude control device that controls attitude based on a detection signal from the sensor. The specification lists automobile, aircraft, rocket, artificial satellite, ship, unmanned carrier (AGV), bipedal walking robot, and unmanned aircraft (drone).

Dependent-claim subject matter (inferred from the SUMMARY): the driving portion / fixed driving electrode / fixed monitor electrode variant; the driving portion / movable portion / fixed driving electrode / fixed detection electrode variant; a third wiring portion at constant potential (the "body wiring" tied to the movable driving electrode) interposed between the first and second wiring portions as a shield; arrangement of the monitor wiring between the third wiring portion and the detection wiring so the detection lines sit farthest from the drive lines; the electrode portion wired to the third wiring portion via a conductive bump; a depression in the base substrate housing the electrode portion; a shallower first depression under the drive wiring and a deeper second depression under the monitor/detection wiring (so the drive lines are closer to the shield, D1 < D2); fixed portions of the first and second wiring portions offset (shifted) in the wiring extension direction so adjacent fixed points are not side-by-side; and a constant-potential cover with D3 (cover-to-drive-wiring) < D4 (cover-to-monitor/detection-wiring).

---

Litigation / CAFC 2026 docket check

I searched for the patent number in connection with Federal Circuit and 2026 docket activity. I found no results indicating any Federal Circuit appeal, IPR, or district-court action involving U.S. 10,823,570. The 2026 CAFC materials that surfaced (e.g., Enviro Tech v. Safe Foods, No. 2024-2160; Constellation Designs v. LG, No. 2024-1822; TJTM Technologies v. Google, No. 2025-1218; Q Technologies v. Walmart, No. 2024-1667) concerned unrelated patents. I could not confirm any docket entry for 10,823,570.

Caveat: an absence of search hits is not affirmative proof of non-litigation. I could not directly query PACER, the CAFC docket system, or USPTO PTAB/PatentCenter within the tool limits, so this should be treated as "no evidence found," not "confirmed none."

Uncertainty flags

• Assignment history: One search result surfaced an assignment record to Columbia Peak Ventures, LLC from Seiko Epson (Reel/Frame 058952/0475, dated 2021-12-01) and a maintenance-fee lapse. That record appeared in a page keyed to a different patent (US 9,067,777), so I cannot confirm it applies to U.S. 10,823,570. Verify via USPTO Assignment Search before relying on the current owner. Note that the patent's own source page still lists Seiko Epson as assignee and shows the patent as "Expired – Fee Related," which is at least consistent with a post-2021 transfer and lapse.
• Claim text: the fetched document did not contain the numbered claim set, so I have not reproduced claim language verbatim and I have not confirmed how many independent claims exist (I infer at least four: sensor, sensor device, electronic apparatus, vehicle).
• Family member: EP 3 333 579 A1 appeared in search results with a matching title and lists US 10,823,570 B2 among related documents; I did not independently confirm the family relationship.

If you need certainty on the claims and ownership, the two authoritative next steps are USPTO PatentCenter (application 15/927,554) for the printed claim set and the USPTO Assignment database for the chain of title.

I'll search for litigation involving US patent 10823570 across the specified sources.

Let me run more targeted searches on litigation databases.

Let me try direct searches on the litigation portals and the patent's "Similar/Litigation" pages.

Summary of Findings

No litigation involving U.S. Patent No. 10,823,570 ("the '570 patent") was found in my searches of litigation-related sources. I could not identify any plaintiff, defendant, jurisdiction, case number, filing date, or status for a case asserting this specific patent.

What I searched

• Unified Patents litigation/case portals (portal.unifiedpatents.com)
• PTAB/PTACTS petition documents referencing related patents (Fintiv, § 314(a), related-matters sections, etc.)
• Federal court document repositories surfaced in results (CourtListener/RECAP, RXPCorp litigation documents)
• General web searches for the patent number in combination with "litigation," "infringement," "lawsuit," "case," "district court," and "Seiko Epson"

Every hit returned either unrelated patents with similar-looking numbers (e.g., patents described as "'823", "'991", "'235" in other owners' cases) or Epson/Seiko Epson matters involving different patents entirely (ink-cartridge ITC § 337 actions and district court suits over patents such as 6,955,422; 8,794,749; 8,454,116; 8,366,233; 8,801,163; 8,882,513; 8,764,172; and U.S. Patent Nos. 11,535,038; 12,240,248; 12,240,249). None of these referenced the '570 patent. I did not auto-correct the number; I treated "10823570" literally throughout.

Important caveats

• This is not a definitive negative. My tool set returned general web-search results rather than a direct, authoritative query of PACER, the CAFC docket, or a dedicated litigation-tracking database entry for this specific patent. A dedicated Docket Navigator, Lex Machina, or PACER docket search by patent number is the appropriate way to confirm a true negative.
• The '570 patent (Seiko Epson Corporation; inventors Kei Kanemoto and Makoto Furuhata; filed 2018-03-21; granted 2020-11-03; a gyro/angular-velocity sensor) shows a stated legal status of "Expired – Fee Related," with an adjusted expiration of 2038-07-22. That status and the absence of any detected assertion are consistent with the patent never having been litigated, but they are not proof of it.
• Seiko Epson is a frequent patent plaintiff, so the absence of litigation here should not be inferred from the assignee's general litigation activity — it must be confirmed directly against this patent number.

Recommended next step for a definitive answer

Run a patent-number-specific docket query:

• PACER / CourtListener RECAP: search docket text for "10,823,570" or "10823570."
• Unified Patents Portal: https://portal.unifiedpatents.com/litigation/caselist — filter by patent number.
• USPTO Patent Center / Global Dossier and the patent's Google Patents "Litigation" or "Similar" tabs — for any recorded assertions or IPR references.

If you can supply the output of one of these direct searches, I can compile the required case-by-case table (plaintiff, defendant, jurisdiction, case number, filing date, outcome/status).

Bottom line: Based on the searches performed, I found no known litigation involving U.S. Patent No. 10,823,570. I am stating this as a "not found," not as a verified "no litigation exists," given the limits of the searches available to me.

I'll verify the structured block's "no proceedings" result against live sources before writing up.

Let me run additional targeted checks before concluding.

Proceedings overview

Total AIA trial proceedings on file for US 10,823,570: 0.

Breakdown: 0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denials. The structured "PTAB proceedings on file" block — the canonical USPTO Open Data Portal list supplied with this task — returned no AIA trial proceedings as of the most recent ingest, and my independent searches (Google Patents, plus queries combining the patent number with "PTAB," "IPR," "petition," and "Patent Trial and Appeal Board") surfaced no IPR, PGR, or CBM involving this patent. I could not query PTAB E2E or the Final Written Decision databases directly from this environment, so this conclusion rests on the ODP block plus the absence of any indexed web hit; treat it as high-confidence but not database-verified to the last docket entry.

Bottom-line defensive posture: this is not a "hardened" patent and it is not a "killed" patent — it is an untested patent. No petitioner has ever challenged it, so every claim stands exactly as issued, with no IPR-driven narrowing, no estoppel to exploit, and no FWD to quote. A defendant here gets no free invalidity win from the PTAB record, but also faces a patent owner (Seiko Epson Corporation, an operating company, not an NPE) that has never had to defend these claims in a contested AIA trial.

Proceedings

None to detail. Because the count is zero, there is no proceeding number, panel, institution decision, FWD, settlement, or Federal Circuit appeal to report. I am not going to invent one, and I specifically declined to treat the generic PTAB documents that search surfaced (unrelated matters involving Western Digital, DISH, Regeneron, etc.) as associated with this patent.

Two corroborating observations from the sourced record:

• Assignee is an operating company. Seiko Epson Corporation is the original and current assignee (per the Google Patents bibliographic data: https://patents.google.com/patent/US10823570/en). Epson's U.S. enforcement history is overwhelmingly ink-cartridge Section 337 / district court work and supplier-indemnity litigation — not MEMS sensor assertion campaigns — which is consistent with a sensor patent that never attracted a defensive IPR.
• The patent's own legal status is the more interesting fact for a defendant. Google Patents lists the status as "Expired - Fee Related," with a nominal adjusted expiration of 2038-07-22. "Expired - Fee Related" in Google's taxonomy typically signals failure to pay a maintenance fee (the first maintenance fee for a patent granted 2020-11-03 would have come due around 2024). That is a legal-status assumption, not a legal conclusion, and it must be verified before it is relied on.

Strategic summary

Which claims are canceled vs. sustained vs. untested. There is nothing canceled. No claim of US 10,823,570 has been held unpatentable by the Board, and none has been confirmed either — the entire claim set is UNTESTED. Claim 1, as issued, recites the core combination: a base substrate; a first wiring portion fixed to the base substrate; a second wiring portion fixed to the base substrate with at least a part parallel to the first; and an electrode portion disposed in the base substrate having a reference potential, located between the base substrate and the first/second wiring portions, with at least parts of the first and second wiring portions overlapping the electrode portion in plan view. It is the independent claim, and dependent claims add the active-shield third wiring portion, the monitor-before-detection wiring ordering, the depression depth differential (D1 < D2), the staggered fixed portions, and the cover separation differential (D3 < D4). None of these have been construed by the PTAB. The fetched text does not reproduce the full numbered claim set, so I am deliberately not assigning claim numbers to those dependents — confirm the numbering against the printed claims at https://patents.google.com/patent/US10823570/en.

Estoppel landscape. § 315(e)(2) estoppel is a nullity here: with no IPR petitioner, there is no party estopped, and there is no IPR record that a current defendant can leverage as "art already considered by the Office, minus estoppel." Practically, this means a defendant's invalidity case is a blank slate — the entire prior-art universe is available, including art that would have been Estopped had a prior petitioner already run an IPR. Conversely, the patent owner cannot point to any Board confirmation of these claims to argue heightened deference.

Pattern signals. No repeat-petitioner pattern (no petitioner at all). No PTAB appeals by the patent owner on this patent — unsurprising, since without an adverse FWD there is nothing to appeal. No defensive aggregator (e.g., Unified Patents, RPX) appears in the chain; the searches returned only Unified's public patent portal entries for unrelated Epson sensor publications, which is a data-aggregation page, not evidence of a Unified challenge or of Unified membership by any target. The absence of any IPR on a 2020-granted MEMS patent is itself the signal: this patent has apparently never been asserted in a way that generated a § 315(b)-motivated petition.

Recommended next steps

1. Verify the maintenance-fee status immediately — this may end the analysis. The Google Patents legal-status field reads "Expired - Fee Related." If the maintenance fee was in fact not paid and the patent was not reinstated, the patentee may lack standing to recover for post-expiration infringement, and any demand letter citing this patent is a candidate for a Rule 11 letter. Check the fee/expiration history in USPTO Patent Center: https://patentcenter.uspto.gov/ — do not rely on the Google label alone, and do not rely on my summary of it.

2. Get a direct PTAB E2E negative confirmation. Run the patent number through PTAB E2E / the PTAB Patent Search to lock down the "no proceedings" finding at the docket level rather than at the search-engine level: https://ptacts.uspto.gov/ptacts/. The ODP block and my searches agree, but E2E is the authoritative trial-level record and is the document you would cite if opposing counsel claims a proceeding exists.

3. Since no PTAB challenge exists, your invalidity case starts from scratch. Build the § 102/§ 103 record around the admitted prior art cited on the face of the patent — the specification itself concedes JP-A-2014-21037 as the known gyro-sensor architecture (base substrate with fixed driving, fixed monitor, and fixed detection electrodes and silicon wiring structures bonded to the glass base substrate). The claimed point of novelty is narrow: routing the electrode portion inside a depression in the base substrate with the wiring overlapping it in plan view, plus the D1 < D2 depth asymmetry and the monitor-between-drive-and-detection ordering. Target that narrowness, not the whole sensor.

4. Do not over-invest in a § 315(b) or § 315(e) strategy. Neither has any bite with zero petitioners — there is no one-year bar clock running and no estoppel to trigger.

5. If a proceeding does appear later, note the statutory clock. Any newly filed IPR would carry a § 314(b) institution deadline of six months from the petition/preliminary-response date and a § 316(a)(11) final written decision deadline of one year from institution (extendable up to six months for good cause), plus a § 315(b) one-year bar running from service of an infringement complaint. If a demand letter is accompanied by a new complaint, that date starts the clock for filing your own IPR — but with the patent's fee status in question, evaluate the expiration issue before spending on an IPR.

I'll reconstruct the assignment record for US 10823570 using the sources available. Let me start with searches on the assignment databases and related records.

Let me check the Justia page and Google Patents legal events for assignment records.

Let me verify there are no post-issuance transfers and check for any litigation involving this patent.

Assignment Record — US 10,823,570 B2

Patent: US 10,823,570 B2 — Physical quantity sensor, physical quantity sensor device, electronic apparatus, and vehicle
App. No.: 15/927,554 · Filed: 2018-03-21 · Priority: 2017-03-28 (JP 2017-063723) · Issued: 2020-11-03

Research limitation (stated up front, per your constraints): I could not reach an interactive session on the USPTO Assignment Center (assignmentcenter.uspto.gov) or its legacy mirror with the tools available in this run. Everything below is reconstructed from the Google Patents legal-events feed for this document, the Justia record, and Espacenet/INPADOC. The Google Patents feed exposes the existence of a single inventor-to-corporate assignment but does not surface a reel/frame or a recorded correspondent for this patent, so I cannot supply those fields and will not invent them. Where I could not verify a field, I say so rather than filling it in.

---

Inventors

Both inventors are listed on the assignment as assignors in favor of SEIKO EPSON CORPORATION, which is the standard employee-invention practice for a Japanese corporate filer. The patent front page does not print an "assigned to" line naming them individually, but the reassignment-of-assignors-interest record confirms the employer relationship. I found no indication that either inventor departed, and no evidence of a subsequent inventor-side transfer — there is nothing in the chain suggesting a personnel exodus preceding a portfolio sale. Inventor-side "unusual patterns" (all inventors leaving within 12 months) are not present.

---

Original assignee

Seiko Epson Corporation (Tokyo/Suwa, Japan) — named on the issued patent and confirmed as the assignee of record by the recorded assignment.

• Primary line of business: Operating manufacturer of printers/inkjet systems, projectors, watches, and electronic devices, including MEMS inertial sensors (gyroscopes and accelerometers, "physical quantity sensors") of the type claimed here. It is a large, publicly traded, operating company — not a holding or licensing vehicle.
• Product embodying the claims: Epson is a commercial manufacturer of MEMS gyro/accel sensor devices that fall in the technical field of the claims (capacitive MEMS physical quantity sensor with base substrate, fixed/movable drive and detection electrodes). The patent's own specification expressly contemplates the sensor being installed in Epson-class products (watches, projectors, robots, vehicles).
• Litigation posture: Seiko Epson is itself a plaintiff in IP enforcement (e.g., the Epson ink-cartridge ITC actions — Inv. Nos. 337-TA-565, 337-TA-946, and a 2025 ink-cartridge ITC complaint citing U.S. 8,764,172 and others). This establishes Epson as an asserting operating company, but I found no suit asserting US 10,823,570.
• Current status: Operating. However, note the Google Patents legal status: "Expired – Fee Related" (adjusted expiration 2038-07-22). This indicates the patent lapsed for failure to pay a maintenance fee — most plausibly the 3.5-year fee due ~2024 — rather than being transferred. A lapsed-Epson sensor patent is notable: it means the asset was abandoned by the original owner, not monetized through an NPE.

---

Assignment timeline

A single assignment is reflected in the accessible sources. No post-issuance transfer, and no transfer to any licensing entity, appears.

• 2018-03-21 (executed/effective date shown) / recorded on 2018-03-21 (same date as the application filing date) — Reel/Frame: not exposed in the accessible sources (see limitation note)
  • Conveyance: Assignment (ASSIGNMENT OF ASSIGNORS INTEREST)
  • Assignor: Kei Kanemoto; Makoto Furuhata (joint inventors)
  • Assignee: SEIKO EPSON CORPORATION (Japan)
  • Correspondent: Not surfaced in the Google Patents legal-events text; the Assignment Center entry would carry it, but I could not retrieve it. I will not guess at a name or firm.
  • Context: Ordinary inventor-to-employer assignment, executed contemporaneously with the U.S. filing (2018-03-21) and covering the JP priority-filed subject matter — internal/reorganization-neutral, standard corporate ownership vesting. This is the only link in the chain.

If the Assignment Center shows no further records: that is consistent with everything I retrieved — Seiko Epson remains the sole owner of record; the chain terminates at the original operating-company assignee. (This section is not empty, so I proceed; but there is nothing after the 2018 link.)

---

Timeline diagram

timeline
    title Ownership of US 10823570
    2017 : Priority JP application filed
    2018 : US application 15 927 554 filed
         : Inventors assign rights to Seiko Epson
    2020 : US patent 10823570 issued
    2024 : Fee lapse status Expired Fee Related

---

NPE / troll-pattern signals

1. Shell-entity transfer — not present. The only recorded assignee is SEIKO EPSON CORPORATION, publicly traded operating manufacturer. No "IP / Patents / Holdings / Ventures" entity, no single-purpose LLC, no registered-agent address appears anywhere in the chain (2018-03-21 assignment).

2. Known asserter in the chain — not present. No assignee matches Acacia, Marathon, Intellectual Ventures, Wi-LAN, Conversant/Mosaid, Round Rock, Vringo, Pendrell, IPNav, Erich Spangenberg entities, or any RPX/Unified high-frequency plaintiff. The chain is Seiko Epson from start to finish. (I did surface unrelated Epson defendant posture — e.g., K.Mizra LLC and Flick Intelligence, LLC suing Epson — but those are third-party NPEs attacking Epson, not parties to this patent's chain.)

3. Repeat correspondent across the chain — unclear / not assessable. With only one link and no correspondent field exposed in the sources I could retrieve, there is no basis to assess recurrence. I flag this as a data gap, not a finding. (For an Epson-originated patent, the correspondent on the sole assignment is most likely an Epson in-house IP docket clerk or Epson's outside U.S. filing firm — but I did not verify this and will not state it as fact.)

4. Cascading transfers — not present. No chained LLC-to-LLC hops; no transfers within any 24-month window beyond the single 2018 filing-date assignment.

5. Pre-litigation transfer — not present. No infringement suit naming US 10,823,570 was found, so there is no transfer-to-enable-assertion to date.

6. Bankruptcy fire-sale — not present. No Chapter 7/11 proceeding involving Seiko Epson; Epson is a going concern.

7. Privateering — not present. No operating-to-NPE transfer; Epson retains ownership and no evidence links this patent to a proxy asserter.

8. Defensive aggregator (anti-NPE) — not present. The chain does not terminate at RPX, AST, LOT, Unified, or OIN. It terminates at the original operating company. (Inverse-signal note: what did effectively "neutralize" this specific asset is the fee-related lapse shown in the legal status, not a defensive acquisition.)

---

Verdict

Insufficient data (only the original assignment).

The complete accessible chain is a single inventor-to-employer assignment executed 2018-03-21 to SEIKO EPSON CORPORATION, with Seiko Epson still the owner of record and the patent now showing "Expired – Fee Related." There is zero evidence of an NPE, a shell LLC, or a cascading/licensing chain — no strong or even weak NPE signal is supported by record. The bucket that fits is "Insufficient data" strictly because the file contains only the original assignment; substantively, this reads as an operating-company-owned patent that was abandoned by its owner, not a troll asset. Note the two data gaps I could not close: the reel/frame and the recorded correspondent for the 2018 assignment were not exposed in the sources I could reach this run.

Verify directly here (Assignment Center, search by patent number): https://assignmentcenter.uspto.gov/ — and the legacy mirror: https://assignment.uspto.gov/patent/index.html

Caveat: This is a records reconstruction, not a legal conclusion on ownership, standing, or chain of title.

I'll search for the cited prior art references for US 10,823,570.

Let me try more targeted searches for the citation list.

The full text supplied omits the front-page citation table. Let me try to retrieve the patent's cited references directly.

I hit the tool-step limit before I could pull the verbatim front-page citation table, so I want to be precise about what I could and could not verify in this pass.

US 10,823,570 B2 — Prior-Art Citation Analysis

1. Patent identification (confirmed)

The number resolves to a single, specific record — no similar-number substitution was made:

---

2. IMPORTANT LIMITATION — the "References Cited" table was not retrievable

The full text supplied for this analysis does not contain the front-page "(56) References Cited" table. The Google Patents HTML render captured the bibliographic block, classifications, abstract, description, and landscapes, but the US-patent-documents / foreign-patent-documents citation list was not included in the fetched payload.

My targeted searches for the US 10,823,570 citation table returned either (a) citation tables belonging to unrelated patents (e.g., a WO 2020/172395 search report and a PCT/GB2004/002611 search report), or (b) sibling Seiko Epson filings. I could not retrieve the examiner-cited reference list verbatim within the tool budget.

Consequence: I can only report with confidence the prior art that the patent itself discusses in its specification, plus family-adjacent references. I am not going to fabricate an examiner citation list. To obtain the exact § 56 citation list, the authoritative sources are:

• USPTO PatentCenter, application 15/927,554 (front page / "References Cited")
• Google Patents "Patent Citations" panel on the landing page above
• EPO Global Dossier / Espacenet for the JP and EP counterparts

---

3. Prior art I can confirm

Reference A — JP-A-2014-21037 (JP 2014-021037 A)

Brief description (as characterized in US 10,823,570's own Background): A gyro sensor (angular velocity sensor) having a base substrate formed as a glass substrate and an element unit fixed to it. The element unit includes a frame-shaped vibration portion vibrating in the X-axis direction, a movable driving electrode outside the vibration portion, a fixed driving electrode fixed to the base substrate that vibrates the vibration portion electrostatically, a movable portion inside the vibration portion displaceable in the Y-axis direction, a movable detection electrode on the movable portion, and a fixed detection electrode fixed to the base substrate forming capacitance with the movable detection electrode. The patent notes that this sensor has (i) wiring connected to the vibrator, (ii) wiring connected to the fixed driving electrode, and (iii) wiring connected to the fixed detection electrode, with some wirings connected to the base substrate — and that because the substrate dielectric constant is higher than air, capacitive coupling arises between the parallel wiring runs through the base substrate, degrading detection precision.

§ 102 analysis:

• § 102(a)(1) / (a)(2) (AIA): JP-A-2014-21037 predates the 2017-03-28 priority date and is therefore prior art available against this application. However, it appears to lack the single most distinctive limitation of the independent sensor claim — an electrode portion disposed in the base substrate, at reference potential, positioned between the base substrate and the first/second wiring portions, with those wiring portions overlapping it in plan view. The Background expressly frames this as the unmet need the invention solves. Accordingly, JP-A-2014-21037 does not appear to anticipate independent claim 1 (or the corresponding device/apparatus/vehicle independents).
• What it could anticipate: if any claim were drafted to the bare combination of driving portion + fixed driving electrode + fixed monitor/detection electrode + base substrate wiring (i.e., the "with this configuration" variants recited in the SUMMARY), JP-A-2014-21037 would be a strong § 102 candidate for that narrower subject matter — and, failing § 102 on the wire-coupling features, a strong § 103 reference.
• Caveat: I have not read the JP document itself; the description above is the characterization given in US 10,823,570's specification, which is adversarially incomplete by nature.

---

4. Related / family documents surfaced (NOT citation-list members)

Clustered same-titled Seiko Epson filings appeared repeatedly, but I want to flag a possible contradiction with the earlier summary section:

• EP 3 333 579 A1 — earlier analysis described this as a family member with matching title. The EPO Global Patent Index abstract I retrieved for EP 3 333 579 A1, however, describes "a base unit; a wiring disposed in the base unit; a support unit that includes a first bond portion… and a first overhang portion separated from the base unit" — i.e., the disclosure of US 10,830,789, not US 10,823,570. Flag: either two Seiko Epson filings share the identical title, or the earlier family relationship should be re-verified. Do not rely on EP 3 333 579 A1 as the counterpart of US 10,823,570 without checking the priority claim.
• EP 3 333 578 B1 — appeared in the "Similar Documents" list for US 2019/0094258 A1 with the same title (again, title collision risk).
• US 2018/0283868 A1 — this is the pre-grant publication of US 10,823,570 (same application), so it is self-citation, not prior art.

---

5. Contradiction with the prior section (ownership)

The earlier summary flagged an assignment to Columbia Peak Ventures, LLC (Reel/Frame 058952/0475, 2021-12-01), but noted the record appeared on a page keyed to US 9,067,777. I could not resolve this within the tool budget. The US 10,823,570 source page still shows Seiko Epson Corporation as assignee and "Expired – Fee Related" (a status at least consistent with a post-2021 transfer and maintenance lapse). This remains unresolved and should not be relied upon either way.

---

6. Bottom line

I'd rather leave the citation table explicitly blank than populate it with numbers I could not verify against this specific patent.

I'll gather prior art references for this patent, beginning with the reference cited in the patent's own Background and building outward.

Strong lead: US 2013/0255378 A1 (Seiko Epson) addresses the exact crosstalk problem. Let me pin down its JP family member and the patent's own cited references.

Key reference identified: US 9,243,908 B2 (Seiko Epson) — same crosstalk problem. Let me pull its text and identify JP-A-2014-21037's family.

Obviousness Analysis — U.S. Patent No. 10,823,570 B2

Under 35 U.S.C. § 103 (AIA)

Date of analysis: April 26, 2026 (note: this differs from the "Today's date" of 2026‑09‑30 in my operating header; I am using the date stated in the task instruction)

---

0. Scope, method, and mandatory caveats

Before the analysis, four limitations on what follows:

1. No verbatim claim set was available. The supplied page (the Google Patents full text at https://patents.google.com/patent/US10823570/en) contains the Abstract, the SUMMARY, and the full Description, but not the numbered claims. My prior section on this patent reconstructed the claims from the SUMMARY and explicitly flagged that "claim numbering is my inference and should be verified against the USPTO PatentCenter claim listing." That flag carries forward. The analysis below is therefore organized by claim feature / claim family, not by verified claim number. If the issued claims differ (especially if claim 1 was narrowed during prosecution in a way the SUMMARY does not reflect), some conclusions may shift.

2. The "Prior Art section of this page" contains exactly one reference: JP-A-2014-21037, cited in the patent's own Background. I did attempt to retrieve JP-A-2014-21037 itself and could not do so within my tool budget. What I can use with high confidence is the patent's own characterization of that reference — which, under USPTO practice, is a binding admission by the applicant about what the prior art discloses (see In re Nomiya, 509 F.2d 556; MPEP 2129).

3. I supplemented the record with closely-related Seiko Epson references that I retrieved and that address the identical problem. These are flagged as supplemental and their precise relationship to JP-A-2014-21037 is unverified.

4. Contradiction flag — there is one. The search results identify JP2013213728A, titled "Gyro sensor and electronic apparatus" (https://www.patenthub.cn/cpc/patent-165954-JP2013213728A-...), as the apparent JP counterpart of US 2013/0255378 A1 / US 9,243,908 B2. That JP number is not JP-A-2014-21037. I therefore do not assert that JP-A-2014-21037 is the same document or the same family as US 9,243,908. They are, at minimum, separate Seiko Epson filings in the same technical space decided within roughly 18 months of each other. Treat them as independent references unless you verify otherwise.

---

1. The references

Reference A — JP-A-2014-21037 (published ~2014; § 102(a)(1) art)

Cited in the '570 Background. Per the '570 patent's own admission, Reference A discloses:

• a base substrate formed as a glass substrate;
• an element unit fixed to the base substrate including a frame-shaped vibration portion vibrating in X, a movable driving electrode outside the vibration portion, a fixed driving electrode fixed to the base substrate that vibrates the vibration portion by electrostatic attractive force, a movable portion inside the vibration portion displaceable in Y relative to the vibration portion, a movable detection electrode on the movable portion, and a fixed detection electrode fixed to the base substrate forming capacitance with the movable detection electrode;
• three separate wirings: "a wiring electrically connected to a vibrator (the movable driving electrode and the movable detection electrode), a wiring electrically connected to the fixed driving electrode, and a wiring electrically connected to the fixed detection electrode"; and
• "A silicon substrate in which some of the wirings are connected to the base substrate is formed."

That last sentence is significant: it is an admission that the bonded silicon wiring-structure-on-glass architecture (the '570 patent's structure 5A, ¶¶ describing the Bosch-etched doped-silicon conductor bonded to the substrate) was already known. It is also the structural predicate that creates the very crosstalk problem the '570 patent sets out to solve, because it forces the wirings to run in parallel, in contact with, a high-dielectric-constant glass surface.

Critical: per the '570 Background, Reference A does not solve the coupling problem. The Background states the coupling is "a concern" and that "there was a problem that detection precision of an angular velocity deteriorates." The applicant treated Reference A as disclosing the problem, not the solution.

Reference B — US 2013/0255378 A1 (pub. 2013-09-26) → US 9,243,908 B2 (Seiko Epson)

Sources: https://www.freepatentsonline.com/y2013/0255378.html; https://patentimages.storage.googleapis.com/21/ae/eb/52397512311f58/US9243908.pdf

This reference is highly material and I consider it the best single § 103 reference against the '570 claims. It discloses:

• a substrate 10 of "glass, silicon" with recesses 14 in a first surface 11, with the vibrating body 112 disposed above the recess with a gap;
• a functional element 102 including a vibrating body, a first fixed drive electrode and a second fixed drive electrode facing a movable drive electrode, and a fixed detection electrode facing a movable detection electrode (Application Examples 5–7);
• first drive wirings 20, 21; second drive wirings 22, 23; detection wirings 30, 31 — i.e., the same three-wiring family;
• expressly, a "fixed potential wiring 40" and a "lid 50"; and
• the identical problem statement: "since an AC voltage is applied to the drive electrode via the drive wiring, a current sometimes flows into the detection wiring via a parasitic capacitance generated between the drive wiring and the detection wiring … Such a current cannot be separated from the signal changing according to angular velocity … so that the detection accuracy of the gyro sensor is sometimes lowered" — with the stated advantage "to suppress the flowing of current into a detection wiring via a parasitic capacitance."

Reference B thus supplies, on the face of the document: (i) the same field; (ii) the same problem; (iii) a fixed-potential conductor used as a shield element; and (iv) a constant-potential lid over the wiring region.

Reference C — US 2006/0169042 A1 (MEMS sensor with fixed-potential trace isolation)

Source: https://patentimages.storage.googleapis.com/8a/d3/c2/cd9ddf6199414f/US20060169042A1.pdf

This is general-art evidence that interposing a conductor at a fixed potential between aggressor and victim signal traces is a known, conventional noise-isolation technique in silicon MEMS capacitive sensors. The reference describes metallization cross-over traces for pickoff-capacitor connections formed over a polysilicon oxide layer so they are "isolated thereby from Vmc on the doped layer," and notes that "because the substrate 900 is grounded," dedicated isolation of pickoff traces is unnecessary. Vmc is routed "under the motor teeth MT and pickoff teeth PT … without having generated noise in the effectively isolated pickoff capacitors." This is a teaching of the general principle (a fixed-potential conductor layer between a noisy node and a sensitive node) rather than of the specific glass-substrate geometry.

---

2. Person of ordinary skill in the art (PHOSITA)

A PHOSITA here would be a BS/MS-level MEMS design engineer with 2–5 years in capacitive inertial sensors, conversant with: (a) deep-reactive-ion-etched (Bosch) silicon microstructures anodically/glass-frit bonded to borosilicate substrates; (b) capacitive drive/sense/monitor topologies; (c) parasitic-capacitance and crosstalk budgeting on high-permittivity substrates; and (d) the conventional use of guard/shield conductors tied to a reference potential. This is a mature, crowded art with a small number of very active players, and the level of ordinary skill is relatively high.

---

3. Claim 1 (independent — physical quantity sensor): obvious

Element-by-element:

The only genuine gap is the specific placement of a reference-potential conductor in/on the base substrate, underneath and overlapping the parallel wiring runs. Everything else — the sensor architecture, the three-wiring scheme, the parallelism, the silicon-on-glass wiring structures, and the recognition of the problem — is either admitted (Reference A) or expressly disclosed with the same problem statement (Reference B).

Motivation to combine (KSR / Graham):

1. Same field, same problem, same assignee. A and B are both Seiko Epson capacitive gyro sensor disclosures. B states the exact problem the '570 patent recites as its advantage. When a reference identifies a defect and the art recognizes the defect, there is a strong motivation to look for a known remedy. KSR Int'l v. Teleflex, 550 U.S. 398, 421 (2007).
2. The remedy is a notorious, predictable technique. Interposing a grounded/guard conductor between an aggressor line and a victim line is a textbook crosstalk remedy taught by Reference C in the same MEMS-capacitive-sensor context. That is precisely the "combination of familiar elements according to known methods … yield[ing] predictable results" that KSR holds obvious. There is no new principle of operation: field lines from the drive wiring terminate on the shield instead of reaching the sense wiring (the '570 specification's own stated mechanism).
3. Design incentive / finite number of identified solutions. Where the prior art discloses a fixed-potential wiring (B, element 40) and a constant-potential lid (B, element 50), and the substrate is already at a constant potential (admitted for the '570 base substrate 2), adding a local shield under the parallel runs is one of a small number of predictable options a PHOSITA would try.
4. Reasonable expectation of success. Because the shield merely terminates field lines in a region the applicant itself concedes is coupling-dominated, success is expected with a high degree of confidence.
5. No criticality or unexpected result. The '570 specification recites no threshold, no unexpected magnitude of improvement, and no result that would not be predicted from elementary electrostatics.

Conclusion on claim 1: obvious over JP-A-2014-21037 in view of US 2013/0255378 A1 (US 9,243,908 B2), further in view of the general shielding teaching of US 2006/0169042 A1. A single-reference rejection over US 9,243,908 B2 alone would likely fail only on the "electrode portion in the base substrate between substrate and wirings" limitation; that is why the second and third references are needed.

---

4. Dependent claim features: individually and collectively obvious

4.1 Driving portion / fixed driving electrode / fixed monitor electrode; and movable portion / fixed detection electrode variants

These are directly disclosed by Reference A as admitted in the '570 Background (frame-shaped vibration portion = driving portion; fixed driving electrode; movable portion inside; fixed detection electrode), and by Reference B (Application Examples 5–7, movable drive electrode + fixed drive electrodes; displacement portion + movable detection electrode + fixed detection electrode). Adding a fixed monitor electrode is standard gyro practice (the '570 specification itself treats monitoring as conventional) and is a design choice within the art. Obvious.

4.2 Third wiring portion (body wiring 51) at constant potential between the first and second wiring portions, functioning as a shield layer

Reference B discloses a "fixed potential wiring 40" alongside the drive and detection wirings, plus a constant-potential lid 50. Interposing a fixed-potential line between two lines of different noise character is the same shielding principle and is squarely a predictable arrangement. Reference C teaches the identical isolation principle. Obvious.

4.3 Monitor wiring between the third wiring portion and the detection wiring

Once the artisan has the three wiring groups and a shield line, ordering them to maximize separation between the strongest aggressor (drive) and the weakest victim (detection) is a routine engineering optimization — exactly the kind of "arrangement of parts" that is a design choice absent evidence of unexpected results. Notably, the '570 specification itself concedes the order is not critical, stating: "the arrangement order of the wiring portions 51, 52, 53, 54, 55, 56, and 57 is not particularly limited," and even discloses the inverse arrangement as an alternative ("the first detection wiring portion 56 and the second detection wiring portion 57 may be disposed between the first driving wiring portion 52 and the second driving wiring portion 53, and the first monitor wiring portion 54 and the second monitor wiring portion 55"). A limitation the applicant expressly describes as optional and reversible is weak § 103 material. Obvious.

4.4 Electrode portion electrically connected to the third wiring portion

Merely tying the shield to the existing constant-potential line instead of running a separate reference net. Design choice; obvious. The specification even concedes the alternative ("the electrode portion 6 may not be electrically connected to the body wiring portion 51").

4.5 Depression in the base substrate overlapping the wirings, with the electrode portion in the depression

Forming a recess in a glass substrate to house or clear a conductor is admitted base-substrate practice: Reference B has recesses 14 in the substrate surface; Reference A's silicon wiring structures are bonded to the substrate; the '570 specification itself describes a depression 21 as a clearance portion, and its own prior-art discussion attributes the wiring-on-glass arrangement to Reference A. Obvious.

4.6 D1 < D2 (shallower first depression under the drive wirings; deeper second depression under the monitor/detection wirings)

This is the dependent feature most likely to draw a "results-effective variable / optimization" or "design choice" rejection. The trade-off is explicit and predictable in the specification itself: bring the aggressor close to the shield (better termination of its field lines), and hold the victim further from the shield (avoid capacitively loading the sense node). A PHOSITA optimizing shield effectiveness versus sense-node capacitance would arrive at asymmetric spacing through routine experimentation, and would expect exactly the recited benefit. Absent evidence that the recited range (D1 = 0.1–1 µm; D2 = 1–10 µm) is critical (i.e., produces a result not obtainable outside the range), this is obvious. Note the specification's own hedge that "the separation distances D1 and D2 may be the same as each other," undermining any argument of criticality. Obvious (at minimum, obvious to try with a reasonable expectation of success).

4.7 Fixed portions of the first and second wiring portions shifted (offset) in the wiring extension direction

This is a pure geometric/design-choice limitation: it merely lengthens the dielectric path between adjacent bond regions that are not overlapped by the shield. The specification concedes the alternative ("the fixed portions … may be disposed to be adjacent to each other"). Obvious.

4.8 Constant-potential cover, with D3 (cover-to-drive wiring) < D4 (cover-to-monitor/detection wiring)

Reference B discloses a lid 50; Reference A's family and the '570 specification both treat a constant-potential cover (cover 3 at GND) as conventional. Applying the same asymmetric-spacing logic to the ceiling shield that the artisan would apply to the floor shield is the same solution applied to an analogous shielding surface, and thus obvious. Obvious.

---

5. The device / electronic apparatus / vehicle claims

These are aggregation claims whose only substantive addition is the sensor of the independent sensor claim:

• Physical quantity sensor device (mounted on a support substrate, IC on the opposite side, transfer-molded in epoxy, or ceramic package): each packaging option is described in the '570 specification as an alternative ("the physical quantity sensor 1 may be configured to be accommodated in a ceramic package"), which signals a routine design choice, not invention. Obvious.
• Electronic apparatus (notebook PC, mobile phone, digital camera, wearable): incorporating a gyro/acceleration sensor into these host devices was ubiquitous long before 2017. Obvious.
• Vehicle with an attitude control device controlling attitude based on the sensor's detection signal: vehicle attitude control using inertial sensors (ESC, ABS, airbags, ECU) is decades-old and is acknowledged in the '570 specification itself. Obvious.

§ 103 conclusion: every claim, independent and dependent, is likely obvious.

---

6. Secondary considerations (Graham factor 4)

I found no evidence of nexus-bearing secondary considerations:

• No unexpected results are asserted in the specification — only the expected electrostatic benefit.
• No licensing or industry-acclaim evidence surfaced.
• No litigation or IPR involving the '570 patent was found in my prior searches (stated then, and repeated here, as "not found," not "confirmed none").
• The patent is recorded as "Expired – Fee Related" with an adjusted expiration of 2038-07-22 (the source itself states this status is an assumption, not a legal conclusion). A lapsed patent is consistent with low commercial significance, though it is not proof of obviousness.

---

7. Bottom line

The patent's principal vulnerability is that its Background admits Reference A discloses the entire sensor architecture and the silicon-wiring-on-glass arrangement, while its characterization of Reference A frames the problem rather than the solution. That leaves the claimed "electrode portion" as the only point of novelty — and a reference-potential shield conductor interposed between noisy and sensitive parallel lines is one of the most familiar crosstalk remedies in the electrical arts, taught in the same MEMS-capacitive-sensor context by Reference B and by Reference C.

---

8. What I could not verify — do not treat as settled

1. The verbatim, numbered claim set of US 10,823,570 B2 was not available to me. All claim-level conclusions above are keyed to the SUMMARY-derived feature set. Verify against USPTO PatentCenter for application 15/927,554.
2. JP-A-2014-21037's actual text was not retrieved. I relied on the '570 Background's characterization, which is an applicant admission of what the applicant says the reference discloses. If the reference is broader or narrower than that characterization, the analysis moves.
3. The relationship between JP-A-2014-21037 and US 2013/0255378 A1 / US 9,243,908 B2 is unresolved and appears to be not the same document — the JP counterpart of US 2013/0255378 appears to be JP2013213728A. If, contrary to appearances, JP-A-2014-21037 is a divisional/member of that family, the two references should be treated as one reference, and the case becomes a single-reference-plus-general-knowledge obviousness (or a § 102 question) rather than a three-way combination.
4. I did not retrieve the face-of-patent "References Cited" list for the '570 patent. The examiner's actual cited art may include closer references (e.g., other Seiko Epson or Denso MEMS shielding filings) that would strengthen or simplify the rejection.
5. Reference C (US 2006/0169042 A1) was retrieved only in fragmentary form; I use it only for the general proposition that interposing a fixed-potential conductor between signal traces is conventional. Verify its full disclosure before relying on it in a written rejection.

Recommended next steps: (a) pull the printed claims from PatentCenter; (b) pull JP-A-2014-21037's full text and its English machine translation; (c) pull the "References Cited" page of the '570 patent; (d) if preparing an IPR or a defense, run a targeted search for shield-electrode-in-substrate references from Denso, Bosch, Panasonic, and Analog Devices filed 2013–2016.