US 10712220

I'll verify the identifiers and check for any litigation/docket activity before summarizing.

Let me run a few more targeted checks on the docket/litigation side.

US Patent 10,712,220 — Verification & Summary

Search coverage note: I checked Google Patents (including the full authoritative text), patent-family/assignee profiles, and general web sources, plus a targeted sweep for CAFC 2026 opinions/orders and district-court dockets citing this number. I do not have direct PACER/Docket Navigator or USPTO Patent Center API access from this session, so the litigation check is a negative finding based on open-web sources only — treat it as no evidence found, not as proof of no litigation.

Bibliographic data (US 10,712,220 B2)

Two caution points on the data as presented: the Patent Center/Google record lists "Prior art date" and "Priority date (assumption)" both as 2015-12-21, and the "expires 2037-06-13" figure is derived from the PCT-based 20-year term plus any adjustment — it is a database assumption, expressly so labeled by the source.

Abstract (as issued)

"Aiming to more easily perform calibration of a sensor output from a pressure sensor, the calibration being necessitated due to the occurrence of sedimentation, a resonance point measurement unit (122) measures a resonance point of a diaphragm (112) on the basis of the result obtained by performing measurement of a constant pressure using the pressure sensor while a power supply frequency is changed, a characteristic calculation unit (123) calculates, on the basis of the measured resonance point, an elastic modulus of the diaphragm (112) at the time of the measurement of the resonance point, and a correction unit (124) calculates a corrected sensor sensitivity resulting from correcting a sensor sensitivity of a sensor chip (101) on the basis of the elastic modulus calculated by the characteristic calculation unit (123)."

Plain-language claim overview

Claim 1 (the only independent claim)

A capacitive pressure sensor in which a pressure-receiving portion (a diaphragm) deflects under pressure and the pressure is derived from the resulting capacitance change. It must include four cooperating elements:

1. Pressure value output unit — converts the capacitance change into a pressure value using a set sensor sensitivity and outputs it.
2. Resonance point measurement unit — finds the resonance point of the diaphragm by holding pressure constant and sweeping the power supply frequency (the applied excitation frequency between movable and fixed electrodes), watching for the output to change beyond a threshold.
3. Characteristic calculation unit — from that measured resonance point, computes a physical characteristic of the pressure-receiving portion: either its thickness or its elastic modulus.
4. Correction unit — uses that physical characteristic to compute a corrected sensor sensitivity, and overwrites the sensor sensitivity set in the pressure value output unit with it.

Practical effect: the sensor self-recalibrates for the mass-loading/"sedimentation" effect (process deposits building up on the diaphragm in CVD/ALD film-forming tools), which stiffens/changes the diaphragm's mechanical behavior and therefore shifts sensitivity — without removing the sensor from the process tool. Supporting math in the spec: f_n = αtE^(1/2) (1); E′ = (f_n/α/t)² (2); S = β⁻¹/Et³ (3); S′ = β⁻¹/E′t³ (4).

Claim 2 (depends on claim 1)

Narrows element 3: the calculated physical characteristic is specifically the elastic modulus of the pressure-receiving portion. (This is the embodiment the specification and abstract emphasize.)

Claim 3 (depends on claim 1)

Adds a detection unit that compares the freshly measured resonance point against a reference value defined as the resonance point used when the previously computed corrected sensitivity was calculated, and tests whether the difference exceeds a specified value. Only if it does do the characteristic-calculating and sensitivity-updating steps actually run — i.e., a recalibration trigger / hysteresis threshold so the sensor doesn't recalibrate for trivial drift. Per the spec the reference value is held in a reference value storage unit (125) and is rewritten each time a correction is made; the as-manufactured resonance point can serve as the initial value.

Literal-text caution: the granted claim 3 text as it appears on the source page reads "the pleasured resonance point," which appears to be an OCR/transcription artifact for "measured." I am flagging it rather than silently correcting it; a certified copy should be checked if this wording matters to you.

Claim 4 (depends on claim 1)

Adds an alarm output unit that detects when the corrected sensor sensitivity falls outside a set allowable range and outputs an alarm. The spec's rationale: if sensitivity drops too far, the output dynamic range and measurement resolution degrade — so the alarm prompts replacement of the sensor chip rather than continued recalibration.

Dependency structure note: claims 3 and 4 each depend from claim 1 directly — they are not cumulative on claim 2.

Claim scope observations

• The independent claim is drafted at the functional/block level (units), not tied to a particular diaphragm material, electrode geometry, or algorithm — consistent with the spec's statement that "a manner of deriving the corrected sensor sensitivity may be set as appropriate in conformity with the characteristics of the sensor chip to be used and measurement environments," and that β may be a function of pressure where the pressure–capacitance relation is non-linear.
• The frequency-sweep-at-constant-pressure step is the point of novelty relative to the cited prior art, including the applicant's own earlier PTL 1 (JP 2014-109484 A, Azbil, "Capacitance type pressure sensor") and JP 2005-337924 A (Tokyo Electron), which address pressure-gauge structure/deposition but not resonance-based self-recalibration.

Prior art cited on the face of the patent (5 examiner/other citations plus 1 family citation): JP 2005337924 A; US 2007/0068266 A1 (Fujimori); JP 2013124947 A (Panasonic); JP 2014109484 A (Azbil); US 2014/0253219 A1 (Silicon Laboratories); US 2014/0306623 A1 and US 9,000,833 B2 (Silicon Laboratories); ATE 510201 T1 (Freescale). Foreign search report: ISR dated 2016-12-20 in PCT/JP2016/082926.

Litigation / PTAB / CAFC check

• No evidence found of any federal district court action, PTAB proceeding (IPR/PGR), reexamination, or Federal Circuit appeal involving US 10,712,220. Searches for CAFC 2026 opinions and orders returned unrelated matters (e.g., J.R. Simplot Co. v. McCain Foods USA, Nos. 24-1845, decided Feb. 6, 2026), and hits on the string "220" resolved to different patents (US 10,732,220 in Natera v. Genosity).
• The patent does appear in later Azbil-family filings — most notably JP 7662405 B2 (2025-04-15, Azbil, "Diaphragm Vacuum Gauge") and WO 2020/121576 A1 (Ulvac, "Vacuum gauge and pressure measurement system comprising vacuum gauge") — which is citation/family activity, not litigation.

Confidence: High on all bibliographic, claim, and abstract content (it is taken verbatim from the granted patent text). Low-to-moderate on the negative litigation finding, because it rests on open-web sources rather than a direct PACER, Docket Navigator, or USPTO Patent Center query; if you need a defensible clearance-style result I'd want to confirm directly in those systems.

I'll search for litigation involving US patent 10712220.

Let me search more targeted litigation databases.

Result: No litigation found for US Patent 10,712,220

I searched for district court litigation, CAFC appeals, PACER/CourtListener dockets, ITC actions, and PTAB proceedings referencing US Patent No. 10,712,220 ("Pressure sensor," Azbil Corporation). I found no known litigation involving this patent — no district court case, no ITC investigation, no PTAB post-grant proceeding, and no CAFC appeal.

What the patent is

• Patent: US 10,712,220 B2 — "Pressure sensor"
• Application: US 16/064,229 (filed 2016-11-07; national-stage of PCT/JP2016/082926)
• Priority: 2015-12-21 (JP 2015-248582 / JP 2016556616B2 family)
• Assignee: Azbil Corporation (Japan)
• Grant date: 2020-07-14; status Active, adjusted expiration 2037-06-13
• Subject matter: Capacitive (diaphragm) pressure sensor with self-calibration — resonance-point measurement of the diaphragm, calculation of a physical characteristic (elastic modulus or thickness), and correction of sensor sensitivity to compensate for sedimentation/buildup on the pressure-receiving surface (claims 1–4).

Basis for the "no litigation" conclusion

• Targeted searches on the patent number, on the assignee, and on litigation-tracking sites (Unified Patents portal, CourtListener, PACER-adjacent dockets, CAFC docket feeds) returned no case naming US 10,712,220.
• Search hits that superficially resembled the number were unrelated to this patent and were discarded as required (they do not reference US 10,712,220):
  • Bear Creek Technologies v. RCN et al. (E.D. Va. 2011) — a different "'722" patent, no relation to US 10,712,220.
  • EP 2 421 845 B1 (Abbott) — a chemical patent.
  • A UK Companies House gazette list containing the string "10712220" (a company number, not a patent).
  • A Hong Kong exchange filing URL fragment containing "10712220".
• The Google Patents page for US 10,712,220 lists citing documents and family members (JP 6556616B2, CN 108431571B, WO 2017110270A1, US 2019/0025146 A1) but no litigation events (no "Litigation" section), which is consistent with the searches.

Caveats

• Azbil is a Japanese industrial-automation company; it is not a frequent US patent plaintiff, and no US enforcement campaign involving this patent appears in public records. That is consistent with a clean result, but it is not affirmative proof of zero disputes (e.g., confidential settlements or foreign proceedings would not necessarily surface).
• I could not verify against a paid docket service (e.g., Lex Machina, Darts-IP, Westlaw) within this session; my conclusion rests on public search coverage. If you need a litigator-grade clearance, run the number through Lex Machina/Darts-IP and a PACER docket search (all district courts, "patent" nature of suit) to confirm.

Bottom line: To the best of the publicly available record as of April 26, 2026, there is no known litigation involving US 10,712,220 — no plaintiff/defendant pair, jurisdiction, case number, filing date, or outcome exists to report.

I'll verify against the canonical ODP block with a few targeted searches before reporting.

Proceedings overview

Total AIA trial proceedings on US 10712220: 0. The canonical USPTO Open Data Portal block in this prompt reports no AIA trial proceedings on file, and independent web searches (Google Patents litigation/related-proceedings data, PTAB decision repositories, and general docket searches for "Azbil" + IPR/PGR) surfaced nothing — no petition, no institution decision, no FWD, no settlement, no appeal. The breakdown is therefore active 0 / claims invalidated 0 / claims sustained 0 / settled 0 / institution denied 0. Bottom-line defensive posture: the patent is completely untested at the PTAB — it has been neither hardened nor wounded. All four issued claims stand exactly as examined, and a defendant has a full, unblemished menu of invalidity theories available with no § 315(e)(2) estoppel and no adverse PTAB record to argue around. That is a neutral-to-favorable position for a challenger: no petitioner before you has burned the good art, but you also get no free roadmap from a Board decision.

Proceedings

None to report. I am deliberately not constructing per-proceeding sections, because doing so would require inventing proceeding numbers, panels, and dispositions. For completeness, here is what was checked and what was found:

Two things worth flagging as non-PTAB context (not proceedings on this patent):

1. Co-pending family members exist but are separate patents. The same priority filing (JP 2015-248582, 2015-12-21) produced JP6556616B2, CN108431571B, and WO2017110270A1. If you are facing a global assertion campaign from Azbil, any invalidity or opposition work done on those counterparts has no estoppel or preclusion effect on the '220 patent, and vice versa — but the same prior art will generally read across, so art developed for one is reusable here.
2. No defensive aggregator in the chain. No Unified Patents, RPX, or similar entity appears anywhere in the prosecution, assignment, or citation record for this patent. The 2018 assignment (Azbil Corporation, recorded 2018-06-20, Reel/Frame 046144/0670) reflects a straightforward inventor-to-employer transfer by Hidenobu Tochigi and Takuya Ishihara — no third-party interest.

Strategic summary

Claim status. Because there has been no PTAB trial, all of claims 1–4 are UNTESTED and intact. Nothing is canceled; nothing has been adjudicated patentable over prior art. For precision: claim 1 is the sole independent claim (pressure value output unit + resonance point measurement unit + characteristic calculation unit + correction unit, with the physical characteristic being "a thickness or an elastic modulus"); claim 2 narrows to the elastic modulus specifically; claim 3 adds the detection unit and the reference-value/specified-value trigger; claim 4 adds the alarm output unit. Every one of those limitations is live and was never construed by the Board. There is no surviving-claims list to report because nothing was lost.

Estoppel landscape. With no IPR or PGR petition ever filed and no FWD entered, § 315(e)(2) estoppel is a non-issue. You are not restricted to grounds that "a petitioner would have raised" — you may file a first IPR on any § 102/§ 103 ground and, critically, you also retain the option of running § 112 challenges in district court or an ITC action without a PTAB estoppel shadow. The only constraints on a new petitioner are the ordinary ones: the § 315(b) one-year bar measured from service of a complaint on you, § 311(b)'s limitation to patents and printed publications, and § 325(d)/Advanced Bionics discretion if you rely on art the examiner already considered. On that last point, the examination record here is unusually thin — the citations of record are only five references (JP2005337924A, US20070068266A1, JP2013124947A, JP2014109484A, US20140253219A1) plus the Silicon Laboratories MEMS-transduction family (e.g., US9000833B2), so there is meaningful room for non-cumulative art.

Pattern signals. No petitioner has filed once, let alone repeatedly. No PTAB appeal pattern exists because there is no PTAB case. The absence of any IPR against a 2015-priority patent that has been on the books since 2020-07-14 — and against an assignee (Azbil) that is a well-known process-instrumentation supplier with the resources to litigate — suggests one of two things: the patent is asserted narrowly, if at all, or alleged infringers have assessed the claim as difficult to invalidate cheaply. The absence is a signal worth weighing against your own art search. Notably, all three patent documents citing '220 are later Azbil-family filings, which is a sign the owner is building a portfolio around it rather than that third parties are attacking it.

Recommended next steps

• No PTAB FWD to link or quote. Because no trial was ever instituted, there is no disposition to cite, no IPR number to reference, and no Board reasoning to leverage. Do not let opposing counsel characterize the silence as a merits win or a loss — it is simply an empty record.
• Where to verify, and how to re-check later: run the patent through USPTO Patent Trial and Appeal Board E2E / PTAB Center and the USPTO ODP API on a rolling basis, and set a CourtListener docket alert for the patent number and for Azbil as a party. A petition can be filed the day you are served in district court; the PTAB would then post a Notice of Filing Date Accorded and, if instituted, a statutory FWD due within 12 months of institution (extendable to 18 months for good cause under 35 U.S.C. § 316(a)(11)) — but none of those clocks are running today.
• Practical posturing: if you have been served with a demand letter or complaint, your § 315(b) one-year window is the binding constraint, and it is running from the date of service. The good news is that with zero prior petitions you have an entirely clean slate: no estoppel, no Fintiv duplication with a prior Board proceeding, and no risk of a second-petition joinder problem under § 315(c). Build the invalidity case on the merits as if the patent had never been challenged — because it has not.
• Caveat on confidence: I found no PTAB activity through the structured ODP block and my searches, and I am not aware of any proceeding that would be missing from those sources. If Azbil or a petitioner filed something very recently that post-dates the ODP ingest, I cannot confirm it here; verify against PTAB E2E before relying on the zero-count for a filing-deadline decision.

I'll search for the assignment record for this patent.

Let me verify the assignment record details and check for any post-issuance transfers.

Let me try to pin down the correspondent of record for the lone assignment.

Ownership & Assignment Analysis — US 10,712,220 B2 ("Pressure sensor")

Verification note: The authoritative full-text I was given (Google Patents legal events for US 10,712,220) shows only one recorded assignment in the entire chain — reel 046144/0670 — plus routine USPTO administrative events (FEPP, STPP, STCF, MAFP). Cross-checks against Azbil's patent portfolio listings confirm no successor owner. I could not retrieve the Assignment Center's per-record correspondent field in this session (the change-of-record text embedded in the Google Patents event log omits it); I note that gap explicitly rather than guess at an attorney name.

Inventors

Both inventors appear together on a series of Azbil diaphragm-gauge filings (e.g., US 10,161,821; US 10,520,385; EP 2921837 family), i.e., they are members of Azbil's in-house pressure-sensor R&D team, not one-off filers.

Unusual-pattern check: No departure/sale pattern observed. There is no evidence either inventor left Azbil around the filing date, and no sign of a later portfolio fire-sale (see timeline — the patent stays with Azbil through at least 2023). The two-inventor, single-employer, single-assignment pattern is the opposite of the "inventors scatter, rights get flipped" precursor to a fire-sale.

Original assignee

• Entity on the issued patent: AZBIL CORPORATION (Japan). Legal-event text: "AZBIL CORPORATION, JAPAN — ASSIGNMENT OF ASSIGNORS INTEREST" (reel 046144/0670).
• Identity: Azbil Corporation (formerly Yamatake Corporation, renamed 2012), Tokyo; listed on the Tokyo Stock Exchange; established 1949. Business: building automation, advanced (industrial) automation, and life automation — it makes measurement and control equipment including capacitive diaphragm pressure sensors / vacuum gauges — exactly the subject matter of this patent.
• Ships a product embodying the claims: Yes, highly likely. Azbil is an operating sensor manufacturer (its "Advanced Automation" segment sells pressure gauges/vacuum gauges). The patent is an internal R&D self-calibration feature for its diaphragm vacuum gauges, not a standalone licensing asset. (I state this as well-supported but not product-catalog-verified to a specific model number in this session.)
• Current status: Operating, active. No bankruptcy, dissolution, or acquisition event appears in the record; maintenance fee paid (4th year, large entity). Google Patents status: Active, adjusted expiration 2037-06-13.

Assignment timeline

2018-06-18 / 2018-06-19 (executed) / recorded 2018-06-20 — Reel 046144 / Frame 0670

• Conveyance: Assignment (inventor-to-employer)
• Assignors: Hidenobu Tochigi; Takuya Ishihara (individually, signing 2018-06-18 and 2018-06-19 respectively)
• Assignee: Azbil Corporation (Japan)
• Correspondent: Not captured in this session — the reel's correspondent field was not returned by the sources available. Flagged for confirmation via Assignment Center. (Note: this is the only assignment record, so there is no "repeat correspondent" question to answer — see Signal 3.)
• Context: Standard employment/confirmatory assignment perfecting title in the employer for the PCT national-stage entry (PCT/JP2016/082926, filed 2016-11-07; US app 16/064,229). Recorded after the 2018-01-24 US publication date, as is routine for national-stage filings.

Post-issuance: None. No assignment, security agreement, merger, change of name, license, release, or correction is recorded after the 2018-06-20 entry. The subsequent legal events are purely administrative (issue-fee/allowance notices 2020; patent grant 2020-06-24; maintenance fee 2023-12-27).

Because the Assignment Center shows only the original inventor assignment and nothing afterward, the plain reading is: Azbil Corporation still owns US 10,712,220 outright.

Timeline diagram

timeline
    title Ownership of US 10712220
    2015 : Priority filing JP 2015-248582
    2016 : PCT application filed
    2018 : Inventors assign to Azbil Corp
         : Reel 046144 frame 0670
    2020 : US patent issued
    2023 : Maintenance fee paid

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. The only assignee in the chain is Azbil Corporation, an operating manufacturer (reel 046144/0670, recorded 2018-06-20). No "IP / Holdings / Ventures / Licensing" suffix, no registered-agent-service address, no single-purpose LLC.

2. Known asserter in the chain — Not present. Neither current nor prior assignee matches any entity on the Acacia / Marathon / IV / Wi-LAN / Conversant / Pendrell / Round Rock etc. lists, or any Unified Patents / RPX high-frequency-plaintiff directory. Sole assignee = Azbil Corporation.

3. Repeat correspondent across the chain — Not present / N/A. There is only one recorded assignment, so recurrence cannot exist by definition. (Correspondent name itself not retrieved this session — a data gap, not a signal.)

4. Cascading transfers — Not present. No chained LLC transfers at any time; a single link (inventors → Azbil) in the entire history.

5. Pre-litigation transfer — Not present. No litigation exists involving this patent (consistent with the prior litigation section), so no assignment can be timed against a suit. The sole assignment predates issuance by ~2 years and is dated to the national-stage perfection timeline, not to an assertion date.

6. Bankruptcy fire-sale — Not present. No Chapter 7/11 involvement by Azbil; assignee remains an active listed operating company (maintenance fee paid 2023).

7. Privateering — Not present. No transfer to any third party, NPE or otherwise, that could constitute asserting on Azbil's behalf.

8. Defensive aggregator (anti-NPE) — Not present. Chain does not terminate at RPX, AST, LOT, Unified, or OIN; it terminates at the original operating manufacturer.

Summary: 0 of 8 NPE signals present.

Verdict

Insufficient data — the record contains only the original inventor-to-employer assignment (reel 046144/0670, executed 2018-06-18/19, recorded 2018-06-20) and no post-issuance transfers, so there is nothing in the assignment chain to characterize as an NPE pattern. What the record does establish affirmatively is a stable operating-company ownership: Azbil Corporation — the entity that shipped the accused-embodying diaphragm vacuum gauges — has held the patent continuously since 2018 and paid maintenance fees through 2023, with no licensing vehicle ever interposed. If I had to choose a substantive label it would be "Operating-company ownership, non-asserting," but because Azbil is not known to be asserting this patent against competitors, the option-set's closest fit is Insufficient data (only the original assignment).

Verify at: USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (query patent 10712220 / app 16/064,229) and the indexed mirror https://assignment.uspto.gov/patent/index.html. The expected single hit is reel 046144/0670, assignee AZBIL CORPORATION.

Data gaps to close if a litigator-grade ownership opinion is required: (1) retrieve the correspondent of record on reel 046144/0670 to confirm it is Azbil's ordinary prosecution counsel (not an asserter-side filer); (2) confirm no unrecorded license or covenant exists via Azbil's SEC-equivalent (Tokyo Stock Exchange / EDINET) disclosures — noting Azbil files with the TSE, not the SEC, so 10-K/8-K coverage does not apply here.

I'll analyze the prior art cited against US 10,712,220. Let me gather details on each cited reference.

Let me gather details on the remaining cited references.

Prior-Art Analysis — US 10,712,220 B2 ("Pressure sensor," Azbil Corp.)

Scope / method note. I confirmed the target is US 10,712,220 B2 (App. US 16/064,229, priority 2015-12-21, granted 2020-07-14) — not any near-number such as US 10,732,220 or US 10,912,321. The reference set below is the citation list printed on the face of that patent (examiner/other citations plus the family citation), as reproduced in the authoritative document text. Web verification was run on the principal references. Two references (JP 2013124947 A and ATE 510201 T1) could not be pulled in full within this session; those two entries are flagged as lower-confidence and are described only to the level supported by the citation-of-record and general knowledge, not asserted as verified.

Legal-standard caveat up front: "Anticipation" under 35 U.S.C. § 102 requires a single reference to disclose every element of a claim, arranged as claimed. As shown below, none of the cited references anticipates any claim of US 10,712,220. They are properly characterized as § 103 (obviousness) references or as background art. I state that explicitly rather than overclaiming a § 102 rejection.

The citation list (citation of record)

ISR of record: International Search Report dated 2016-12-20 in PCT/JP2016/082926. Family: WO 2017110270 A1, CN 108431571 B, JP 6556616 B2, US 2019/0025146 A1.

Reference-by-reference analysis

1. JP 2005337924 A — Tokyo Electron Ltd.

Full citation: JP 2005337924 A, "Pressure gauge manufacturing method, gas processing apparatus manufacturing method, pressure gauge, and gas processing apparatus," appl. JP2004158047A, filed 2004-05-27, published 2005-12-08; inventors Yoneda/Kasai/Kubodera. (Granted as JP 4678752 B2.)
URL: https://patents.google.com/patent/JP2005337924A/en

Brief description. Diaphragm-type (capacitance manometer) pressure gauge for film-forming chambers. It expressly recognizes that reaction gas deposits a thin film on the diaphragm, and states that this "changes the deflection characteristic" / "the elastic modulus of the diaphragm also changes substantially," producing both a zero-point offset (compressive/residual stress) and a sensitivity/measurement error. The disclosed remedy is directed at the gauge's structure/design to reduce deposition, and to manufacturing methods — not at resonance-based self-recalibration.

Claims potentially affected: None anticipated. It does not disclose element (c) (resonance-point measurement by constant-pressure frequency sweep), element (d) (calculation of thickness/elastic modulus from a measured resonance point), or element (e) (computing a corrected sensitivity and overwriting the set value). It is significant as a § 103/background reference because it supplies the motivation and the physics (deposit → changed elastic modulus → changed sensitivity) that the challenged patent exploits.

2. US 2007/0068266 A1 — Fujimori et al. (Assignee n/a on face; JP-origin)

Full citation: US 2007/0068266 A1, "Sensor and sensor module," pub. 2007-03-29, filed 2006-07-26, priority 2005-09-26; inventor Tsukasa Fujimori et al.
URLs: https://pubchem.ncbi.nlm.nih.gov/patent/US-2007068266-A1 ; https://www.patents-review.com/a/20070068266-sensor-sensor-module.html

Brief description. A capacitive sensor (TPMS-oriented) whose capacitor has a frequency–capacitance characteristic with a resonant frequency. In normal measurement the capacitance is read at a control-signal frequency "much higher or much lower" than resonance; in self-diagnosis the capacitance is read at a frequency "equal or close to" the resonant frequency. The spec teaches sweeping the frequency of the measurement pulse to obtain the frequency spectrum of the movable/diaphragm capacitance and thereby identify the resonant frequency, then determining whether the resonant frequency has varied (indicating a failed/abnormal movable electrode), and comparing the measured capacitance against a stored value.

Claims potentially affected — this is the closest reference to the novelty point of claim 1.

• Claim 1: Discloses elements partially: a capacitive sensor with a physical quantity (pressure) capacitance change; and element (c), resonance-point measurement by frequency sweep. But it does not disclose the constant-pressure sweep limitation, the characteristic calculation unit (d) computing thickness or elastic modulus from the resonance point, or the correction unit (e) computing/updating a corrected sensor sensitivity. Reasonable examiner position: § 103 combination — Fujimori (resonance measurement) in view of Tokyo Electron ('922) (deposit changes modulus/sensitivity). Not anticipation of claim 1, and therefore not of claims 2–4 either (all depend from claim 1).
• Claims 2, 3, 4: No independent § 102 bearing; only via the combination above.

3. JP 2007086002 A — Hitachi, Ltd.

Full citation: JP 2007086002 A, "Sensors and sensor modules" / "Sensor and sensor module," appl. JP2005277895A, filed 2005-09-26, published 2007-04-05; Hitachi. (Granted 2013-06-26.) This is the JP counterpart/sibling of US 2007/0068266 A1 (same priority 2005-09-26).
URL: https://patents.google.com/patent/JP2007086002A/en

Brief description. The same subject matter as Fujimori, plus an explicit discussion of prior art self-diagnosis using resonance phenomena: it cites JP-A-2003-121457 for "performing self-diagnosis by measurement using a frequency higher than the resonant frequency, avoiding abrupt changes in sensor values caused by resonance," and cites the electrostatic-force simulation approach (JP-A-5-322921). It characterizes failures of the movable electrode (pinholes, adhered foreign matter) that produce only slight measurement changes.

Claims potentially affected: None anticipated. Same analysis as Reference 2 — disclosed for element (c)/the concept of monitoring resonance in a capacitive pressure sensor; silent on (d) thickness/elastic-modulus calculation and on (e) sensitivity correction. § 103 material only.

4. JP 2013124947 A — Panasonic Corp.

Full citation: JP 2013124947 A, "Semiconductor pressure sensor," appl. JP2011-273455, priority 2011-12-15, published 2013-06-24; Panasonic Corp.
Confidence: Lower — not retrieved in full this session. Title/date/assignee are as cited of record on the patent face.

Brief description (as cited): a semiconductor (piezoresistive/capacitive-type) pressure sensor. On the citation of record alone it appears to be a structural/semiconductor pressure-sensor reference rather than one addressing resonance-based recalibration.

Claims potentially affected: None apparent. No basis to assert § 102 against any of claims 1–4. If you need a firm statement, the JP text should be pulled directly (JPO/J-PlatPat).

5. JP 2014109484 A — Azbil Corp. (the patent's own PTL 1)

Full citation: JP 2014109484 A, "Capacitance type pressure sensor," appl. JP2012263689, priority 2012-11-30, published 2014-06-12; Azbil Corp.
URL: https://patents.google.com/patent/JP2014109484A/en

Brief description. Applicant's own earlier capacitive diaphragm pressure sensor. Its whole premise is that deposition of the measured medium's components on the diaphragm causes zero-point shift (compressive/tensile internal stress bending the diaphragm), that a filter cannot eliminate deposition completely, and the flexure grows with the amount of deposit. Its solution is structural: a cover plate with a pressure-introduction hole positioned at 50–70 % of the diaphragm radius to suppress the zero shift. Related Azbil structural art includes US 8,656,787 B2 (baffle/rigidity profiling) and US 2015/0040674 A1 (baffle structure).

Claims potentially affected: None anticipated. It discloses the problem (deposit-driven zero shift) but not the claimed solution (resonance-based characteristic calculation + sensitivity update). It is the "admitted prior art"/background that frames the invention and would combine with a resonance reference under § 103.

6. US 2014/0253219 A1 / 7. US 2014/0306623 A1 / 8. US 9,000,833 B2 — Silicon Laboratories Inc.

Full citations:

• US 2014/0253219 A1, "Compensation of changes in MEMS capacitive transduction," pub. 2014-09-11, filed 2013-03-06 (Ser. 13/786,686); inventors Caffee & Quevy; granted as US 9,000,833 B2 (2015-04-07).
• US 2014/0306623 A1, "Integrated MEMS design for manufacturing," pub. 2014-10-16, filed 2013-03-06, CIP of the '219 application (incorporated by reference).
• US 9,000,833 B2 is the grant of US 2014/0253219 A1.
  URLs: https://patents.google.com/patent/US9000833 ; https://www.freepatentsonline.com/y2014/0253219.html ; https://patentimages.storage.googleapis.com/c5/6d/86/f872f07db7cbff/US20140306623A1.pdf

Brief description. MEMS device compensation. Explicitly: a change in electrode capacitance affects equivalent spring stiffness (electrostatic pulling), "which affects the resonant frequency of the MEMS device." The method compares an indicator of a resonant frequency of the MEMS device to a predetermined resonant frequency and generates a compensation signal applied to signal-processing logic to compensate the output for a strain-induced resonant-frequency shift; alternatively, strain-sensitive vs. strain-insensitive electrodes/capacitors are compared, and the compensation is applied via a bias voltage or in signal processing. Compensation can be repeated periodically in a "compensation mode" and updated to track aging.

Claims potentially affected: The most substantive § 103 reference against the calculation/correction elements. It teaches (i) using a resonant frequency of the moving MEMS structure as a measured diagnostic variable, and (ii) deriving a compensation and applying it to correct the sensor's output, including repeating/updating over time. This maps toward elements (c)→(e). But it does not disclose calculating a thickness or elastic modulus as an explicit intermediate physical characteristic (d), nor a capacitive diaphragm pressure sensor with a power-supply-frequency sweep at constant pressure as in (c), nor overwriting a "sensor sensitivity" set in a pressure value output unit (e).

• Claims 1–4: No single-reference anticipation. Reasonable § 103 combination: Silicon Labs ('219/'833/'623) + Fujimori ('266/Hitachi '002) + Tokyo Electron ('924)/Azbil ('484) — resonance measurement of the moving element + deposit-changes-modulus motivation + sensitivity-compensation update.

9. ATE 510201 T1 — Freescale Semiconductor, Inc.

Full citation: ATE 510201 T1 (Austrian/EP-type translation of an EP grant), "Improvements in capacitive sensor diagnostics," pub. 2011-06-15, priority 2007-04-06; Freescale Semiconductor Inc. (Listed in the "Family Cites Families" set, i.e., cited by a family member rather than an examiner citation.)
Confidence: Lower — not retrieved in full this session.

Brief description (as cited): directed to diagnostics of capacitive sensors — i.e., detecting/self-testing a capacitive sensor's health. It does not, on the record available, address resonance-based physical-characteristic calculation or sensitivity recalibration of a diaphragm pressure sensor.

Claims potentially affected: None apparent. Peripheral § 103/background reference at best; not a § 102 reference to claims 1–4.

Why the prior art does not anticipate claims 1–4 (element map)

Claim 1 requires all of: (a) capacitive pressure sensor with a movable pressure-receiving portion; (b) a pressure value output unit converting the capacitance change to pressure per a set sensor sensitivity; (c) a resonance point measurement unit measuring the diaphragm resonance point from constant-pressure measurement while the power-supply frequency is swept; (d) a characteristic calculation unit computing thickness or elastic modulus from that resonance point; and (e) a correction unit computing a corrected sensor sensitivity and overwriting the set value.

No single cited reference bridges (d) and (e) as claimed. Each top reference discloses at most one of the two novelty-bearing elements, so the reference set supports an obviousness challenge, not anticipation.

Related (non-cited) art worth flagging for completeness

These are not on the citation list but are Azbil sibling filings and appear in the "Similar Documents"/family record. They are highly relevant to validity/scope and to the "same-problem, same-assignee" picture:

• US 10,161,821 B2 (Ishihara et al., Azbil) — "Pressure sensor" — adjusts the pressure sensor's temperature characteristic by varying sensor-chip temperature, without removing the device; discusses deposit-driven changes in mechanical/thermal characteristics of the diaphragm.
• US 2017/0248487 A1 (Ishihara, Soeda, Sekine, Tochigi; Azbil) — "Pressure sensor state detection method and system" — detects accumulation on the diaphragm early by obtaining a sensor characteristic (via temperature change) and comparing to a reference characteristic — note this is the reference-value concept that reappears in claim 3 of the '220 patent.
• WO 2020/121576 A1 (Ulvac) and JP 7662405 B2 (Azbil, 2025) cite the '220 family.

If you are building a § 103 posture, the pairing of US 2017/0248487 (reference-value comparison for deposit detection, same assignee) with US 2007/0068266 / US 9,000,833 (resonance measurement + compensation) is the most probative combination to evaluate.

Bottom line

• No cited reference anticipates US 10,712,220 under § 102 — none discloses the full (c)+(d)+(e) combination, and the two elements that carry novelty, calculating thickness/elastic modulus from a swept-frequency resonance measurement (d) and computing and overwriting a corrected sensor sensitivity (e), are not jointly present in any single reference.
• Closest references: US 2007/0068266 A1 (Fujimori) and its Hitachi sibling JP 2007086002 A (resonance-point identification by frequency sweep, for self-diagnosis); US 9,000,833 B2 / US 2014/0253219 A1 / US 2014/0306623 A1 (Silicon Laboratories) (compare resonant frequency to a predetermined value and apply a compensation).
• Problem-statement/background references: JP 2005337924 A (Tokyo Electron) and JP 2014109484 A (Azbil, PTL 1) — deposit → changed diaphragm elastic modulus/sensitivity.
• Peripheral: JP 2013124947 A (Panasonic) and ATE 510201 T1 (Freescale) — no § 102 bearing identified; both are lower-confidence here because the full texts were not retrieved this session.
• Practical takeaway: the defensible challenge to claims 1–4 is an obviousness challenge built from the resonance-measurement references combined with the deposition/compensation references — not a § 102 anticipation.

Confidence: High on the reference identities, dates, and the content of references 1, 2, 3, 5, 6, 7, 8 (verified from the sources cited above). Lower on references 4 and 9 (JP 2013124947 A, ATE 510201 T1), which I could not pull in full within this session — treat those two descriptions as provisional and confirm against J-PlatPat/Espacenet before relying on them.

I'll ground the obviousness analysis in the actual prior-art references cited on the face of US 10,712,220. Let me pull details on the most technically relevant ones.

Obviousness Analysis — US 10,712,220 B2 under 35 U.S.C. § 103

Scope note. This analysis builds on the claim-by-claim overview and the prior-art listing already produced (see the "Claim scope observations" and "Prior art cited on the face of the patent" sections above) and does not repeat them. Where I rely on a reference's disclosure, the passage or snippet is quoted and linked. All references discussed below are of-record citations on the face of US 10,712,220 and all predate the 2015-12-21 priority date, so each qualifies as prior art under § 102(a)(1)/(b) / pre-AIA § 102(b). I flag where my read of a reference rests on its abstract/ISR snippets rather than full-text verification.

Caveat up front: the examiner allowed these claims over these references, and the record shows a mix of examiner-cited and third-party-cited art (the "*"/"†" provenance flags on the source page). Nothing below disturbs the presumption of validity — it is a validity-challenge-style, post-hoc combination analysis, which is a different exercise from the examiner's claim-by-claim prosecution. Treat this as an assessment of exposure, not a prediction of litigation outcome.

1. Governing framework and level of ordinary skill

• Standard: Graham v. John Deere, 383 U.S. 1 (1966) — scope/content of claims, differences over prior art, PHOSITA level, secondary considerations; plus KSR Int'l v. Teleflex, 550 U.S. 398 (2007) (predictable combinations, "obvious to try," finite number of identified solutions).
• Claim construction: the independent claim is drafted as a set of functional "units" (pressure value output unit, resonance point measurement unit, characteristic calculation unit, correction unit). None uses the word "means," so a Williamson v. Citrix § 112(f) presumption is weak; the terms are best read as the corresponding algorithms/structures disclosed in the spec (capacitance-to-pressure conversion per set sensitivity; frequency sweep at constant pressure; f–E–t computation; sensitivity overwrite).
• Level of ordinary skill (proposed): a bachelor's degree in electrical engineering, mechanical engineering, or applied physics, plus 2–5 years of experience in capacitive/MEMS pressure-sensor or diaphragm-vacuum-gauge design, including familiarity with diaphragm resonance/mechanical-property relationships and closed-loop sensor calibration. Every reference below is directed to that same skill set.

2. The references and what each actually teaches

The pivot. Fujimori already supplies every mechanical step of claim 1 except the sensitivity-recalculation: (i) a capacitive sensor with a movable electrode; (ii) measurement of a constant condition while the excitation/power-supply frequency is swept; (iii) identification of a resonance point from the output change; and (iv) comparison of the resonance point against a pre-measured value to decide sensor state. Silicon Laboratories supplies the missing use: convert a measured resonant-frequency shift into a correction applied to the sensor's output, and do so periodically/automatically. Azbil's PTL 1 supplies the problem and the field: deposition on the diaphragm of a capacitive pressure sensor degrades measurement.

3. Claim 1 — proposed grounds of rejection

Ground 1 (primary): Fujimori + Silicon Labs '833/'219 + Azbil PTL 1

Claim-element mapping:

Motivation to combine.

1. Same field, same problem. All three are capacitive pressure/diaphragm sensors in the vacuum-metrology / semiconductor-film-forming environment (Azbil's PTL 1 and the '220 background both target ALD/CVD film-forming tools). Fujimori is expressly a "sensor and sensor module" with capacitive sensing; Silicon Labs is expressly about capacitive transduction.
2. The problem is the applicant's own admitted starting point. PTL 1 (Azbil) already states that deposition on the diaphragm shifts/degraded measurement. The '220 specification itself acknowledges PTL 1 as the base art. A PHOSITA seeking to avoid "demounting the pressure sensor from the apparatus" (spec, Background) would look to in-situ resonance characterization.
3. Fujimori teaches the exact measurement technique claimed. Frequency sweep of the sensor drive signal to locate the resonance point, at a constant (0-point) condition, is not merely analogous — it is the same step, in the same class of device. This is a use of a known technique (resonance-frequency sweep) to improve a similar device (capacitive pressure sensor) in the same way (KSR rationales (A), (C), (D)).
4. Silicon Labs teaches the exact corrective use and the automation. "Compar[ing] an indicator of a resonant frequency… to a predetermined resonant frequency" and then generating a compensation that is applied to the output — plus the express teaching to repeat the measurement periodically and update the compensation "to compensate for changes… due to… aging." That is the claim's "correction unit … updating the sensor sensitivity" in all but name; the only delta is where the number is applied (bias voltage vs. the sensitivity constant in the pressure-value conversion), an implementation choice with no patentable weight (In re Carlson; KSR).
5. Reasonable expectation of success. f = αt√E and deflection ∝ p·a⁴/(E·t³) are textbook diaphragm relations; the spec's own formulas (1)–(4) are the standard relations, not novel physics. Deriving E (or t) from a measured resonance and then rescaling sensitivity is the mechanical inverse of a known forward relation — a predictable, finite-solution step.
6. KSR "obvious to try." Given (a) a known need to detect deposition-induced sensitivity change, (b) a known in-situ proxy (resonance), and (c) a known way to convert that proxy into an output correction, the set of identified, predictable solutions is small; the claimed arrangement is one of them.

Ground 2 (alternative/independent): Fujimori + Silicon Labs '623

The '623 application (Silicon Labs CIP) adds an explicit "device parameter … geometric dimension of the MEMS device" monitor and a control circuit that supplies a compensating signal — i.e., a discrete parameter (including a dimension related to thickness) is measured and used to compensate. That maps more tightly onto claim 1's "thickness" alternative and onto the "characteristic calculation unit." Combined with Fujimori's frequency-sweep resonance measurement and Azbil's pressure-sensor context, this ground reaches claim 1 without the '833 patent's strain-specific framing.

Ground 3 (secondary): Azbil PTL 1 as base + Fujimori + Silicon Labs

Framed as an improvement of the admitted prior art: take the deposition-prone capacitive diaphragm sensor of PTL 1 and add Fujimori's resonance self-diagnosis, then Silicon Labs' resonant-frequency-based output compensation. This is the cleanest KSR "improve a known device ready for improvement" formulation, because PTL 1 itself identifies the defect (deposition → zero-point/sensitivity error) that the combination cures.

4. Claim 2 — elastic modulus as the physical characteristic

Claim 2 depends from claim 1 and merely chooses elastic modulus as the computed quantity. This is highly vulnerable: once claim 1's resonance-measurement-and-correct step is established, selecting which mechanical property to invert the resonance relation for is a design choice between two disclosed alternatives ("a thickness or an elastic modulus," as claim 1 itself recites). Fujimori and Silicon Labs both describe resonance as a function of the movable structure's stiffness — the very quantity E governs. Choosing E over t (or computing both) is not a patentable distinction; it is the routine selection of one of a finite number of identified, predictable parameters (KSR; In re Larson; In re Kuhle — the limitation is a recognized design expedient).

5. Claim 3 — detection unit, reference value, specified value (recalibration trigger)

Claim 3 adds: (a) a reference value = the resonance point at the time of the previous correction; (b) a detection unit testing whether |measured − reference| exceeds a specified value; and (c) only then recalculating and rewriting sensitivity.

This is closely met by the art of record:

• Fujimori discloses precisely a reference/threshold comparison on the resonance point: the resonant frequency is "pre-measur[ed]," and a self-diagnosis decides "whether there is a variation in the resonant frequency"; claim 10 runs the resonance measurement only "when a result of the first measurement is beyond a standard range or a deviation thereof is beyond a standard range." That is a reference value + specified-value test + conditional second operation.
• Silicon Labs '833/'219 discloses updating the reference over time and periodically refreshing the compensation — i.e., the "resonance point at the time of the previous correction" being rolled forward. The '219 specification states the parameter is measured at wafer probe, after packaging, after board assembly, and then repeatedly during compensation mode with the compensation updated accordingly.
• Silicon Labs '623 discloses setting a predetermined range and acting (rejecting the die) only when the parameter falls outside it.

Motivation. The rationale for a hysteresis/threshold gate — don't recalibrate on trivial drift — is exactly the rationale Silicon Labs gives for periodic rather than continuous re-compensation, and Fujimori gives for the two-tier (normal vs. resonance) measurement. A PHOSITA automating the Ground 1 combination would naturally (i) store the last-used resonance as the baseline and (ii) gate recalibration on a threshold, because both behaviors are disclosed as beneficial in the same art. Claim 3 is, on this record, the weakest of the four claims.

Note on claim text: the granted claim 3 recites "the pleasured resonance point," which appears to be a transcription/OCR artifact for "measured" (flagged already in the prior claim-overview section). Nothing in this analysis turns on it; a certified copy should be verified if the exact wording matters.

6. Claim 4 — alarm output unit on out-of-range corrected sensitivity

Claim 4 adds an alarm when the corrected sensitivity falls outside an "allowable range." The strongest teaching of record is Silicon Labs '623: reject/flag when the sensed parameter "falls outside of a predetermined range." Read together with claim 3's threshold comparator (already in Fujimori), using an out-of-range flag to trigger an operator alarm is the conventional automation of a comparison the art already performs. This ground is moderate-strength and more dependent on "well-known alarm/annunciation" reasoning than Grounds 1–3.

Weaker aspect to watch: claim 4 ties the alarm specifically to the corrected sensitivity (a computed downstream quantity) exceeding a range, rather than to a raw parameter. A patentee could argue the art only alarms on raw parameters. Counter: the corrected sensitivity is a deterministic function of the measured resonance and the known constants (formulas (2)–(4)), so testing it is mathematically equivalent to testing the resonance against a corresponding threshold — a mere change in the form of the comparison, which is not inventive (In re Rose; Gardner v. TEC). I rate the § 103 case on claim 4 as moderate, versus strong on claims 1–3.

Corroborating background art (not a ground by itself): the same field recognizes automatic recalibration of capacitance diaphragm gauges (e.g., US 8,965,725 B2, "Automatic calibration adjustment of capacitance diaphragm gauges…," listed among "Similar Documents" on the source page). I have not independently verified its full disclosure or dates, so I use it only as context for the state of the art.

7. The best non-obviousness arguments (and why they likely fail)

Potential secondary considerations: the record shows no evidence of commercial success, licensing, or long-felt-need attributable to the claimed combination (the Azbil sapphire-CDG commercial story, per the 2024 Azbil literature and the 2016 J. Vac. Soc. Jpn. article, is about diaphragm shape/baffle hardware, not resonance-based self-recalibration). In the absence of a nexus, secondary considerations are unlikely to save the claims.

8. Conclusion

• Claim 1: strong prima facie obviousness over Fujimori (US 2007/0068266 A1) in view of Silicon Labs (US 9,000,833 B2 / US 2014/0253219 A1), further in view of Azbil JP 2014-109484 A — and, alternatively, over Fujimori + Silicon Labs US 2014/0306623 A1. All elements are disclosed or are predictable mechanical applications of known diaphragm physics; the motivation (in-situ correction of deposition/aging-induced sensitivity drift in a capacitive pressure sensor) flows from the references and the field.
• Claim 2: likely obvious as the selection of one of two alternatives (E vs. t) expressly offered by claim 1 and reflected in the art's stiffness-based reasoning.
• Claim 3: likely obvious over Fujimori's pre-measured-resonance / "beyond a standard range" gate plus Silicon Labs' reference refresh; the weakest claim on this record.
• Claim 4: moderately vulnerable; the out-of-range alarm is the conventional annunciation of a comparison already performed in the art, though a patentee has a colorable "alarm on a computed downstream value" argument.

Confidence: high on the claim-element mapping and on the existence/dates of the references; moderate on the ultimate legal conclusion, because (i) the examiner allowed the claims over precisely this art, (ii) my reading of Fujimori's and the Silicon Labs references leans on abstracts/spec snippets and the ISR rather than full claim-by-claim full-text verification, and (iii) a genuine § 103 conclusion requires the full intrinsic record (prosecution history, any applicant remarks distinguishing these references), which I have not reviewed. Before relying on this for an invalidity contention, I would (a) pull the full texts of US 2007/0068266 A1, US 9,000,833 B2, US 2014/0306623 A1, and JP 2014-109484 A, and (b) read the US 16/064,229 prosecution history for applicant statements about these references.