US 9939317

I'll run targeted searches for this specific patent number, including litigation checks.

Let me run a few more targeted checks—litigation, PTAB, and the patent's own record.

US Patent 9,939,317 — Summary

Important caveat up front: I searched for this patent number and for any litigation. My searches surfaced no CAFC 2026 docket, no district-court complaint, and no PTAB proceeding naming US 9,939,317. Note two numeric "false friends" that a literal ID search pulls up — WO1999039317A1 (Immersion Medical's surgical-simulation interface device) and the various "'717 patent" references in Lone Star v. Zebra — neither is this patent; they are unrelated coincidental overlaps. I cannot confirm from live sources what USPTO/CAFC may hold that isn't indexed publicly; treat absence of litigation as "none found," not "none exists."

Bibliographic data (from the authoritative patent text)

Abstract

A sensitivity-parameter storing portion pre-stores, as known sensitivity parameters of a flame sensor, a reference received light quantity, a reference pulse width, a probability of regular discharge, and probabilities of non-regular discharge. Discharge probability is computed from the number of drive pulses applied to the flame sensor and the number of discharges detected. The computed discharge probability and the known sensitivity parameters yield a received light quantity per unit time. A pulse-width correcting portion corrects the pulse width of the drive pulses generated by an applied-voltage generating circuit so that the discharge probability equals a target discharge probability.

Plain-language overview of the claims

Claim 1 (the only independent claim) — A flame detecting system with six elements, all structurally tied to a UV/electron-tube flame sensor having a pair of electrodes:

1. Flame sensor — a pair of electrodes that detects light from a flame.

2. Applied voltage generating device — periodically generates a pulsed voltage and applies it across the electrodes as "drive pulses."

3. Current detecting device — senses current flowing through the flame sensor.

4. Storage device — stores, as pre-known "sensitivity parameters" of that specific sensor: a reference received light quantity Q₀, a reference pulse width T₀, a probability P₀ of regular discharge (the photoelectric-effect discharge at Q₀ and T₀), and a probability P_N of discharge other than regular discharge (i.e., noise/non-regular discharge).

5. Number-of-discharges counting portion — counts n, how many times discharge is judged to have occurred across the electrodes based on the detected current, over the applied drive pulses.

6. Received light quantity calculating portion — computes the received light quantity Q per unit time from: the stored sensitivity parameters, the pulse width T and pulse count N of the applied drive pulses, and the counted number n of discharges.

   Plus the key limitation: a pulse width correcting portion that corrects the pulse width T of the drive pulses so that the discharge probability P (derived from N and n) equals a predetermined target discharge probability P_TG.

   The practical point: rather than integrating analog current (slow, and the subject of prior art JP-A-2011-141290) or monitoring an analog voltage waveform (complex, per JP-A-2012-210284), the system counts discrete discharge events and converts them into a light quantity using stored calibration parameters — including noise-discharge probabilities — and then continuously re-tunes the pulse width to hold the discharge probability at a target. Holding P constant keeps electrode temperature (and thus sensor sensitivity) stable, improving accuracy.

Claim 2 (dependent on claim 1) — Adds (a) an electrode temperature detecting device for the flame sensor and (b) a pulse width temperature correcting portion that further corrects the already-corrected pulse width based on the difference between the detected electrode temperature and a predetermined reference electrode temperature. (This corresponds to "embodiment 2" of the specification.)

Claim 3 (dependent on claim 1) — Specifies that the pulse width correcting portion calculates a correction value T_new for the drive-pulse pulse width using the stored known sensitivity parameters, the received light quantity Q calculated by the received light quantity calculating portion, and the predetermined target discharge probability P_TG. (The specification gives this as Equation 9, derived from Equation 8.)

Technical context worth noting

• The mathematical core is Equations 7–9 of the specification: regular-discharge probability corrected by three noise-discharge types — N₁ (constant per unit time, independent of photons), N₂ (constant per pulse, depends on photon count), and N₃ (constant per pulse, independent of photons) — with probabilities P_N1, P_N2, P_N3. The specification notes any one, or any two, of these may be used (the unused ones set to 0).
• Prior art cited: US 5,227,640 (Nittan), US 5,256,057 (Protection Controls), US 2005/0174244 (Grow), JP 2011-141290 (Yamatake), JP 2013-210284 (Azbil).
• This patent is part of an Azbil family of flame-detection patents using discharge-probability computation (e.g., related US 9,625,311; US 9,746,181; and later-cited-by US 10,648,857 Honeywell, plus Azbil's US 11,280,672, US 11,428,575, US 11,346,711, US 11,359,962, US 11,402,261, US 12,153,166).

Uncertainty statement: All bibliographic and claim information above comes from the authoritative full patent text you supplied (source: patents.google.com/patent/US9939317). My independent searches produced no evidence of litigation, PTAB review, or a CAFC 2026 appeal for this specific patent, and I found no live USPTO record beyond what's shown here. I did not retrieve a live USPTO PatentCenter or CAFC docket record directly, so I cannot rule out non-indexed proceedings — but nothing found contradicts the record above.

I'll search for litigation involving US patent 9939317 across patent litigation sources.

The initial results returned unrelated patents (e.g., WO9939317 and US 8,593,931 — different numbers). Let me run more targeted searches.

Let me run more targeted queries on litigation databases and PTAB.

Summary

No litigation involving U.S. Patent No. 9,939,317 was found.

I searched for litigation, PTAB proceedings, and enforcement activity specifically tied to this exact patent number, checking patent-litigation aggregators and general sources (Unified Patents portal references, CourtListener, Docket Navigator, Justia, RPX Insight, PTAB/PTActs pages, and general web). None of the results identified any lawsuit, ITC investigation, or PTAB proceeding naming U.S. Patent No. 9,939,317 as an asserted or challenged patent.

What the searches did and did not surface

Important false-positive warning (interpreted literally, not auto-corrected): Several results involve superficially similar but legally distinct identifiers, and none of these is U.S. 9,939,317:

• WO 99/39317 A1 ("WO9939317") — a PCT publication for a medical procedure simulation interface device (EP 1103041). This is a 1999 PCT application, not U.S. 9,939,317.
• U.S. Patent No. 8,593,931 (the "'931 Patent") — a Blu-ray media patent (inventors Edwards/Kerfeld, originally assigned to Legger Col. A.B. LLC). The RPX and RPX-linked complaints that mention "'931" refer to 8,593,931, not 9,939,317.

Neither of these has any relationship to U.S. 9,939,317 beyond a coincidental numeric string.

About U.S. Patent No. 9,939,317 (context, not litigation)

• Title: Flame detecting system
• Patentee/Assignee: Azbil Corporation (original assignee; assignment from inventor Raita Mori recorded 2017-11-17)
• Application: US 15/816,234, filed 2017-11-17
• Priority: JP 2016-225892, filed 2016-11-21
• Grant/publication date: 2018-04-10
• Family/publications: JP 6782612 B2; CN 108088558 B; KR 101860632 B1 (per the "similar documents"/family listings)
• Post-grant activity shown on the patent record: Only maintenance-fee payments (4th year 2021; 8th year 2025). No reissue, no disclaimer, and no PTAB (IPR/PGR/CBM) filings appeared.
• Citations/continuations: The record lists later Azbil applications citing this family (e.g., US 11,280,672; US 11,428,575; US 11,346,711; US 11,359,962; US 11,402,261; US 12,153,166) and a Honeywell patent (US 10,648,857 B2). These are patent citations, which are not evidence of litigation.

Confidence and caveats

• My searches were run on public web sources and could not query paywalled dockets (e.g., full Docket Navigator or PACER) directly, so I cannot rule out a very recent or low-profile filing that has not been indexed publicly.
• I ran out of allotted search steps before exhausting every docket-specific query, so this should be treated as "no litigation found in the sources searched," not as an absolute negative.
• I found nothing indicating that Azbil Corporation has ever asserted U.S. 9,939,317 against any party, and nothing indicating any party has challenged it (IPR/PGR/CBM/declaratory judgment).

Bottom line: Based on the sources searched, there is no known litigation, ITC action, or PTAB proceeding involving U.S. Patent No. 9,939,317. If a definitive negative is required (e.g., for a freedom-to-operate or clearance opinion), I recommend a dedicated pull from a subscription docket service (Docket Navigator, Lex Machina) and PACER/PTAB docket checks, which I could not access directly here.

I'll verify the structured ODP data against live web sources before reporting.

Let me run targeted checks for any challenge activity or litigation that might have spawned (or failed to spawn) an IPR.

US 9939317 — PTAB Proceedings Analysis

Proceedings overview

Zero AIA trial proceedings.

The structured USPTO ODP block returns no AIA trial proceedings for US 9939317 (no IPR, PGR, or CBM), and my independent web checks surfaced nothing the ODP might not yet have indexed — no petition, no institution decision, no FWD, no termination, no Federal Circuit appeal. Breakdown by status: active 0, claims invalidated 0, claims sustained 0, settled 0, institution denied 0.

Bottom-line defensive posture: the patent is completely untested rather than "hardened." That cuts both ways. There is no FWD canceling claims you can lean on, but there is also no § 315(e)(2) estoppel, no § 325(d) adverse record, and no prior petitioner who has burned through the obvious art. All of claims 1–3 are live, and every invalidity theory you can build is still on the table — including theories a prior petitioner would have exhausted.

⚠️ Confidence caveat: I could not locate an ODP/PTAB E2E record directly to confirm the empty result independently; my conclusion rests on (a) the structured block supplied in this prompt, which is the canonical source, and (b) the absence of any IPR/PGR/CBM hit for this patent number or for Azbil flame-detection patents in web search. That is strong but not equivalent to reading the E2E docket. See "Recommended next steps."

Proceedings

None to report. Per the constraint against fabricating proceeding numbers, I am not listing any. For completeness, the related activity that does exist on the public record is non-PTAB:

Source: Google Patents US9939317B1.

Two observations that matter more than they first appear:

1. The examiner's cited art is unusually thin for a § 103 case, and two of the five references are the patent owner's own applications. PTL 1 (JP 2011-141290, Yamatake — Azbil's former name) and PTL 2 (JP 2013-210284, Azbil) are the very references the specification disparages as the starting point for the invention. That is a signal the inventive concept was distinguished over the applicant's own prior work, not over a crowded third-party field.
2. No litigation- or competitor-driven IPR has ever been filed. That is the classic fingerprint of a patent that is asserted narrowly (if at all) or that the accused industry — industrial burner controls, a small, standard-driven market (UL/FM/CE/CSA, JIS B 8415) — has chosen to design around rather than attack.

Strategic summary

Claim status: everything is UNTESTED. Claims 1, 2, and 3 of US 9939317 have never been construed by the PTAB and have never been cancelled. Claim 1 is the sole independent claim, and it is narrow in an important respect: it requires not merely a flame sensor and pulsed drive, but specifically a storage device holding the four "known sensitivity parameters" (Q₀, T₀, P₀, P_N), a received-light-quantity calculating portion that computes Q from those parameters plus pulse width T, pulse count N, and discharge count n, and a pulse-width correcting portion that drives P to a target P_TG. Claim 2 adds an electrode-temperature sensor and a pulse-width temperature correcting portion. Claim 3 requires the pulse-width correction value T_new be computed from the stored sensitivity parameters, the calculated Q, and P_TG. Because no claim has been through an IPR, there is no claim-level survival map to hand a defendant — the only narrowing on record is whatever prosecution history exists in App. 15/816,234 and its JP/CN counterparts, which you would need to pull from the file wrapper.

Estoppel landscape: wide open. § 315(e)(2) estoppel attaches only to a petitioner (and its privies/real parties in interest) that obtains an FWD. Here there has been no petition, therefore no estoppel runs against anyone, and there is no § 325(d) record of the PTAB having previously considered and rejected your art. Two practical corollaries: (i) § 315(b)'s one-year bar has never started for any alleged infringer, because that clock runs from service of a complaint alleging infringement of this patent — so if you are newly served, you have a full year to file, and if you are not yet served, indefiniteness of that clock is an advantage; (ii) you face no risk of being accused of raising "substantially the same prior art or arguments previously presented to the Office" under § 325(d). For a defendant today, the entire universe of § 102/§ 103 art is available, including combinations that a first-mover petitioner might have "reasonably could have raised."

Pattern signals: none of the usual ones. No serial petitioner on this patent (there can't be — zero petitions). No defensive aggregator (Unified Patents, RPX, etc.) appears anywhere in the chain. No PTAB appeal activity by Azbil on this patent, because there has been nothing to appeal. The one pattern worth flagging runs the other way: Azbil has a deep, actively-prosecuted flame/UV-detection portfolio around this disclosure — US 9746181 B2, US 9625311 B2, US 10247417 B2, US 10415829 B2, US 11280672 B2, US 11428575 B2, US 11346711 B2, US 11359962 B2, US 11402261 B2, and US 12153166 B2 all appear as similar documents or cited-by art — and this patent sits in a family with live JP (JP 6782612 B2) and CN (CN 108088558 B) members. A patent owner with a dense continuation strategy and no PTAB scars is a patent owner that has never had its claim-drafting tested adversarially. My web searches also did not surface any district-court assertion of US 9939317 (the Azbil litigation hits I found were unrelated — a 2010 Photon Control trade-secret/false-marking action against Azbil North America regarding optical temperature sensors, and Azbil's tangential appearance as a third party in the valsartan antitrust MDL, N.D. Cal./D.N.J. No. 397935). That absence is itself data: without infringement litigation, there has been no engine to generate an IPR, and the patent's offensive use to date appears minimal. I did not exhaustively verify litigation dockets, so treat that as a search result, not a clearance opinion.

Portfolio-defense caveat. Because the untested claim set sits inside a thick, still-growing family, an IPR on this patent alone may be strategically incomplete: Azbil has repeatedly filed follow-on applications (the 2020-priority "Light detection system" family, US 11346711 B2 / US 12298_* / US 12153166 B2, all cited as citing this patent) whose claims Azbil would likely reassert if this patent fell. Budget for a family-level, not patent-level, invalidity strategy.

Recommended next steps

If you are a defendant or a target of a demand letter citing US 9939317:

1. There is no FWD to cite — do not let anyone tell you otherwise. No claim of US 9939317 has been cancelled, and there is no PTAB disposition you can quote. Any assertion that "claims 1–5 were invalidated" would be wrong on both the claim count (this patent has three claims, 1–3) and the outcome (none invalidated). Verified against the ODP structured block and Google Patents US9939317B1.
2. Confirm the empty docket yourself before relying on it. Open the patent on PTAB E2E and search PTAB proceedings by patent number, then cross-check the USPTO PTAB Decisions library and CourtListener for any IPR/PGR/CBM or CAFC appeal naming 9939317. My search found nothing, but a docket read is the authoritative check and costs you minutes.
3. Preserve the full invalidity runway — and watch the § 315(b) clock. No estoppel and no § 325(d) exposure means you can run the strongest art first without fear of a "could have raised" bar arising from someone else's petition. If you have been served with a complaint alleging infringement of this patent, the § 315(b) one-year petition deadline is now running; calendar it immediately.
4. Attack the narrow claim 1 with § 102/§ 103 art aimed at the diagnostic-parameter architecture, not the flame-sensor hardware. The electron-tube flame sensor and pulsed-drive concepts are old (see US 5227640 A and JP 2011-141290 A on the face). The likely pivot points are (a) storing Q₀/T₀/P₀/P_N as factory "known sensitivity parameters" and (b) the feedback loop forcing measured discharge probability P to a target P_TG by adjusting drive pulse width. Art directed at probabilistic / statistical UV-flame-sensor diagnostics and at closed-loop pulse-width control in UV detectors is the highest-value place to search.
5. Because no IPR exists, consider whether you file first. As the first petitioner on this patent you get the PTAB's construction, you shape the record, and you impose estoppel on yourself only as to what you raise. But weigh that against the portfolio point above: with a live JP 6782612 B2 / CN 108088558 B family and an actively-continuing US family, a single-patent IPR may produce a win Azbil can route around. A family-level validity opinion should precede any petition decision.
6. Pull the file wrapper (App. 15/816,234) and the JP 2016-225892/JP 6782612 prosecution in parallel. With no PTAB FWD to constrain construction, the prosecution history — especially any narrowing amendments around Equations 6–9 (the "non-regular discharge" P_N1/P_N2/P_N3 corrections) — is your best source of prosecution-history estoppel and indefiniteness arguments. The dependent claims also import a temperature-correction element (claim 2) and a specific T_new computation (claim 3), both of which are potentially rich § 112 territory precisely because no tribunal has ever reviewed them.

If you are the patent owner: the clean PTAB record is a genuine asset — all three claims are intact, and no institution decision has ever hinted at weak claims. But "untested" also means "no confirmation." An ex parte reexamination or a proactive validity study before assertion is worth considering if you intend to enforce, because the first IPR filed against this patent will be the one that defines its value.

I'll start by verifying the assignment record and the assignee's corporate status.

Now let me check for any litigation, additional assignments, or NPE-related activity involving this patent.

Let me confirm there's no litigation history naming this patent and no NPE assertion activity.

Assignment & Ownership Analysis — US 9,939,317 B1 ("Flame detecting system")

Patent snapshot: US 9,939,317 B1 · App. No. 15/816,234 · filed 2017-11-17 · granted 2018-04-10 · priority JP 2016-225892 (2016-11-21) · current assignee Azbil Corporation · status Active (anticipated expiration 2037-11-17).

Verification note: The chain below is reconstructed from the Google Patents legal-events/assignment block for US 9,939,317, which mirrors the USPTO Assignment Center record. One field — correspondent of record — is not exposed in the fetched record; the Assignment Center entry for reel 044164/0265 must be opened directly to read the recording attorney/firm. I have not guessed it.

Inventors

Pattern notes:

• Single-inventor patent — common in Japanese corporate filings where one engineer is designated on a specific control-algorithm improvement.
• No inventor-departure signal. Mori is still filing Azbil-assigned work well after this patent: he is the named inventor on US D1,047,725 ("Vacuum gauge"), filed 2020-09-30 and assigned to Azbil Corp. (Tokyo). A continuing, same-employer inventor relationship is the opposite of the "all inventors exit within 12 months" fire-sale precursor.
• No inventor has ever appeared as an assignor on any non-Azbil recording for this patent.

Original assignee

Azbil Corporation (Tokyo, Japan; TSE Prime/First Section listed) — named as assignee on the face of the issued patent.

• Primary line of business: building automation, advanced/industrial automation, and life automation — i.e., a measurement-and-control operating company, not a licensing vehicle.
• Products embodying the claims: Yes. Azbil commercializes an entire flame-safeguard product family directly on this technology: AUR880/AUR890/AUR255 burner controllers, AUD100/110/120 and AUD300 ultraviolet flame detectors, and flame rods. Azbil's own literature describes flame-detector technology as "always a real focus for azbil" (see Azbil Technical Review 2022 and Azbil North America product/profile coverage at us.azbil.com). The patented subject matter — pulsed drive voltage across a UV electron-tube sensor, discharge-probability-based received-light-quantity estimation — maps onto exactly this product line.
• Current status: Operating. No bankruptcy, no dissolution, no acquisition of the patent-owning entity. Corporate lineage: Yamatake Shokai (1906) → Yamatake-Honeywell → Yamatake Corporation → Azbil Corporation (name change effective 2012-04-01; group rebranded "azbil Group" in 2008). The Honeywell equity relationship ended in 2002. The 2012 name change is a change-of-name only and post-dates the 2016 priority filing, so it does not appear as a recorded assignment on this patent.
• Azbil is also the assignee of the sibling family members on this invention: JP 6782612 B2 and CN 108088558 B.

Assignment timeline

Exactly one recorded assignment exists for this patent. There are no post-issuance transfers, security interests, mergers, licenses, releases, or corrections on record.

• 2017-11-09 (executed) / recorded 2017-11-17 — Reel 044164/0265
  • Conveyance: Assignment
  • Assignor: Mori, Raita (inventor)
  • Assignee: Azbil Corporation (Japan)
  • Correspondent: Not disclosed in the fetched Google Patents record. The reel/frame entry must be opened in the USPTO Assignment Center to read the recording attorney/firm and address. Flagging as a verification gap, not as a finding — no recurrence can be assessed without it.
  • Context: Routine in-house inventor-to-employer assignment executed on the standard corporate invention-agreement track (executed 8 days before the US filing, recorded the same day the application was filed). Matches the ASSIGNMENT OF ASSIGNMENT OF ASSIGNORS INTEREST / assignor MORI, RAITA entry.
  • Note: The "2017-11-17 Assigned to AZBIL CORPORATION" and "2016-11-21 priority date" entries are also what the record shows; the priority-date line is a priority claim, not an assignment event.

Post-issuance activity on record: none other than maintenance-fee payments (4th year, paid 2021-09-22; 8th year, paid 2025-09-24) and the grant event (2018-03-21). A chain with only these events is a chain in which the original operating-company assignee still owns the patent.

Timeline diagram

timeline
    title Ownership of US 9939317
    2016 : Priority application filed in Japan
    2017 : Inventor assigns rights to Azbil
         : US application filed and recorded
    2018 : US patent issued
    2021 : 4th year maintenance fee paid
    2025 : 8th year maintenance fee paid

NPE / troll-pattern signals

Verdict

Insufficient data — and the insufficiency is itself the finding: the record is clean, not opaque.

Justification: the sole recorded assignment is Reel 044164/0265, executed 2017-11-09 and recorded 2017-11-17, conveying rights from inventor Raita Mori to Azbil Corporation — a TSE-listed operating manufacturer that ships the accused-class hardware (AUR-series burner controllers, AUD-series UV flame detectors) and remains the assignee of record today, with maintenance fees paid 2021-09-22 and 2025-09-24. Because the chain consists of only the original inventor assignment, zero of the eight NPE signals are present (one is unclear solely due to the unread correspondent field), so no NPE verdict — high or moderate — can be supported. If the rubric requires a non-NPE characterization when the original operating-company assignee retains ownership: this is an operating-company-owned patent with no assertion activity and no defensive-aggregator involvement, i.e., the diametric opposite of a troll pattern.

Verify at: USPTO Assignment Center — search patent number 9939317, then open reel/frame 044164/0265 to capture the correspondent of record → https://assignmentcenter.uspto.gov/ (indexed mirror: https://assignment.uspto.gov/patent/index.html). Cross-check the post-grant legal-events tab against the Google Patents page for US 9,939,317.

One open item to close out: pull the correspondent name/firm for reel 044164/0265. If it turns out to be a known NPE-recording firm, that would still be explained by routine filing-time prosecution counsel and would not, standing alone, flip any of the eight calls above.

I'll research this patent and its cited prior art. Let me start with searches to verify the specifics.

Now let me research the other cited references.

Prior-Art Analysis — US 9,939,317 B1

1. Patent under review (as identified literally)

Source of record: https://patents.google.com/patent/US9939317/en

I ran targeted searches against the USPTO/Google Patents record for 9939317 only; no substitute or nearby numbers were returned or substituted. The reference set below is exactly the "Citations (5)" / "Patent Citations (5)" listed on the face of US 9,939,317.

2. The five examiner-cited references

All five were cited against this patent and all pre-date the 2016‑11‑21 priority date, so each is available as prior art under AIA 35 U.S.C. § 102(a)(1)/(a)(2).

Note: #4 and #5 are the two references the specification itself identifies as PTL 1 and PTL 2 (see the "CITATION LIST" in US 9,939,317).

3. Per-reference detail

3.1 US5227640A — Nittan

• URL: https://patents.google.com/patent/US5227640A/en
• Description: A flame detector using an ultraviolet sensor (UV detection tube). A signal-processing unit divides the time base into successive spans, accumulates the UV-sensor discharge pulse count in each span, stores the per-span counts, and judges flame occurrence by weighting the current span's count against counts from previous spans. It also self-tests sensor sensitivity within the apparatus. Abstract: "an accumulating means for dividing a time base into several time spans … and a calculation means for judging the occurrence of flame by taking into consideration the accumulated values of previous spans, as well as the accumulated value of the latest span."
• § 102 relevance: Discloses (a) a UV electron-tube flame sensor with a pair of electrodes, (b) periodic application of voltage via a "sensitivity adjustment unit," and (c) accumulating/counting discharge pulses over time. It does not disclose storing Q₀/T₀/P₀/P_N sensitivity parameters, computing a received light quantity Q from them, or a pulse-width correcting portion. Its "accumulated values vs. threshold" logic is of the same genus as the integrating method criticized at PTL 1 in the spec.
• Claims potentially anticipated: None in full. It is a § 102 reference only against isolated elemental concepts of claim 1 (flame sensor, applied voltage, discharge counting). Not anticipatory of claim 1, 2, or 3.

3.2 US5256057A — Protection Controls

• URL: https://patents.google.com/patent/[US5256057](/patent/US5256057)#1
• Description: A six-stage, fail-safe fuel control circuit. A flame scanner produces a raw count signal of pulses at a rate proportional to UV energy near the flame; a buffer sharpens/amplifies; a capacitive pulse-height discriminator counts pulses in a timing window and signals flame presence if the count exceeds a desired rate; a capacitor-discharge oscillator, resonant amplifier, and relay hold the fuel valve open. Explicitly emphasizes fail-safe design, not quantitative photometry.
• § 102 relevance: Discloses a flame scanner, pulse counting in a fixed timing window, and a valve decision — but no flame-sensor sensitivity-parameter model and no drive-pulse-width feedback/correction.
• Claims potentially anticipated: None in full. Elemental teaching relevant to claim 1's "current detecting / counting" concepts only.

3.3 US20050174244A1 — Grow (granted as US7088253B2)

• URL: https://patents.google.com/patent/US20050174244A1/en
• Description: A flame detector and burner fuel-valve control using an ultraviolet-sensitive cold-cathode gas discharge (UV) tube in a DC quench circuit. In the presence of a flame the circuit generates a pulse signal at a frequency related to flame intensity; a contaminated tube produces a higher frequency. A frequency-discriminator/band-pass (nominally 10–500 Hz, nominal tube pulse rate 250 Hz) distinguishes flame-responsive pulses from contamination or no-flame pulses; the flame detector is calibrated by attenuating UV radiation (mesh screens) so the tube oscillates at the nominal frequency. Abstract: https://pubchem.ncbi.nlm.nih.gov/patent/US-[7088253](/patent/7088253)-B2
• § 102 relevance: Closest of the US references to the "discharge event processing" theme. It discloses pulse-rate discrimination and a calibration (attenuation) step, but the calibration is optical attenuation — it is not correction of the electrical drive-pulse width T, and there is no received-light-quantity computation from stored sensitivity parameters.
• Claims potentially anticipated: None in full. Elemental teaching relevant to claim 1 (UV discharge sensor, pulse signal, discriminator).

3.4 JP2011141290A — Yamatake (spec's PTL 1)

• URL: https://patents.google.com/patent/JP2011141290A/en (cited in US 9,939,317 as PTL 1)
• Description (per the US 9,939,317 specification): A flame detection device in which the current flowing between the sensor electrodes is integrated, and the presence of a flame is determined when the integrated value ≥ a predetermined threshold, absence when < threshold. The US patent's Background criticizes exactly this: "since this method integrates the current flowing between the electrodes, the time for integration is required even in flame extinction … it takes time until flame extinction is detected."
• § 102 relevance: Discloses the integrating/threshold flame-detection approach that US 9,939,317 expressly characterizes as the prior art and departs from. It contains no discharge-probability model, no Q₀/T₀/P₀/P_N parameters, and no pulse-width correction.
• Claims potentially anticipated: None. It is background art establishing the state of the art, useful for § 103 motivation, not § 102 anticipation of claims 1–3.

3.5 JP2013210284A — Azbil (spec's PTL 2) — closest reference

• URL: https://patents.google.com/patent/JP2013210284A/en (cited in US 9,939,317 as PTL 2; also cited as "专利文献2 / Patent Document 2" in sibling CN108088559A, e.g. https://patents.google.com/patent/CN108088559A/en)
• Description (per the US 9,939,317 specification): A flame detecting device comprising an electron tube with a pair of electrodes that emit electrons when irradiated with UV; an applying portion that applies a periodically changing voltage across the electrodes; a detecting portion that detects a voltage waveform representing the temporal change of the voltage between the electrodes; and a determining portion that determines presence/absence of a flame on the basis of the detected voltage waveform. The US patent states that PTL 2 avoids the integration-time problem but "needs to monitor the voltage waveform … and to perform signal processing in an analog manner … so the implementation … is not easy."
• § 102 relevance: This is the most relevant cited reference. It discloses the flame sensor, the periodic drive-voltage applying portion, the current/voltage detection, and a presence/absence determination. It does not disclose:
  • a storage device holding Q₀, T₀, P₀ and non-regular-discharge probability P_N;
  • a received-light-quantity calculating portion computing Q from those parameters plus T, N, n; or
  • a pulse-width correcting portion that corrects T so the measured discharge probability P converges on a target P_TG.
• Claims potentially anticipated: None in full — but it is the strongest candidate for an obviousness attack on claim 1's front-end elements, and it is the reference over which the claimed invention's characterizing feature (pulse-width feedback to pin P at P_TG, to stabilize electrode temperature and thus keep Q well-defined) is defined.

4. Claim-by-claim § 102 bottom line

Summary judgment: None of the five references cited on the face of US 9,939,317 is a § 102 anticipation of any pending claim. They function as (i) background/state-of-the-art (§ 3.4, § 3.5) and (ii) elemental § 103 starting references (§ 3.1–§ 3.3). The patent's own specification concedes JP2013210284A (PTL 2) as the point of departure and frames the invention as the pulse-width-correction/electrode-temperature stabilization improvement. The non-regular-discharge probability P_N parameterization and the T→P_TG feedback loop appear to be the distinguishing limitations that preserved patentability.

5. Related records worth noting (not § 102 art)

• Family / also published as: JP2018084422A (2018‑05‑31), JP6782612B2 (2020‑11‑11), CN108088558A (2018‑05‑29), CN108088558B (2020‑03‑24) — the JP and CN counterparts of this same invention.
• Later Azbil continuations citing this family (examiner/third‑party citations, all post‑priority): US10648857B2 (Honeywell, 2020‑05‑12); US11280672B2 (2022‑03‑22); US11428575B2 (2022‑08‑30); US11346711B2 (2022‑05‑31); US11359962B2 (2022‑06‑14); US11402261B2 (2022‑08‑02); US12153166B2 (2024‑11‑26); CN113340437A and CN113340413A (2021‑09‑03). These post-date the 2016 priority and cannot be § 102 art against it.
• Two-family forward citations (JP): JP2015200575A (Azbil, 2015‑11‑12) and JP5922452B2 (Hochiki, 2016‑05‑24) are the only family-cited documents published before the 2016‑11‑21 priority; they were not among the five references applied against US 9,939,317 but are chronologically available if a broader § 102/§ 103 sweep is desired. I would need to pull their full texts before asserting any mapping.

6. Confidence and limitations

• The five references and their bibliographic data are taken directly from the authoritative full text of US 9,939,317 and confirmed by live search (Google Patents / FreePatentsOnline / PubChem for US5256057A, US20050174244A1→US7088253B2, US5227640A).
• Descriptions of JP2011141290A and JP2013210284A are drawn from how US 9,939,317 itself characterizes them (PTL 1 / PTL 2) and corroborated indirectly via sibling CN108088559A; I was not able to retrieve their full native-language texts within this session, so the JP descriptions above are second-hand (via the patent's own citation list) rather than independently verified from the JP documents.
• § 102 anticipation is asserted only where a single reference discloses all claimed features arranged as claimed. On this record, that threshold is not met by any of the five references; the strongest available challenge would be a § 103 combination led by JP2013210284A, with the remaining references supplying secondary elements.

I'll ground this analysis in the actual prior-art references listed on the patent page. Let me verify what those references disclose before building the § 103 analysis.

Let me verify the two Japanese references (PTL 1 and PTL 2) and the closely related Azbil family members that appear in the "Similar Documents" list, since these bear directly on the pulse-width-correction feature.

Obviousness Analysis — U.S. Patent No. 9,939,317 B1 ("Flame Detecting System")

Assignee: Azbil Corporation · Inventor: Raita Mori · Priority: 2016-11-21 (JP 2016-225892) · Filed: 2017-11-17 · Granted: 2018-04-10

Grounding note. This analysis is built from (a) the authoritative full text of US 9,939,317 B1 on Google Patents (https://patents.google.com/patent/US9939317/en) and (b) the prior-art references listed in that page's own Prior Art / Citations / Similar-Documents sections. Where I could not confirm a reference's full disclosure, I say so explicitly rather than asserting it. All identifiers are reproduced literally as they appear on the page.

1. Bottom Line Up Front

The single most vulnerable feature is the "pulse width correcting portion" that drives measured discharge probability P to a target P_TG. It is the only element not squarely taught by the art of record, and it is nonetheless the element most readily attacked as an obvious design choice once one appreciates (as the specification itself does) that discharge probability is a known function of the product of received light quantity and pulse width.

2. Governing Framework

US 9,939,317 was filed 2017-11-17 with a 2016-11-21 foreign priority, so the AIA §§ 102/103 framework applies. Obviousness is assessed under Graham v. John Deere, 383 U.S. 1 (1966), and KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007):

1. Scope and content of the prior art;
2. Differences between the prior art and the claims;
3. Level of ordinary skill; and
4. Secondary considerations.

The AIA-art references in the section below are prior art under § 102(a)(1) (patented or printed-published before the 2016-11-21 effective filing date):

3. The Prior Art of Record and What It Discloses

JP 2013-210284 A ("Flame detector," Azbil) — PTL 2. Verified from the patent's own background section and family records: discloses an electron tube with a pair of electrodes, an applying portion that applies a periodically changing voltage across the electrodes, a detecting portion that detects a voltage waveform representing the temporal change of voltage between the electrodes, and a determining portion that determines flame presence/absence from that waveform. PTL 2 is the closest art: it already contains the sensor, the pulsed/periodic application means, the electrical-signal detection means, and flame determination.

JP 2011-141290 A ("Flame detection device," Yamatake) — PTL 1. Discloses the integration method: current flowing between the electrodes is integrated and compared to a threshold. Teaches the sensor/detection/determination architecture and, critically, supplies the problem the invention addresses — integration is slow to respond to flame extinction.

US 5,227,640 A (Nittan). Verified: an ultraviolet sensor (71) with a "sensitivity adjustment unit 72 for applying voltage to the ultraviolet sensor," means for counting discharge pulses per time span and comparing the count to threshold levels (MAX₁/MAX₂/MAX₃), and — importantly — means for testing and determining the "sensitivity rank" of the sensor. This maps to two claim elements: (i) counting discharge events over a defined interval, and (ii) treating sensor-specific sensitivity as a stored/derived parameter. The adjustable applied-voltage unit also supplies a motivation hook for adjusting a drive parameter to control the sensor's operating point.

US 5,256,057 A (Protection Controls). Verified: a flame scanner generating a count signal comprising pulses at a rate proportional to flame UV energy, and a capacitive pulse-height discriminator that counts pulses "during a given period" and produces a signal when the count exceeds a desired rate. Reinforces counting-pulses-over-N-applications as a known flame-detection primitive.

US 2005/0174244 A1 (Grow). Verified: a UV flame-sensor tube in a quench circuit that generates a desired pulse signal when a flame is present and undesired pulse signals when the tube is contaminated or when no flame is present, with a discriminator that distinguishes the desired flame-responsive signal from the undesired signal. This is directly analogous to the claimed distinction between "regular discharge" (P₀) and "discharge other than regular discharge" (P_N) — i.e., separating a signal-of-interest from a noise / false-discharge component.

US 9,746,181 B2 (Azbil) (appears in the page's Similar Documents list). Discloses a flame detecting system with an applied-voltage generating portion that generates a pulse to drive the flame sensor, a detecting portion measuring the electrical signal in the sensor, a storing portion that stores sensitivity parameters of the flame sensor in advance, and a central processing unit that obtains a quantity of received light using the parameters of known received light quantity, pulse width, and discharge probability, together with a discharge probability derived from an actual pulse width and the measured number of discharge times. This is, in substance, Claim 1's "received light quantity calculating portion" minus the pulse-width-correction and the P_N term. Caveat: US 9,746,181 B2 shares the Azbil assignee, so its availability may turn on § 102(b)(2)(C) (common-ownership exception). Its priority (JP 2015-106035, 2015-05-26) and grant (2017-08-29) both sit close to this patent's 2016-11-21 / 2017-11-17 dates and must be pinned down before relying on it.

US 9,625,311 B2 (Azbil) and JP 2015-200575 A (Azbil) also appear in the page's Similar-Documents / Family-Cites lists and are the same general subject matter; their disclosures should be confirmed before use (see § 9).

4. Claim 1 — Element-by-Element Mapping

Difference summary. The claim does not read on any single reference. The gap is the confluence of (a) computing received light quantity in the digital discharge-probability domain, (b) storing a non-regular-discharge probability P_N as a calibration constant, and (c) servoing the pulse width T to hold P at a target.

5. The Two Principal Combinations

Combination I (primary): JP 2013-210284 + US 5,227,640 + routine calibration

Where it gets you. JP 2013-210284 supplies the sensor, the periodic pulsed drive, the current/voltage detection, and flame determination. US 5,227,640 supplies counting of discharge pulses over a defined interval and treating sensor sensitivity as a characterized, adjustable quantity. The remaining "known sensitivity parameters" (Q₀, T₀, P₀, P_N) are calibration constants obtained by delivery inspection — a fact the specification itself states ("need to be measured by, for example, delivery inspection"). Populating a store with the results of routine factory characterization is a design choice / predictable use of a known technique, not patentable weight.

Deviation you must bridge. This combination still needs the arithmetic linking stored parameters + measured P to Q. If US 9,746,181 B2 is available, that bridge is exact. If it is not (ownership), the bridge is supplied by the "obvious to try" rationale: once PTL 2's analog waveform monitoring is rejected as too complex (the patent's own stated motivation), the artisan naturally moves to counting the discrete discharge events PTL 2 already detects.

Combination II: + US 2005/0174244 A1 (for the P_N element)

US 2005/0174244 A1 expressly distinguishes a desired flame-responsive pulse from undesired pulses arising from contamination or no-flame. A POSITA seeking to prevent false flame indications would be motivated to store a probability for that undesired class — i.e., the claimed P_N — so it can be subtracted out of the discharge probability used to infer received light quantity. The rationale is "use of a known technique to improve a similar device in the same way," the classic KSR rational (C).

6. Motivation to Combine (KSR Rationales)

1. Art-recognized problem — (F). The patent's own background admits PTL 2 "needs to monitor the voltage waveform… and to perform signal processing in an analog manner… so the implementation of the flame detecting device is not easy." That admission is a built-in motivation: the artisan would prefer a digital/discrete-event implementation of the same physical insight as PTL 2. This is the strongest single motivation in the record.

2. Known dependence of P on pulse width — (C). The specification's own governing relation (its Equation 6, (1−P) = (1−P₀)^(QT/Q₀T₀)) makes explicit that the number of photons contributing to discharge is the product of Q and T. Anyone who derives the discharge-probability model thereby also learns that T is the actuator that sets the operating point. Servoing T to hold P at a target P_TG is the direct, predictable engineering corollary — a controller wrapped around a known monotonic relationship.

3. Adjusting a drive parameter to control the sensor is known — (B). US 5,227,640 already provides a "sensitivity adjustment unit for applying voltage to the ultraviolet sensor." Adjusting pulse width instead of (or in addition to) applied voltage to set the same operating point is a simple substitution of one known parameter for another to achieve the same result.

4. Desirability of a stable operating point — (D). The patent concedes the physics: heat from discharge raises electrode temperature, which de-fixes the Q↔P relationship, degrading accuracy. Keeping a sensor's temperature/operating point constant is a recognized goal in detector design (the Honeywell flame-safeguard literature surfaced in this family states plainly that "Flame-detector performance is affected by… Hot refractory… [and] Ambient temperature at detector"). Holding the drive parameter constant to stabilize the measurement is applying a known technique to a known device ready for improvement.

5. Desired "fail-safe/fast-response" design incentives — (F). All five references sit in the same field of endeavor (UV flame detection / combustion safeguards) and address the same art-recognized problem — reliably and rapidly distinguishing flame from no-flame under noise. This satisfies the reasonably pertinent to the problem test and makes cross-combination natural rather than hindsight-driven.

7. Claim 2 (Electrode-Temperature Correction)

Claim 2 adds an electrode temperature detecting device and a pulse width temperature correcting portion that further corrects T based on the difference between measured electrode temperature and a reference electrode temperature.

Assessment: the weakest case for the petitioner on the cited art alone. None of JP 2013-210284, JP 2011-141290, US 5,227,640, US 5,256,057, or US 2005/0174244 A1 clearly discloses detecting the electrode temperature of the flame sensor and closing a loop on it. The claim is therefore vulnerable only through:

• Official Notice that temperature compensation of photodetectors/discharge tubes is notoriously old and well-known; or
• an additional secondary reference teaching sensor-temperature feedback.

Because the specification itself treats the electrode-temperature/operating-point stability issue as the problem behind Claim 1, a petitioner can argue that adding an explicit temperature sensor to the Claim 1 architecture is the natural, predictable completion of the same stability goal (KSR rationale (C)/(D)). But absent a concrete teaching of electrode-temperature feedback, this is a weaker, notice-dependent position than Claim 1.

8. Claim 3 (Correction-Value Calculation)

Claim 3 recites only that the pulse width correcting portion calculates T_new using the stored sensitivity parameters, the calculated Q, and P_TG. This is a method-of-calculation dependent claim: it adds nothing structural beyond Claim 1, and the "using… parameters, Q, and P_TG" limitation is simply the ordinary way one would solve for T given the model the specification itself supplies (its Equation 8 solved for T with P = P_TG).

Once Claim 1 is found obvious, Claim 3 is obvious a fortiori: rearranging a known equation to solve for a different variable is not an inventive act. The arithmetic form of the limitation is not a patentable distinction.

9. Counterarguments the Patentee Will Raise (and how they fare)

10. Confidence, and What Must Be Verified

• High confidence on the content of JP 2013-210284 and JP 2011-141290, because the patent's own Background reproduces their disclosures, and on US 5,227,640, US 5,256,057, and US 2005/0174244 A1, which I verified from their original documents.
• Moderate confidence on the mapping to US 9,746,181 B2 — it is highly material (it appears to recite the stored-sensitivity-parameter/received-light-quantity computation) but I have not confirmed whether it is available as prior art given the shared Azbil ownership (potential § 102(b)(2)(C) exception). Confirm its exact U.S. filing and 18-month publication dates. Its earliest effective date (JP 2015-106035, 2015-05-26) precedes this patent's priority, which makes the § 102(a)(2)/§ 102(b)(2)(C) analysis decisive.
• Must verify the disclosures and dates of US 9,625,311 B2, JP 2015-200575 A, US 10,247,417 B2, and US 10,415,829 B2 before relying on them. The last two granted in 2019, after this patent's filing, so they are only usable if they qualify as § 102(a)(2) art and are not excepted by common ownership — likely unusable against Claim 1.
• I have not read the full claim-by-claim text of every reference; the mappings above are provisional element matches, not final claim charts.

11. Conclusion

Claim 1 is, on the art of record, susceptible to a strong § 103 challenge built on JP 2013-210284 as the primary reference, supplemented by US 5,227,640 (discharge-pulse counting and drive-parameter/sensitivity adjustment), US 2005/0174244 A1 (desired vs. undesired discharge — the P_N analogue), and — subject to an ownership check — US 9,746,181 B2 (stored known sensitivity parameters and discharge-probability-based received-light computation). The motivation to combine is unusually clean because the patent's own specification supplies the problem statement (PTL 2's analog implementation is hard) and the model itself (Equation 6) exposes pulse width T as the control variable, making the pulse-width servo an obvious design choice rather than an inventive step.

Claim 3 stands or falls with Claim 1 and adds only a rearrangement-of-known-equation limitation. Claim 2 is the most defensible claim and would require either an additional temperature-feedback reference or reliance on Official Notice that flame-sensor temperature compensation is well known.

Prepared as a technical/legal-opinion-style analysis for internal use; not a formal validity opinion. All references and identifiers are stated literally as they appear in the source record.