US 9945722

I'll search for this specific patent number, including any CAFC 2026 docket activity.

Let me run a couple more targeted searches to confirm litigation/docket status.

Let me do one final check for any PTAB or Federal Circuit proceedings specifically naming this patent.

US Patent 9,945,722 — Analyst Summary

Sourcing note: I do not have direct authenticated access to USPTO PatentCenter or the CAFC docket system; I ran open-web searches (uspto.report, Google Patents, general docket material). The patent full text provided in my working set is treated as authoritative and was used for all claim/abstract/date data below. Any litigation status is reported only to the extent I could find evidence.

Bibliographic Data

Minor anomaly worth flagging (not auto-corrected): the recorded assignment documents list an assignor "CHALUPA, LEO M." who does not appear in the granted-inventor list or on the Google Patents inventor field. I am reporting this discrepancy as-is rather than reconciling it.

Abstract (as granted)

A spectral radiometer system measures incoming light intensity and spectral distribution in different wavelength-bands. An additional data storage device allows recording of measured data. The sensor system yields very high sensitivity to incoming light, with outstanding detector-response linearity over several orders of magnitude of incoming light. Additional benefits are ultra low power consumption and minimum size. The system is usable in remote solar radiation monitoring (e.g., mobile solar power units) and long-term environmental monitoring where high precision and low power consumption are required.

Independent Claims — Plain-Language Overview

The patent has 10 claims; independent claims are 1, 6, and 8.

Claim 1 — System/apparatus claim (broadest).
A spectral radiometer that measures light intensity, detects spectral distribution across wavelength bands, and stores/transfers the data, comprising three parts:

1. a spectrally sensitive circuit element section using photon-dependent charge-generation/recombination elements plus additional charge-characteristic components (diode components and resistance–capacitance components);
2. a CPU with processor, memory-stored program code and multiple I/O ports, coupled to the sensor section, a timing apparatus (clock), and a data storage device — the code measures incoming light intensity and calculates, records, and processes the light data; and
3. a case with a removable lid housing the sensor section, CPU, wiring, timing apparatus, and storage.
   The sensor section must deliver linear sensitivity to incoming light across more than three orders of magnitude, and its elements include pn-diodes, spectrally sensitive resistors, Schottky diodes, and other photon-dependent elements — which may be mixed and used simultaneously.

Claim 6 — Diode-based apparatus with adjustable external capacitance.
Like Claim 1 in architecture, but the sensing element is specifically a photon-dependent charge-generation/recombination semiconductor diode with a given bandgap energy and given reflection/absorption properties (defining its absorbing wavelength), whose charge characteristic includes the diode's inherent capacitance, encapsulated in a transparent housing. The distinguishing feature is an adjustable external capacitance switchably connected in parallel with, in series with, or open across the diode. That external capacitance tunes the voltage-decay time to match expected light-intensity levels.

Claim 8 — Diode-based apparatus with multiple different-bandgap diodes in one housing.
Again the CPU/case/transparent-housed-diode architecture, but the sensor section contains a plurality of diodes (at least two, at least three, or at least four), each from semiconductor material with different bandgap, reflection, and absorption properties, and all of them encapsulated together in a single transparent housing large enough to contain them.

Dependent claims in brief:

• Claim 2: adds a coin-cell battery and ultra-low-power operation averaging 100 µW.
• Claim 3: case dimensions/weight — width ~1½–3 in, depth ~1–2 in, height ~2½–5 in, weight ≤ ~0.1–0.2 lb.
• Claim 4: data storage device capacity up to at least 128 MB integrated in the apparatus.
• Claim 5: apparatus is free of shock sensitivity with long-term remote monitoring capability of <2 up to ~10 years.
• Claim 7: transparent housing may have different optical acceptance angles via modification of the housing's optics.
• Claim 9: the multi-diode sensor sections are inserted in openings on different sides of the case, providing directional light-intensity detection.
• Claim 10: adds bandgap selection and switchably adjustable diode sensitivity (different capacitance), selectable before, after, and during measurement cycles.

CAFC 2026 Docket / Litigation Findings

No CAFC 2026 docket activity referencing US 9,945,722 was found. My searches for "9945722" together with Federal Circuit, 2026, PTAB, IPR, and appeal terms returned no case, petition, or docket naming this patent. The only substantive third-party hit was the uspto.report grant record (https://uspto.report/patent/grant/9,945,722), which corroborates the bibliographic data above (grantee "The United States of America, as represented by the Secretary of the Navy").

Uncertainty statement: I cannot rule out a CAFC or PTAB proceeding that is not indexed by open-web search, and I did not query the CAFC's official PACER-adjacent docket or USPTO PatentCenter/PTAB API directly. Based on available evidence, the patent appears to be unlitigated and active, with its only known post-grant activity being routine maintenance-fee events. I am not asserting a definitive "no litigation exists" conclusion.

Cited Prior Art of Record (from the patent's front page)

US 7,135,848 B1 (Xytrans); US 7,250,747 B1 (Xytrans); EP 1,923,679 A1 (Pixartis SA); US 2008/0259318 A1 (Everfine Photo-E-Info); US 2012/0280114 A1 (Rodrigues); US 2015/0051498 A1 (Hypermed Imaging). Non-patent citation: Parry, "NRL Develops Low Cost, High Efficiency Solar Sensor," NRL News Release (Aug. 12, 2013).

I'll search patent-litigation-focused sources for US 9,945,722 / "9945722" specifically.

Let me run additional targeted searches on litigation databases and the patent family.

Litigation Search: US Patent 9,945,722

Bottom line: No litigation, PTAB proceeding, or Federal Circuit appeal naming US Patent 9,945,722 was found.

Search performed

I queried open patent-litigation and docket-facing sources for the exact identifier:

Result: no known cases

I found no case — in any U.S. district court, the Court of Federal Claims, the ITC, or the PTAB (IPR/PGR/CBM) — in which US 9,945,722 is a patent-in-suit, and no CAFC appeal arising from such a case. The only third-party post-grant activity of record is the routine maintenance-fee event reported in the file (4th-year fee paid 2021-08-13; reminder mailed 2025-12-08).

Because there is no case, I cannot supply plaintiff(s), defendant(s), jurisdiction, case number, filing date, or outcome/status — these fields are absent, not merely unavailable.

Important disambiguation — do NOT conflate

The searches surfaced a different patent with a similar number, which I am flagging rather than adopting:

• US 9,896,722 ("the '722 patent" in that reporting) — asserted by Bio-Rad Laboratories against 10x Genomics in 2018 (later reported court activity in 2019). This is not US 9,945,722. The digits differ (9,896,722 vs 9,945,722), the assignee differs, and the technology (life-science/biotech instruments) is unrelated to the Navy radiometer. Per your instruction not to return results for similar numbers, I exclude this as a hit for 9,945,722.
• Sanofi v. Eli Lilly documents also reference a "'722 patent," but that is an Orange Book insulin-glargine patent — likewise unrelated.
• IdeaHub Inc. v. Unified Patents, LLC, No. 2024-1684 (Fed. Cir. Apr. 10, 2026, Rule 36 affirmance; appeal from PGR2022-00044) appeared in CAFC search results but does not involve 9,945,722 or any Navy patent in this family.

Confidence and caveats

• High confidence that no publicly indexed litigation names this patent. This matches the patent's profile: it is U.S. Government–owned (Secretary of the Navy / Office of Naval Research, as confirmed by the grant record at https://uspto.report/patent/grant/9,945,722), which materially reduces NPE-style assertion and commercial infringement-suit exposure.
• Explicit limitation (not a definitive negative): I ran open-web searches only. I did not query the CAFC's official docket, PACER, USPTO PatentCenter, or the PTAB API directly, and my tool calls reached their step limit before I could run the last batch of supplementary checks. A sealed, recently filed, or poorly indexed action cannot be entirely excluded on this evidence. I am therefore reporting "no litigation found," not "no litigation exists."
• For a citable negative, the appropriate next steps would be a direct PACER/Docket Alarm docket search on "9:945:722", a PTAB API query, and a CourtListener/RECAP full-text query — none of which I could complete here.

Cross-reference note: This is consistent with the earlier-generated "Patent summary" section, which likewise reported no CAFC 2026 docket activity for this patent. No contradiction between sections.

I'll verify the structured data (no AIA proceedings on file) against open-web sources before drafting.

The initial searches surfaced several unrelated "'722" patents (U.S. 8,407,722 — asynchronous messaging; U.S. 9,483,722 — IPR2025-00640). Let me run targeted checks to rule out confusion and confirm no AIA proceeding names this patent.

PTAB Proceedings on US 9,945,722 — Analyst Report

Data provenance note: The canonical source for this report is the structured "PTAB proceedings on file" block supplied in the prompt, which reflects the USPTO Open Data Portal (ODP) AIA-trial ingest. I ran independent open-web searches (PTAB petitions portal at ptacts.uspto.gov, Google Patents, uspto.report, and general docket material) to corroborate. No AIA trial proceeding naming US 9,945,722 was found in either source. I did not query PTAB E2E or the CAFC docket directly, so I flag the residual uncertainty at the end rather than asserting a categorical negative. No proceeding numbers are invented below, because there are none to report.

Proceedings overview

Total AIA trial proceedings on file for US 9,945,722: zero (0 IPRs, 0 PGRs, 0 CBMs; no active proceedings, no claims invalidated, no claims sustained through trial, no settlements, no institution denials). The bottom-line defensive posture for a defendant is therefore the inverse of the "hardened patent" story: this patent has never been tested at the Board, so there is no PTAB estoppel record, no claim-cancellation ruling, and no prior-art roadmap in a public FWD to hand you. Conversely, there is also no unfavorable FWD foreclosing your invalidity theories — the patent is an untested, un-narrowed instrument sitting on all 10 originally-issued claims.

Proceedings

None. There is no proceeding to enumerate. No case number, petitioner, panel, institution decision, Final Written Decision, settlement, or appeal exists to report for this patent.

False-positive search hits — explicitly not this patent

My searches surfaced several documents referencing a "'722 patent," none of which concern US 9,945,722. I list them here so this report is not misread later as finding activity that does not exist:

None of these should be cited as authority about US 9,945,722.

Strategic summary

Claim status. Because no AIA trial has ever been instituted against this patent, all 10 claims stand exactly as granted on 2018-04-17: independent claims 1, 6, and 8, and dependent claims 2, 3, 4, 5, 7, 9, and 10 are all UNTESTED at the Board — neither canceled nor sustained-through-trial, simply never challenged. There is no narrowing claim amendment, no certificate of correction affecting scope, and no adverse judgment. A defendant today faces the full, un-narrowed claim set.

Estoppel landscape. There is no § 315(e)(2) estoppel on any party — estoppel attaches only to a petitioner that reaches an FWD, and no petitioner has. That cuts both ways: (a) you are not blocked by anyone else's prior IPR, and you may raise any § 102/§ 103 ground on patents and printed publications, or any § 112/enablement/indefiniteness ground in district court (IPR cannot reach § 112); but (b) you also get no free roadmap — the strongest prior art was never vetted by the examiner against a motivated challenger, so the invalidity case must be built from scratch. Note the prosecution record is potentially fertile: the sibling continuation US 10,852,188 B2 carries an unusually dense Forrest M. Mims III prior-art citation set (LED-as-spectrally-selective-detector sun photometers, Applied Optics 1992; "Using LEDs as Light Detectors," Popular Electronics 1977; etc.). That art targets the core "LED/photodiode as spectrally selective detector + decay-time readout" concept and is the obvious starting point for a § 102/§ 103 attack — and it was flagged in the family's prosecution history, not in any PTAB record.

Pattern signals. No petitioner has filed even once, so there is no serial-petition or multi-petitioner pattern; no defensive aggregator (Unified Patents or similar) appears anywhere in the chain; and the patent owner (the U.S. Navy / Office of Naval Research) has never pursued a PTAB appeal on this patent. The owner profile is the key signal: this is a government-owned, non-practicing-entity-free patent. Government labs generally do not run assertion programs, which is the most plausible structural explanation for the total absence of IPRs — well-asserted commercial patents attract IPRs; a Navy research-tool patent that is not being enforced does not. Treat the absence as "not commercially asserted," not "weak."

Recommended next steps

1. If you have received a demand letter or complaint citing US 9,945,722: do not assume a prior IPR exists to lean on — none does. You have a clean slate to file the first-ever IPR. Remember the statutory gate: under 35 U.S.C. § 315(b) a petition is time-barred if filed more than one year after you are served with a complaint alleging infringement (service of a demand letter alone does not start that clock). If litigation is already underway, scope your § 314(a)/Fintiv and § 325(d) posture against the Director's current discretionary-denial practice.

2. Build the invalidity theory around the family prosecution record. Start with the Mims LED-sun-photometer corpus cited in US 10,852,188 B2's IDS/prior-art listing, then add the examiner-cited references of record here — US 7,135,848 and US 7,250,747 (Xytrans radiometer sensor/linearization), EP 1,923,679 A1 (Pixartis autocalibrating multichannel radiometer), US 2008/0259318 A1 (Everfine multi-channel array spectrometer), US 2012/0280114 A1 (Rodrigues), US 2015/0051498 A1 (Hypermed) — and the non-patent Parry NRL news release (2013-08-12). Several of these address the same "linearity over orders of magnitude" problem the claims recite.

3. Exploit the § 112 angles IPR cannot reach. The granted claims contain functional/result-oriented language that is vulnerable to § 112 grounds in district court — e.g., claim 1's requirement that the sensor "provides linearity of measurement sensitivity to incoming light … at over three orders of magnitude of incoming light," and claim 5's "free of shock sensitivity" and "long-term monitoring operational capability … ranging from less than about 2 years up to about 10 years." These read as statements of result/purpose with potentially indefinite boundaries, and an IPR cannot reach them. Pair a § 112 indefiniteness/enablement attack with the § 102/§ 103 IPR.

4. Set expectations on timing. No proceeding is pending, so there are no trial-stage milestones (no institution-decision deadline, no oral hearing date, no statutory FWD deadline) to track for this patent. Any timeline would begin only if you file.

Uncertainty statement. I relied on the structured ODP "no proceedings on file" data plus open-web corroboration, and found no AIA trial for US 9,945,722. I could not directly query PTAB E2E or the CAFC docket in this session; a very recently filed petition not yet indexed by ODP would not have been caught. Subject to that caveat, the patent has no AIA trial history of any kind.

I'll reconstruct the assignment chain. The authoritative patent text already contains Google Patents legal-event records with reel/frame numbers; let me verify and look for any post-issuance transfers or correspondent data.

Ownership Reconstruction — US Patent 9,945,722 B2

Sourcing note. The records below are drawn from the authoritative full patent text in my working set (Google Patents "Legal Events" and Assignment records, which mirror the USPTO Assignment Center reel/frame data). I could not load the Assignment Center detail pages directly (assignmentcenter.uspto.gov / assignment.uspto.gov), so I have not fabricated the correspondent field — I state plainly where it is unavailable. Google Patents does not expose a correspondent-of-record field, and no open-web mirror I could reach published one.

Assignment records do exist for this patent, so I proceed past the "no records" checkpoint.

Inventors

Pattern assessment: All five are federal employees of a single institution (NRL). This is the classic government-invention fact pattern, not a startup/operating-company pattern:

• No inventor departed the assignee; there is no fire-sale precursor. The re-recordings in 2020 simply re-paper the same 2014 execution dates (2014-05-12 to 2014-05-19), i.e., no change of beneficial ownership.
• Titles/assignments were executed before the 2014-07-02 filing (assignment signing began 2014-05-12), consistent with 37 CFR 3.1 statutory assignment of federal-employee inventions.

⚠️ Anomaly flagged (not auto-corrected): Reel 033385/0273 lists an assignor "CHALUPA, LEO M." who is not a named inventor on US 9,945,722 (nor on its continuation US 10,852,188). Leo M. Chalupa is a neuroscientist (vision science), not a co-inventor of record here. Because the assignment is recorded as one instrument covering "HOHEISEL, RAYMOND; CHALUPA, LEO M.," this is most consistent with either (a) a multi-invention omnibus Navy assignment recorded under a shared reel/frame, or (b) a recording error. I report it as-is; it is not evidence of an ownership defect, since the Navy's ownership is independently established by reels 033384/0889 and 052973/0917.

Original assignee

The United States of America, as represented by the Secretary of the Navy — Department of the Navy / Office of Naval Research; current-assignee string also styled "US Gov't Represented by Secretary of the Navy, Chief of Naval Research" and (2020 recording) "The Government of the United States of America, as represented by the Secretary of the Navy, Virginia."

• Primary line of business: National defense / naval research and development (federal government), not a commercial manufacturer.
• Did they ship a product embodying the claims? No commercial product. The patent describes an in-house low-cost (~<$20) research/military instrument (spec: dynamic range 0.01–2 suns, 100 µW operation, ~6×4×2.5 cm case). Its public footprint is a government news release (Parry, "NRL Develops Low Cost, High Efficiency Solar Sensor," NRL News Release, Aug. 12, 2013 — the sole non-patent citation of record). The Government commercializes federal inventions by licensing under 35 U.S.C. § 207 / 37 CFR Part 404 (Federal Register notices of intent to grant licenses), not by selling products.
• Current status: Operating (the U.S. Government is a perpetual, solvent entity). Patent Active, adjusted expiration 2036-04-24; 4th-year maintenance fee paid 2021-08-13; fee reminder mailed 2025-12-08.

Assignment timeline

Correspondent (attorney/agent of record) for each recording: NOT EXPOSED in any source I could retrieve. The granted continuation US 10,852,188 B2 lists its prosecution attorney/agent/firm of record as "US Naval Research Laboratory," which implies the recordings were likewise handled internally by a Navy patent attorney — but I will not assert a named individual correspondent without a reel/frame citation.

1. Executed 2014-05-12 → 2014-05-19 / recorded 2014-07-24 — Reel 033384/0889

• Conveyance: Assignment (ASSIGNMENT OF ASSIGNORS INTEREST)
• Assignor: SCHEIMAN, DAVID A.; LORENTZEN, JUSTIN R.; JENKINS, PHILLIP P.; AND OTHERS (Walters, Robert J.)
• Assignee: US Gov't Represented by Secretary of the Navy, Chief of Naval Research
• Correspondent: not exposed (see note above)
• Context: Initial statutory assignment of federal-employee inventors to the Government — original ownership vesting, not a transfer.

2. Executed 2014-05-12 → 2014-07-10 / recorded 2014-07-24 — Reel 033385/0273

• Conveyance: Assignment (ASSIGNMENT OF ASSIGNORS INTEREST)
• Assignor: HOHEISEL, RAYMOND; CHALUPA, LEO M.
• Assignee: US Gov't Represented by Secretary of the Navy, Chief of Naval Research
• Correspondent: not exposed
• Context: Companion original ownership assignment for the remaining named inventor; contains the Chalupa anomaly flagged above.

3. Executed 2014-05-12 → 2014-05-19 / recorded 2014-07-28 — Reel 033428/0567

• Conveyance: Assignment — CORRECTIVE ASSIGNMENT (corrects prior recording at Reel 033384/0889)
• Assignor: SCHEIMAN, DAVID A.; LORENTZEN, JUSTIN R.; JENKINS, PHILLIP P.; AND OTHERS
• Assignee: US Gov't Represented by Secretary of the Navy, Chief of Naval Research
• Correspondent: not exposed
• Context: Administrative correction only — deletes the word "MAY" (leaving filing date "July 2, 2014") and corrects Jenkins's mid-initial from "R" to "P." Not a change of ownership.

(2018-03-28 — STCF "Information on status: patent grant" is a prosecution/status event, not an assignment; excluded from the chain.)

4. Executed 2014-05-12 → 2014-05-19 / recorded 2020-06-18 — Reel 052973/0917

• Conveyance: Assignment (ASSIGNMENT OF ASSIGNORS INTEREST)
• Assignor: SCHEIMAN, DAVID A.; LORENTZEN, JUSTIN R.; JENKINS, PHILLIP P.; AND OTHERS
• Assignee: The Government of the United States of America, as represented by the Secretary of the Navy, Virginia
• Correspondent: not exposed
• Context: Confirmatory / re-recording of the same 2014 inventors' assignment (identical signing dates), restyling the assignee name for the post-issuance record. Not a new transfer, and not a cascade — the beneficial owner never changed.

(2021-08-13 "MAFP" maintenance-fee payment and 2025-12-08 "FEPP" fee reminder are administrative fee events, not assignments.)

Net ownership chain: five federal inventors → U.S. Government (Navy). No outward transfer in ~12 years.

Timeline diagram

timeline
    title Ownership of US 9945722
    2013 : Provisional filed by five NRL inventors
    2014 : Filed by US Navy
         : Assigned to US Navy
         : Reel 033384 frame 0889
         : Reel 033385 frame 0273
         : Corrected by Reel 033428 frame 0567
    2018 : Patent issued
    2020 : Confirmatory re-recording
         : Reel 052973 frame 0917
    2021 : Fourth year fee paid
    2025 : Fee reminder mailed

NPE / troll-pattern signals

Additional check — litigation/assertion footprint: My searches for "9945722," "9,945,722," and title terms combined with infringement/lawsuit/license returned no litigation, no IPR/PGR, and no CAFC docket naming this patent. The only substantive third-party record is the grant entry at uspto.report (https://uspto.report/patent/grant/9,945,722), corroborating Navy ownership.

Verdict

Defensive / non-asserting.

The chain is a clean, single-owner, inventor-to-Government vesting path — Reel 033384/0889 (2014-07-24), Reel 033385/0273 (2014-07-24), corrected at Reel 033428/0567 (2014-07-28), and confirmatorily re-recorded at Reel 052973/0917 (2020-06-18, bearing the same 2014 execution dates) — all naming the U.S. Government / Secretary of the Navy as assignee. There is zero evidence of a shell-entity transfer, NPE assignee, cascading chain, or pre-suit (or any) litigation, and the 12-year absence of any outward assignment is itself the finding: the original assignee still owns the patent outright. I do not score this as an NPE of any confidence level, and I note explicitly that the "defensive aggregator" label is a closest-fit classification only — the Navy is a government owner, not an RPX/AST-style defensive aggregator; the operative conclusion (no assertion, no NPE) is unchanged.

Verification links: USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (search patent number 9945722); legacy mirror — https://assignment.uspto.gov/patent/index.html. Caveat: I could not load these pages myself, and the correspondent-of-record field remains unverified for every recording above; treat signal #3 as unresolved rather than negative.

Cross-reference note (per your instruction to flag contradictions): This analysis is consistent with the previously generated patent summary. The one item the earlier summary flagged — the "CHALUPA, LEO M." assignor who is not a named inventor — recurs here at Reel 033385/0273 and remains unresolved in the public record; I have not reconciled it and report it as-is.

I'll search the USPTO database and verify each cited reference for US 9,945,722. Let me run these in parallel.

Prior-Art Analysis — US 9,945,722 B2 ("Ultra low power solid state spectral radiometer")

Sourcing note: I searched for the literal identifier 9945722 and confirmed it resolves to US 9,945,722 B2 (the Navy radiometer), not to any similar number. The authoritative patent full text in my working set was used for claim language, dates, and the front-page citation list. I then ran open-web searches (Google Patents, FreePatentsOnline, Justia, ipqwery, patenthub) against each cited reference. Some supplementary searches for the last three references (US 2012/0280114, US 2015/0051498, and the Parry NPL) hit the tool step-limit and returned no content; for those I rely on the authoritative front-page data and clearly label where I could not independently corroborate.

Minor date note (not auto-corrected): the task header says "Current Date: April 26, 2026," while my operating context states today is 2026-09-30. I flag the discrepancy rather than reconcile it; it does not affect the analysis below.

Cross-reference: This is consistent with the earlier "Patent summary" and "Litigation summary" sections (same patent identity, same 6 patent citations + 1 NPL, no litigation). No contradiction to flag.

A. The cited references of record (front page of US 9,945,722)

Applicable law: US 9,945,722's priority date (2013-09-20) post-dates the AIA's March 16, 2013 change, so AIA 35 U.S.C. § 102(a)(1)/(a)(2) governs. Note the front page marks US 2015/0051498 A1 with the "*" (cited by the examiner) indicator; the others are listed without that marker.

Overarching observation that drives the entire § 102 analysis: References 1–3 (Xytrans, Pixartis) are microwave/millimeter-wave (MMW) radiometers. In that literature "radiometer" means an instrument that measures thermal radio-frequency noise power using square-law detector diodes and amplifiers — not an optical/visible/UV spectral radiometer measuring light by photon absorption across wavelength bands. US 9,945,722 is squarely an optical spectral radiometer in which a semiconductor junction absorbs light whose wavelength is set by its bandgap energy (Claims 1, 6, 8). That technology gulf is decisive for novelty: an RF radiometer cannot, in a single reference, disclose the claimed photon-absorption/bandgap-selected sensing architecture.

B. Reference-by-reference analysis

1. US 7,135,848 B1 — Xytrans, Inc., "Highly integrated radiometer sensor cell"

• Dates: filed 2005-12-08 (app. US 2006046882; pub. US 2007/070383); granted 2006-11-14; priority 2005-12-12. Inventor: Danny F. Ammar.
• Description: A miniaturized MMW radiometer sensor cell: a surface-mountable base with an RF feed and at least one MMIC containing a quadrature hybrid, an LNA, and a detector circuit producing a detected signal. (Corroborated at ipqwery and Justia's Xytrans assignee profile.)
• § 102 mapping: Could at most be argued against the generic "spectrally sensitive circuit element section … diode components" language of Claim 1 and the "transparent housing encapsulated diode" language of Claims 6/8 (the sensor cell is a packaged, integrated component). It discloses a detector diode and a housing, but:
  • the diode is an RF square-law detector, not a photon-dependent charge-generation/recombination element absorbing light within a bandgap-defined wavelength;
  • no CPU charging a diode capacitance and timing a voltage decay;
  • no case with a removably fastened lid.
• Conclusion: Does not anticipate Claims 1, 6, or 8. Best treated as background art; at most a § 103 secondary reference for "small integrated diode sensor package."

2. US 7,250,747 B1 — Xytrans, Inc., "Radiometer measurement linearization system and method"

• Dates: priority 2005-12-12; granted 2007-07-31. Inventor: Danny F. Ammar. (Google Patents: https://patents.google.com/patent/US7250747)
• Description: A dual-channel quadrature-hybrid amplifier MMW radiometer with a detector circuit, a log-ratio amplifier that linearizes the detected output to broaden dynamic range, an integrator summing over an observation period, and a digital acquisition/processing circuit.
• § 102 mapping: The only textual overlap with US 9,945,722 is the linearity / wide-dynamic-range element of Claim 1 ("linearity … at over three orders of magnitude of incoming light"). But the reference's linearity is achieved by a log-ratio amplifier on an RF signal, and its "incoming signal" is antenna thermal noise, not light. It lacks: light/bandgap spectral selection, the CPU-driven charge/discharge decay-timing measurement, and the case-with-lid.
• Conclusion: Does not anticipate any claim. Useful only as a § 103 reference for the concept of extending radiometer dynamic range via linearization — a different mechanism from the claimed capacitance/decay-time approach.

3. EP 1,923,679 A1 — Pixartis SA, "Autocalibrating Multichannel Radiometer"

• Dates: priority 2006-11-16; published 2008-05-21 (PCT WO 2008/059444 A3). Inventors incl. Boisbouvier, Butz, Seetharamdoo, Vellas, Werquin. (Google Patents: https://patents.google.com/patent/EP1923679A1/en)
• Description: A multichannel microwave radiometer (temperature sensing of an object by its thermal noise power): parallel independent amplification channels, a correlator, square-law detector, integrator, A/D converter, an auto-calibration framework with noise diodes and a precision temperature sensor, operating across n frequency bands (via band-pass filters/power dividers).
• § 102 mapping: It shares the concepts of (a) a multichannel, multi-band radiometer and (b) calibration/data processing. These superficially touch Claim 8's "plurality of … different wavelength bands." Critically, however, Pixartis achieves multi-band operation by splitting a broadband RF signal with power dividers and band-pass filters, not by using diodes of different bandgap energies in one transparent housing, and it measures microwave noise, not light. The transparent-housing/bandgap/decay-time architecture is absent.
• Conclusion: Does not anticipate Claims 1, 6, 8 (or dependents). Possible § 103 reference for multi-channel/multi-band radiometer architecture and calibration only.

4. US 2008/0259318 A1 — Everfine Photo-E-Info Co., Ltd., "Multi-channel array spectrometer and method for using the same"

• Dates: filed 2007-12-27; published 2008-10-23 (granted as US 7,978,324 B2, 2011-07-12). Inventors: Jian-gen Pan; Qian Li.
• Description: An optical multi-channel array spectrometer with a bandpass filter wheel, grating, array detector (CCD/PDA), a silicon-photodiode reference detector, a microcomputer, and an A/D converter. It expressly claims "wide dynamic range with high linearity" and states that good silicon photodiodes "can reach <0.2% in 8 orders of magnitude." (Google Patents / patents-review.)
• § 102 mapping: This is the only cited reference that is genuinely an optical/spectral instrument and the only one that expressly ties a silicon photodiode (a photon-dependent charge-generation element) to linearity over multiple orders of magnitude — the exact numeric limitation recited in Claim 1 ("linearity … at over three orders of magnitude of incoming light"). It also has a microcomputer and data interface, plausibly reading on Claim 1's "CPU … timing apparatus … data storage device."
• Gaps that defeat anticipation of Claim 1: (i) sensing is via a grating + array detector + mechanically rotated filter wheel — a moving-part, dispersive architecture contrary to the patent's "no moving parts / no sophisticated optics" gist, and it does not teach charging a diode/capacitor and measuring a voltage-decay time as the intensity metric; (ii) no bandgap-selected diode; (iii) no case with a removably fastened lid as claimed.
• Conclusion: Everfine is the strongest § 102 candidate on Claim 1's core (photodiode + wide-dynamic-range linearity + microcomputer), but because anticipation requires every element in a single reference and the claim's distinctive charge/decay-timing and case-with-lid limitations are missing, a clean § 102 rejection is unlikely. It is far better positioned as a § 103 reference. It does not anticipate Claims 6 or 8 (no external-capacitance decay tuning; no multiple different-bandgap diodes in one transparent housing).

5. US 2012/0280114 A1 — Djalma Luiz Rodrigues, "Multifunctional radiometer, hospital equipment, multiuse measurement tool, system and method for measuring irradiance in phototherapy"

• Dates: priority 2009-11-30; published 2012-11-08.
• Description (from title/front page): An irradiance-measuring radiometer for phototherapy, described as a multifunctional/multiuse hospital measurement tool and system. (I could not retrieve the specification text within the tool budget; the description here derives from the title and front-page listing only.)
• § 102 mapping: An optical-irradiance radiometer is at least in the right domain (unlike refs 1–3). It could conceivably touch Claim 1's broad "spectrally sensitive circuit element + CPU + data storage" concept if it discloses a photodiode front end with digital readout. I found no evidence it discloses: multi-band spectral distribution, bandgap-selected diodes, the charge/decay-time measurement, an adjustable external capacitance, or the case with removable lid.
• Conclusion: Not shown to anticipate any independent claim. Treat as background only; I explicitly flag that I could not perform a full-text § 102 element-by-element comparison due to the search step-limit.

6. US 2015/0051498 A1 — Hypermed Imaging, Inc., "Methods and apparatus for coaxial imaging of multiple wavelengths"

• Dates: priority 2012-06-05; published 2015-02-19 (marked "*" = cited by examiner).
• Description (from title/front page): Apparatus/methods for coaxially imaging multiple wavelengths (e.g., tissue oximetry/spectral imaging). (Full text not retrieved within tool budget.)
• § 102 category — important timing point: Its publication date (2015-02-19) is after US 9,945,722's effective filing date (2013-09-20 priority / 2014-07-02 filing), so it is not available under § 102(a)(1). Its only potential § 102 role is § 102(a)(2) (a published US application effectively filed 2012-06-05, before the Navy patent's effective filing date). Even so, a multi-wavelength coaxial imaging device is architecturally remote from the claimed single/multi-diode charge-decay spectral radiometer.
• Conclusion: Does not anticipate any claim; at most a § 102(a)(2)/§ 103 reference on the "multiple wavelengths" concept. Flagged for completeness.

7. NPL — D. Parry, "NRL Develops Low Cost, High Efficiency Solar Sensor," NRL News Release, Aug. 12, 2013

• Dates: published 2013-08-12 — ~5 weeks before the 2013-09-20 provisional-priority date.
• Description: A US Naval Research Laboratory press release announcing a low-cost, high-efficiency solar sensor developed at NRL — i.e., a public disclosure by the inventors' own institution covering the same general subject matter.
• § 102 mapping: This is the most potentially dangerous § 102(a)(1) reference, because it is a printed publication/public disclosure that predates the priority date by only five weeks and concerns the very same NRL sensor technology. If the release discloses the claimed matter (a multi-band diode spectral radiometer with CPU readout and data storage), it could be novelty-destroying under § 102(a)(1) as the inventors' own disclosure. Its relevance maps to Claims 1 and 8 (multi-channel/bandgap-diode spectral sensor architecture). Caveat: I could not retrieve the release text within the tool budget, so I cannot confirm the level of technical disclosure; a five-week margin also raises grace-period / § 102(b)(1)(A) exception considerations (the inventors' own disclosure ≤ 1 year before filing may be excepted under AIA).
• Conclusion: Potentially relevant to Claims 1 and 8 if it discloses the inventive architecture; needs full-text verification. This is the reference I would prioritize for follow-up.

C. Bottom line

• No cited reference anticipates (35 U.S.C. § 102) independent Claims 1, 6, or 8. Anticipation requires a single reference disclosing every element, and each reference is missing at least one essential limitation — most decisively the charge-then-measure-voltage-decay-time measurement (Claims 1/6), the switchably adjustable external capacitance in parallel/series/open (Claim 6), or the multiple different-bandgap diodes co-encapsulated in one transparent housing (Claim 8).
• Technical-domain mismatch is the key theme: Refs 1–3 are microwave/MMW noise radiometers, not optical spectral radiometers; only Refs 4 (Everfine), 5 (Rodrigues) and 6 (Hypermed) are optical, and only Everfine explicitly couples a silicon photodiode to wide-dynamic-range linearity (8 orders of magnitude) — the closest textual overlap with Claim 1's "over three orders of magnitude" limitation.
• Best § 102 candidates, in order: (1) Parry/NRL release (same technology, predates priority by 5 weeks — but needs full-text confirmation and is subject to the inventors'-own-disclosure exception); (2) Everfine US 2008/0259318 (optical photodiode + linearity + microcomputer). Both still fall short of full anticipation of the independent claims as written.
• Most of these references are better characterized as § 103 obviousness art (Xytrans '747 for linearization, Pixartis for multichannel/multi-band architecture, Everfine for photodiode linearity) rather than § 102 anticipation.

Caveats (stated explicitly, not hedged silently):

1. My tool budget was exhausted before I could pull the full specifications of US 2012/0280114 A1, US 2015/0051498 A1, and the Parry NPL; their § 102 element mappings are therefore provisional and based on title/front-page metadata.
2. I did not query the USPTO PatentCenter/PTAB API directly; this is an open-web analysis. A definitive anticipation opinion would require full-text element-by-element charts against each reference's specification and claims.
3. Per operating rules I did not auto-correct any identifier; all numbers are as recorded.

I'll pull details on the six references cited of record so the § 103 combinations are grounded in their actual disclosures.

Now the two remaining references of record, plus technology-specific prior art on the capacitance-decay sensing technique.

Obviousness Analysis — US 9,945,722 B2 under 35 U.S.C. § 103

Note on scope of record. This analysis uses the six references cited of record on the face of US 9,945,722 and the single non-patent citation (Parry, NRL News Release, Aug. 12, 2013), as instructed. My searches retrieved substantive disclosure text for five of the six references; US 2015/0051498 A1 (Hypermed Imaging) I could only characterize bibliographically — my tool calls hit their step limit before I could pull its specification. I flag that gap rather than infer its content. All attribution of subject matter to the other five references below is grounded in the retrieved text and the URLs cited.

1. Legal framework applied

Obviousness is assessed under Graham v. John Deere Co., 383 U.S. 1 (1966), and KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), considering: (a) the scope and content of the prior art; (b) the differences between the prior art and the claims; (c) the level of ordinary skill; and (d) objective evidence of non-obviousness. Under MPEP 2143, a prima facie case requires an articulated reason to combine, drawn from the non-exhaustive KSR rationales — (A) combining known elements according to known methods to yield predictable results; (B) simple substitution of one known element for another; (C) use of a known technique to improve a similar device in the same way; (D) applying a known technique to a known device ready for improvement; (E) "obvious to try"; and (F) design incentives and market forces.

Level of ordinary skill (proposed). A person having ordinary skill in the art ("PHOSITA") here would hold at least a B.S. in electrical engineering, applied physics, or optics, plus roughly 2–5 years of experience designing photodetector-based instrumentation — i.e., someone comfortable with (i) photodiode/photoconductive detector front ends, (ii) microcontroller-based data acquisition with ADCs, clocks, and nonvolatile storage, and (iii) radiometric calibration and spectral band selection. This is a modest skill level: the patent's own specification concedes the invention uses "no sophisticated electronics or optics" and "direct readout from CPU 101 without external circuitry."

Claim construction points that drive the analysis.

• "spectrally sensitive circuit element section" — broad; the patent defines a "diode 102" functionally as "an element with spectrally dependent charge generation or recombination properties and/or spectrally dependent electrical resistance properties." That captures ordinary photodiodes, photoconductors (CdS cells), and Schottky photodiodes — all long-known.
• "linearity … at over three orders of magnitude of incoming light" — a result-oriented limitation. This is critical: the record contains an express statement that photodiodes already achieve far more.
• "timing apparatus" — a clock/oscillator, functionally a microcontroller crystal or internal oscillator.
• "charge characteristic components … consisting of diode components and resistance capacitance components" — an RC/diode network; not narrowly limited to any particular topology.

2. Scope and content of the prior art of record

3. Claim 1 — Primary § 103 combination

Proposed rejection: Claim 1 is obvious over Everfine (US 2008/0259318 A1) in view of Pixartis (EP 1,923,679 A1) and Rodrigues (US 2012/0280114 A1), optionally further in view of Xytrans (US 7,135,848).

Element-by-element mapping:

Motivation to combine:

1. Same field, same problem — Everfine, Pixartis, and Rodrigues are all radiometric/photometric measurement devices; the field of endeavor is identical, and the references are a fortiori combinable. KSR rationale (A).
2. Everfine already supplies the "linear photodiode spectrometer" core; Pixartis supplies the missing multi-band channelization with one detector per band and simultaneous/consecutive readout — a known technique applied to a known device "ready for improvement." KSR rationales (C)/(D). Pixartis's own words ("measures simultaneously through parallel channels or consecutively through a switching cycle the calibrated temperatures at n different frequency bands") supply the very function claim 1 recites.
3. Rodrigues supplies the packaging/handheld integration and the compute-irradiance-from-sensor-data control unit, addressing the same stated objective of a single, low-cost, field-portable instrument.
4. Predictable results. Nothing in claim 1 requires anything more than aggregating known photodiode sensing, known multi-channel band splitting, known microcontroller digitization, and a known enclosure. KSR, 550 U.S. at 417 ("combination of familiar elements according to known methods … does not become nonobvious by virtue of … yield[ing] only predictable results").

Secondary observation undermining any "unexpected linearity" argument. The patent's principal asserted advantage — linearity "at over three orders of magnitude of incoming light" — is squarely anticipated in the art. Everfine teaches that a good silicon photodiode is linear to <0.2% over eight orders of magnitude, i.e., the claimed three orders are not merely known but conservative. Absent a showing that the claimed apparatus as a whole achieved some surprising increase attributable to the claimed structure, this limitation cannot support patentability.

4. Claim 6 — Diode with adjustable external capacitance

Proposed rejection: Obvious over the Claim 1 combination above further in view of the Xytrans '747 integrator/observation-period teaching and the general knowledge of RC time-constant / integration-time tuning.

Claim 6 recites: (i) a photon-dependent charge-generation/recombination semiconductor diode with inherent junction capacitance; (ii) encapsulated in a transparent housing; (iii) an external capacitance switchably connected in parallel with, in series with, or open across the diode; (iv) the external capacitance adjusts the decay time "according to expected levels of light intensity."

• (i)–(ii): An encapsulated photodiode in a transparent package is the most conventional element in the art; Everfine's silicon photodiodes and Xytrans '848's encapsulated, cover-sealed miniature cell with its "similar in concept to light sensing diodes" remark together map to this.
• (iii)–(iv): The concept of varying integration time / detector time constant to match expected signal level is a classic radiometric design practice, and the record supports it. Everfine expressly enlarges dynamic range by switching density filters into the optical path and by band-by-band versus full-spectrum measurement modes — i.e., changing the effective measurement sensitivity to handle different light levels. Pixartis teaches a defined "observation time interval" over which signal is integrated, and switching cycles with multiple phases. Xytrans '747's integrator "sums the signal over an observation period," and its whole point is adapting the response to the input signal's dynamic range. A PHOSITA seeking to operate one instrument across widely varying illumination would predictably implement a switchable capacitor bank to change the RC decay/integration constant — switchable capacitor banks across a detector node are routine circuit design.

Candid weakness / where this ground is exposed. My verified record does not contain a reference that expressly discloses a switchable capacitor placed across a photodiode to tune its discharge decay time. The combination therefore relies on: (a) the general knowledge of RC time-constant selection, and (b) an inference from "integration period" and "density filter" teachings. If the applicant can show the art did not contemplate using the detector's own capacitance as the integrating element with a switchable parallel/series capacitor for range selection, claim 6 is the strongest candidate for survival among the independent claims. Rejecting claim 6 robustly would require adding an art reference teaching variable/switchable detector time-constant or integrating-capacitance selection (e.g., a light-to-frequency converter with sensitivity-select inputs, or a switched-integrator readout IC). I was unable to pull such a reference here; this should be verified before relying on this ground in an office action.

5. Claim 8 — Plurality of different-bandgap diodes in a single transparent housing

Proposed rejection: Obvious over Pixartis (EP 1,923,679) in view of Everfine (US 2008/0259318) and Xytrans (US 7,135,848), further in view of the admitted/known multi-diode LED-style transparent package.

Claim 8 recites a plurality of diodes (at least two / three / four), each of different bandgap, reflection, and absorption properties, all encapsulated in a single transparent housing "large enough to fully contain" them.

• Multiple detectors, one per spectral band: Pixartis teaches n detectors, one per frequency sub-band, with reference to "the n desired frequency bands." Everfine teaches a multi-channel array with per-band filters and states the number of filters "can be varied from 2 to 20."
• Different semiconductor bandgaps as the band-selection mechanism: Everfine expressly motivates this. It teaches that optical filters are the conventional way to make a photodiode spectrally selective, and that this "technique is complicated and the cost is high," and that the filter approach "can not prevent producing of stray light." A PHOSITA reading Everfine would be directly motivated to seek an alternative intrinsic band-selectivity — i.e., choosing photodiodes whose semiconductor bandgap already defines the absorption edge (Si, InGaAs, GaP, SiC, GaN, CdS, etc.), eliminating filters. KSR rationale (F) — design incentive and cost pressure — plus (C).
• Single transparent multi-die package: The patent specification itself concedes that a three-diode transparent package "is commercially available and used for LEDs." Multiple semiconductor dies (including different-bandgap dies) co-encapsulated in one transparent molded package is the ordinary construction of multi-color LED and photodetector packages. Xytrans '848 teaches the analogous miniaturized, encapsulated, multi-cell architecture ("a plurality of radiometer sensor cells can be surface mounted on a printed wiring board to form a radiometer device"; "designed similar in concept to light sensing diodes"), supplying the miniaturized-package motivation and the express transfer across the RF/optical boundary.
• Group of alternatives is permissive. Claim 8's "at least two … and at least three … and at least four" is a Markush-style enumeration; a rejection needs to reach only one member (two diodes in one package).

Motivation summary: cost and complexity of filters (Everfine) + demonstrated need for multiple simultaneous spectral channels (Pixartis) + known miniature multi-die transparent packaging (admitted; Xytrans) = the claimed multi-bandgap, single-package sensor with predictable results.

6. Dependent claims

7. Objective evidence and counter-arguments considered

No objective evidence of non-obviousness is presently established in this record. The patent is U.S. Government–owned (Secretary of the Navy / ONR), and the prior sections of this analysis found no litigation, no PTAB proceeding, and no CAFC appeal involving US 9,945,722. That means there is (a) no adjudicated validity finding, (b) no licensing/acquisition nexus probative of non-obviousness, and (c) no presumption of validity pressure-testing these claims. Absent evidence of unexpected results, long-felt but unmet need with a nexus, or industry praise, the Graham factor (d) is neutral.

Likely applicant rebuttals and their vulnerability:

1. "Everfine is a grating spectrophotometer; a filter/grating spectrometer is not a bandgap-selective radiometer." Weak — Everfine also teaches a reference photodiode and photodiode-based radiometry; Pixartis and Rodrigues bridge the gap; and the patent itself disparages sophisticated optics ("no sophisticated electronics or optics"), removing a distinction it might otherwise claim.
2. "The references are RF/microwave, not optical." Weak — Xytrans '848 expressly frames its sensor cell as "designed similar in concept to light sensing diodes," and Xytrans '747's purpose is generic detector linearization, a function agnostic to wavelength. Under KSR, a PHOSITA would consult these for their linearization/dynamic-range solutions.
3. "Nothing suggests using the photodiode's own capacitance as an integrating element." This is the strongest point and is where Claim 6 (and Claim 10's "during measurement" adjustment) most credibly differentiate. Asserting § 103 on Claims 6 and 10 requires art teaching switchable/variable detector integration capacitance or time-constant selection, which I could not confirm from the six-reference record.

Also worth flagging (not auto-corrected): the recorded assignment documents for this patent list an assignor "CHALUPA, LEO M." who does not appear among the five named inventors — an unresolved discrepancy already noted in the bibliographic section of this analysis and relevant only insofar as it bears on § 102(b)(1)(A) inventor-disclosure questions respecting the Aug. 12, 2013 NRL news release.

8. Bottom line

• Claim 1 — strong prima facie § 103 case over Everfine + Pixartis + Rodrigues (optionally + Xytrans '848). Every element, including the asserted three-orders-of-magnitude linearity, is taught or rendered predictable.
• Claim 6 — prima facie case is weaker and record-dependent. The adjustable external capacitance and decay-time tuning are the most plausible non-obvious features in the patent. A rejection should not be relied on without a reference (or admission) teaching switchable detector integration capacitance / variable time-constant selection.
• Claim 8 — strong case over Pixartis + Everfine + Xytrans '848, reinforced by the specification's own admission that multi-diode transparent packages are commercially available for LEDs.
• Claims 2, 3, 4, 5, 7, 9 — routine design choices, predictable results, per KSR/MPEP 2144.05 / 2144.04 (result-effective variables and aesthetic/arbitrary dimensional changes).
• Claim 10 — depends on Claim 6; obvious only if the Claim 6 ground holds (with Pixartis's switching cycle supplying the "before/during/after" timing).

Confidence and limitations. High confidence in the mapping of Everfine, Pixartis, Rodrigues, Xytrans '848, and Xytrans '747 to the claim language, based on the retrieved texts cited above. Lower confidence on US 2015/0051498 A1 (not retrieved substantively) and on the Claim 6/Claim 10 grounds, where the six-reference record appears insufficient. I did not have access to a claims-construction order, prosecution history, or a search of non-of-record art (e.g., light-to-frequency converters with sensitivity selection, switched-integrator readout ICs, or the NRL technical publications underlying the Parry release); any final rejection on Claims 6 and 10 should be supplemented on that basis.