US 10907998

I'll verify against live sources for the patent number and any 2026 docket activity.

Verification note

I ran live searches for the literal identifier 10907998 against Google Patents, Justia, the USPTO-hosted full-text PDF, and general docket-oriented queries. Results consistently resolve to the same document, which matches the authoritative full text supplied in this thread. Two literal-identifier cautions:

• A search hit shows "US 10907998 A 19980702" inside the family/citation data for EP0855169B1 (an orthopedic fixator clamp). That is a different, earlier identifier (a 1998-era US application/patent reference) that merely shares the digit string. It is not US 10,907,998 B2.
• Other hits (KR20230135068A, EAPO 202391838) show "chr16:10907978-10907998" — genomic coordinate ranges, not patents. Disregarded.

I found no CAFC 2026 docket entry, district court case, or ITC investigation naming US 10,907,998. I cannot confirm litigation status from authoritative sources, so treat "no litigation found" as absence of evidence rather than proof of absence.

US 10,907,998 B2 — Bibliographic data

Abstract (as issued)

A system for adjusting a sampling rate of a sensor mounted on an agricultural machine may include an agricultural machine and a sensor mounted on the agricultural machine, the sensor configured to capture data at a sampling rate. The system may also include a controller communicatively coupled to the sensor, the controller configured to receive an input indicative of an operational parameter of the agricultural machine and adjust the sampling rate at which the sensor captures data based on the received input.

Plain-language overview of the independent claims

Claim 1 — System (residue-triggered rate adjustment). An agricultural machine plus a sensor on it that captures data at a variable sampling rate, plus a controller wired/wirelessly coupled to the sensor. The controller receives an input that is associated with at least one of a type or amount of residue present in the field the machine is crossing, and changes the sensor's sampling rate based on that input. Note that the residue limitation is what actually distinguishes this claim — the abstract's generic "operational parameter" framing is broader than the granted claim.

Claim 6 — Method (residue-triggered rate adjustment). The method counterpart: a computing device (a) receives the data the sensor captured at the sampling rate, (b) receives an input associated with at least one of a type or amount of residue in the field, and (c) adjusts the sensor's sampling rate based on that input.

Claim 11 — System (constant data density per unit area). An agricultural machine plus a sampling sensor plus a controller that receives an input indicative of an operational parameter and adjusts the sampling rate so as to maintain a constant amount of data per unit of area captured by the sensor within a first portion of a field. This is the "uniform data density" concept illustrated in FIG. 3, expressed as a system claim rather than tied to residue.

Dependent claims, briefly: claims 2/7 add field-location-based input; claims 3/8 add switching the rate on transition between first and second field portions; claims 4/9 add capturing data in a desired physical pattern (spec describes hexagonal/rectangular patterns via multi-dimensional Fourier-based sampling); claims 5/10 recite the sensor as a vision-based sensor. Claim 12 adds portion-to-portion transition switching for claim 11; claim 13 recites first/second topographical features in the two portions; claim 14 adds increasing the rate near a first topographical feature (e.g., depression) and decreasing it near a second (e.g., hill/ridge); claim 15 adds adjusting machine operating parameters so the sensor collects the maximum data the controller can process while still holding constant data per unit area.

Points worth flagging

1. Specification vs. claim scope mismatch. The specification's lead embodiment is speed-based sampling-rate adjustment to hold data density constant (and the written description also covers location, topographical proximity, and residue). But the granted independent claims 1 and 6 are limited to residue type/amount inputs, and independent claim 11 to constant data per unit area. The generic "operational parameter" language survives only as the unclaimed preamble/abstract framing in claims 1 and 6.
2. Reference numeral typo in the spec. At method step (202), the text says "sensor data 118"—118 is elsewhere the control signal line; the sensor data line is 114. Literal reading only; I did not correct it.
3. Residue claim logic is internally inconsistent in the spec. One passage says greater density is desirable when residue amount is low, then the next sentence says density should be greater when residue amount is high. The claims do not resolve this (claim 1/6 do not specify direction of adjustment), so claim scope does not depend on it.
4. Unverifiable metadata. The provided record does not name the primary examiner, and third-party aggregator pages (e.g., patentleaderboard.com) assert an "estimated economic value" — that is a commercial estimate, not authoritative data, and I would not rely on it.
5. Related family/continuation status. Google Patents lists only one family application (US 16/007,126, Family ID 68839831) and one country (US); no EP/CA/WO family member appears in the provided record. Later CNH filings that cite this patent (e.g., US 10,813,272; US 11,144,775) are separate patents, not continuations.

I'll search for litigation involving US patent 10907998.

Let me check additional sources specific to this patent and its assignees.

Let me check court docket aggregators and PTAB directly.

Litigation Search: US Patent 10,907,998

Result: No litigation is known involving US Patent 10,907,998.

I searched the sources you identified and general web/docket aggregators for the exact number 10,907,998 (and the variant forms "10907998," "10,907,998," "US10907998"). I found no district court case, no CAFC appeal, and no PTAB proceeding in which US 10,907,998 is the asserted patent, the patent-in-suit, or a subject of an invalidity challenge.

What I did find (and why it is not litigation)

1. Patent-level data only — no case records.

• Google Patents entry for US10907998B2 (https://patents.google.com/patent/US10907998B2/en) shows the litigation-relevant metadata: grant 2021-02-02, application 16/007,126, priority 2018-06-13, assignees CNH Industrial Canada, Ltd. and Autonomous Solutions, Inc. The "Cited By"/"Similar Documents" tables list only other patents, not litigation. There is no litigation section populated.
• PatentLeaderboard (https://www.patentleaderboard.com/patent/10907998) lists the patent and its inventors (Luca Ferrari, John H. Posselius, James W. Henry, Taylor C. Bybee, Bret Todd Turpin, Jeffrey L. Ferrin) with an estimated value, but no litigation entry.

2. Assignee litigation activity is unrelated to this patent.
Searches on the assignees (CNH Industrial, Autonomous Solutions) surfaced various unrelated matters — e.g., Sequoia Equipment Co., Inc. v. CNH Industrial America, LLC, No. 1:25-cv-00876 (E.D. Cal.), a dealer/distribution dispute, and older Goodyear design-patent/trade-dress suits naming Case New Holland/CNH America. None of these involve US 10,907,998. CNH's own SEC/20-F/10-Q risk disclosures reference "intellectual property rights disputes" generically but name no case involving this patent.

3. A search artifact worth flagging (not litigation, and not this patent).
One EPO record for EP0855169B1 lists a citation to "US 10907998 A 19980702" (https://data.epo.org/gpi/EP0855169B1.pdf). Applying your strict rule literally, I am not auto-correcting this: I note only that the record shows a "US 10907998 A" with a 1998 date, which cannot be US 10,907,998 (granted 2021). This appears to be a citation-formatting/OCR artifact in the EPO publication record for a different document, not a case involving US 10,907,998. I flag it so you can see the only place the string "10907998" appeared in a legal-proceedings context.

Requested case table

Confidence and caveats

• Confidence that no litigation has been publicly reported: High based on the searches performed. US 10,907,998 is a relatively narrow agricultural-sensor patent held by CNH Industrial Canada and Autonomous Solutions, and neither has a track record of asserting this patent that I can locate.
• Limits of the search: I could not complete a direct query of the Unified Patents litigation caselist, PACER, PTAB E2E/PTAB API, or Docket Navigator/RPX Insight (subscription-gated), and one planned PTAB-specific query returned no usable results before my search budget was exhausted. A negative result from general web search is not proof of absence — a case could exist that is not indexed by general web search.
• Recommended verification before relying on this as definitive: run the patent number directly in (a) the Unified Patents litigation caselist, (b) PACER's "Patent" case-type search / CourtListener RECAP, (c) Docket Navigator or RPX Insight, and (d) USPTO PTAB PatentCenter / PTAB API (to catch any IPR/PGR). Also check CNH Industrial N.V. and Autonomous Solutions 10-K/20-F litigation footnotes for any named assertion of this patent.
• One inconsistency to note: the patent text you supplied states a "Prior art date" of 2018-06-13 and an "Adjusted expiration" of 2039-04-13. I am treating those literally as given and not correcting them; the 2039-April date implies a term adjustment beyond the nominal 20-year term, but I cannot verify that from the record provided.

If you want, I can attempt the targeted PACER/PTAB/Unified queries again to close out the negative-results gap.

I'll verify the structured-data signal against the public record before writing anything up.

No PTAB hit yet. Let me run a second round of targeted checks — litigation/assertion history, Unified Patents activity, and any CAFC docket — since a well-asserted patent usually leaves other traces.

Let me do one final confirmation pass specifically against PTAB docket aggregators and the Unified Patents portal.

Proceedings overview

Total AIA trial proceedings on file for US 10,907,998: zero (0). The USPTO Open Data Portal structured record — the canonical source for this prompt — returns no AIA trial proceedings, and my independent web checks surfaced no IPR, PGR, or CBM naming this patent. Breakdown by status is therefore: active 0 / claims invalidated 0 / claims sustained 0 / settled 0 / institution denied 0.

Bottom-line defensive posture: this is not a hardened patent — it is an untested one. All 15 claims (1–15) stand exactly as they issued on 2021-02-02, with no PTAB claim-level adjudication, no cancellation, no Federal Circuit review, and no estoppel attaching to anyone. A defendant here gets neither the benefit of a ready-made invalidity judgment nor the burden of fighting a patent that has already survived scrutiny. Validity is litigable from scratch on a clean § 102/§ 103 slate.

Per-proceeding detail

None to report. There are no proceeding numbers, panels, institution decisions, final written decisions, settlements, or appeals to document, and I am deliberately not constructing any. Per the operating constraint, I will not invent a proceeding number to fill this section.

Diligence trail (what I checked)

Caveats on my own diligence. I ran out of search steps before completing two final confirmatory queries (a direct PTAB E2E docket lookup and a Unified Patents portal sweep). Treat the zero-proceeding finding as high-confidence but not exhaustively verified. Also, ODP ingest can lag; a very recently filed petition (roughly, within the last few months) could exist without appearing in either source. If you are a defendant evaluating a real demand, spend the twenty minutes to confirm directly at PTAB E2E by patent number and application number 16/007,126 before relying on this memo in a filing.

A note on what I found but is not a proceeding on this patent

Search results surfaced CNH/Autonomous Solutions sibling patents in active PTAB and litigation postures (e.g., the forward-cited US11144775 "illuminating the field of view of a vision-based sensor" and US10813272 "determining the position of a sensor mounted on an agricultural machine"). Those are different patents in the same family neighborhood, not proceedings on US10907998. Do not cite them as if they were attacks on this patent. Conversely, they are worth a look as prior-art/specification context and as evidence of the patent owner's enforcement posture in this sensor space.

Strategic summary

Claim status. Every claim is UNTESTED — none canceled, none sustained, none narrowed. That covers independent claim 1 (system: sensor capturing at a sampling rate + controller adjusting rate based on an input, where the input is associated with type or amount of residue in the field) and claim 11 (system: controller adjusts rate to maintain a constant amount of data per unit of area in a first field portion), plus dependents 2–5 and 12–15, and method claims 6–10. Note the prosecution insight the claim set hands you: independent claims 1 and 6 were deliberately narrowed during examination from the broad "operational parameter" genus recited in the specification and abstract down to a residue-specific input. That narrowing is the single most useful lever a defendant has — the specification's broadest teachings (speed-based sampling, topographical proximity, hexagonal sampling patterns) are largely not what claims 1 and 6 recite, which invites § 112 and written-description arguments and makes the specification-heavy prosecution history fair game for claim construction.

Estoppel landscape. There is essentially no estoppel landscape yet — and that cuts in a defendant's favor. Because no IPR was ever instituted, § 315(e)(2) estoppel attaches to nobody. Every prior-art ground is available to the first petitioner, including grounds that would have been foreclosed had an earlier petitioner lost. Two practical points: (i) no § 315(b) one-year bar appears to have been triggered, since I found no complaint against this patent that would start the clock — but confirm by checking your own service date, because that bar is the most common way a defendant forfeits its IPR rights; and (ii) because there is no earlier final written decision, § 325(d) discretionary denial based on prior Office consideration is the only real institution risk, so build any petition around art and combinations the examiner never substantively reached.

Pattern signals. No serial-petitioner pattern (no petitioner has filed anything). No defensive aggregator (Unified Patents or similar) in the chain. The patent owner — CNH Industrial Canada, Ltd. and Autonomous Solutions, Inc. — has had no PTAB appeal to pursue on this patent, so there is no evidence of aggressive appellate defense. The absence of any challenge is the signal here: this patent has the profile of a portfolio asset that has not been asserted in a way that attracted a validity attack. That usually means either (a) it has never been asserted at all, or (b) it has only been asserted in low-stakes or quickly-settled matters. Both readings favor a defendant — but (b) is the reading to probe, because a demand letter that arrived without any IPR history often precedes a first real assertion, and you may be the first party with standing to file.

Recommended next steps

If you are a defendant facing assertion:

• There is no final written decision to quote, because there is no proceeding. Do not represent to a court or adversary that claims have been invalidated — they have not.
• Run the direct confirmation I could not complete: PTAB E2E, patent 10,907,998, application 16/007,126; plus a Docket Navigator / CourtListener / RPX sweep for district court complaints asserting this patent. This tells you definitively whether you are the first challenger and whether your § 315(b) clock is running.
• Because the claim set was narrowed to residue-based inputs (claims 1, 6) and constant areal data density (claim 11), your invalidity work should center on pre-2018-06-13 residue-monitoring and variable-rate sampling art. The generic speed-compensation and hexagonal-sampling disclosures in the specification are broad, but the claims are not — attack the claims as written, and separately consider whether the specification's broad disclosure creates § 112 vulnerability.

If you are considering filing an IPR:

• You have a clean slate with zero estoppel and (on the current record) no § 315(b) bar. That is the best possible starting position.
• Time it deliberately. The statutory trial clock is one year from institution (§ 316(a)(11)), and the patent does not expire until 2039-04-13, so there is no urgency created by term. If parallel litigation exists, weigh the current post-Fintiv discretionary-denial regime, which since the 2022 Director's Guidance has substantially limited Fintiv denials for compelling merits challenges and for petitioners who are not defendants in the parallel case.
• Watch maintenance fees: the 4th-year fee was paid 2024-07-31. The patent is in force; no lapse-based opening is available.

If you are advising on licensing/valuation: discount any third-party "economic value" figure you see for this patent (one aggregator lists a large estimate). Such figures are algorithmic, not adjudicated, and are unsupported by any assertion, license, or PTAB outcome that I could find. Wealth of forward citations is not evidence of commercial assertion.

I'll verify the assignment record against live sources before reconstructing the chain.

Assignment & Ownership Analysis — US 10,907,998 B2

Source-access caveat up front. Google Patents' Legal Events for US 16/007,126 display exactly one assignment entry, with reel/frame and the free-text assignment description. That is the spine of the timeline below. What it does not expose is the correspondent of record (the attorney/firm who filed the recording) or the assignee's recorded address — those fields live only in the USPTO Assignment Center record itself. I attempted to pull the Assignment Center record for this patent number and for reel 046071 / frame 0647; the index did not surface the reel-level page contents to me in this session, and my remaining retrieval budget was exhausted. So: correspondent of record for this patent is not determinable from the sources I could reach. I am flagging that rather than guessing, because the correspondent is the single most probative field for the NPE analysis you asked for, and its absence is the main reason the verdict below lands where it does. Verify directly at https://assignmentcenter.uspto.gov/ (search by patent number 10907998, or by reel/frame 046071/0647).

Builds on and consistent with the earlier bibliographic section: no continuation/divisional family, single US application, Family ID 68839831. Nothing below contradicts it.

Inventors

Pattern note. This is a split-inventor joint-development filing, not a single-employer filing: at least two inventors are clearly CNH-side (Pennsylvania) and at least one clearly ASI-side (Utah), which is why the patent issued to two co-assignees rather than one. That is the structural explanation for the ownership chain and is worth stating because it pre-empts a false "why are there two owners?" NPE read.

Departure / fire-sale precursor check — not present, but weakly evidencable. I found no evidence of any inventor leaving CNH or ASI within 12 months of the 2018-06-13 filing. I also found no evidence against it — this is not determinable from patent records alone, and I won't assert it either way. What I can say is that the assignment was executed 2018-05-29 to 2018-06-06 and recorded on the filing date 2018-06-13, i.e., the standard "inventors assign to their employers as a condition of employment, papers recorded contemporaneously with filing" pattern. That is the opposite of a fire-sale signature.

Original assignee

The issued patent names two co-assignees, both recorded on the same cover:

1. CNH Industrial Canada, Ltd. (Canada)

• Line of business: the Canadian operating subsidiary of CNH Industrial N.V. (now marketed as "CNH"); the agricultural-equipment OEM behind Case IH, New Holland, Steyr and (post-2021) Raven. Canadian operations center on Saskatoon, Saskatchewan (the entity is the legacy CNH Canada, Ltd.).
• Product embodying the claims: Yes, in substance. CNH sells tillage and seeding implements with sensor-driven control (Advanced Farming Systems, and after the Raven Applied Technology acquisition, the Raven/CNH precision-ag portfolio). The patent's own FIG. 1 embodiment — tractor plus towed tillage implement carrying vision sensors fore and aft — is a CNH product architecture.
• Current status: Operating. No bankruptcy, no dissolution. Parent CNH Industrial N.V. legally persists (the on-highway business, Iveco Group, was spun off in January 2022; that restructuring did not remove the agriculture patent estate, and I found no recorded assignment reflecting it for this patent).

2. Autonomous Solutions, Inc. (Utah, US)

• Line of business: vehicle-autonomy technology developer (founded 2000; Petersboro/Mendon, Utah; ~140+ employees at the time of the CNH autonomous-tractor program). It is CNH's autonomy technology supplier — the 2016 cabless Case IH Magnum / New Holland T8 concept tractors were ASI–CNH joint demonstrators. It also supplies John Deere and others.
• Product embodying the claims: Yes — ASI's Mobius fleet-management / autonomy stack and sensor suites embody the subject matter, though the claimed system is the agricultural machine as a whole.
• Current status: Operating and independent. 2025 evidence: ASI launched a construction business unit with SoftBank Robotics America in an equity partnership, and separately divested ASI Mining to Epiroc (a business-unit sale, not a bankruptcy, and not touching this patent's chain). SoftBank Robotics America holds a strategic equity stake in ASI. Not a shell, not an NPE.

Co-ownership consequence to note for enforcement. The two assignees hold what appears to be an undivided joint interest. Under US law, absent a written agreement to the contrary, each co-owner may independently practice and license the patent, and neither can be sued for infringement by the other. This means the patent's practical assertion profile depends on the joint-development agreement between CNH and ASI, which is not a publicly recorded document. That is a real limitation on any assertion analysis — but it is an agreement question, not a chain-of-title question, and it is not an NPE signal.

Assignment timeline

Recorded conveyances of record for US 16/007,126 / US 10,907,998 B2 — one entry.

• 2018-05-29 to 2018-06-06 (inventor signing dates) / recorded 2018-06-13 — Reel 046071 / Frame 0647
  • Conveyance: Assignment (ASSIGNMENT OF ASSIGNORS INTEREST)
  • Assignor: Ferrari, Luca; Posselius, John H.; Henry, James W.; Bybee, Taylor C.; Turpin, Bret T.; Ferrin, Jeffrey L. — all six named inventors
  • Assignee: CNH Industrial Canada, Ltd. (Canada) and Autonomous Solutions, Inc. (Utah, US) — joint
  • Correspondent: Not determinable from the sources available to me. The reel/frame is confirmed; the correspondent field was not retrievable. I will not infer an attorney name. On your recurrence test this is precisely the field to pull first — if this patent's recording shares a correspondent with other CNH/ASI joint filings, it tells you the parties used one coordinating firm for the whole joint-development estate. It does not, by itself, indicate an NPE.
  • Context: Ordinary employment/contractor assignment to two operating companies under a joint-development relationship; executed immediately pre-filing. Not a fire-sale, not a reorg, not a securitization.

No post-issuance assignment of any kind is recorded. The only post-grant Legal Event is a maintenance-fee payment (4th year, large entity, 2024-07-31), which is a fee event, not a conveyance.

Two gaps you should close manually, because my sources structurally cannot:

1. Security agreements / liens. USPTO assignment records include security interests (conveyance type: Security Agreement / Grant of Security Interest), and these frequently do not appear in Google Patents' Legal Events feed. I saw none, but "none visible in a feed that often omits them" is not "none exist." Search Assignment Center filtered to Security Agreement conveyances for both CNH Industrial Canada, Ltd. and Autonomous Solutions, Inc. — the latter in particular is a venture-backed/PE-held private company (SoftBank Robotics America equity stake) and private-company credit facilities are the classic source of unindexed patent liens.
2. Corporate name changes. CNH's group has done a lot of renaming. If CNH Industrial Canada, Ltd. was ever the record owner under a prior name, a Change of Name record may exist elsewhere in the estate. Note that for this patent the chain shows no such record, so a name-change filing would post-date this patent or sit on other assets.

Timeline diagram

timeline
    title Ownership of US 10907998
    2018 : Assignments signed 29 May to 6 June
         : Application filed 13 June
         : Recorded reel 046071 frame 0647
    2021 : Patent issued 2 February
    2024 : 4th year maintenance fee paid 31 July

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT.
The patent has never left the two operating assignees. Reel 046071/0647 (2018) is the first and only link, and it runs to operating companies, not from them. No "IP / Licensing / Holdings / Ventures" successor appears anywhere in the chain. No registered-agent-service address is in play because there has been no transfer to evaluate.

2. Known asserter in the chain — NOT PRESENT.
Neither current assignee (CNH Industrial Canada, Ltd.; Autonomous Solutions, Inc.) appears on the public NPE lists you enumerated (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, DGC, Spangenberg entities) or, to my knowledge, as a high-frequency plaintiff in Unified Patents' or RPX's directories. Caveat: I could not re-verify against those directories in this session, so treat this as "no match found" rather than "verified absent."

3. Repeat correspondent across the chain — UNCLEAR / NOT DETERMINABLE.
There is only one link in the chain, so a recurrence test has nothing to recur across. A single appearance is not a finding by your own stated rule, and I could not retrieve even that single correspondent. This signal is unanswerable from my sources and is the one to close manually. If the 046071/0647 correspondent also appears on the sibling CNH/ASI joint filings (e.g., the US 15/179,703 and US 15/178,805 families), the correct read is coordinated joint-development prosecution, not NPE coordination — recurrence alone would not flip the analysis here given signals 1, 2, 4–8 all clear.

4. Cascading transfers — NOT PRESENT.
Zero subsequent assignments. There are no chained LLCs, no shared correspondent addresses, no common principals to test, because there is no chain beyond the original grant.

5. Pre-litigation transfer — NOT PRESENT.
There is no transfer at all, let alone one within 6 months of a suit. Consistent with the earlier section's finding: no CAFC 2026 docket entry, district court case, or ITC investigation naming US 10,907,998 was located. The only chain event precedes issuance by nearly three years.

6. Bankruptcy fire-sale — NOT PRESENT.
Neither assignee has been in bankruptcy. CNH Industrial N.V. is a public company that completed an orderly spin-off of its on-highway segment (Iveco Group, January 2022) — a corporate separation, not a Chapter 7/11. ASI's 2023–24 sale of ASI Mining to Epiroc is a voluntary business-unit divestiture by a solvent company and does not involve this patent or its chain.

7. Privateering — NOT PRESENT.
No operating company has transferred this patent to an NPE. To be explicit about the inverse structural risk: because ASI holds an undivided co-owner interest, ASI could in principle assign that interest to a licensing vehicle without CNH's consent in some circumstances — but no such record exists, and I will not treat a hypothetical as a finding. Flagging the structural possibility is not the same as reporting a signal.

8. Defensive aggregator (anti-NPE) — NOT PRESENT.
The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. The patent remains with the two original operating co-owners.

Verdict

Insufficient data — and specifically, insufficient in the narrow sense your taxonomy defines: only the original assignment is recorded, with no post-issuance conveyances to analyze, and the key NPE-diagnostic field (correspondent of record) was not retrievable from the sources available to me. Substantively, however, this is a clean non-NPE record: the sole chain event is Reel 046071/0647, executed 2018-05-29 to 2018-06-06 and recorded 2018-06-13 on the filing date, assigning from all six named inventors to two operating co-assignees — CNH Industrial Canada, Ltd. and Autonomous Solutions, Inc. Both assignees are live operating businesses with product lines embodying the claims (CNH's tillage/seeding precision-ag portfolio and ASI's Mobius autonomy stack), neither appears on any enumerated NPE list, there is no litigation naming this patent, and none of the eight NPE signals is present. The one item that would upgrade this from "insufficient data" to a fully verified "non-asserting operating-company ownership" finding is the Assignment Center correspondent and recorded-address fields for Reel 046071/0647 — plus a filtered Security Agreement search for the private-company co-owner, ASI, whose venture financing is the most plausible source of an unindexed lien.

Verify at: https://assignmentcenter.uspto.gov/ (search patent number 10907998; or reel/frame 046071/0647).

I'll research the cited prior art references to build the anticipation analysis. Let me search for the key references.

Prior-Art Analysis — US 10,907,998 B2

Scope note. I searched the cited-reference set for the literal number 10907998 and its citations. Per the authoritative record in this thread, US 10,907,998 B2 issued 2021‑02‑02 from Application 16/007,126, effective filing date 2018‑06‑13, with no earlier priority. It is therefore a post‑AIA patent, and the governing provisions are 35 U.S.C. § 102(a)(1) (patented / printed publication / public use / on sale before the effective filing date) and § 102(a)(2) (a U.S. patent or published application effectively filed before the effective filing date). I have not auto-corrected any identifier or date; where the record and my search results conflict I follow the record/search.

Two framing cautions before the tables:

• The 20 references below are the patent citations listed on the face of US 10,907,998. Being cited is not the same as being anticipatory — most are background art the examiner used to frame the field.
• The granted independent claims are narrower than the specification. Claims 1 and 6 require the input to be associated with a type or amount of residue, and claim 11 requires maintaining a constant amount of data per unit of area. Anticipation must be measured against those limitations, not against the abstract's generic "operational parameter" language.

1. The complete cited-reference set

2. Detailed § 102 treatment of the most relevant references

A. The Deere residue family — #15 (US 9,282,688 B2), #16 (US 9,554,098 B2), #18 (US 2017/0112043 A1)

What they disclose. All three come from Deere's residue-monitoring program and share a common teaching: one or more camera assemblies (visible, IR, or stereo) image the field ahead of and/or behind a tillage implement; a processor analyzes those images to determine an indicator of residue coverage — percent coverage and/or residue size ("an indicator of residue coverage may include an indicator of percent residue coverage, or of residue size"). US 9,554,098 adds that the residue indicator "may be associated with a location of the implement on the field," and US 2017/0112043 expressly assesses "the amount and type of residue detected" and adjusts the implement "to respond to the amount and type of residue." (Sources: https://patents.google.com/patent/US9282688; https://patents.google.com/patent/[US9554098B2](/patent/US9554098B2)/en; https://patents.google.com/patent/US20170112043.)

Overlap with the patent. These references map almost element-for-element onto the residue-input limitation of claims 1 and 6 and the vision-based sensor limitation of claims 5 and 10:

• "sensor mounted on the agricultural machine… configured to capture data at a sampling rate" → the camera assemblies;
• "input… associated with at least one of a type or amount of residue present within a field across which the agricultural machine is moved" → the residue-coverage/type determinations of #15/#16/#18;
• "vision-based sensor" (claims 5/10) → expressly disclosed.

Where anticipation fails under § 102. Anticipation requires a single reference to disclose each and every element, arranged as claimed. The intervening element none of the three discloses is the adjustment of the sensor's own sampling rate ("adjust the sampling rate at which the sensor captures data"). In the Deere references the controller adjusts the implement's operating configuration — "adjusting a depth or down-pressure of a plurality of tillage tools," changing disk/gang angles, harrow down-pressure, tractor speed — not the image-capture rate of the camera. Under the strict § 102 standard, these references therefore do not anticipate claims 1 or 6, even though they are the closest art and would be the primary § 103 references.

Where they do bite. They are good § 102(a)(1) art against the sub-elements recited in dependent claims 2/7 (location-based input) and 5/10 (vision-based sensor), and they disclose the residue concept that the specification itself describes.

B. Olympus #8 (US 2007/0195182 A1) — the "adjust the sampling rate" concept

This is the cited reference that speaks most directly to the rate-adjustment concept: an imaging apparatus "for setting image areas having individual frame rates." Frame rate is the temporal sampling rate of an image sensor, so it is the closest cited art on the claim element the Deere family lacks. Its defect for § 102 purposes is context and completeness: it is a general imaging apparatus, discloses no agricultural machine and no residue input, so it can at most anticipate the isolated "capture data at a sampling rate / adjust the sampling rate" element — not the full claim 1, 6, or 11 combinations. Expect it to be relied on in a § 103 combination with a Deere residue reference, not as standalone § 102 art.

C. Location/field-mapping references — #2 (US 5,870,689), #4 (US 2002/0022929), #5 (US 2002/0169558), #13 (US 2011/0240730)

These teach spatially-variable field data capture with location correlation:

• US 5,870,689 (Case) — a sensing circuit generates characteristic signals "representative of the work performed by the tool at a plurality of locations within the field," a DGPS circuit generates location data, and the controller "correlate[s] the characteristic data with the respective location data."
• US 2002/0022929 (AGCO) — creates spatially-variable field attribute maps.
• US 2002/0169558 (Smith) — mobile soil-conductivity sensor with DGPS, capturing data "in association with the determined location."
• US 2011/0240730 (Covely) — soil-sample tracking by location.

§ 102 exposure: These are the best cited art for the "input associated with a location of the agricultural machine within a field" limitation of dependent claims 2/7. They do not disclose any modification of the rate at which a working sensor captures data (they change what is recorded/mapped, not how often it is sampled), so they do not anticipate claims 1, 6, or 11. Against claim 3/8 (adjusting the rate on transitioning between first and second field portions) they establish location-awareness but not rate-switching.

D. Remaining references (#1, #3, #6, #7, #9, #10, #11, #12, #14, #17, #19, #20)

These are background: medical imaging (#1 US 5,993,393), crop-moisture sensing (#3 US 6,185,990), product tracing (#6 US 6,963,881), condition assessment (#7 US 7,191,097), mapping-display (#9 US 7,317,988; #12 ES 2,371,972), video display (#10 US 2010/0066836), harvester mechanics (#11 US 2010/0071335), biological-signal irrigation (#14 US 2012/0123817), agronomic recommendations (#17 US 2016/0078375), sensor calibration (#19 US 2017/0251600), and map verification (#20 US 10,584,971). None discloses adjusting a working agricultural sensor's sampling rate responsive to an operational parameter, and none is anticipatory of any claim. #20 warrants the § 102(a)(2) note above (its pre-2018 effective filing makes it available even though it published in 2020).

3. Bottom-line anticipation table (claim → best cited art → verdict)

4. Confidence and caveats

• Confidence: high for the identification and dates of the cited set — they are drawn verbatim from the patent's own front-page/Google Patents citation table, and I independently verified the three Deere residue references (#15, #16, #18), the Case scouting reference (#2), the Smith soil-mapping reference (#5), and the ISO-date/priority facts for those via live search.
• Confidence: medium-to-lower for the one-line descriptions of the non-agricultural background references (#1, #7, #10, #11, #20) — these rest on the patent's own titles plus the citation table; I reached the search-step limit before individually opening every one, so treat those descriptions as title-level, not abstract-level, readings.
• Strict § 102 point I want to flag: If any third party asserted that a cited reference anticipates claims 1, 6, or 11, the decisive gap is that the cited art modifies machine/implement operating parameters or image-display frame rate, not the sampling rate of the field-facing sensor in response to residue/location — the very element the patent's independent claims require. The references are best characterized as § 103 art, not § 102 anticipation.
• Consistency check with the prior sections: this analysis agrees with the earlier "Points worth flagging" that the residue limitation (not the generic "operational parameter" language) is what distinguishes claims 1/6 — which is precisely why the Deere residue references are the most relevant cited art even though they fall short of anticipation.

If you want, I can (a) pull the full claim charts for the three Deere references against claims 1, 6, and 11 element-by-element, or (b) run the remaining unverified background references to upgrade their descriptions from title-level to abstract-level.

I'll ground this analysis in the actual prior art of record. Let me pull the key references.

Strong residue-control material. Now the sampling-rate-adjustment and variable-density arts of record.

Obviousness Analysis — US 10,907,998 B2 under 35 U.S.C. § 103

Note on scope: This builds on the earlier bibliographic and claim-construction sections. The critical takeaway from the claim-scope section drives everything below: the granted independent claims are narrow. Claims 1 and 6 are limited to residue type/amount inputs; claim 11 is limited to constant data per unit area. The generic "operational parameter" language survives only in the preamble/abstract. An obviousness attack must therefore target those specific limitations, not the broad spec disclosure.

Framework applied: Graham v. John Deere Co., 383 U.S. 1 (1966) (scope/content, differences, PHOSITA level, secondary considerations) plus KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) (motivation may come from any source; predictable combination of known elements; "obvious to try").

Critical date: All references must predate 2018-06-13 (the '998 filing/priority date — no earlier priority is claimed). I verified the dates of record below.

1. The prior art of record (all pre-date 2018-06-13)

Two date nuances worth stating precisely:

• US 10,584,971 granted 2020-03-10 (after the '998 filing) but its effective filing is 2016-10-28, so it qualifies as prior art under § 102(a)(2) for what it describes. The record flags it as examiner-cited.
• Per the Google Patents legend in the supplied record ("* Cited by examiner"), the Deere residue patents US 9,282,688 / US 9,554,098 are unmarked, i.e., appear as applicant-submitted art, while US 2007/0195182 (Olympus), US 10,584,971 (Zoox), and US 2017/0112043 (Deere residue detection) are examiner-cited.

2. What the allowed claims actually require (and what the art must supply)

The record's own "Prior art keywords" — sensor, agricultural machine, sampling rate, controller, data — summarize the examiner's search. The decisive question is whether any combination teaches using a measured field condition to change the sensor's own sampling rate (as opposed to using the measurement to change an implement setting, which is what the Deere residue patents do).

3. Primary obviousness grounds

Ground 1 — Claims 1 and 6 (residue-triggered sampling rate)

Combination: US 9,282,688 (or its CIP, US 9,554,098) in view of US 2007/0195182.

• Deere teaches everything except the sampling-rate adjustment: an agricultural implement, one or more cameras mounted fore/aft, image analysis to determine an indicator of residue coverage including percent coverage and residue size (i.e., amount and type of residue), and control of a subsequent operation based on that indicator. Its own text notes residue "may vary from location to location even within a single field."
• Olympus teaches the missing element: a controller that sets frame rates individually for different image areas and controls read timing accordingly — i.e., a sensor whose sampling rate is not fixed but is driven by device logic.
• Motivation: (a) The '998's own Background admits the problem — sensors "capture data at a predetermined constant interval," generating "large quantities of data," while on-board "processors and other computing devices... have limited computing power." An admission of the problem in the patent itself supplies the motivation (KSR; also reflected in the examiner-cited Zoox reference, which expressly frames data collection as a processor/bandwidth-load problem). (b) Deere itself explains that residue characteristics affect the reliability of the analysis (accuracy depends on coverage/size) — giving a POSITA a reason to allocate more/fewer image samples where residue makes the data more/less informative. (c) Both are in the same field (machine vision on agricultural machines), a "predictable use of prior art elements according to their established functions."

Ground 2 — Claims 2, 3, 7, 8, 12 (location/portion-transition switching)

Combination: Ground 1 + US 5,870,689 (Case) and/or US 2002/0022929 (AGCO).

• Case teaches a control circuit that generates characteristic field data "at a plurality of locations," receives DGPS signals, correlates the data with locations, and stores it — and expressly states that manual sampling "provides only a limited number of data samples." That is the location-aware, variable-density data acquisition concept.
• US 9,554,098 independently associates residue indicators with location readings and uses location-specific data in a later operation.
• Motivation: A POSITA operating a field with known spatially varying conditions would naturally sample more densely where conditions differ — a routine design choice to which Case and AGCO directly point. Claims 3/8/12 (switching on transition between portions) are merely the necessary consequence of correlating sampling with location.

Ground 3 — Claims 11, 13, 14, 15 (constant data density; topography; max processing throughput)

Combination: US 9,282,688 / 9,554,098 + US 5,870,689 + CROPWATCH NPL, optionally + US 2002/0169558 (terrain) and/or US 2002/0022929 (attribute maps).

• Claim 11's "constant amount of data per unit of area" is a ground-speed-compensated sampling rate — the classical result of making the interval distance-based rather than time-based. The SensorsONE data-logger NPL and CROPWATCH yield-mapping NPL establish configurable sampling and interval-based mapping as conventional.
• Claims 13–14: US 2002/0169558 teaches analyzing and mapping terrain, and US 2002/0022929 teaches field attribute maps; Deere's '098 notes coverage varies with "local terrain." Selecting a higher sample density near depressions and lower near hills (claim 14) is an optimization a POSITA would reach by applying known agronomic reasoning (low areas capture/redistribute residue and water differently) to a known mapping capability.
• Claim 15: Deere's references already teach adjusting machine/implement operation on the go; Zoox's examiner-cited teaching expressly motivates reducing processor load; and the '998 background admits finite on-board processing capacity. Constraining machine speed/operation so that the sensor collects no more than the controller can process is an obvious engineering boundary condition, not an inventive step.

4. Why a PHOSITA would combine (motivation, stated affirmatively)

1. Same field, same problem. All references address sensing and control on agricultural (or autonomous) machines; the shared problem of excess data vs. limited compute is named in the '998 background and in Zoox.
2. Established functions, predictable result. Varying a camera's read/frame rate (Olympus) is a known, routine technique with a predictable effect on data volume; nothing unpredictable occurs when it is applied to Deere's residue camera.
3. Direct design incentive. The '998's own residue passages supply the incentive: data accuracy varies with residue type (lower accuracy → take the higher sampling rate to compensate), so residue is a natural control variable for sample density.
4. No teaching away. Deere's use of residue to control implement depth/down-pressure is complementary to, not inconsistent with, also using residue to control sampling rate.
5. Obvious to try. Where the art discloses a finite menu of sampling rates and a known set of field variables (location, topography, residue), selecting among them is the paradigm of KSR's "finite number of identified, predictable solutions."

5. Weak links and honest limitations of this analysis

I want to be explicit about where the obviousness case is strong vs. contested:

• Claims 4 and 9 (desired physical pattern; spec describes hexagonal/rectangular patterns via "multi-dimensional Fourier-based sampling techniques") — weakest link. I did not locate, in the prior art of record, a reference that expressly teaches spatially hexagonal sampling of a field sensor. Olympus 2007/0195182 is about per-image-area frame rates, not spatial lattice geometry; AGCO/Case are about map layers and location correlation. A hexagonal-sampling limitation could therefore survive a § 103 challenge on this record unless the examiner/defendant supplies a multidimensional-sampling reference or general-knowledge evidence (multidimensional sampling theory treats rectangular and hexagonal lattices as the two standard options — but I cannot ground that in a specific reference of record, so I flag it as unverified rather than assert it).
• US 10,584,971 (Zoox) is post-filing publication date (2020). It works only as § 102(a)(2) art based on its 2016-10-28 effective filing, and only for the subject matter actually described. I have only partial text (crowd-sourced map verification), so I would not lean on it as a primary reference.
• I did not retrieve full text for every one of the 20 citations. My characterizations of US 2002/0022929 (AGCO), US 2002/0169558 (Smith), US 7,317,988, US 7,191,097, US 6,963,881, US 6,185,990, US 2012/0123817, ES 2371972 and US 2016/0078375 rest on their titles/abstracts of record, not on a full read. Any of these could be stronger or weaker than my mapping suggests.
• Prosecution tension worth noting. The Deere residue patents (US 9,282,688 / US 9,554,098) were before the examiner (they appear unmarked in the record, consistent with an IDS), yet claims 1/6 issued with the residue limitation. The most plausible reconciliation is that the examiner treated the difference as the feedback target: Deere feeds residue into implement control, whereas claims 1/6 feed residue into the sensor's own sampling rate. That is precisely the gap Ground 1 attacks, and it is a gap a defendant could close. Conversely, the fact that the examiner cited Olympus 2007/0195182 and Zoox 10,584,971 suggests he was already looking at variable-rate sensing art — which cuts against a defendant's ability to argue the combination was non-obvious, because the motivation was evidently on the examiner's own mind.
• No secondary considerations data. I found no objective evidence (commercial success, unexpected results, licensing, copying) in the record or via search. Absent such evidence, the KSR/"predictable combination" analysis controls against the broader claims.

6. Bottom line

• Claims 1 and 6 are, on this record, the most vulnerable: the combination of Deere US 9,282,688/9,554,098 (residue amount/type via camera) with Olympus US 2007/0195182 (individually adjustable frame rates) supplies every element, and the motivation is supplied by the admitted processor/memory limitation and by Deere's own recognition that residue characteristics govern data reliability.
• Claims 11–15 are vulnerable to Deere + Case US 5,870,689 (+ AGCO US 2002/0022929 / Smith US 2002/0169558 + NPL) for location-based, topography-aware, constant-density sampling and machine-parameter adjustment.
• Claims 4 and 9 are the strongest part of the patent on this record, because the spatial-lattice limitation is not clearly met by any reference I could verify.
• Claims 2/3/5/7/8/10/12/13/14 are all straightforwardly obvious as routine integrations of the above.

If you want, I can attempt to retrieve the full text of the remaining unread citations (AGCO '929, Smith '558, Ag Leader '988, Gov't '097, Deere '881, New Holland '990) to firm up the Ground 2/3 mappings and to test specifically whether any of them supplies the spatial-sampling-pattern teaching that claims 4/9 need.