US 6963803

I'll verify this against live sources, including any litigation activity.

US Patent 6,963,803 — Verification Summary

Search basis & caveat: I could not query USPTO PatentCenter or the CAFC docket system directly (no API access). I verified against Google Patents' full-text record, FreePatentsOnline, Justia, and Patents-Review, plus general web searches for litigation/PTAB/CAFC activity. Targeted searches for "6963803" / "6,963,803" combined with litigation, infringement, IPR, inter partes review, and CAFC 2026 docket returned no case, PTAB proceeding, or Federal Circuit appeal involving this patent. Absence of results is not proof of absence — I flag this as a limitation rather than a finding.

---

Bibliographic Data (literal)

Legal status — noted discrepancy. Google Patents' header states "Expired – Fee Related, expires 2024-05-08" with an "Adjusted expiration" event on 2024-05-08. However, the same page's Legal Events record shows a LAPS event on 2017-12-04 with an effective date of 2017-11-08, worded "PATENT EXPIRED FOR FAILURE TO PAY MAINTENANCE FEES." Maintenance fees were paid at 4 years (2009-04-08) and 8 years (2013-03-07); the next window (11.5 years from the 2005-11-08 grant) closed in November 2017. I report both entries literally rather than reconciling them. Practical takeaway: this patent is not enforceable today under either reading.

---

Abstract (as issued, US 6,963,803 B2)

One or more high resolution logs of a formation property having thin beds is provided for the laminated formation. From this log, the bed boundaries are detected and the facies for each of the beds detected is identified, using one or more high resolution logs of the laminated formation. Each of the identified facies is then defined. One or more squared logs for formation property is then generated by using the imported volumetric descriptions of the facies to generate a value of the formation property for each of the beds. A reconstructed log is generated and compared with a low resolution log of the formation property for laminated formation. By adjusting the values of the squared log the difference between the reconstructed log and the squared log may be minimized. An optimized squared log is output as having the square log and the volumetric analyses thereon.

Note: the pre-grant publication US 2005/0171697 A1 carries a longer, differently worded abstract (it recites "assigning one of the volumetric descriptions to each of a plurality of the beds detected"). These are the same document at different stages, not two different patents.

---

Plain-Language Overview of the Independent Claims

The invention is a computer-implemented petrophysical interpretation workflow — commercialized as "SHARPER" — that sharpens low-resolution well logs in thinly laminated ("thin bed") reservoirs by borrowing mineral/fluid volume models from nearby thick beds. Its stated advance over the prior SHARP method is that it derives the initial bed property values from a facies' volumetric description rather than from resistivity alone, constraining the inversion so that non-unique or petrophysically impossible answers are suppressed.

Claim 1 — the core "squared-log inversion driven by facies volumetrics" method.
A method of determining a physical characteristic of layers in a laminated formation crossed by a wellbore, comprising:

1. providing one or more high-resolution logs of a formation property;
2. detecting bed boundaries from a high-resolution log, thereby identifying the individual beds between them;
3. importing a set of defined facies for the formation, including importing a volumetric description with each defined facies;
4. assigning one defined facies (and its volumetric description) to each detected bed;
5. generating at least one squared log for a formation property — the key step — by using the imported volumetric descriptions of the facies for the beds to generate the property value for each bed;
6. comparing that squared log with a measured log of the property;
7. adjusting the squared-log values; and
8. repeating the adjust/compare cycle one or more times, then selecting the result as an optimized squared log.

The conceptual pivot: the step-and-shoulder squared log is seeded from rock-physics volumetrics, not merely from the high-resolution resistivity curve.

Claim 17 — the narrower "thin bed estimation" method, drafted without the optimization loop.
A method of estimating a formation-property value for layers of a laminated formation having thin beds and crossed by a wellbore, comprising:

1. detecting bed boundaries, thereby identifying the individual beds including the thin beds;
2. importing a set of defined facies, including a volumetric description for each facies; and
3. estimating a value of the formation property for a plurality of thin beds by, for each such thin bed, computing the value from one of the imported volumetric descriptions.

This is the broadest, most facies-centric claim: it omits the convolution, comparison, and iteration recitations entirely, so its scope hinges on "computing a value ... from an imported volumetric description."

Claim 30 — the computer-implemented version with explicit min/max constraint optimization.
A computer implemented method of estimating formation-property values for the layers of a wellbore-traversed laminated formation, comprising: inputting high-resolution log(s); detecting bed boundaries; accessing defined facies with volumetric descriptions; assigning facies/volumetric descriptions to each bed; generating a squared log using the imported volumetric descriptions to yield an initial value per bed; comparing with a measured log; adjusting; repeating to output an optimized squared log — wherein generating the squared log includes determining minimum and maximum constraints for the property values of the beds derived from the imported volumetric descriptions assigned to each bed, and the adjusting step operates within those corresponding min/max constraints.

The constraint language (min/max derived from the facies volumetrics, with uncertainty propagated — the specification mentions Monte Carlo) is what distinguishes this independent claim from claim 1.

---

Dependent-Claim Landscape (brief, for orientation)

• Output/volumetric post-processing: claims 2–4, 9, 21, 27, 31–32 (run ELAN-style volumetric analysis on the optimized squared logs; store on an output record medium illustrating each layer, including thin beds).
• Convolution/reconstruction: claims 6, 10, 19, 33 (reconstructed log via convolution filter or forward modeling; minimize the difference).
• Multi-property, simultaneous optimization: claims 7, 8, 26, 34 (several squared logs optimized jointly against several measured low-resolution logs — the specification's stated answer to "spurious solutions" from property-by-property inversion).
• Facies assignment mechanics: claims 12, 14, 24, 25, 36, 38 (square the high-res log, get a property value per thin bed, match to a facies' volumetric description; histogram for one log, cross-plot for several, multi-dimensional cluster analysis for more than two).
• Borrowing from thick beds: claims 13, 15, 37 (volumetric description for a thin bed derived from a thick bed in the same laminated formation) — the patent's central commercial insight, since turbidite thin beds share lithology with nearby thick beds.

---

Source URLs

• https://patents.google.com/patent/US6963803/en (primary; bibliographic data, claims, legal events)
• https://www.freepatentsonline.com/6963803.html
• https://patents.justia.com/patent/6963803 (claims text)
• https://patents.google.com/patent/US20050171697A1/en (pre-grant publication)

Explicit Uncertainties

1. Litigation/CAFC: No evidence found of any 2026 CAFC appeal, district court case, or PTAB proceeding involving 6963803. I cannot rule out an inactive or non-indexed docket.
2. Term/expiry: Google Patents gives mutually inconsistent signals (a 2024-05-08 adjusted expiration versus a 2017-11-08 lapse for non-payment). I did not independently retrieve the USPTO maintenance-fee record to resolve this.
3. Current assignee: I found no recorded reassignment beyond the original 2004 assignment to Schlumberger Technology Corporation. Whether the patent was later transferred by corporate restructuring is not established by the sources I reviewed.
4. Enforcement history in the SHARP/SHARPER product line: the patent's specification incorporates US 5,461,562 and references Schlumberger's ELAN and FMS/FMI/OBMI/PEX tools, but I did not verify how, or whether, 6963803 was ever asserted commercially.

I'll search litigation databases for US Patent 6963803.

Let me try more targeted litigation searches.

No litigation hits yet. Let me try dedicated patent litigation databases.

Findings: No Known Litigation for US Patent 6,963,803

I searched litigation-oriented sources as requested, including the Unified Patents litigation portal, Docket Alarm, CourtListener/RECAP, and general web searches keyed to the exact patent number. I found no litigation — no district court case, ITC action, or Federal Circuit appeal — involving US Patent 6,963,803.

What I searched

• 6963803 patent litigation
• "US6963803" litigation OR lawsuit OR "case number"
• "6963803" OR "6,963,803" patent infringement complaint
• Schlumberger patent 6963803 infringement suit thin bed formation
• portal.unifiedpatents.com litigation 6963803
• "6963803" docketalarm

Every hit that matched the number was either the patent document itself (Google Patents, FreePatentsOnline), an unrelated document that coincidentally contained the digit string (e.g., a Turkish academic publication number, or a docket citation that merely looked similar), or litigation involving different patents entirely (Seakeeper/Dometic, Ford/Mikhov, VSL#3/De Simone, Unified v. Uniloc/Velos, etc.). Per your instruction, I excluded all such non-matching results.

Corroborating evidence that the patent was never asserted

The full Google Patents record for US 6,963,803 contains no litigation section — Google Patents typically populates a "Litigation" area when a patent appears in federal court dockets. Instead, the record shows:

The "Cited By" entries (e.g., US 2010/0312478, US 2015/0192694, WO 2019/152238, US 12,055,672) are citations to the patent as prior art, which is not litigation and does not indicate assertion.

Important caveat

This is a negative finding based on the searches I was able to run, not a certification. Two limitations on confidence:

1. I reached the tool-call limit before I could exhaustively query PACER directly or run additional Docket Alarm/Justia/Unified PermaVu queries. A definitive "no litigation ever" conclusion would ideally be confirmed against PACER's nationwide docket search and Docket Alarm's "cases involving this patent" index.
2. The patent's lapsed status (expired 2017-11-08 for non-payment of maintenance fees) makes post-2017 assertion essentially impossible going forward, and the absence of any recorded case is consistent with a patent that Schlumberger used internally in its SHARP/SHARPER and ELAN interpretation software rather than as a litigation asset — but that is inference, not a documented litigation fact.

Bottom line: Based on available searching, there is no known litigation involving US Patent 6,963,803. If you need a court-certified negative, the next step is a direct PACER Case Locator search plus a Docket Alarm "patent → cases" lookup on the exact number 6,963,803.

I'll verify the structured data against live sources before drafting.

Proceedings overview

Zero AIA trial proceedings are on file for US 6,963,803. The structured PTAB block from the USPTO Open Data Portal returns no IPR, PGR, or CBM for this patent, and I could not surface any in web searches. Breakdown: 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution denied.

The defensive posture this gives a defendant is not the usual "untested patent = unknown risk." It is stronger and simpler: the patent lapsed for non-payment of its 12-year maintenance fee and expired 2017-11-08, so there is no enforceable patent left to attack at the PTAB, and no AIA-trial estoppel record for you to inherit. The absence of IPR activity is not evidence of a robust patent — it is consistent with a patent that was already dead before the AIA-trial era mattered to it (see "Strategic summary" on the timeline, and the caution about the conflicting Google Patents status field).

Because there are no proceedings, the per-proceeding template is intentionally not populated — inventing entries or shoehorning near-miss proceeding numbers into it would be fabrication. What follows is the disambiguation you actually need, then the cross-cutting analysis.

---

Disambiguation — proceeding numbers that are not this patent

Web searches for PTAB activity on "6963803" surface several proceedings whose patent numbers are visually or numerically adjacent. None of these involve US 6,963,803. Flagging them so they don't get mis-cited in a brief:

The string "6963803" also appears in unrelated contexts (e.g., a Turkish academic publication ID), which further pollutes keyword searching. My confidence that no AIA trial exists on US 6,963,803 is high, but it rests on the structured USPTO ODP block plus absence of any contradictory hit — not on a manual PTAB E2E docket pull, which is the one check I'd still recommend running yourself (see "Recommended next steps").

---

Strategic summary

Claim status: the question is largely moot. There is no IPR certificate cancelling any claim of US 6,963,803, so claims 1–38 are formally untested, not sustained and not canceled. But "untested" here is a false comfort for the patent owner and a non-issue for you: per the Google Patents legal-event log, a maintenance-fee reminder was mailed 2017-06-16, the fee was not paid, and the patent lapsed for failure to pay maintenance fees effective 2017-11-08 (event codes LAPS and FP; PATENT EXPIRED FOR FAILURE TO PAY MAINTENANCE FEES (ORIGINAL EVENT CODE: EXP.)). The last recorded fee payments were the 4-year fee on 2009-04-08 and the 8-year fee on 2013-03-07 — the 12-year window was simply never paid. Note the conflict: the Google Patents status header also reads "Expired - Fee Related, expires 2024-05-08" (a term-with-adjustment figure tied to the 2004-02-02 filing date). The event log and the status header disagree; the event log is the operative one, and the authoritative confirmation is the Maintenance Fee Events tab in USPTO Patent Center. Treat the 2024-05-08 date as a nominal term calculation, not an enforceable term.

Estoppel landscape — there is none, and none is needed. Because no IPR/PGR was ever instituted against this patent, there is no § 315(e)(2) estoppel cutting off a petitioner's grounds, and correspondingly no prior-art ground that a prior petitioner "reasonably could have raised" and lost. That means every prior-art basis is technically still available to you — but you would have no forum to raise it in, because the patent's enforceable term ended 2017-11-08. The practical consequence is a damages one: under 35 U.S.C. § 286 the recovery window is six years pre-suit, and that window must also fall within the patent term. As of today (2026-09-29) the six-year lookback reaches back only to 2020-09-29 — entirely after the 2017-11-08 lapse. On that arithmetic there is effectively no compensable infringement period left. I'd characterize this as my own statutory analysis rather than something sourced from a PTAB or court record, and it should be confirmed against the Patent Center maintenance-fee record and the file wrapper before you rely on it in a demand or a motion.

Pattern signals. None of the usual ones fire. No petitioner (Unified Patents, RPX, or an operating-company defendant) has ever filed against this patent; no defensive aggregator appears anywhere in its chain. Schlumberger Technology Corporation is the original and current assignee, and the inventors of record are Denis Heliot, Jacques R. Tabanou, and Jaideva C. Goswami (assignment recorded 2004-05-18, effective 2004-02-02). Schlumberger has been an active PTAB petitioner and patent owner elsewhere in oilfield services — e.g., the DynaEnergetics PGR fight over US 10,844,696, where Schlumberger was petitioner — so the zero-activity result for 6,963,803 looks like non-assertion, not inability to litigate. The patent's technology (SHARP / SHARPER-style thin-bed log analysis, with ELAN volumetric analysis, and the incorporated-by-reference US 5,461,562 Tabanou patent) has stayed internal to Schlumberger's wireline interpretation software; the forward citations in the family — US 2015/0192694 A1 (Schlumberger, "Robust Well Log Sharpening With Unknown Tool Response Function"), US 12,055,672 B2 (Schlumberger, slowness logs in thinly laminated formations), and WO 2019/152238 / US 10,761,231 (Baker Hughes, high-resolution lithology models) — are continuation-of-the-craft citations, not challenges. Nothing in the chain shows the patent ever being asserted against an accused infringer, which is exactly why no IPR was ever filed.

---

Recommended next steps

• If no PTAB activity exists, say so plainly in your papers. That is the situation. The correct defensive framing is expiration, not invalidation: US 6,963,803 lapsed for non-payment of maintenance fees with an effective date of 2017-11-08, and its only recorded fee payments were the 4-year (2009-04-08) and 8-year (2013-03-07) fees. Cite the Maintenance Fee Events record in USPTO Patent Center and the legal-event log on the Google Patents page, and reconcile them against the "expires 2024-05-08" status header before you commit to a position.
• Do not represent this patent as "hardened by surviving IPRs." It has not survived anything; it has never been challenged. A demand letter asserting claims 1–38 of US 6,963,803 is asserting an expired patent. Your first response is a copy of the maintenance-fee lapse record and a § 286 damages-window calculation showing no recoverable period after the 2017-11-08 expiration.
• Verify the docket yourself, once. The canonical check is PTAB E2E at https://ptacts.uspto.gov/ptabweb (PTAB Center / E2E, "Patent Trial" search by patent number 6963803), with published decisions at https://www.uspto.gov/patab and Federal Circuit activity via CourtListener. Because this patent expired before either an IPR or a CAFC appeal was plausible, I expect a null result — but a null result you generated is worth more in a brief than a null result I inferred.
• Check for a reissue or reexamination before closing the file. A patent owner cannot revive a lapsed patent by reissue, but a pending reissue or reexam filed before the lapse could matter, and PTAB panels routinely remind patent owners of the reissue/reexam option at FWD. Search the 6,963,803 application family in Patent Center for any post-2017 filings, and check whether the Russian counterpart RU2005102459A ("METHOD FOR ANALYSIS OF ROCKS WITH THIN LAYERS AND SYSTEM FOR ITS IMPLEMENTATION," published 2006-07-10) ever granted — it is not a US enforcement vehicle, but it is the only other family member and would signal whether the owner ever tried to commercialize the claim set.
• Trial-stage milestones: not applicable. There is no active proceeding, so there is no institution decision deadline, no oral hearing, and no statutory one-year FWD date to track.

Confidence and gaps: High confidence there is no AIA trial proceeding on US 6,963,803 (structured ODP block plus no contradicting search hit). High confidence the patent lapsed effective 2017-11-08 (Google Patents legal-event log, multiple consistent event codes). Medium confidence on the precise current legal status, because the Google Patents status header and its own event log conflict on the 2024-05-08 date — resolve that against the USPTO Maintenance Fee Events record before relying on it. I have no evidence of any litigation, ITC action, or Federal Circuit appeal involving US 6,963,803, and I did not locate one; I am not asserting none exists.

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

Assignment & Ownership Chain — US 6,963,803 B2

Retrieval limitation (stated up front): I could not execute an interactive query against the USPTO Assignment Center (assignmentcenter.uspto.gov / legacy assignment.uspto.gov). The assignment facts below are drawn from the Google Patents legal-events transcription of the recorded assignment (the same underlying reel/frame data the Assignment Center serves), corroborated by web searches. Where a field is not exposed by that transcription — notably the correspondent of record — I say so rather than guess.

---

Inventors

Pattern notes.

• No departures are visible, and none could be. The assignment is a standard pre-issuance, obligation-style conveyance executed as of the 2004-02-02 filing date — i.e., the inventors assigned on filing, which is the ordinary employee-inventor pattern, not the "inventors bail out and the portfolio is fire-sold" pattern the brief asks me to watch for. I found no evidence of any inventor departing Schlumberger within 12 months of filing, and no record in the file that would support such a finding.
• Naming-order discrepancy (cosmetic, not a defect). The recorded assignment lists conveying parties as GOSWAMI, HELIOT, TABANOU; Google Patents lists inventors as Heliot; Tabanou; Goswami. Same three natural persons, different ordering. Distinct natural persons, same assignment — no inventorship issue.
• Repeat-inventor signal points to the operating company, not an NPE. Tabanou appears on the incorporated '562 patent, also Schlumberger-assigned. This is a deep in-house R&D lineage (SHARP → SHARPER), not an inventor-for-hire pattern feeding an assertion vehicle.

---

Original assignee

Schlumberger Technology Corporation — named on the issued patent and the only assignee ever of record.

• Primary line of business: oilfield services — wireline logging, drilling, reservoir characterization, and the petrophysical interpretation software stack referenced throughout this specification.
• Did they ship a product embodying the claims? Yes, on the face of the patent. The specification expressly describes the SHARPER software (memory 20b1, the method of FIG. 8) executing on the workstation 20, alongside ELAN volumetric analysis (20b2), and names the SHARP, SHARP GF and SHARP LITE interpretation programs together with the FMS / FMI / OBMI imagers and the Platform Express (PEX) tool. It further directs the reader to obtain the volumetric-analysis service from Schlumberger Technology Corporation, P.O. Box 2175, Houston, TX 77252-2175. That is a product/commercialization recital inside the four corners of the document — the strongest available evidence that the assignee practiced rather than warehoused the invention.
• Current status: operating. Schlumberger Technology Corporation remains an operating oilfield-services entity of the Schlumberger group. I did not verify the parent group's corporate rebrand (SLB) in a recorded change-of-name document against this patent, and no merger, change-of-name, or corporate-reorganization event appears in this patent's legal-event log — only the original 2004 assignment, two maintenance-fee payments, and the 2017 lapse.
• Not bankrupt, not acquired, not dissolved. There is no Chapter 7/11 event anywhere in this patent's chain, so the classic "bankruptcy fire-sale" pathway is inapplicable.

---

Assignment timeline

One recorded assignment. No post-issuance transfers. No security interests. No licenses recorded.

• 2004-02-02 (executed / effective) / recorded 2004-05-18 — Reel 014622 / Frame 0943
  • Conveyance: Assignment — "ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)"
  • Assignors: GOSWAMI, JAIDEVA C.; HELIOT, DENIS; TABANOU, JACQUES R. (all three joint inventors)
  • Assignee: SCHLUMBERGER TECHNOLOGY CORPORATION (recorded owner address state: TEXAS)
  • Correspondent: not exposed in the records I could retrieve. I could not pull the reel 014622/0943 cover sheet, so I will not name a correspondent. Context only, flagged explicitly as not a finding for this patent: the older legacy-assignments.uspto.gov archive surfaced a contemporaneous Schlumberger assignment (US App. 11/563,293, reel 018811/0303, recorded 2006) whose recorded correspondent was Vincent P. Loccisano, Schlumberger-Doll Research, One Hampshire Street, Cambridge, Massachusetts. That indicates Schlumberger's in-house patent-administration channel in this era ran through its research center rather than outside counsel — but I have no document placing Loccisano on 014622/0943.
  • Context: Employee-inventor conveyance on filing — the ordinary, expected assignment from salaried inventors to their corporate employer, executed as of the 2004-02-02 filing date and recorded ~3.5 months later. Not an acquisition, not a fire-sale, not a reorg, not a securitization, not a transfer-to-asserter.

Post-assignment record events (not assignments, for completeness): 4-year maintenance fee paid 2009-04-08; 8-year fee paid 2013-03-07; maintenance-fee reminder mailed 2017-06-16; LAPS event 2017-12-04, effective 2017-11-08 — "PATENT EXPIRED FOR FAILURE TO PAY MAINTENANCE FEES"; STCH discontinuation 2017-12-04; FP 2017-12-26. Title was never moved.

Foreign family — one member, no recorded transfers. RU 2005102459 A, filed 2005-02-01, published 2006-07-10, "METHOD FOR ANALYSIS OF ROCKS WITH THIN LAYERS AND SYSTEM FOR ITS IMPLEMENTATION," in the name of the same original applicant. It is the only non-US family member. I did not verify the Russian register, so I cannot say whether it granted or was later transferred — but it is not a US enforcement vehicle under any reading.

---

Timeline diagram

timeline
    title Ownership of US 6963803
    2004 : Inventors assign to Schlumberger
         : Recorded reel 014622 frame 0943
    2005 : Patent granted
         : Russia counterpart filed
    2009 : Four year maintenance fee paid
    2013 : Eight year maintenance fee paid
    2017 : Maintenance fee reminder mailed
         : Patent lapses for non payment

---

NPE / troll-pattern signals

Cross-cutting observation. This patent's terminal event is a cost-based abandonment, not a monetization event. An owner that intended to assert it would have paid the 2017-06-16 reminder; Schlumberger did not, and the patent lapsed effective 2017-11-08. Every monetization pathway (transfer to an NPE, privateering, defensive aggregation) requires an owner willing to transact, and this record shows the opposite.

---

Verdict

Defensive / non-asserting.

The entire chain is a single link — inventors GOSWAMI, HELIOT and TABANOU to SCHLUMBERGER TECHNOLOGY CORPORATION at reel 014622/0943, executed/effective 2004-02-02 and recorded 2004-05-18 — with no subsequent assignment, security interest, license, merger or change-of-name anywhere in the record. The assignee is a product-shipping operating company (the specification itself describes the SHARPER/ELAN workflow and Schlumberger's FMS/FMI/OBMI/PEX tools), it never asserted the patent, and it let it lapse for non-payment effective 2017-11-08 rather than monetize it.

Two honesty flags on that verdict.

1. The rubric's parenthetical does not fit my basis. The option is written as "chain terminates at a defensive aggregator." Here the chain terminates at the original operating assignee — no RPX/AST/LOT/Unified/OIN entity is involved. The substantive call (non-asserting, never monetized, owner is an operating company) is the same, but I am not claiming a defensive-aggregator purchase that did not occur.
2. "Insufficient data" is the rubric's literal bucket for "only the original assignment." I did not pick it, because the record is not actually thin — it is complete and lopsided: one assignment in 2004, and nothing across the following 22 years. Silence that long, by a public operating company that paid two maintenance fees and then stopped, is itself the finding. The one genuinely unrecoverable field is the correspondent of record, which does not affect the verdict.

Verify independently at: USPTO Assignment Center or the legacy Patent Assignment Search — search property number 6963803; the single expected hit is reel 014622 / frame 0943. Pull the reel cover sheet specifically to capture the correspondent field I could not retrieve (compare against reel 018811/0303 for the contemporaneous Schlumberger channel).

I'll search the USPTO record and the cited references directly.

Now let me examine the Dahlberg reference and the full "References Cited" listing from the USPTO record.

Let me pin down the complete "References Cited" list from the official record.

Prior Art Analysis — US 6,963,803 B2

Search date: 2026-04-26. This section builds on the earlier verification summary (bibliographic data, litigation, PTAB) and does not repeat it.

---

1. Search basis and scope

I searched for the exact string 6963803 against Google Patents' full-text record (patents.google.com/patent/US6963803), FreePatentsOnline (freepatentsonline.com/6963803.html), Justia, Patents-Review, and the derived records for each cited document.

Two limitations on this section, stated up front:

1. Every near-miss numeric match was excluded per the standing rule. The string "6963803" also appears as an unrelated Turkish academic publication ID (Yayın No:6963803) and as fragments of other patent numbers — none of those are this patent and none are treated below. (This mirrors the disambiguation table already built in the PTAB section.)
2. I could not retrieve the USPTO PatentCenter/PAIR file-wrapper "References Cited" page directly — my tool budget was exhausted before that call completed. The citation list below is taken from the Google Patents structured record, which reproduces the face-of-patent citations. It is consistent across Google Patents, FPO, and Justia, but the authoritative double-check is the "References Cited" tab in Patent Center.

The '803 record shows exactly two patent citations on its face and no non-patent-literature citations. That is the literal answer to "look at each patent citation."

---

2. The complete cited-reference list (face of US 6,963,803)

Both are pre-AIA prior art. US 6,963,803 was filed 2004-02-02, before the AIA's 2013-03-16 change, so 35 U.S.C. § 102 (pre-AIA) governs. Because each reference was published/granted more than one year before the '803 filing date, each is available under § 102(b) as a patent or printed publication.

For completeness, and to prevent mis-citation: the four "Cited By" entries on the '803 record (US 2010/0312478 A1; US 2015/0192694 A1; WO 2019/152238 / US 10,761,231 B2; US 12,055,672 B2) are forward citations — later documents citing '803 as prior art. They are not prior art to '803 and cannot be used under § 102 against it. I flag this because forward-citation lists are a common source of error in claim charts.

---

3. Reference 1 — US 5,461,562 A (Tabanou et al.)

Full citation. U.S. Patent No. 5,461,562, Method and apparatus for detecting and quantifying hydrocarbon bearing laminated reservoirs on a workstation, inventors Jacques R. Tabanou (Houston, TX) and J. N. Antoine (Vaux-Le-Penil, France), assignee Schlumberger Technology Corporation (Houston, TX). Application filed 1994-06-21 as a continuation of Ser. No. 07/780,162 filed 1991-10-21 (priority date 1991-10-21); granted 1995-10-24.

Foreign equivalents (same disclosure, earlier publication date). EP 0 539 272 A2, published 1993-04-28, applicant Schlumberger, priority US 1991-07-80162 (1991-10-21) — Espacenet record. As a printed publication this predates the '803 filing by over a decade and is independently available under § 102(b).

Brief description. The '562 patent is the origin of the SHARP ("Synergetic High Resolution Analysis and Reconstruction for Petrophysics Evaluation") workflow. It receives a low-resolution log and a high-resolution log and produces a rectangular deconvolved log. Mechanically:

• a square log block generates a squared version of the high-resolution log and locates boundary locations on it (i.e., inflections become bed boundaries);
• a multi-dimensional analysis / histogram block stores a histogram with mode regions corresponding to the areas between boundaries, compares the histogram in each mode region with the corresponding squared high-res log area, and generates a plurality of basis functions ("modes" = possible formation types such as sand/shale);
• the modes are solved in a least-squares sense by minimizing the sum (over a depth interval) of the squared residual between the tool measurement and the reconstructed log — "equivalent to solving a system of N equations and N unknowns";
• the rectangular deconvolved log is convolved with the vertical response function of the tool to produce a reconstructed log;
• the reconstructed log undergoes interactive refinement by an operator;
• finally, a volumetric analysis (ELAN) is performed on the refined rectangular deconvolved log to produce an output record medium illustrating and quantifying each layer.

§ 102 analysis — does it anticipate?

No. Under § 102 a single reference must disclose every limitation arranged as in the claim (Net MoneyIN v. VeriSign). The '562 discloses the post-squaring half of the '803 workflow and omits the pre-squaring half that the '803 claims make mandatory.

The third, fourth, and fifth rows are the '803 patent's entire stated point of novelty. As the '803 specification itself says, the '562/SHARP approach is limited because "several distributions of properties (square logs) may be generated that can be convolved to compare well and thus 'match' with the same low resolution measurement data log" — i.e., the answer is non-unique because the solution space is seeded from the log rather than constrained by volumetrics. '562 cannot anticipate a claim whose novelty is the cure for '562's own admitted defect.

Which claims does it potentially anticipate? Literally: none of the 38 claims, because claims 2–16 all depend from claim 1, claims 18–29 depend from claim 17 (or 18), and claims 31–38 depend from claim 30 — and each of the three independent claims (1, 17, 30) contains at least one facies-volumetric limitation absent from '562.

Where it is nonetheless the closest and most dangerous art — § 103:

Net: '562 is the primary reference and the natural lead in any obviousness combination, but it is an anticipation near-miss, not an anticipation.

---

4. Reference 2 — US 2002/0133323 A1 (Dahlberg)

Full citation. U.S. Patent Application Publication No. 2002/0133323 A1, Volumetric laminated sand analysis, inventor Kenneth E. Dahlberg. Filed 2001-01-31; published 2002-09-19. Published ~16.5 months before the '803 filing date → § 102(b) printed publication (not § 102(e), since it published more than a year before filing).

Family (for completeness). The disclosure also appears as EP 1 368 777 A2 (European counterpart to the Dahlberg application; the Espacenet/Google record shows it proceeded to a grant-intent communication, then lapsed for want of national translations and non-payment of renewal fees). I note this as a family-mate of the citation, not as separate '803 prior art.

Brief description. Dahlberg discloses Volumetric Laminated Sand Analysis ("VLSA") — a deliberately low-resolution alternative to the high-resolution thin-bed approach. Key steps:

• the thinly bedded interval is modeled as a sequence of approximately parallel, planar beds classified into bed types (e.g., sandstone beds, shale beds);
• each bed type is characterized by estimates of petrophysical properties — porosity, capillary-pressure behavior/water saturation, aspect ratio — plus frequency of occurrence, drawn from core data, regional knowledge, conventional or high-resolution logs, or core image analysis;
• theoretical models combine these to produce estimated average log responses over the interval;
• estimated log responses are compared to measured log averages for consistency;
• the model is refined by Monte Carlo inversion: random earth models are sampled from the parameter distributions, log responses computed, and consistent models retained; distribution statistics for HPV, sand porosity, water saturation, and net-to-gross are then computed;
• Dahlberg expressly touts that all logs (not just resistivity) are used, that shale fraction is extracted from all available logs while integrating a priori information such as a high-resolution image log, and that "a completely general solution technique … provides stable solutions and uncertainty assessments."

§ 102 analysis — does it anticipate?

No. Dahlberg is the closest prior art to the philosophy behind '803 — constrain the log inversion with a rock model and quantify uncertainty — but it does not teach the claim structure.

Which claims does it potentially anticipate? Again none of the 38 claims, and for a cleaner reason than with '562: Dahlberg never detects bed boundaries or produces a squared log, so it cannot meet independent claims 1, 17, or 30, and every dependent claim inherits that failure.

Where it is nonetheless relevant — § 103 and the "constraint" claims:

Dahlberg is the most pertinent art against the minimum/maximum constraint features, because VLSA propagates parameter uncertainty through a Monte Carlo procedure to produce distributions, which is the same technique the '803 specification recites for deriving constraints ("these minimum/maximum values are computed from uncertainties known from the volumetric description of the facies (e.g., employing the Monte Carlo method)"). Specifically, Dahlberg bears on:

• Claim 5 — min/max constraints for property values derived from the volumetric descriptions, with adjustment confined within them;
• Claim 28 — same, phrased as constraints "for various facies … derived from imported volumetric analysis";
• Claim 30 — the independent computer-implemented claim in which the min/max-constraint language is the only feature distinguishing it from claim 1.

That last point is the sharpest § 103 attack available on this patent: claim 30 = claim 1 + Monte-Carlo-derived volumetric constraints, and claim 1's core loop is '562 while Monte-Carlo-derived volumetric uncertainty is Dahlberg. A '562 + Dahlberg combination maps onto claim 30 almost element-for-element, with the residual gap being the per-bed facies-assignment step. Dahlberg also bears on:

• Claims 13, 15, 37 — "volumetric description derived from a thick bed located in the same laminated formation." Dahlberg's use of core/regional data and a priori high-resolution information for bed-type parameters is analogous but does not literally teach the thick-bed-to-thin-bed borrowing.

---

5. A structural point that changes the § 102 framing

US 5,461,562 is expressly incorporated by reference into the '803 specification: "U.S. Pat. No. 5,461,562 … is incorporated by reference for all purposes and made a part of the present disclosure." The '803 specification also recites the SHARP method — and "more current versions … entitled 'SHARP GF' and 'SHARP LITE,' each of which, including documentation, is available from Schlumberger Technology Corporation" — as the admitted prior art against which the invention is defined.

Two consequences:

1. The incorporated '562 disclosure is part of the '803 written description. The patentee has, in effect, conceded that everything in '562 is prior knowledge. The novelty of '803 therefore reduces entirely to the facies-volumetric seeding of the squared log (claims 1, 30) and the per-thin-bed value computation from an imported volumetric description (claim 17).
2. If SHARP GF and SHARP LITE were publicly available or on sale more than one year before 2004-02-02, they are independent § 102(b) prior art in their own right — and the specification's own admission of their availability is evidence of that public accessibility. I flag this as a potential § 102(b) on-sale/public-use theory not on the face of the patent, which is exactly the kind of art an ex parte reexamination requester or an invalidity defendant would pursue.

---

6. Additional prior art identifiable in the record (beyond the two face citations)

These are my identifications from the specification and the '562 disclosure, not entries in the '803 "References Cited" list. Flagged as such.

Note that the "Similar Documents" and forward-citation lists on the Google Patents record are not prior art to '803 unless they qualify under § 102(a)/(b)/(e) by date. Several of them (e.g., US 2010/0312478, US 2015/0192694, WO 2019/152238, US 12,055,672) post-date '803 and are therefore irrelevant to § 102. Do not let those into a claim chart.

---

7. Bottom line on anticipation

Neither cited reference anticipates any of the 38 claims, and I can state that with high confidence based on the claim text and the reference disclosures.

• US 5,461,562 is the most relevant prior art to US 6,963,803. It discloses the entire squared-log → convolution → compare → iterate machinery, the ELAN volumetric post-processing, and the output-record-medium claims. It fails to disclose only the three limitations at the heart of '803: importing defined facies with volumetric descriptions, assigning one such description to each detected bed, and generating the per-bed property value from that imported description. Because claims 2–16, 18–29, and 31–38 all depend from independent claims 1, 17, or 30, that single gap insulates the whole claim set from anticipation.
• US 2002/0133323 A1 (Dahlberg) is the most relevant prior art against the min/max-constraint features (claims 5, 28, 30) via its Monte Carlo uncertainty propagation on facies/bed-type parameters. It fails to disclose bed-boundary detection and squared-log generation entirely, so it anticipates nothing — but it is a strong § 103 second reference against the claim-30 distinguishing limitation.

The patentability question therefore turns on § 103, not § 102: a '562 + Dahlberg combination aimed at claim 30, plus the § 102(b) SHARP-product theory, is where the real invalidity exposure lies. That is also consistent with the examiner having allowed the claims with both references of record.

Practical relevance: the anticipation question is moot for enforcement purposes, for the reasons already established in the earlier PTAB and litigation sections — the patent lapsed for non-payment of the 12-year maintenance fee effective 2017-11-08. Anticipation analysis remains useful only for (a) reverse-engineering a prior-art story if the status record is ever reversed, or (b) portfolio/historical study.

---

Explicit uncertainties in this section

1. The "References Cited" list is sourced from the Google Patents structured record, not pulled directly from USPTO PatentCenter/PAIR. I reached my tool limit before completing that check. Any non-patent-literature citations on the face of the patent (if the examiner listed any) would not appear in the Google Patents structured list and are therefore a possible omission from Section 2. Confirm against Patent Center's "References Cited" tab.
2. I inferred, rather than read, the § 102(b) status of the SHARP/SHARP GF/SHARP LITE software. The '803 specification states these are "available from Schlumberger Technology Corporation," which is strong evidence of public accessibility, but I did not locate a publication date or product-availability date for them.
3. The EP family member for Dahlberg (EP 1 368 777) — I saw grant-intent and lapse events on the record but did not confirm the grant number or exact publication date; treat the family reference as indicative, not verified.
4. Claim-element mapping tables above are my own comparison, not an examiner's or a court's. They reflect the claim language as issued and the reference disclosures as reproduced in the sources cited; a full-claim-text element-by-element chart against the '562 and Dahlberg PDFs would be the next step if this were for a brief.

I'll ground this § 103 analysis in the actual cited references before drafting.

Obviousness Analysis — US 6,963,803 B2 under 35 U.S.C. § 103

Building on the prior sections. The claim-construction map, the claim-class breakdown (output post-processing / convolution / multi-property / facies-assignment mechanics / thick-bed borrowing), and the enforceability findings (lapsed for non-payment effective 2017-11-08; no litigation, no PTAB proceeding) are taken as given and are not repeated. This section addresses only the § 103 merits.

Source discipline. The "Prior Art" section of the Google Patents record for US 6,963,803 lists exactly two references: US 5,461,562 A (Tabanou et al., Schlumberger, issued 1995-10-24, "…on a Workstation") and US 2002/0133323 A1 (Dahlberg, ExxonMobil, published 2002-09-19, "Volumetric laminated sand analysis"). I retrieved and read substantial text of both, plus the published European counterpart of the '562 patent, EP 0 539 272 A2 (published 1993-04-28), whose disclosure is materially broader than the U.S. claim set and which is independently available as § 102(b) art. I have not read either reference in full; my mappings below are grounded in the passages quoted with URLs, and I flag where a mapping rests on inference rather than express text.

---

1. Bottom line

Claims 1, 17, and 30 are each obvious under § 103 over US 5,461,562 A in view of US 2002/0133323 A1, and the dependent claims add nothing that is not either (a) expressly disclosed in the '562 patent's own specification and figures, or (b) expressly disclosed in Dahlberg's Monte Carlo / bounded-distribution treatment. The single element that the '562 patent does not expressly supply — seeding the step-and-shoulder ("squared") log property values from a facies' volumetric description rather than from a histogram of the high-resolution log — is supplied by Dahlberg, and the motivation to make that substitution is supplied by the '562 patent's own text.

Two points sharpen the conclusion:

1. Both references are § 102(b) art. US 5,461,562 issued 1995-10-24 and US 2002/0133323 A1 published 2002-09-19 — each more than one year before the 2004-02-02 filing date. That matters because US 5,461,562 is commonly owned with the '803 patent (both Schlumberger). A patent owner may reach for the § 103(c) common-ownership exception, but § 103(c) removes only art that qualifies solely under § 102(e), (f), or (g). Because the '562 patent is § 102(b) art, § 103(c) is unavailable. (Dahlberg is ExxonMobil-assigned, so no common ownership exists in any event.)
2. The '803 specification is riddled with admissions that the surrounding workflow was known: facies/lithofacies classification from high-resolution resistivity; squaring by second-derivative inflection detection; convolution with the tool vertical response function; ELAN volumetric analysis; and iterative manual adjustment to minimize the residual. Those admissions shrink the § 103 inquiry to the one or two features that are actually new.

Practical context. Per the earlier sections, the patent expired for non-payment of maintenance fees effective 2017-11-08, no AIA trial was ever filed, and the § 286 damages lookback now begins after the lapse. This analysis is therefore most useful for a validity opinion, a retrospective invalidity defense, or an invalidity contention in a case concerning pre-2017 conduct — not for an IPR petition, which would be moot in economic terms.

---

2. Proposed level of ordinary skill (Graham factor 1)

Not established by the record; proposed for argument:

A person of ordinary skill in the art as of 2004-02-02 would have a bachelor's degree in petroleum engineering, geophysics, electrical engineering, or physics, plus approximately 2–3 years of experience in open-hole well-log interpretation (or a master's degree with approximately 1 year), and would be familiar with: (a) convolution-based vertical-resolution enhancement of logs and vertical response functions; (b) ELAN-style volumetric (mineral/fluid-volume) analysis; (c) laminated/thin-bed reservoir evaluation; and (d) routine statistical inverse methods, including Monte Carlo sampling with bounded parameter distributions.

Every reference relied on is addressed to this person. The '562 patent is a workstation-based implementation of exactly this workflow in the same company; Dahlberg's is the same problem in a competitor's shop.

---

3. The references — what each discloses

3.1 US 5,461,562 A (Tabanou et al.) — primary reference; the "SHARP" method

The '803 patent states that it "provides both improvements and modifications to the various aspects of the system, apparatus, and method described in the '562 patent" and incorporates it by reference. The '562 patent therefore functions as admitted framework art.

Two additional disclosures appear in the '562 patent's specification and in EP 0 539 272 A2, and are important because they map onto the '803 patent's "facies" language:

• "Modes" are expressly extended to "electro-facies": "A high resolution log such as SHDT, FMS, [OBDT] and EPT is used to delineate the location of boundaries, and to classify the beds into a finite set of modes… It is only when several high resolution logs are available, each sensitive to different petrophysical parameters, that one should extend the concept of modes to electro-facies." And: "a histogram of the high resolution measurement, or a cluster analysis of several high resolution logs is used to assign each layer to a specific mode or 'electro-facies'." — http://data.epo.org/publication-server/rest/v1.2/patents/EP0539272NWA2/document.pdf
• The stated motivation for importing external knowledge and for a good first guess: "Very often, a more physical solution can be obtained if a good first guess is assumed and if the problem is 'regularized' by taking into account external a priori knowledge." — same PDF.

That second passage is the hinge of the whole obviousness case: the primary reference itself teaches that the accuracy of the deconvolution depends on (i) a good initial guess and (ii) external a priori knowledge. Dahlberg is a published, indexed source of exactly those two things.

Also note that the '562 patent's own workflow performs ELAN volumetric analysis on the same formation, producing volumes — so the "volumetric description of a facies" is not foreign to the primary reference; it is already computed downstream in the same tool. The '803 patent effectively reverses that data flow and uses the ELAN volumes as the input to the squared log.

3.2 US 2002/0133323 A1 (Dahlberg) — secondary reference; "VLSA"

Cross-reference to the '803 specification. The '803 patent tells us its constraints "are computed from uncertainties known from the volumetric description of the facies (e.g., employing the Monte Carlo method)." That sentence is, functionally, Dahlberg's Fig. 8 table and claim 4. The overlap is close to verbatim in substance, and it is the feature the patent uses to distinguish independent claim 30 from claim 1.

---

4. Claim 1 — element-by-element

Result: claim 1 is obvious over '562 in view of Dahlberg. The only gap in '562 standing alone is the source of the per-bed property value. '562 takes it from a histogram of the high-resolution measurement; Dahlberg takes it from a forward calculation off a bed type's volumetric/petrophysical description. Both are methods of initializing the same quantity for the same purpose in the same workflow, and the primary reference expressly says the answer depends on getting a good first guess and importing external a priori knowledge.

Dependent claims 2–16

---

5. Claim 17 — element-by-element

Claim 17 is the broadest claim on the face of the patent and drops the convolution/compare/iterate recitations entirely.

This is the cleanest § 103 theory in the patent. Given that Dahlberg independently performs the last two elements in a thinly laminated sandstone/shale reservoir, the only thing supplied by '562 is boundary detection from a high-resolution log — and '562's use of a high-resolution FMS log to locate boundaries in laminated reservoirs is its core teaching. The objective technical problem ('803 col. 1: low-resolution tools average properties over regions much larger than the thin bed) is the identical problem Dahlberg states in its own background.

Related § 102 flag (secondary). If one were to accept that Dahlberg's "frequency of occurrence of the different bed types… may be estimated from conventional or high-resolution logs or from core image analysis" implies detecting bed boundaries from a high-resolution log, claim 17 could be argued as anticipated rather than obvious. I would not lead with that theory: Dahlberg's preferred embodiment deliberately operates on interval averages, and its "high-resolution" reference is to bed-type frequency, not to boundary detection. The § 103 combination is the stronger, safer theory.

---

6. Claim 30 — element-by-element

Claim 30 = claim 1 + (a) "computer implemented," and (b) min/max constraints for the formation property values derived from the imported volumetric descriptions, with adjustment confined within those constraints.

• (a) "Computer implemented" is not a patentable distinction. The '562 patent is a "workstation based system" claim set (claim 11) and its method claim 9 is expressly "[a] method practiced by a computer processor." Dahlberg's Monte Carlo inversion is claimed "carried out using a programmed digital computer" (EP1368777B1 claim 6). Both references anticipate the computer-implementation recitation.
• (b) The constraint feature is the point of alleged novelty over claim 1 — and it is disclosed by Dahlberg with unusual precision:

The '803 specification confirms the intended correspondence by tying its constraints to the same statistical machinery Dahlberg uses: "these minimum/maximum values are computed from uncertainties known from the volumetric description of the facies (e.g., employing the Monte Carlo method)." A patent owner arguing that claim 30 is non-obvious over Dahlberg must argue that Dahlberg applies the Min/Max bounds to petrophysical parameters while claim 30 applies them to the formation property value. That is a thin distinction, because Dahlberg's forward model is precisely the mechanism that maps one to the other, and the '803 specification does not describe any non-routine step in that mapping.

Dependent claims 31–38

• 31, 32 — volumetric analysis on the optimized log; storing on an output record medium illustrating each layer, including thin beds → '562 claims 9(f)–(l).
• 33, 34 — reconstructed log by convolution; multiple properties optimized against multiple measured logs → '562 (convolution means, vertical response function) + Dahlberg (joint probability index across log types).
• 35–38 — thin beds; squaring the high-res log; per-bed value matched to a facies; application to a plurality of high-resolution logs → '562 histogram/crossplot/multi-dimensional analysis; EP 0 539 272 electro-facies from multiple high-resolution logs of differing petrophysical sensitivity.

---

7. Motivation to combine — the enumerated rationales

A petition or invalidity contention should plead these cumulatively; each is independently sufficient under KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007).

1. Same field of endeavor, same problem, same artisan. Both references analyze thinly laminated sand/shale reservoirs in which low-resolution tools average over intervals larger than the beds. Dahlberg's background frames the very "high-resolution vs. low-resolution" dichotomy on which the '803 patent relies, and criticizes the high-resolution approach's limitations. Combination of references within a single recognized field to solve a shared problem is the paradigm case.
2. The primary reference expressly calls for what the secondary reference supplies. EP 0 539 272 A2: "a more physical solution can be obtained if a good first guess is assumed and if the problem is 'regularized' by taking into account external a priori knowledge." Dahlberg supplies (i) a principled source of external a priori knowledge (a priori statistical earth model from core data or regional knowledge) and (ii) a mechanism for generating the first guess. This is not hindsight — it is the primary reference's own stated desideratum.
3. The '803 specification itself identifies the problem the combination solves. The patent states that the prior SHARP method "may not be unique" — several squared logs convolve to match the same measured log — and that property-by-property minimization "may result in spurious solutions" because rock properties are interdependent. Dahlberg addresses exactly this: its joint probability index fits all log types jointly, and its acceptance test retains only models that fit within tolerance. A skilled artisan seeking to fix the non-uniqueness the '562 workflow concedes would look to a published, same-problem reference that does so.
4. Known automation incentive / elimination of a known manual burden. '562 claim 9 requires "manual interaction with a human operator" at both the comparing and the modifying steps, and the '803 specification criticizes the prior art because "users have to manually depth match the logs and manually adjust the properties in many layers," making the interpretation too slow to affect operations. Dahlberg's automated Monte Carlo accept/reject loop is a known route to automating precisely that task. KSR: "the desire to enhance commercial opportunities by improving a product or service" and "the mere fact that a combination is automated" are recognized motivations.
5. Design incentive within the existing toolchain. The '562 workflow already runs ELAN volumetric analysis on the same formation. Using the resulting volumes to seed the squared log is a reversal of an existing data flow within the same software, requiring no new instrumentation and no new data acquisition. The '803 specification concedes the interchangeability of value sources, stating that thick-bed values "may be obtained by taking the median of the high resolution log measurements for that bed. Alternatively, the values of the formation property for the thick beds may be obtained from the facies definition, and more particularly, from the volumetric description of the facies." Where a patent's own specification presents two sources of the same quantity as alternatives, selecting one or the other is a design choice within the ordinary skill in the art.
6. Predictable result, finite set of identified solutions. The field had two established thin-bed analysis families — the high-resolution route ('562, sharpen the logs) and the low-resolution route (Dahlberg, model the interval) — and the combination of the two is the finite, predictable pairing. Where "a finite number of identified, predictable solutions" is presented, KSR teaches that the combination is obvious to try with a reasonable expectation of success.

---

8. Reasonable expectation of success (Graham factor 3)

High. Neither reference reports a technical obstacle to the combination:

• '562 already produces a squared (rectangular deconvolved) log with constant per-bed values — the exact data structure claim 1 requires. Dahlberg already produces per-bed-type property values by forward computation. Placing Dahlberg's computed value into '562's step function is a substitution of input values into an existing structure, not a re-architecture.
• '562 already convolves and compares and iterates, and already performs ELAN volumetric analysis downstream. The combined method therefore retains every operative step of the primary reference.
• Dahlberg demonstrates the combination works, at least in the interval-average limit, on a synthetic dataset and on a real cored interval between two massive sandstones ("the method recovered a good answer to the thin-bed problem that was posed"). The artisan would have had confirmatory evidence that forward-modeled, volumetrically described bed types reproduce measured logs within tolerance.

---

9. Where the patent owner will push back, and how strong each argument is

---

10. Non-asserted but available art

Not in the "Prior Art" section of the record, but surfaced in the '562 patent's own background and in the Google Patents "Similar Documents" list — useful as corroboration that each individual step was routine, and potentially as primary art if a reference's date or content is later disputed:

• Trouiller et al. (1989), "Thin-bed reservoir analysis from borehole electrical images" — thin-bed analysis from borehole images (supports the facies-from-image step).
• Nelson et al. (1990), "Improved vertical resolution of well logs by resolution matching" — supports claim 6/10's convolution-and-match step.
• Flaum & Galford (1987), "Enhanced vertical resolution processing of dual detector gamma-gamma density logs" and Galford et al. (SPE 15441, 1986) — "alpha processing," a two-step resolution-enhancement technique disclosed within the '562 patent's own background.
• Barber (1988), "Induction log vertical resolution enhancement — physics and limitations" — also discussed in the '562 background.
• Suau et al. (SPWLA 25th Symposium, 1984), "Interpretation of very thin gas sands in Italy" — pattern recognition and cross-correlation of EPT/neutron/density/sonic logs for thin gas sands; discussed in the '562 background and directly supportive of the multi-log facies-classification claims (14, 25, 38).
• Schlumberger's published SHARP / SHARP GF / SHARP LITE documentation. The '803 specification states that this documentation "is available from Schlumberger Technology Corporation, Houston, Tex." That is a potential § 102(b) printed publication describing the admitted prior-art workflow. I have not located or verified it. If it predates 2003-02-02, it could operate as primary art and simplify the case considerably; obtaining it should be the first file-wrapper / discovery step.

---

11. Confidence and gaps

• High confidence that claims 1, 17, and 30 are obvious over US 5,461,562 A in view of US 2002/0133323 A1. The mapping of every element except the facies-volumetric initialization is express in '562's own claims and figures; the missing element is express in Dahlberg; and the motivation is express in the primary reference's specification.
• High confidence that the constraint feature of claim 30 (Min/Max from bounded distributions, Monte Carlo) is disclosed by Dahlberg at [0087] and claim 4 — the alignment with the '803 specification's own "Monte Carlo" sentence is close to dispositive.
• Medium confidence on the "change in principle of operation" rebuttal, because I have not read the '562 patent's full specification and cannot rule out a passage in it that more firmly ties the mode values to the high-resolution measurement in a way a patent owner could characterize as critical. This is the one place a fuller read of the reference could change the framing (though not, in my assessment, the outcome).
• I have not reviewed the file wrapper, so I cannot rule out a Rule 132 declaration, a terminal disclaimer, or an applicant argument that shaped the claim scope in a way that matters to claim construction of "squared log" or "volumetric description."
• I have not verified any secondary-considerations evidence, because I found none. The absence of a litigation record (per the earlier section) means there is no developed record of commercial success, copying, or industry praise to rebut — which itself favors the obviousness conclusion.
• Dates relied on: '803 filing/priority 2004-02-02; US 5,461,562 issued 1995-10-24; EP 0 539 272 A2 published 1993-04-28; US 2002/0133323 A1 published 2002-09-19 (filed 2002-01-09; provisional 60/265,659 filed 2001-01-31). All are more than one year before the '803 filing date, making them § 102(b) art and rendering § 103(c) inapplicable.

Primary URLs consulted for this section:

• https://patents.google.com/patent/US5461562
• http://patentimages.storage.googleapis.com/1c/af/20/84288971cacfe3/US5461562.pdf
• https://patents.justia.com/patent/5461562
• http://data.epo.org/publication-server/rest/v1.2/patents/EP0539272NWA2/document.pdf
• https://patents.google.com/patent/US20020133323
• https://patentimages.storage.googleapis.com/77/3d/15/623d0d590603e8/US20020133323A1.pdf
• https://patents.google.com/patent/EP1368777B1/en
• https://patents.google.com/patent/US6963803/en