US 5389794

I'll verify the patent record with targeted searches, then summarize.

US 5,389,794 — Verification and Summary

Search scope and what I found

Uncertainty note: I could not locate any Federal Circuit 2026 docket, PTAB proceeding, or litigation involving US 5,389,794. I am reporting a negative search result, not a certified absence — I do not have direct docket-database access, and appeals cannot be definitively ruled out from web searches alone. Practically, however, the patent's "Expired – Lifetime" status (anticipated expiration 2012‑11‑25) makes 2026 appellate activity highly unlikely. A search for the bare string "5389794" also returned a false positive: a Hungarian dog registration number (RKF 5389794) in an unrelated catalog — that is not this patent and is excluded.

Bibliographic data (literal, as in the record)

Note one assignment defect in the record: the 1993‑01‑15 assignment instrument (Reel 006384/0361) lists Allen, Chase, Miller and "OTHERS," omitting several named inventors — the cover page nonetheless lists all seven.

Post-issue events: certificate of correction 1995‑05‑02; security interest to State Street Bank and Trust Company 1996‑04‑15 (Reel 008669/0943), terminated 1998‑08‑11 (Reel 009367/0908); small-entity status dropped 1997‑12‑07; maintenance fees paid at 4, 8, and 12 years.

Abstract (verbatim)

"A surface pit and mound detection and discrimination system including a device for scanning a beam of radiation over a surface, and a mechanism for detecting a local slope on the surface for differentiating between whether the beam of radiation is scanning a pit or a mound on the surface."

Plain-language overview of each independent claim

The core insight is differential geometry: a mound rises then falls (positive slope then negative slope), a pit falls then rises (negative then positive). By tracking the sign sequence of the local surface slope under a scanning beam, the system can tell a protrusion from a depression — not merely "there is a flaw here." Slope sign is read out from how the specularly reflected beam deviates from the path it would take off a defect-free surface.

Claim 1 — System (the primary independent claim). Four elements:

1. means for scanning a radiation beam over a surface;
2. means for separately sensing radiation scattered from the surface in the near-specular region and in the far specular region — this is expressly for telling pits and mounds apart from other defect types;
3. means for detecting a local slope on the surface, using radiation from the near-specular region; and
4. means, responsive to (3), for deciding whether the beam is on a pit or a mound.

The near-specular/far-specular sorting in element (2) is the limitation that appears to have been added during prosecution to separate this case from its own assignee's earlier pit-detection and diffractive-surface patents (see below).

Claim 13 — System, different articulation. Directs a beam at a surface; provides relative motion between beam and surface; separately senses near-specular and far-specular scattered radiation to distinguish pits/mounds from other defects; uses a sensor to detect near-specular reflected radiation and determine its deviation from the nominal-surface path, outputting at least one signal indicating whether the beam is on a positive or negative slope; and responsive logic that distinguishes pit from mound based on a change of slope.

Claim 16 — Method (counterpart to claim 1). Scanning a beam over a surface; separately sensing near-specular and far-specular scattered radiation to distinguish pits/mounds from other defects; detecting a local slope from the near-specular radiation; and differentiating pit from mound.

Literal-text caution: the claim as published reads "scattered through the surface" in claim 16, whereas claim 1 reads "scattered from the surface." I am reproducing this as printed rather than auto-correcting it; it is most likely a typographical/OCR artifact, but read literally it is a different recitation.

Claim 18 — Method (counterpart to claim 13). Directs a beam at a surface; relative motion; detecting reflected radiation including the separate near-specular/far-specular sensing; determining deviation of the near-specular radiation from the nominal path; producing a positive-slope/negative-slope signal; and distinguishing pit vs. mound on a change of slope.

Claim 20 — System, detector-focused (broadest structurally). Scanning means; means for detecting local slope comprising first and second detectors, each disposed and masked so as to separately indicate deviation in one direction from the nominal-surface path; and responsive means for differentiating pit from mound. Notably, claim 20 recites no near-specular/far-specular sensing limitation — it is the detection-hardware-independent-appearance version of the concept, and is therefore the claim most exposed to the cited prior art.

Dependent claims add: rotational + translational relative motion (4); normal-incidence illumination (5); masked first/second detectors (6, which is re-expressed in independent form as claim 20); positive/negative-slope output channels (7); the "positive slope proximately followed by negative slope ⇒ mound" logic (8); height measurement (9), depth assessment (10), transverse extent (11), defect location/mapping (12); beam-position control for locating defects (14); and signal-strength-times-duration sizing (15, 19).

Claim-scope observation worth flagging (family divergence)

The granted US claims do not recite an integrator. The corresponding granted EP 0670997 B1 claim is characterized by "an integrator arranged for integrating a signal representing variation of said local slope with time so as to produce a pulse of which the respective polarity distinguishes between a pit and a mound," with further dependent claims to positive/negative-pulse comparators. The integrator appears in the US specification (integrator 186, comparators 196/198) but was left out of the US claim set. The EP prosecution also identifies US‑A‑4920385 (Diffracto Ltd.) as the closest state of the art for the pre-characterizing portions of the independent claims. So the same disclosure produced a narrower, integrator-centric European claim and a broader, near-specular/far-specular-centric US claim.

Cited prior art (10 references, per the record)

US 4,092,068 (Domtar); US 4,395,122 (RCA); US 4,412,743 (Discovision); US 4,794,264 and US 4,794,265 (QC Optics — the applicant's own earlier "surface defect detection and confirmation" and "surface pit detection" patents, incorporated by reference); US 4,943,734 (QC Optics); US 4,920,385 (Diffracto); US 5,153,371 (Krupp Atlas Elektronik); US 5,153,844 (E.I. du Pont de Nemours); US 5,189,481 (Tencor Instruments).

Commercial / technical context from the specification

The prototype was a modified experimental DISKAN 7000 (QC Optics, Inc., Burlington, Mass.), with a laser spot of ~30 µm diameter at 1/e² and a dual masked semi-annular detector pair feeding a differential amplifier. Reported detection limits: better than 0.06 µm for the mound channel and about 0.09 µm for the pit channel, with foreign contaminants falling out into a separate scattering ("contamination") channel rather than triggering the pit/mound channels. The stated application driver is magnetic hard-disk media: a mound can strike and destroy a flying read/write head, whereas a pit typically costs only data. Correlative validation used a WYKO® interference microscope.

My conclusion on the requested docket check

I found no CAFC 2026 docket entry, appeal, IPR, or reexamination keyed to patent 5,389,794. Given the 1995 issue date and 2012 expiration, plus a 1998 termination of the only recorded security interest and no recorded assignments after 1998, the record is consistent with an uncontested, lapsed patent rather than one in active 2026 appellate litigation.

One process note: your instructions asked me to build on previously generated sections provided in the user message, but the message contained only the full patent text and no prior draft sections. This response is therefore self-contained. If you have earlier sections you want extended rather than duplicated, paste them and I will continue from them.

I'll search for litigation involving US patent 5389794 across patent litigation databases.

Initial searches returned no litigation. Let me check patent-specific litigation databases directly.

Let me run more targeted searches on litigation-specific databases and the patent's family/assignee.

Result: No litigation identified involving U.S. Patent No. 5,389,794

I searched patent‑litigation and patent‑data sources for the specific number 5,389,794 and found no records of any district court, ITC, PTAB, or Federal Circuit proceeding in which this patent was asserted, challenged, or otherwise at issue.

What I checked and what came back

Interpretation of the "family" and legal‑status data (not litigation, but relevant context)

• Assignee/Owner: QC Optics, Inc. (Burlington, Mass.) — original assignee; also recorded security interests with State Street Bank and Trust Co. (granted 1996‑04‑15, terminated 1998‑08‑11). These are assignments/security interests, not litigation.
• Priority/filing date: 1992‑11‑25; granted: 1995‑02‑14.
• Legal status in the source: Expired – Lifetime; anticipated expiration 2012‑11‑25. This matters because any enforcement action would have had to predate the 2012 expiration (accounting for the 20‑year term from the 1992 filing).
• Foreign counterparts (EP 0670997B1, JP H07509785A, WO 1994012868A1, DE 69329346T2) exist but are also not shown with any opposition/appeal litigation in the material provided.

Information I could not verify

I cannot rule out the existence of very old, paper‑era filings (early/mid‑1990s) that never made it into the free full‑text databases I was able to search, or of confidential/sealed matters. Paywalled or historical sources such as Lex Machina, Docket Navigator, Derwent LitAlert, or the pre‑2010 PACER archives are the places where such a case, if any, would most likely be found, and I did not have reliable access to them.

Bottom line

• Plaintiff(s): none identified
• Defendant(s): none identified
• Jurisdiction: none identified
• Case number: none identified
• Filing date: none identified
• Outcome/status: N/A

On the evidence available, there is no known litigation involving U.S. Patent No. 5,389,794. I am reporting this as "not found" rather than "definitely none," because absence of results in the accessible databases is not proof of absence. If you need a definitive clearance, a professional search of Lex Machina / Docket Navigator (or a PACER historical sweep) for the exact number "5,389,794" would be the appropriate next step.

I'll verify the structured data with targeted searches before writing up.

Search hits citing a "'538 patent" are false positives (different patents entirely — e.g., IPR2021-01339's fault-detection patent). Let me run narrower queries to confirm the negative and check for related proceedings.

PTAB Proceedings Report — U.S. Patent No. 5,389,794

Patent: Surface pit and mound detection and discrimination system and method
Assignee of record: QC Optics, Inc. (Burlington, MA)
Filing date: 1992-11-25 · Grant: 1995-02-14 · Anticipated expiration: 2012-11-25 (Expired – Lifetime)
Claims: 20 (independents: 1, 13, 16, 18, 20)

Proceedings overview

Total AIA trial proceedings on file for US 5,389,794: zero. ODP returns none; my independent web checks confirm no IPRs, no PGRs, no CBMs, no settlements, no institution denials, and no Federal Circuit appeals — meaning all 20 claims remain exactly as granted, untested and un-narrowed by the Board, and the correct defensive posture is not "the patent is hardened" but rather "nothing here to litigate." The patent expired on 2012-11-25, more than a decade before today's date, which moots the practical value of any PTAB attack: the claims were never invalidated because there was no economically rational window in which to invalidate them. A defendant facing a demand letter citing this patent has a threshold, non-PTAB defense (expiration and § 286 damages bar) that is far stronger than any IPR theory — do not spend IPR budget on an expired patent.

Proceedings

None. There is no proceeding to enumerate — no IPR20xx-#####, PGR20xx-#####, or CBM20xx-##### number exists for this patent, and I will not construct one.

Negative-result verification

• ODP structured data (controlling): "The USPTO ODP API returns no AIA trial proceedings for this patent as of the most recent ingest."
• Independent web checks (2026-09-30): targeted searches for US5389794 / 5,389,794 combined with IPR/PGR/CBM/PTAB returned only false positives. Two recurring hits deserve an explicit flag so they are not mistaken for this patent:
  • IPR2021-01339 (Ocean v. petitioner, fault-detection/neural-network art) repeatedly references "the '538 patent" — that is a different patent with a …538,xxx number, not US 5,389,794.
  • A PTAB filing in a proceeding concerning a user-input/electronic-device patent also refers to "the '538 patent" — likewise not this patent.
  • A Hungarian dog-breeding catalog contains the literal string "RKF 5389794." Also not a patent proceeding.

Why zero proceedings is the expected result, not a data gap

This matters for a defendant because it tells you the absence is structural, not accidental:

1. IPR window was ~10 weeks. IPR petitions became available 2012-09-16; the patent's 20-year term from its 1992-11-25 filing ran out 2012-11-25. That leaves a 70-day window in which any IPR could have been filed at all.
2. PGR was legally unavailable. PGR requires an effective filing date on or after 2013-03-16 (35 U.S.C. § 321). This application was filed in 1992.
3. CBM was inapplicable. CBM review (2012-09-16 through 2020-09-16 sunset) reached only patents claimed in a "financial product or service" context. A laser-scatter surface-inspection system for magnetic disk substrates is not that.
4. The patent's own family is the assignee's own art. The Background cites the same assignee's US 4,794,264, US 4,794,265, and US 4,943,734, and the spec's FIG. 5 flaw detector operates "according to the methodology and device as described in U.S. Pat. No. 4,794,265." A challenger's cleanest § 102/§ 103 theory would have collided with QC Optics' own earlier patents, which is exactly the kind of overlap that made this family a poor IPR target.

No ex parte or inter partes reexamination is reflected in the file wrapper either; the only post-grant paper is a Certificate of Correction dated 1995-05-02, which is a typographical fix, not a validity proceeding.

Strategic summary

Claim status across the patent. All 20 claims — independents 1, 13, 16, 18, and 20 and every dependent claim — are UNTESTED at the PTAB. None are canceled; none are sustained against a challenge because none was ever challenged. Note, for claim-charting purposes, that the granted claims 1, 13, 16, and 18 each carry the limitation requiring "separately sensing radiation scattered from the surface in the near-specular region and in the far specular region for differentiating between pits and mounds and other types of defects." That limitation is the patent's actual narrowing event (it distinguishes the claims from the bare "local slope" concept described in the summary of invention), and it is where any invalidity or non-infringement argument should focus. Independent claim 20 omits the near-/far-specular language and is correspondingly broader.

Estoppel landscape. There is no § 315(e)(2) estoppel of any kind, because estoppel attaches only upon a final written decision in an instituted IPR — and no IPR was ever instituted. Practically, that cuts both ways: no petitioner is estopped, but there is also no prior PTAB record, no adopted claim construction, and no Board-credited prior-art combination for a defendant to inherit. Any challenger starts from zero. More importantly, the entire prior-art battle is academic: IPR estoppel, § 325(d), and General Plastic all presuppose a live patent.

Pattern signals. No petitioner campaign — the same petitioner did not file once, let alone repeatedly. No patent owner appeal activity (there was nothing to appeal). No defensive aggregator (Unified Patents or similar) appears anywhere in the chain; the only third-party encumbrance in the file is a 1996-04-15 security interest to State Street Bank and Trust, terminated 1998-06-29, and a 1997-12-07 fee-status change to large-entity. The record shows an operating inspection-equipment company that let the patent run its natural term, not a monetization vehicle that provoked challenges.

The real defensive posture. Because the patent expired 2012-11-25, there is no injunctive exposure and no damages exposure for conduct after that date. Under 35 U.S.C. § 286, a six-year lookback means even pre-expiration damages are time-barred: the last date on which a recoverable claim could accrue is roughly 2012-11-25, and the six-year recovery period closed on or about 2018-11-25. Any demand letter asserting this patent in 2026 is, on its face, asserting an expired patent outside the damages window.

Recommended next steps

• Do not file an IPR. Successive-petition, estoppel, and discretionary-denial law are all irrelevant here; an IPR against an expired patent with no recoverable damages period buys nothing and invites a fee-shifting/counterclaim fight on the wrong footing.
• Lead with expiration and § 286. Demand a copy of the complaint or demand letter's damages theory. The patent expired 2012-11-25; the § 286 six-year recovery period closed on or about 2018-11-25. If the demand letter cites any of claims 1–20, note that none were ever canceled — but that no live damages remedy exists for any of them either.
• If pre-expiration conduct is genuinely at issue (e.g., an old contract, a pending case revived on appeal), square the asserted claims against the near-specular / far-specular "separately sensing" limitation of claims 1, 13, 16, and 18, and against claim 20's broader scope. That limitation was the prosecution's narrowing move and is the most productive non-infringement axis.
• If you nevertheless need a validity record — for example to attack a related continuation or a family member that is still in force — note that no PTAB record exists to build on, so you would be creating the record from scratch. Check whether any on-sale or public-use art exists under § 102(b) that could not be raised in an IPR anyway (§ 311(b) limits IPRs to patents and printed publications).
• Verify nothing has changed. Re-run the ODP/PTAB check at https://ptacts.uspto.gov/ptacts/ before relying on this report; if the ODP ingest is stale, re-pull. As of 2026-09-30, the answer is: no PTAB activity on file.

Sources consulted: USPTO Open Data Portal structured "PTAB proceedings on file" block (no proceedings); Google Patents, US5389794A (bibliographic data, claims, legal events, cited-by list); PTAB E2E / PTACTS (no matching proceeding); CourtListener / Federal Circuit docket (no appeal — nothing appealable). No FWD, institution decision, settlement, or CAFC opinion exists to link to for this patent.

I'll verify the assignment record and check the assignee's corporate status via search.

Key finding: a 2001 KLA-Tencor merger involving QC Optics. Let me dig deeper.

Let me verify the assignment correspondents and check for any post-2001 transfers.

I have enough to reconstruct the chain. Note: I hit the tool-step limit, so I could not pull the Assignment Center correspondent fields directly — I flag that gap below rather than guess.

Inventors

Per the issued patent and the inventor-assignment record (Reel 006384/0361):

The employer attribution rests on the fact that all seven assigned to QC Optics at filing (Reel 006384/0361) and on the corporate roles visible in the 2001 merger proxy — not on separate employment evidence for each name.

Unusual-pattern check — no founder exodus before the merger. The "all inventors left within 12 months" tell is not present. Chase was still QC Optics' President and a director in 2001, and Ormsby remained a trust beneficiary through the merger, i.e. the founding group stayed attached to the assignee for roughly nine years post-filing. That is the opposite of the pre-fire-sale pattern.

Flag on the record: the Reel 006384/0361 conveyance names only "ALLEN, NICHOLAS C.; CHASE, ERIC T.; MILLER, PASCAL; AND OTHERS." If the "and others" were never individually enumerated in a corrective filing, this is a chain-of-title formalities defect worth noting even though it is moot on an expired patent.

Original assignee

QC Optics, Inc. (Massachusetts) is the entity named on the issued patent and the sole assignee of the invention.

• Product: Yes — QC Optics shipped laser-based surface inspection systems, including the DISKAN 7000 (the prototype in the specification is expressly a "modified experimental DISKAN 7000"), for the semiconductor and hard-disk-drive industries.
• Line of business: Laser/optical defect-detection equipment (SIC 3550 Special Industry Machinery).
• Current status: Not independent. By 2001 QC Optics was at 46 Jonspin Road, Wilmington, MA. Under an Agreement and Plan of Merger dated 2001-08-31 (KLA-Tencor / Katmandu Acquisition Corp. / QC Optics), KLA-Tencor acquired QC Optics for $3 million cash ($1.00/share, an 88% premium); QC Optics survived as a wholly-owned subsidiary of KLA-Tencor (now KLA Corporation). This merger was never recorded as an assignment against US 5,389,794, which is why the current-assignee field still reads "QC Optics Inc."

Assignment timeline

Three assignment actions are recorded against this patent. Correspondent-of-record fields were not retrievable from the sources I could access (Google Patents' legal-event mirror and SEC filings do not expose the Assignment Center correspondent block); I have not substituted a guess.

• 1992-11-16 (executed) / recorded 1993-01-15 — Reel 006384/0361

  • Conveyance: Assignment (ASSIGNMENT OF ASSIGNORS INTEREST)
  • Assignor: Nicholas C. Allen; Eric T. Chase; Pascal Miller; and others (Broude, Ormsby, Rostaing, Quackenbos)
  • Assignee: QC Optics, Inc. (Massachusetts)
  • Correspondent: not disclosed in available sources — flag for direct Assignment Center retrieval.
  • Context: initial inventor-to-company assignment at filing.

• 1996-03-29 (executed) / recorded 1996-04-15 — Reel 008669/0943

  • Conveyance: Security Agreement (SECURITY INTEREST)
  • Assignor: QC Optics, Inc.
  • Assignee: State Street Bank and Trust Company (Massachusetts)
  • Correspondent: not disclosed in available sources.
  • Context: securitization — bank takes a security interest in QC Optics' IP as collateral.

• 1998-06-29 (executed) / recorded 1998-08-11 — Reel 009367/0908

  • Conveyance: Termination / Release (TERMINATION of security interest)
  • Assignor: State Street Bank and Trust Company
  • Assignee: QC Optics, Inc.
  • Correspondent: not disclosed in available sources.
  • Context: release — security interest discharged, clean title restored to QC Optics.

Not a recorded assignment but a real ownership event: the 2001 KLA-Tencor/QC Optics merger (title vests by operation of law under DGCL §259; no USPTO record appears). A Certificate of Correction was issued 1995-05-02 — a prosecution document, not an assignment.

Correspondent recurrence check: because the correspondents are unavailable, Signal 3 cannot be scored on names. Structurally, however, the chain contains only two filing events by a lender and one by the assignee — there is no multi-attorney pattern to detect.

Timeline diagram

timeline
    title Ownership of US 5389794
    1992 : Filed 25 Nov
    1993 : Inventors assign to QC Optics Inc
    1995 : Patent issued 14 Feb
         : Certificate of Correction
    1996 : Security interest to State Street Bank
    1998 : Security interest terminated
    2001 : KLA-Tencor acquires QC Optics
    2012 : Patent expires

NPE / troll-pattern signals

1. Shell-entity transfer — not present. No "IP / Patents / Licensing / Holdings / Ventures" entity appears in any of the three records. The only post-issuance filings are a bank security interest (Reel 008669/0943) and its termination (Reel 009367/0908); neither is a transfer of ownership to a new holder.
2. Known asserter in the chain — not present. Assignees of record are QC Optics, Inc. and State Street Bank and Trust Company. Neither appears on any NPE list; State Street acted as a secured lender (and later as Paying Agent in the 2001 merger), not as an assertion vehicle.
3. Repeat correspondent across the chain — unclear. The correspondent block was not retrievable (see above). No factual basis exists in the material I reviewed for a recurrence finding, so I decline to score it.
4. Cascading transfers — not present. Three records span 1993–1998; no chained LLCs, no sub-24-month cascade, no shared registered-agent address.
5. Pre-litigation transfer — not present. No litigation is known against this patent (consistent with the earlier litigation section), and no assignment clusters within six months of any suit.
6. Bankruptcy fire-sale — not present. QC Optics exited via a solvent cash merger (KLA-Tencor, 2001, ~$3M), not a Chapter 7/11 sale. (Contrast the Kodak/Nortel/Polaroid pattern.)
7. Privateering — not present. The 2001 transaction moved the asset to a strategic operating acquirer in the same industry (KLA-Tencor), which kept QC Optics intact as a subsidiary. No evidence the patent was used as an assertion vehicle against KLA competitors.
8. Defensive aggregator — not present. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN.

Verdict

Defensive / non-asserting.

This is the non-assertion limb of the category, not the defensive-aggregator limb: the chain never left operating-company hands — Reel 006384/0361 vested title in QC Optics, Inc., and the only other records are a lender's security interest (Reel 008669/0943, 1996) and its release (Reel 009367/0908, 1998). Title economically passed to KLA-Tencor by the 2001 merger (a strategic operating-company acquisition, unrecorded at USPTO), and with no litigation identified and expiration on 2012-11-25, there is no NPE footprint anywhere in the chain. Caveat: because correspondents were not retrievable, Signal 3 is unresolved; if a direct Assignment Center pull shows a single attorney filing all three records, that would not change this verdict (there is still no ownership transfer to an asserting entity).

Verification: USPTO Assignment Center search — https://assignmentcenter.uspto.gov/ (search property number 5389794); mirrored at https://assignment.uspto.gov/patent/index.html.

Cross-reference note (no contradiction): the earlier litigation section correctly reported "no litigation" and listed QC Optics as assignee. It did not capture the 2001 KLA-Tencor acquisition, so the "current assignee = QC Optics Inc." reading should be understood as assignee of record only — beneficial ownership since 2001 has sat with KLA-Tencor/KLA Corporation.

I'll search the USPTO/patent databases for the cited prior art references of US 5,389,794, verifying each citation literally.

Let me search for the QC Optics family references and the remaining citations.

Prior-Art Analysis — U.S. Patent No. 5,389,794

1. Target patent (verified)

The examiner cited 10 U.S. patents against this application (the "Patent Citations" list). Those ten are analyzed below. Because all ten are U.S. patents, the relevant pre‑AIA anticipation provision is 35 U.S.C. § 102(b) (publication more than one year before the 1992‑11‑25 filing), § 102(a) (publication before the filing/invention date), or § 102(e) (U.S. patent effective as prior art as of its own filing date).

Caveat on completeness: I was able to retrieve and verify full‑text/abstract content for references 1–7 below. For references 8–10 (US 5,155,371; US 5,153,844; US 5,189,481) I was able to confirm the bibliographic data from the patent's own citation list and general knowledge, but I could not fully retrieve the specifications within this session. I flag each of those explicitly rather than overstate the mapping.

2. Summary table of the ten cited references

Note on reference 10: US 5,189,481 published after the 1992‑11‑25 filing date, so it is not § 102(a)/(b) art; it can only be § 102(e) art as of its own 1991‑07‑26 filing date (pre‑AIA § 102(e)).

3. Per‑reference analysis

Reference 1 — US 4,092,068 (Domtar Inc.), "Surface sensor" — filed 1976‑05‑05, published 1978‑05‑30 (§ 102(b))

Description. A light source illuminates a small area of a moving web; at least two angularly spaced detectors sense reflected/scattered light. Signals are processed to give (a) a topographic indication while suppressing reflectivity variation, or (b) a dirt indication. With the illumination beam perpendicular to the mean surface, equal light reaches the two detectors; if the local area is tilted, the detector outputs become unequal, the imbalance being "indicative of the slope of the surface relative to the light beam." Scanning a "hump," the two detectors produce mirror‑image curves (intensity decreasing then increasing past the neutral level) — i.e., a positive‑then‑negative slope signature. Dirt instead reduces total scattered light symmetrically.

Potential § 102 read‑on: This reference is directed to slope/topography sensing and explicitly distinguishes a raised feature (hump) from dirt. It is the closest art to the slope‑detection and mound‑signature concepts of the claims:

• Claim 2 and Claim 8's "mound = positive slope proximately followed by negative slope" concept — Domtar's hump trace (decrease, then increase past neutral) corresponds to this slope order.
• Claim 7's notion of positive‑ vs negative‑slope outputs from separate detector outputs — Domtar's pair of detectors give opposite‑sense signals on a slope.
• Element (c) of claim 1 ("means for detecting a local slope").

Gap: Domtar does not disclose element (b) of independent claims 1/13/16/18 — sensing scattered radiation separately in the near‑specular and far‑specular regions. It also does not state an express pit‑vs‑mound discriminator keyed to slope ordering. Because every independent claim requires the near/far‑specular limitation, US 4,092,068 does not, standing alone, anticipate any claim in full; its relevance is as § 103 art on the slope‑signature feature.

Reference 2 — US 4,395,122 (RCA Corp.), "Defect detection system" — filed 1981‑04‑29, published 1983‑07‑26 (§ 102(b))

Description. Detects defects (small protuberances/hillocks) on a specular, spirally grooved video‑disc surface that acts as a diffraction grating. A focused coherent beam illuminates a spot spanning several groove convolutions; diffracted order cones are collected by a cylindrical lens and focused onto a split photodetector, whose two photosensitive regions feed a differential amplifier. A feedback network compensates for surface attitude (warp) to avoid false defect signals.

Potential § 102 read‑on: Relevant to claim 6's architecture of first and second detectors, differential comparison, and deviation sensing, and to claim 7's separate "positive‑slope"/"negative‑slope" outputs (split detector + differential amplifier is the classic implementation).

Gap: It is a diffractive/grooved‑surface system (not a smooth‑surface near/far‑specular system); it detects defects generally, not pits vs mounds, and does not recite near/far‑specular separation. No independent claim is anticipated.

Reference 3 — US 4,412,743 (Discovision Associates), "Off‑axis light beam defect detector" — filed 1981‑09‑08, published 1983‑11‑01 (§ 102(b))

Description. A beam‑scanning apparatus for detecting a property of a light beam reflected off a surface (particularly optical/video discs), using an off‑axis incident beam relative to the objective‑lens axis, with a reflected‑beam sensing element; provides accurate identification of the location and size of defects. A related family member (EP 0 074 115) describes an oblique test beam and a grooved disc acting as a diffraction grating.

Potential § 102 read‑on: Relevant background to claims 3/5 (beam direction, and the alternative non‑normal incidence taught in the patent at col. re: angle θ) and to defect location/size determination (claims 11, 12, 14).

Gap: Grating/off‑axis optical‑disc defect detection only; no pit‑vs‑mound slope discrimination and no near/far‑specular sensing. No independent claim anticipated.

Reference 4 — US 4,794,264 (QC Optics, Inc.), "Surface defect detection and confirmation" — filed 1987‑05‑08, published 1988‑12‑27 (§ 102(b)) [family: EP 0 290 228]

Description. Directs a beam at the surface and separately senses radiation scattered into the near‑specular region (indicative of a pit) and the far‑specular region (indicative of a flaw); normalizes the near‑specular signal against the far‑specular signal; discriminates the near‑specular component; and declares a defect (not contamination) upon coincidence of the pit signal and the flaw signal. Includes laser source and rotation + translation of the surface (air spindle + encoder + servo).

Potential § 102 read‑on: This QC Optics patent (same assignee as the '794 patent) is the source of the "separately sensing near‑specular and far‑specular" limitation that appears verbatim in independent claims 1, 13, 16, and 18. It also discloses the scanning/relative‑motion elements (claims 3, 4) and location/mapping (claims 12, 14).

Gap: It discriminates pit vs "other flaw/contamination," not pit vs mound, and does not use slope‑ordering. Because the independent claims additionally require pit‑vs‑mound discrimination, '264 does not anticipate them; it is, however, § 102/§ 103‑critical art against the near/far‑specular limitations themselves (and it is the reason those limitations are not novel over the assignee's own prior work).

Reference 5 — US 4,794,265 (QC Optics, Inc.), "Surface pit detection system and method" — filed 1987‑05‑08, published 1988‑12‑27 (§ 102(b)) [family: EP 0 290 227]

Description. Irradiates a surface area; separately senses near‑specular radiation (pit) and far‑specular radiation (other flaw); normalizes the near‑specular signal with respect to the far‑specular signal; discriminates the near‑specular component to detect pits. Defines near‑specular as ~40–100 mrad and far‑specular as >100 mrad. Includes pit‑sizing (four classes) and pit location/mapping (radius + angle), plus laser, air spindle, encoder, servo translation.

Potential § 102 read‑on: The single most on‑point reference for the near/far‑specular pit‑sensing limitation of claims 1/13/16/18; also relevant to size (claim 11), location (claims 12/14). Note it even refers to "concave depression (or convex elevation)," i.e. it acknowledges convex features.

Gap: Its detection is optimized for pits as a class (and neighboring surface warping); it does not distinguish a pit from a mound by the order of slope changes, which is the core of the '794 claims. It therefore does not anticipate any independent claim, but it is the closest same‑assignee art for multiple claimed elements.

Reference 6 — US 4,920,385 (Diffracto Ltd.), "Panel surface flaw inspection" — filed 1984‑02‑14, published 1990‑04‑24 (§ 102(b)) [family: US 4,629,319 "DiffractoSight"]

Description. A "DiffractoSight" full‑field surface inspection technique. Light from the surface is imaged via a reference (retroreflective) screen to a bicell detector; a local slope distortion shifts the beam‑spot image on the detector (a "geometric local reference system" using the normal surface near the defect as reference). The specification expressly states that "Indents and outdents are also normally identifiable," and that the flaw can be characterized by the width and amplitude of the normalized signal.

Potential § 102 read‑on: This is the strongest conceptual prior art against the core discrimination idea — i.e., that a local slope/deflection detected while scanning reveals whether the feature is a depression (indent) or a protrusion (outdent). It maps to the pit‑vs‑mound differentiating means of claims 1/13/16/18/20 (element (d)) and to claim 2's "change in slope in one manner vs a different manner." Its normalized‑signal width/amplitude language also touches claims 9/10/11/15/19 (height/depth/transverse extent from signal amplitude and duration).

Gap: It uses an imaging/retroreflective‑screen geometry, not "separately sensing scattered radiation in the near‑specular and far‑specular regions"; nor does it recite the ordering of positive‑then‑negative slope for pit vs mound exactly. So it does not anticipate the independent claims as literally worded, but it is the most dangerous § 103 reference on the discrimination feature.

Reference 7 — US 4,943,734 (QC Optics, Inc.), "Inspection apparatus and method for detecting flaws on a diffractive surface" — filed 1989‑06‑30, published 1990‑07‑24 (§ 102(b))

Description. An elongate array of electro‑optical shutters is interposed between a scanning laser beam and a photodetector that collects light scattered along the scan line; the shutters are shifted in synchronism with the scan to block unwanted regular (periodic/pattern) signal components, enabling high‑sensitivity flaw detection on a diffractive surface.

Potential § 102 read‑on: Background on scanning + scatter collection on a patterned surface; relevant to the "means for scanning" environment (claim 3) but not to pit‑vs‑mound slope discrimination or near/far‑specular pit sensing.

Gap: Different mechanism and surface type; anticipates no independent claim.

Reference 8 — US 5,155,371 (Krupp Atlas Elektronik GmbH), "Apparatus for detecting deformations in pressed members" — filed 1989‑12‑18, published 1992‑10‑13 (§ 102(a))

Description (limited verification). An electro‑optical apparatus for detecting deformations (dents/bulges) in pressed/formed members; consistent with slope‑ or fringe‑based sensing of surface deformation. I could not retrieve the full specification in this session; the mapping below is therefore provisional.

Potential § 102 read‑on: Likely relevant as general § 102(a) art on optically detecting local surface deformations (dents vs bulges). Because it published only ~6 weeks before the '794 filing date, it is § 102(a) art only (it cannot be § 102(b), since it is not more than one year earlier).

Gap / confidence: I cannot represent that it discloses near/far‑specular sensing or pit‑vs‑mound slope ordering. Treat the read‑on to the independent claims as not established.

Reference 9 — US 5,153,844 (E. I. du Pont de Nemours), "Method and apparatus for measuring surface flatness" — filed 1990‑01‑23, published 1992‑10‑06 (§ 102(a))

Description (limited verification). Measures surface flatness (optical/deflectometric). Relevant background on slope/flatness metrology; § 102(a) art (published ~7 weeks before filing).

Gap / confidence: No indication of near/far‑specular pit sensing or pit‑vs‑mound discrimination. I could not verify the specification in this session, so no independent‑claim anticipation is asserted.

Reference 10 — US 5,189,481 (Tencor Instruments), "Particle detector for rough surfaces" — filed 1991‑07‑26, published 1993‑02‑23 (§ 102(e), effective as of 1991‑07‑26 filing date)

Description (limited verification). A detector for particles on rough surfaces using scattered‑light collection. Because its publication postdates the '794 filing, it is prior art only under pre‑AIA § 102(e), as of its 1991‑07‑26 filing date.

Potential § 102 read‑on: Relevant to scatter‑based particle/flaw detection generally; the "rough surface" context differs from the '794 smooth‑specular regime.

Gap / confidence: No pit‑vs‑mound discrimination shown; no independent‑claim anticipation asserted.

4. Overall § 102 assessment

• No single cited reference appears to anticipate any independent claim (1, 13, 16, 18, 20) in full. Every independent claim requires the combination of (i) scanning a beam over a surface, (ii) separately sensing scattered radiation in the near‑specular region and the far‑specular region, and (iii) differentiating pit vs mound by slope behavior. The citations generally supply (i) and (iii) or (ii), but not all three:
  • References 4 (US 4,794,264) and 5 (US 4,794,265) (both QC Optics) supply the near/far‑specular pit‑sensing limitation but not pit‑vs‑mound slope discrimination.
  • References 1 (Domtar) and 6 (Diffracto) supply slope‑based topography and, in Diffracto's case, express indent vs outdent characterization, but not the near/far‑specular limitation.
• Most relevant prior art, ranked:
  1. US 4,794,265 and US 4,794,264 (QC Optics) — directly supply the "separately sensing near‑specular / far‑specular" element appearing in claims 1, 13, 16, 18; same assignee, so these are the closest art to the claimed sensing architecture.
  2. US 4,920,385 (Diffracto) — closest art on the pit‑vs‑mound discrimination principle (indents vs outdents via local slope), i.e. the heart of the '794 contribution; strongest § 103 combination candidate with the QC Optics references.
  3. US 4,092,068 (Domtar) — closest art on the mound slope signature (positive‑then‑negative slope) and on separating a protrusion from dirt using multiple angularly‑spaced detectors.
  4. US 4,395,122 (RCA) — split‑detector + differential‑amplifier architecture for slope/defect sensing (claims 6/7 architecture).
  5. US 4,412,743 (Discovision) — off‑axis beam defect detection and defect sizing/location (claims 3/5, 11/12/14).
  6. US 4,943,734 (QC Optics) — scanning/scatter‑collection background.
     7–10. US 5,155,371; US 5,153,844; US 5,189,481 — peripheral background; § 102(a)/(e) status noted; no independent‑claim anticipation established on the information retrieved.

Practical conclusion: The examiner's citation set is dominated by the applicant's own earlier QC Optics near/far‑specular work (which reads on the pit‑sensing limitation) plus slope/deflection metrology art (which reads on the mound/discrimination limitation). The allowance of the '794 claims is consistent with there being no single reference that discloses the full combination, while the near/far‑specular and slope‑discrimination features were each individually known — a classic § 103 "combination of known elements" posture rather than a § 102 anticipation posture.

Items I could not verify (stated explicitly)

• Full specifications of US 5,155,371, US 5,153,844, US 5,189,481 were not retrievable in this session; their descriptions and § 102 read‑ons above are provisional and should be confirmed against the USPTO full‑text records before being relied upon.
• I did not independently re‑fetch the USPTO PatentCenter record for US 5,389,794 itself; the citation list, dates, and claim text were taken from the Google Patents full‑text rendering supplied as authoritative for this analysis, which matches the USPTO‑format data.

I'll ground this analysis in the actual prior-art references cited on the patent page. Let me pull details on the key references.

Let me pull details on the remaining cited references to complete the combination analysis.

Obviousness Analysis — U.S. Patent No. 5,389,794 (35 U.S.C. § 103)

Governing law/date: The application was filed 1992‑11‑25 and granted 1995‑02‑14, so pre‑AIA § 103(a) applies. KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) governs the motivation-to-combine inquiry. All prior art must predate 1992‑11‑25 (the filing/priority date).

One contradiction to flag with the earlier section: the previously generated litigation summary assumes "today" ≈ the date the patent page was fetched (2026‑09‑30), while this task sets "today" at April 26, 2026. This does not affect substance, but note that the source page's "Expired – Lifetime," anticipated expiration 2012‑11‑25 status is confirmed and is relevant here only as context (any § 103 challenge would be a retroactive validity challenge, e.g., an IPR‑era estoppel/§ 282 defense; IPRs cannot be filed against an expired patent's claims in practice, but district-court § 103 invalidity defenses remain available). The litigation summary found no litigation; nothing in this analysis depends on that.

Level of ordinary skill (my assumption, stated explicitly): a B.S. in physics, optics, or electrical engineering plus ~2–3 years' experience in scanning-laser surface inspection for magnetic media or semiconductor substrates.

1. What the claims actually cover

Important structural observation before the charts: the granted U.S. independent claims do NOT recite the integrator/polarity feature that the EPO treated as the point of novelty. Compare:

• U.S. claim 1 (source text): "means for scanning a beam of radiation over a surface; means for separately sensing radiation scattered from the surface in the near-specular region and in the far specular region for differentiating between pits and mounds and other types of defects; means for detecting a local slope on the surface from radiation scattered from the surface in the near-specular region; and means… for differentiating between whether said beam… is scanning a pit or a mound."
• EP 0 670 997 B1 claim 1 (family member, citable at https://patentimages.storage.googleapis.com/04/5b/ea/b0675cd7dc4d8d/EP0670997B1.pdf) adds the characterising feature: "an integrator arranged for integrating a signal representing variation of said local slope with time so as to produce a pulse of which the respective polarity distinguishes between a pit and a mound." Its ¶[0012] states that US‑A‑4920385 "discloses features corresponding to those in the pre-characterising parts of the independent claims."

That is a significant signal: the EPO read Diffracto '385 as disclosing — as pre-characterising (i.e., admitted/known) subject matter — a scanned near-specular beam, local-slope detection, and a means for distinguishing pit from mound. The U.S. claims are broader than the EP claims in that they omit the integrator. This materially raises the § 103 exposure of the U.S. independent claims.

Also note all independent claims are means-plus-function claims (§ 112 ¶ 6), so each "means for…" is limited to the structure disclosed (two semi‑annular masked detectors 150/152, pre‑amps 180/182, differential amp 184, integrator 186, comparators 196/198, computer 200 — FIG. 5) and equivalents.

2. The prior art of record (all cited on the face of the patent page)

3. Claim 1 — the core combination

Combination A (primary): US 4,794,265 (QC Optics) in view of US 4,920,385 (Diffracto)

Motivation to combine (KSR / MPEP 2143 rationales):

1. Same field, same problem, same physics. Both references are optical inspection of reflective engineered surfaces using a light beam and analysis of the reflected/scattered light to locate surface deviations. '265 addresses a magnetic disk substrate; '385 addresses reflective panels — the '794 specification itself says the problem "may pose a problem with respect to any smooth surface."
2. '265 supplies the missing "which channel" architecture; '385 supplies the missing "which sign" logic. The claimed combination is the union of two sets of known elements, each performing its own established function — the classic "familiar elements according to known methods" case (KSR, 550 U.S. at 415–421).
3. An express, articulated problem in the field supplies the design incentive. The '794 background states that the pre-existing QC Optics equipment "would automatically determine that a flaw exists generally, but the operator must still use a microscope to inspect… whether an individual flaw is a pit or a mound… Such inspection is time consuming, subject to error… unreliable." That articulated shortcoming of the assigned-art system is itself the motivation to add the known '385 slope-direction discriminator to '265's system.
4. The subject matter was "ready for improvement." '265 already reported a pit channel with size classification; adding a sign-of-slope discriminator from '385 to the same near-specular channel is a predictable, two‑element combination (KSR; MPEP 2143 (C), (D)).

Combination A′ (secondary/alternative for the same claim): '265 in view of US 5,155,371 (Krupp Atlas)

Even if '385 were distinguished as a full‑field imaging (retro‑reflective "D‑Sight") technique rather than a focused scanned-beam technique, Krupp '371 is a closer substitute for the "local slope" element: it uses a laser transmitter and a position‑sensitive photoreceiver where "if the light strikes above or below the center, the output signal is stronger or weaker," and where a dent produces "a dip or a peak" — i.e., a scanned beam whose detected deviation direction encodes local slope sign. Substituting '371's position‑sensitive receiver for '265's near‑specular sensor is a substitution of one known optical-detection element for another to obtain the predictable result of slope (±) information (MPEP 2143 (B)).

My assessment for claim 1: on this record I consider claim 1 more likely than not obvious over '265 + '385 (with '371 as a reinforcing secondary reference). The two references are both on the face of the patent; the combination requires no change in the principle of operation of either; and the EPO's own characterization of '385 as the admitted pre-characterising art is strong extrinsic evidence.

4. Claim-by-claim

Independent claim 13 — parallel to claim 1 (directing means + relative motion + separate near/far-specular sensing + near-specular sensor detecting deviation + means for distinguishing pit vs mound). Same combination as claim 1; '265 supplies the directing/relative-motion/separate-sensing elements, '385/'371 supply the deviation-based distinguishing means.

Independent claim 16 (method) — parallel method. Combinations A/A′ apply; '265 expressly discloses the method of irradiating, separately sensing near/far-specular scattered radiation and normalizing.

Independent claims 18 (method) / 19 — combination as above; claim 19's "analyzing… the time… for measuring the relative height/depth" is squarely taught by '481 ("the surface profile itself can be reconstructed through integration") and by '371's "amplitude detector… to obtain a measure of the depth of the deformation."

Independent claim 20 — system whose detecting means comprises "first and second detectors… each… masked for separately indicating a deviation in one direction." This is the bi‑cell/masked split-detector arrangement. '371's position-sensitive photoreceiver inherently outputs a signed deviation ("stronger or weaker" above vs below center) — a masked pair of detectors reading one direction each is the routine, predictable way to implement that function. ⚠ Dependent on the unverified detector references (US 4,092,068 / 4,395,122 / 4,412,743) for the specific masking detail.

Dep. 2 ("change in slope in one manner… different manner") — '385 slope-direction teaching. Obvious.

Dep. 3–4 (laser source + relative motion; rotate and translate) — '265 (air spindle 72 + servo drive 78); '481 (chuck simultaneously rotated and translated, spiral scan). Obvious.

Dep. 5 (beam normal to the surface) — '265 notes the beam is "shown as normal to surface 14, [though] this is not necessary"; '481 expressly covers normal-incidence variants. Obvious.

Dep. 6 (first and second detectors, each disposed and masked to indicate a deviation in one direction) — as for claim 20; '371 (signed position-sensitive output); plus the cited detector references. Likely obvious (subject to verifying the masking disclosures).

Dep. 7 (first output for positive slope, second for negative slope) — '371 ("stronger or weaker" depending on whether light strikes above or below center); '385 (slope direction). Likely obvious.

Dep. 8 ("a deviation indicative of a positive slope proximately followed by a deviation indicative of a negative slope" = mound) — this is the temporal ordering insight. It follows directly from the geometry of a bump traversed by a scanning beam (positive slope, apex, negative slope) and is the inherent behavior of any ±slope sensor crossing a mound, which a PHOSITA would immediately recognize. '371 describes a dent producing "a dip or a peak" in the time curve of the position-sensitive output. Obvious, but this is the claim family's strongest candidate for a non‑obviousness argument (see § 6).

Deps. 9–10 (measure relative height of mound / relative depth of pit) — '371 amplitude detector gives "a measure of the depth of the deformation"; '481 teaches that integrating the local-slope signal reconstructs the surface profile; '265's pit-size comparator classifies size. Obvious.

Dep. 11 (transverse extent) — time duration of the signal × known scan velocity; '265's multiple-pit-detection mapping of a single pit's radius span is the same idea. Obvious.

Dep. 12 (denote location) — '265 and '481 both report (r, θ) positions and produce defect maps. Obvious.

Dep. 14 (relative motion controls beam location to locate the defect) — '265/'481. Obvious.

Dep. 15 (monitor surface extent + signal strength → relative height/depth) — '265 (size comparator) + '371 (amplitude detector) / '481 (integration). Obvious.

Dep. 17 — parallel to claim 2. Obvious.

5. Summary of the combinations I would assert

6. Counter-arguments and where the patent still has room

I do not think this is a "slam-dunk" invalidity case. The strongest rebuttals:

1. Different principles of operation (MPEP 2143.01(VI)-(VII)). '265's near-specular channel classifies scattered intensity by angular region (near-specular toroid vs far-specular), and maps pits only. '385 is a full‑field retro‑reflective imaging technique in which slope information is encoded as gray-scale, not a focused scanned beam whose specular deflection is tracked. A PHOSITA might not see a reason to graft '385's image-based slope logic onto '265's spot-scanning near-specular channel. Similarly, '371 is a frequency-analysis system (it deliberately avoids "absolute measuring of the profile") aimed at pressed sheet metal, not at sub-micron pits/mounds on polished disk substrates — a PoP (point-of-purchase) argument that the references are non-analogous or at least not combinable without hindsight.
2. The '794's actual insight is the time-sequence of the sign change, not merely sign detection. Neither '265, '385, nor '371 states the rule "positive-then-negative = mound; negative-then-positive = pit" for a scanned spot on a magnetic disk. That "change-in-direction-of-deviation over time" recognition is the inventor's stated realization and is arguably the non-obvious contribution — but it is also arguably an inherent and predictable consequence of the geometry (claims 1/13 don't even require it; only claim 8 does).
3. '265's own text cuts both ways. The '264/'265 family describes near-specular detection as able to "uniquely detect and distinguish pits or minute warping (concave or convex)." If read as saying the near-specular channel already responds to concave and convex events without discriminating them, it supports the inventor's premise that discrimination was not previously achieved; if read the other way, it is close to teaching the result. Either reading should be tested against the actual '265/'264 specifications.
4. Secondary considerations — evidence gap. I found no evidence of record in this record of long-felt-but-unmet need (though the '794 background asserts the need), unexpected results, failure of others, or commercial success. Notably, the '794's own Examples report a mound detection limit "better than 0.06 micrometer" and a pit limit "around 0.09 micrometer" correlated against WYKO® interferometry — that is potential unexpected-results evidence, but the specification offers no comparison showing that the claimed combination produced results unexpectedly better than what '265 + a slope detector would predict. That evidentiary gap is what a § 103 challenger would exploit.
5. Means-plus-function scope. Because each independent claim is a § 112 ¶ 6 claim, the claim scope is limited to the disclosed structure (masked semi‑annular detectors + differential amp + integrator + comparators) or its equivalents. A challenger must show the prior art structure is the same as, or an equivalent of, that structure — not merely that the prior art performs a similar function.

7. What I could not verify (stated explicitly, per the operating rules)

• I could not retrieve the full text of US 4,092,068 (Domtar), US 4,395,122 (RCA), US 4,412,743 (Discovision), or US 4,943,734 (QC Optics) within my search budget (the searches returned the step-limit message). Any statement above attributing a specific detector/optics disclosure to those four rests on their titles and abstract-level metadata on the patent page, not on verified full text. Combinations D and E in particular must be re‑verified against those documents.
• I did not verify the actual USPTO prosecution history (Office actions/applicant remarks). In particular, I cannot say whether the U.S. examiner ever considered, or required, the integrator/polarity feature that appears in the EP claim — I only know the granted U.S. claims omit it. That is an inference from the granted claim text, not a prosecution‑history finding.
• Whether '385's disclosure reaches the specific "mound" (positive‑elevation) case for a magnetic-disk surface, or only for panels, is not fully established by the excerpt I obtained. The EPO's ¶[0012] characterization supports it, but that is a European, not U.S., determination.
• This is a technical/analyst assessment, not a legal conclusion of invalidity, and not a clearance opinion.

8. Bottom line

• Claim 1 (and its method/system parallels, 13/16/18) is, in my assessment, the most exposed — because, unlike its EP counterpart, it does not recite the integrator/polarity mechanism that both the specification and the European grant treat as the point of novelty. A two-reference combination of record (US 4,794,265 + US 4,920,385) covers every element, and the reference combination is supported by (i) same-field/same-problem reasoning, (ii) '385's express slope-direction/protrusion-vs-indentation teaching, (iii) the EPO's own treatment of '385 as the admitted pre-characterising art, and (iv) the '794 background's articulated motivation (eliminating subjective microscope review).
• Claims 9–11, 15, 19 (height/depth/extent) are substantially weakened by US 5,189,481, which expressly teaches that integrating a local-slope signal reconstructs the surface profile, coupled with '371's amplitude-detector depth measurement.
• Claim 8's "positive-then-negative = mound" temporal ordering is the claim element with the best non‑obviousness story (different from merely detecting slope sign), but it is also arguably an inherent consequence of scanning a protrusion — expect this to be the crux of any real dispute.
• Evidence needed to firm this up: (a) full texts of the four unverified cited references; (b) the U.S. prosecution history; (c) any secondary-considerations evidence (the WYKO‑correlated 0.06/0.09 µm detection limits are the starting point).