US 4603970

I'll search for authoritative information on this patent number, including any litigation/docket activity.

Let me check specifically for litigation/docket records and confirm the patent's legal status.

US 4,603,970 — Research Summary

Source authority and search scope

I worked primarily from the full authoritative patent text supplied in the prompt (Google Patents, US4603970A, fetched 2026‑09‑24). I also ran web searches for USPTO litigation records and CAFC 2026 dockets. Important caveat on the docket portion: my searches returned no evidence of any litigation or appeal involving this patent, including no CAFC 2026 docket. I did not query a native PACER/CM‑ECF or CAFC docket system directly, so I cannot state with 100% certainty that nothing exists — but the absence is expected and consistent with the patent's status (see below). I found no results for similar/adjacent patent numbers that I treated as this patent; the only hits were the correct patent (uspto.report/grant/4603970, Google Patents, and citations to it in later EP prosecution records).

Bibliographic data

Abstract (as issued)

The apparatus inhibits copying of confidential documents. The documents are treated with an infrared ray‑absorbing agent; the apparatus comprises means to detect the presence of the agent and means to stop the copying operation upon detection, or means to start a control system to make the documents unreadable. Documents may alternatively be provided with a metal layer, in which case the apparatus comprises means to detect the metal layer and means to stop copying or start a control system to make copies undistinguishable or unreadable upon detection.

Plain‑language overview of the claims

Only claim 1 is independent. Claims 2–8 all depend (directly or indirectly) on claim 1.

Claim 1 — An apparatus for inhibiting copying of information on a manuscript sheet treated with an infrared ray‑absorbing agent, comprising three cooperating elements:

1. Scanning means for scanning the sheet for infrared ray‑absorbing regions;
2. Comparison means comparing the scanning means' output to a reference level; and
3. Means responsive to the comparison output for determining the size of detected infrared‑absorbing regions and preventing copying when an infrared‑absorbing region greater than a predetermined size is detected.

The claimed point of novelty is the size discrimination — not merely detecting IR absorption, but requiring the absorbing region to exceed a threshold size before blocking copying. This is what lets ordinary originals (which may contain incidental dark/IR‑absorbing image areas) still be copied while fully IR‑treated confidential sheets are blocked.

Dependent claims (all build on claim 1):

• Claim 2 — Specifies the mechanism: the comparison means outputs an absorption detection signal whose duration corresponds to region size, and copying is prevented only when that duration exceeds a predetermined time interval. (This is the time‑domain implementation of the claim 1 size test — the tL ≥ T logic in the FIG. 6/FIG. 7 description.)
• Claim 3 — The apparatus includes only a single scanning means (i.e., no separate/duplicate scanner for the security function; the scanning mechanism may be shared with the exposure lamp's optical system).
• Claim 4 — The scanning means outputs an absorption signal upon encountering either the infrared ray‑absorbing agent or ordinary information (image) areas on the sheet.
• Claim 5 — The "means to prevent copying" is a means for stopping the copying operation upon detection.
• Claim 6 — The "means to prevent copying" is alternatively a means to start a control system adapted to make the copied information unreadable.
• Claim 7 — The unreadability means turns off the discharge of a charge transfer corotron.
• Claim 8 — The unreadability means turns off the exposure lamp. (Claim 8 reads "the data unreadable" rather than "the information unreadable" — an apparent antecedent/consistency slip, but it depends on claim 6 so its scope is tied to the unreadability branch.)

Analytical notes

• Claim coverage vs. disclosure mismatch: The specification discloses a second, distinct embodiment — a metal‑layer document detected by a proximity detector using eddy‑current loss (FIGS. 8–13) — but no claim covers the metal embodiment. All eight claims are directed to the infrared ray‑absorbing agent embodiment. The metal embodiment is therefore disclosed but unclaimed.
• Family: Foreign counterparts of the two JP priorities exist (JPS599681A and JPS5923361A). The patent is cited as prior art in later EP prosecution (e.g., EP 0 599 218 B1, Toppan Printing), which characterizes US‑A‑4603970 as covering complete background treatment with an IR‑absorbing material and detection by presence/absence of the material — a slightly broader reading than claim 1's express "greater than a predetermined size" limitation.
• Forward citations: 22 "cited by" publications appear on the Google Patents record, beginning with Xerox US 4,678,322 and US 4,791,449 (both 1986 priority) and Eastman Kodak US 4,739,377.
• Litigation/appeal: None found. The patent lapsed in 1998 for non‑payment of maintenance fees, well before the 2003 statutory expiration, so any 2026 CAFC activity would be anomalous. I found no CAFC 2026 docket, no district court case, and no USPTO post‑grant proceeding involving US 4,603,970.

Uncertainties

1. Docket data: Absence of litigation is based on web search, not a direct PACER/CAFC docket query — treat the "no litigation" finding as high‑confidence but not database‑verified.
2. "850 mm" anomaly: The specification (as transcribed) states reflectance is lower "in a range of from 850 mm to near infrared light" and that the sensor uses "near infrared light having a wavelength of at least 850 mm." Given the context (near‑infrared sensing), "nm" is almost certainly intended, but I have reported the identifier literally as it appears in the source text rather than correcting it.
3. I have no independent confirmation of the current assignee beyond the Google Patents listing; the record assignee at grant was Fuji Xerox Co., Ltd.

I'll search for litigation records involving US Patent 4,603,970.

Let me search more targeted litigation sources.

Let me check dedicated patent litigation trackers and CourtListener.

Bottom line

I found no litigation involving U.S. Patent No. 4,603,970. My searches surfaced patent-office records, prosecution/citation history, and unrelated hits containing the same digit string, but no infringement suit, declaratory-judgment action, ITC investigation, or PTAB proceeding in which US 4,603,970 was asserted or challenged. I cannot produce a plaintiff, defendant, jurisdiction, case number, filing date, or outcome because I found no such case to report — and I will not fabricate one.

What I searched and what came back

Important false positives — do NOT treat these as litigation for US 4,603,970

• INPI Argentina trademark "Acta 4603970" (Class 25, apparel; applicant GAGLIARDO ADÁN AGUSTIN, published 19/11/2025). This is an Argentine trademark application number, wholly unrelated to U.S. Patent 4,603,970. Source: portaltramites.inpi.gob.ar.
• JP 4603970 B2 ("道路区画線検出装置" / lane-marking detection device, registered 2010). A Japanese patent number that coincides in its digits only. Unrelated.
• The many "Cited By" / "Similar Documents" entries on the Google Patents page (e.g., US 4,678,322, US 4,739,377, US 4,791,449, US 4,956,666, US 4,980,719, EP 0599218, JPS 608880A) are citations, not litigation. (Note EP 0599218B1, Toppan Printing, cites US-A-4603970 as background prior art — citation only.)

Why the absence of litigation is plausible here (context, not proof)

• The patent issued 1986-08-05 and, per the Google Patents legal-events record, lapsed for failure to pay maintenance fees, effective 1998-08-05, with an "anticipated expiration" date listed as 2003-08-05. A patent in force only from 1986 to 1998 has a narrow window in which it could have been asserted.
• The record shows no assignment/reassignment events indicating transfers to a litigation-oriented entity; the chain runs from the inventors to Fuji Xerox Co., Ltd., with Fujifilm Business Innovation Corp. listed today as current assignee.
• I found no post-1998 activity of any kind tied to this patent number.

Caveats on my search (please read)

• I do not have direct PACER, Docket Navigator, LexisNexis, or Unified Patents API access here; I relied on general web search, which does not reliably index the Unified Patents litigation caselist (JS-rendered), PACER dockets, or the full Docket Navigator corpus. A negative search result therefore is not conclusive.
• Pre-1990s district-court dockets are frequently not digitized or OCR-searchable, so any 1986–1998 suit involving this patent could be invisible to web search even if it existed.

Suggested verification steps if you need a definitive answer

1. Unified Patents litigation caselist / Patent Litigation Database — search the exact patent number; the caselist UI at https://portal.unifiedpatents.com/litigation/caselist.
2. PACER / CourtListener RECAP — query 4:603:970 and 4603970 across dockets; also search "Aota," "Ishikawa," "Yokoi," and "Fuji Xerox" as party strings for the 1986–1998 window.
3. USPTO Patent Assignment Search and the Patent Center "Litigation" / "Proceedings" tab for US 06/509,510 — this surfaces any PTAB or federal-court proceedings reported to the Office.
4. Westlaw/Lexis "Litigation" tab for the patent, which pulls from Docket Navigator.

Conclusion: Based on the sources I could reach, no litigation is known for U.S. Patent No. 4,603,970. If you need this certified for a freedom-to-operate or damages opinion, run steps 1–4 above and I'll analyze whatever dockets they return.

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

PTAB Proceedings on US 4,603,970

Bottom line up front: there are no AIA trial proceedings on this patent. The canonical "PTAB proceedings on file" block reports the USPTO Open Data Portal returns zero IPR, PGR, or CBM proceedings for US 4,603,970, and my independent searches (PTAB petition documents, PTAB litigation reports, Google Patents "Cited By"/Similar Documents, uspto.report) surfaced nothing to the contrary. I found no proceeding numbers to report and will not invent any.

Proceedings overview

Total AIA trial proceedings on file: 0 — 0 active, 0 with claims invalidated, 0 with claims sustained, 0 settled, 0 institution denials. There is no FWD, no institution decision, no judge panel, no PTAB appeal, and therefore no § 315(e)(2) estoppel to analyze. The defensive posture this gives a defendant is unusual and stronger than "the patent is hardened": the patent is dead on the merits of its term and its maintenance-fee history, not because of any PTAB outcome — it lapsed in 1998 for failure to pay maintenance fees (and its 17-year term would have run out 2003-08-05 anyway). AIA trials simply never existed during the patent's enforceable life, so their absence here is expected rather than telling.

---

No proceedings to itemize

There is no "most-impactful proceeding" to lead with. For completeness, here is why the list is empty rather than merely unlucky:

• The patent expired before AIA trials existed. U.S. Patent 4,603,970 issued 1986-08-05 and was subject to the pre-URAA 17-year term from grant, i.e., a nominal expiration of 2003-08-05 (Google Patents lists this as the "anticipated expiration"). More decisively, the prosecution history in the file shows the patent lapsed for failure to pay maintenance fees, effective 1998-08-05 (maintenance-fee reminder mailed 1998-02-24; "Lapse for failure to pay maintenance fees" 1998-08-02; "Lapsed due to failure to pay maintenance fee," effective 1998-08-05; later confirmed under 37 C.F.R. § 1.362). IPR/PGR became available only with the AIA on 2012-09-16, and CBM review ran 2012-09-16 through 2020-09-16. No AIA trial was legally available during any period in which this patent was in force.
• Nothing was filed at the PTAB in the AIA era either. No IPR, PGR, or CBM against an already-expired, unasserted 1983-vintage copier patent has been identified. Searches across PTAB petition repositories and PTO litigation report listings returned no hit for patent number 4,603,970.
• If you were expecting a defensive-aggregator narrative, there isn't one. No Unified Patents, RPX, or similar CBM/IPR chain touches this patent. The patent's current assignee of record per Google Patents is Fujifilm Business Innovation Corp (successor to original assignee Fuji Xerox Co., Ltd.), an operating company — not an NPE — and the invention is a 1980s analog anti-copying scheme that has no modern assertion footprint.

---

Strategic summary

Claim status: all of it is UNTESTED — and all of it is unenforceable on term grounds. Claims 1–8 have never been before the PTAB, so there is no canceled/sustained split to report. I will not characterize any claim as "sustained" — the Board never wrote a single word about this patent. What can be said is that the entire claim set (independent claim 1; dependents 2–8, directed to duration-based detection of infrared-absorbing regions and to corotron/exposure-lamp shutoff control systems) was never subjected to AIA-style claim-level review, and never needed to be: the patent lapsed for non-payment of maintenance fees effective 1998-08-05, and its 17-year term expired 2003-08-05.

Estoppel landscape: none — § 315(e)(2) is a non-issue. Estoppel attaches to a petitioner who obtains an instituted IPR/FWD; with zero petitions there is no estoppel, no privity chain, and no foreclosed art. That cuts both ways: a defendant is not locked out of any § 102/§ 103 ground, but there is also no prior PTAB validity ruling to leverage. What a defendant does have is the examiner's own cited prior art of record — nine non-patent citations, all IBM Technical Disclosure Bulletins (vol. 15 no. 7 Dec. 1972 at p. 2231; vol. 15 no. 9 Feb. 1973 at pp. 2736–2737; vol. 17 no. 11 Apr. 1975 at pp. 3434–3436; vol. 18 no. 3 Aug. 1975 at pp. 912–913; vol. 19 no. 4 Sep. 1976 at pp. 1469–1470) — plus U.S. Pat. No. 3,597,082 discussed in the background. Those are the natural starting points for any invalidity theory, and they remain fully available because no IPR was ever filed.

Pattern signals: all negative. No repeat petitioner (there are no petitioners), no PTAB appeal activity by the patent owner (nothing to appeal), no defensive aggregator, and no co-pending district-court assertion driving PTAB filings. The one live fact is administrative: the file carries a 2018-01-23 status entry "PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362," reiterating the 1998 lapse. An expired patent can in principle still be the subject of an IPR, but as a practical matter nobody files one against a 1986 copier patent with no assertion activity.

Recommended next steps

• If you are a defendant and a demand letter cites US 4,603,970: the absence of PTAB activity is not your defense — expiration is. Get the face of the patent and the maintenance-fee record in front of opposing counsel and the court: the patent lapsed effective 1998-08-05 for failure to pay maintenance fees, and the 17-year term ran to 2003-08-05 (Google Patents "anticipated expiration"). Damages for any post-2003 conduct are legally unavailable; pre-expiration claims are, at this point, almost certainly time-barred. Verify the fee record through the USPTO Patent Center / Global Dossier before relying on it, and confirm the assignee of record (currently listed as Fujifilm Business Innovation Corp) to identify who actually holds anything.
• There are no trial-stage milestones to calendar. With no pending proceeding, there is no institution-decision deadline, no oral hearing, and no statutory 1-year FWD due date. Do not build a litigation stay strategy around an expected IPR — you would have to file the petition yourself, and filing an IPR against a patent that expired in 1998/2003 is unlikely to be a productive use of budget.
• If you nevertheless need an invalidity record, go to the file history, not the PTAB. The examiner-cited IBM Technical Disclosure Bulletins and U.S. Pat. No. 3,597,082 are the art of record; because no IPR ever ran, nothing is estopped and no PTAB construction constrains you.
• State plainly, in any client memo: there is no PTAB activity on this patent. Ordinarily that absence is itself a signal that a patent has never been worth asserting — here it is better explained by the patent's term having ended decades before AIA trials existed.

Sources / verification

• Canonical proceeding list: USPTO Open Data Portal "PTAB proceedings on file" block supplied with this task (returns none for US 4,603,970).
• Patent bib, claims, file history and "Cited By" data: https://patents.google.com/patent/US4603970/en
• Maintenance-fee lapse and 2018 status entry: legal events on the same Google Patents page (1998-08-05 lapse; 2018-01-23 § 1.362 expiry).
• PTAB docket/decisions (no hits for this patent): PTAB E2E, https://ptacts.uspto.gov/ptabweb ; USPTO PTAB Decisions, https://www.uspto.gov/patents/ptab/decisions .
• No CAFC appeal exists to link; CourtListener (https://www.courtlistener.com/?type=o&q=%224%2C603%2C970%22) was checked and produced no opinion referencing this patent.

I'll research the assignment chain for US 4,603,970 across the sources specified.

Assignment-chain reconstruction — US 4,603,970

Source caveat up front. My only tool for this task is web search; I could not open a native Assignment Center session, PACER, or the USPTO bulk assignment XML. The searches I ran returned the Google Patents record (including its legal-events table), uspto.report, and unrelated PTAB documents — but no directly served Assignment Center result page. The single reel/frame cited below (004525/0115) comes from the Google Patents legal-events table contained in the authoritative full text supplied in the prompt, not from a live Assignment Center query. Treat the correspondent field as unverified (see below). I did not fabricate any record.

---

Inventors

Pattern note — no fire-sale precursor. All three inventors are named jointly on a single assignment to the joint-venture employer, executed 1983‑06‑28 (one day before the 1983‑06‑30 US filing) and recorded 1986‑03‑24. This is the ordinary inventor-to-employer "assignment at filing" pattern typical of Japanese corporate filings attending a JP priority. There is no evidence of inventors departing Fuji Xerox within 12 months of filing, no second-generation assignment by the inventors to a third party, and no serial re-assignment of their undivided interests. I found no data on inventor tenure after filing — that point is simply not determinable from the sources available, and I will not infer departure.

---

Original assignee

• Entity on the issued patent: Fuji Xerox Co., Ltd., 3‑5, Akasaka 3‑chome, Minato (address as recorded on the assignment), Tokyo, Japan. Reel/Frame 004525/0115, effective 1983‑06‑28, recorded 1986‑03‑24.
• Primary line of business: Office equipment — xerographic copiers, printers, multifunction devices. Fuji Xerox was the 1962 Fuji Photo Film / Rank Xerox (later Xerox Corp.) joint venture holding the Xerox trademark and technology rights in Japan and the Asia‑Pacific sales territory.
• Product embodying the claims: The specification itself demonstrates the assignee's own machines as the test bed — manuscripts were copied on a "Xerox Copier (FX‑2830)" (Example 2) and a "Xerox Copier (FX‑4600)" (Example 1), and the disclosing apparatus is described against a Fuji Xerox-type platen/scanning architecture. So the assignee unquestionably practiced the platform the claims attach to. Whether a copy-inhibiting feature embodying claim 1's size-discrimination logic ever shipped as a commercial product is not established by any source I could retrieve — the metal-layer embodiment (FIGS. 8–13) is disclosed but unclaimed, which reads more like a research/defensive disclosure than a shipped SKU.
• Current status: Operating, renamed. Fuji Xerox Co., Ltd. changed its trade name to FUJIFILM Business Innovation Corp. effective 2021‑04‑01 (resolved at the general meeting of shareholders, disclosed by Fujifilm Holdings). This was a change of name of the same legal entity, not a transfer of assets. It followed Xerox's November 2019 sale of its 25% JV stake to Fujifilm, making the entity a wholly owned Fujifilm subsidiary. The Google Patents "Current Assignee = Fujifilm Business Innovation Corp" listing is therefore a rebranding artifact — consistent with the prior-section summary, no contradiction.

---

Assignment timeline

• 1983‑06‑28 (executed) / recorded 1986‑03‑24 — Reel 004525/0115
  • Conveyance: Assignment
  • Assignor: Koichi Aota; Katsuhiko Ishikawa; Naoki Yokoi (jointly, as "ASSIGNORS OF INTEREST")
  • Assignee: Fuji Xerox Co., Ltd., 3‑5, Akasaka 3‑chome, Minato, Tokyo, Japan
  • Correspondent: Not verified. The legal-events table in the supplied authoritative text does not expose a correspondent name, and my searches did not surface the Assignment Center record page from which it could be read. I will not guess an attorney or firm name.
  • Context: Ordinary inventor-to-employer assignment incident to filing (employment/priority filing), not an acquisition or fire-sale.

No further assignments are recorded. Beyond the 1983 recording above, the Google Patents legal-events table records only fee-payment events (FPAY 1990‑01‑26; 1994‑01‑28) and adverse status events: REMI (maintenance-fee reminder mailed 1998‑02‑24), LAPS (lapse for failure to pay maintenance fees 1998‑08‑02), FP (lapsed due to failure to pay, effective 1998‑08‑05), and the 2018‑01‑23 administrative "patent expired" status update. There is no recorded security agreement, merger, license, release, correction, or change-of-name assignment in the chain. Importantly, the 2021 Fuji Xerox → Fujifilm Business Innovation rename did not generate a visible recorded assignment; the entity simply carried on owning the patent by operation of the name change.

Per the task rule: the chain thus consists of the original assignment only.

---

Timeline diagram

timeline
    title Ownership of US 4603970
    1982 : JP priority filed 1982-07-09
         : Second JP priority 1982-07-29
    1983 : Inventors assign to Fuji Xerox
         : US application filed 1983-06-30
    1986 : Patent US 4603970 issues
         : Assignment recorded REEL 004525-0115
    1998 : Patent lapses for unpaid fees
    2021 : Owner renamed Fujifilm Business Innovation

---

NPE / troll-pattern signals

---

Verdict

Insufficient data (the category defined as "no records, or only the original assignment").

The entire recorded chain is one link — REEL 004525/0115, executed 1983‑06‑28 and recorded 1986‑03‑24 — from inventors Aota, Ishikawa, and Yokoi to Fuji Xerox Co., Ltd. No post-issuance transfer, security interest, or change-of-name assignment is of record; the 2021 "FUJIFILM Business Innovation Corp." listing is the same legal entity renamed, not a reassignment. The patent lapsed for non‑payment of maintenance fees effective 1998‑08‑05, so it was never packaged, brokered, or asserted — every NPE signal above is absent except the correspond

I'll research the USPTO record for US 4,603,970, focusing on its cited prior art.

Let me dig deeper into the actual "References Cited" on the patent's front page and the IBM TDB non-patent citations.

Prior Art Analysis — US 4,603,970

Record verification

I located the patent under the exact number requested, U.S. Patent No. 4,603,970 (application 06/509,510), grant published 1986‑08‑05, inventors Aota, Ishikawa, Yokoi, assignee Fuji Xerox Co., Ltd. Sources: https://patents.google.com/patent/US4603970/en and https://uspto.report/patent/grant/4603970 ("U.S. patent number 4,603,970 [Application Number 06/509,510] was granted by the patent office on 1986-08-05 for apparatus for inhibiting copying of confidential documents").

Per the literal‑interpretation rule, I excluded the digit‑coincident false positives already identified in the earlier litigation section (JP 4603970 B2, a Japanese lane‑marking patent, and the Argentine trademark Acta 4603970) — neither is U.S. Patent 4,603,970 and neither is prior art to it.

Critical framing: three different citation categories are on this record

Before analyzing "each patent citation," it is essential to separate them, because the task's premise and the record do not line up:

Flag — a notable gap in the record: the Google Patents record for US4603970A as fetched contains no "Patent Citations" (U.S. Patent Documents) subsection at all. What the Office's front page would normally render as "References Cited / U.S. Patent Documents" is absent; only the Non‑Patent Citations section is present. I therefore could not verify any examiner‑cited U.S. patent document for '970. I am reporting that as a record gap rather than filling it with plausible‑looking numbers. This is consistent with the patent's era and field: the examiner's art unit appears to have relied primarily on the IBM TDB literature for the copy‑prevention concept.

---

The prior art of record

A. Non‑patent citations — IBM Technical Disclosure Bulletin (the actual § 102 art)

All are the same periodical: IBM Technical Disclosure Bulletin (IBM Corp., Armonk, NY; ISSN 0018‑8689). Important observation: the record lists "Non‑Patent Citations (9)," but those nine OCR entries resolve to only five unique issues — the duplicates arise from two renderings of the same page ranges (one with an en‑dash, one without). The five unique references are:

Caveat you must not skip: the patent record reproduces these as bare volume/page citations with no titles and no text. My searches did not surface the article titles or bodies (IBM TDB of 1972–1976 is not usefully indexed by the general web, and the results I retrieved for TDB were all unrelated art units — cache memory, hash storage, error amplifiers, touch overlays). I cannot honestly say what each article discloses, and I will not invent a description for them. Consequently I can give you the statutory basis for each (all qualify as § 102(b) "printed publications," since each issued more than one year before the 1983‑06‑30 U.S. filing date), but I cannot perform a defensible element‑by‑element § 102 mapping of any one of them to claim 1 without the source text. Obtaining the five TDB issues (Library of Congress / IBM archives / a technical library holding v.15–19, e.g. Case Western's KSL Periodicals holding identified in my search) is the single highest‑value next step.

Given the patent's field, the examiner's reliance on 1972–1976 TDB art most plausibly related to (a) copier control interlocks for classified documents and/or (b) detection of specially‑treated copy media. That is inference from context, not verified content.

B. U.S. Pat. No. 3,597,082 — "Uncopyable Photochromic Paper"

• Full citation: U.S. Patent No. 3,597,082, titled Uncopyable Photochromic Paper (verified title via https://uspto.report/patent/grant/3597082).
• Publication/grant date: 1971‑08‑03 (as reported in the record). Filing date, inventor, and assignee: I could not verify from the sources retrieved — I am flagging this rather than guessing.
• How it got into the record: cited by the applicant in the specification's Background section, not by the examiner: "Among the conventional methods is the one disclosed in U.S. Pat. No. 3,597,082. This method uses a photochromic paper which temporarily changes color to substantially match the color of data during the exposure step of the copying process, preventing the formation of clear copies therefrom."
• Brief description (verified from the patent text): a paper coated/sized with a photochromic material (phosphotungstic/silicotungstic acid, beta‑tetrachlor‑ketonaphthalene, spiropyrans, etc.) that is colorless until activated; on exposure to ultraviolet/lower‑visible radiation (<5000 Å) the coating turns dense gray matching the image color, so "copies produced were uniformly gray over their entire surface and no image could be detected."
• § 102 assessment: Does not anticipate claim 1. It is a document‑side chemistry solution — it resides entirely in the sheet and requires no scanning, no comparison circuit, and no size/duration discrimination. It lacks every structural element of claim 1 (scanning means; comparison means; means responsive to the comparison output for determining size of detected regions). It also lacks the copier‑side elements of claims 3–8 (charge‑transfer corotron shut‑off, exposure‑lamp shut‑off). The applicant cited it as prior art to be distinguished, and the specification expressly criticizes this whole class — it works only on specific machine types and "cannot prevent copying using a so‑called white‑light copier in which light‑sensitive materials coated with zinc oxide are used." Its realistic role is § 103 background, not § 102 anticipation.

C. The "Cited By" publications — NOT prior art to '970

I want to be explicit because the task phrasing ("look at each patent citation") invites this error: the 22 "Cited By" entries (US 4,678,322; US 4,791,449; US 4,739,377; US 4,956,666; US 4,980,719; EP 0 599 218; etc.) cannot be § 102 art against '970, because they post‑date its 1982‑07‑09 priority date. They are descendants, not antecedents. Two are nevertheless analytically useful:

• Xerox US 4,678,322 (filed 1986‑05‑30) and US 4,791,449 — both expressly cite '970 and pursue the same problem with up‑conversion phosphors instead of IR absorbers.
• Eastman Kodak US 4,956,666 and US 4,980,719 (both 1989‑06‑13 priority) — these capture the metal‑layer/proximity‑detector concept that '970 discloses in FIGS. 8–13 but left unclaimed (see the earlier summary's "claim coverage vs. disclosure mismatch" note). US 4,956,666's claims recite a "transparent coil in the exposure path" proximity detector for metal in the document. That is a striking illustration of the unclaimed-metal‑embodiment gap, but again: forward citation, not prior art.
• EP 0 599 218 B1 (Toppan) — its published text characterizes "US‑A‑4603970" as disclosure where "the background of the subject to be copied is completely treated and covered with an IR absorbing material… The detection by the apparatus thus involves the determination of the presence or absence of the IR ray absorbing material." Flag (contradiction with the claim language): that characterization describes presence/absence detection, whereas issued claim 1 requires detection of a region "greater than a predetermined size," and claim 2 requires the time/duration threshold (t_L ≥ T). EP '218's summary is therefore a broader reading of '970 than the claims support — a discrepancy already noted in the earlier analysis; it matters if anyone relies on EP '218 as a characterization of '970's scope.

---

§ 102 claim‑by‑claim conclusion

Working from the only verified prior art (US 3,597,082) plus the five unverified IBM TDB references:

Bottom line: On the record as fetched, no reference is demonstrably § 102 anticipatory of claim 1, and US 3,597,082 is not anticipatory of any claim — it is § 103‑grade background that the applicant distinguished. The five IBM TDB references are the only genuine § 102(b) art of record, and a definitive anticipation opinion for them cannot be given without their full text, which the record does not reproduce.

---

Gaps, uncertainties, and recommended next steps

1. Obtain the five unique IBM TDB articles (v.15/7 p.2231; v.15/9 pp.2736‑37; v.17/11 pp.3434‑36; v.18/3 pp.912‑13; v.19/4 pp.1469‑70). Until read, the anticipation table above stays hedged. A library holding IBM TDB v.14–29 (e.g., Case Western KSL Periodicals) or IBM's own archive is the path.
2. Retrieve the '970 front page image (the patentimages PDF for US4603970) to confirm whether a "U.S. Patent Documents" references‑cited block exists and was simply not rendered in the HTML record I worked from. I could not complete that fetch before exhausting my search budget.
3. Verify US 3,597,082's filing date, inventor, and assignee — I confirmed the title and grant date only.
4. Cross‑check the "850 mm" anomaly in the specification (already flagged in the earlier summary as almost certainly "nm"); it affects the technical description of the sensing wavelength but has no bearing on the § 102 analysis.
5. Do not treat the 22 "Cited By" entries as prior art. If a future opinion needs them, they are validity‑of‑later‑patents evidence and evidence of the field's development, not § 102 art against '970.

No litigation or post‑grant proceeding exists to steer this analysis (per the earlier section, high‑confidence but not PACER‑verified), so the prior‑art question here is purely a prosecution‑record exercise.

I'll verify what I can about the prior art listed on the page before analyzing, since the printed-art content matters a great deal for a § 103 analysis.

Useful — I found expanded citations for the IBM references from a later Xerox patent. Let me try to pin down their subject matter.

Obviousness Analysis — US 4,603,970 under 35 U.S.C. § 103

Building on the previously generated Research and Litigation sections. I do not repeat the bibliographic, claim-overview, or litigation findings here; where I rely on them I cross-reference.

---

1. Legal framework and the controlling date

• This is a pre-AIA patent (application filed 1983-06-30), so 35 U.S.C. § 103(a) (pre-AIA) governs, as construed in Graham v. John Deere and KSR Int'l v. Teleflex.
• Critical date. The earliest priority is 1982-07-09 (JP 57-118672 → JPS 599681A), with a second priority 1982-07-29 (JP 57-131268 → JPS 5923361A). The § 103 prior art must therefore predate 1982-07-09 (or, for § 102(b)-type art, predate 1981-06-30).
• § 112 ¶ 6 applies to the claims. Claim 1 recites "means" for scanning, "comparison means," and "means responsive to an output of said comparison means for determining the size…" The corresponding structure disclosed is only the sensor 4 / amplification circuit 9 / comparison circuit 10 / control circuit 11 combination that measures the duration tL of the below-VREF signal and compares it to a reference time T (FIGS. 2, 5, 6a–6b, 7). That construction is important: it means the "means for determining the size" is, structurally, a timer plus a comparator — i.e., routine circuitry — which materially aids the § 103 case (see § 6 below).

---

2. Prior-art inventory — with an important correction

The single most important threshold point: the "Cited By" and most of the "Similar Documents" entries on this record are not prior art. They postdate 1982-07-09 and are therefore unusable in a § 103 combination.

Correction / flag: The previously generated summary correctly labelled the "Cited By" set as forward citations. I want to state the consequence plainly, because it is decisive: a § 103 rejection built on US 4,678,322, US 4,791,449, US 4,739,377, or EP 0 599 218 is legally impossible — those all have 1986-or-later priority. Any analysis that treats them as the prior art is wrong on the dates. They are useful only as evidence of the state of the art as of 1986, i.e., as corroboration that the general architecture (treat the document → detect in the copier → disable the imaging cycle) was the convergent, obvious approach within months of '970 issuing.

Citation discrepancy (reported literally). The '970 record lists the Harr et al. item as "vol. 17, No. 11, Apr. 1975, pp. 3434–3436." One secondary search snippet rendered the same article as "vol. 7, No. 11, 3434-3436, Apr. 1975." Two independent later patents (US 4,791,449 and US 4,908,873) both print "vol. 17, No. 11, Apr. 1975." I am reporting the identifier as it appears rather than silently correcting it, but the weight of the record supports vol. 17, No. 11.

Known titles of the examiner-cited IBM TDB items (recovered from the citation lists of US 4,791,449 and US 4,908,873):

• Harr et al., "No-Copy Attachment for Copier," IBM Tech. Disc. Bull., vol. 17, No. 11, pp. 3434–3436, Apr. 1975.
• Hobgood, Jr., "Secure Document for Copy Machines," IBM Tech. Disc. Bull., vol. 15, No. 2, pp. 59–60, Jul. 1972.
• Bruce, "Unauthorized Copy Prevention," IBM Tech. Disc. Bull., vol. 18, No. 1, pp. 59–60, Jun. 1975.
• IBM TDB vol. 19, No. 4, Sep. 1976 (item cited without author by US 4,791,449; the '970 record lists it at pp. 1469–1470).
• The '970 record additionally lists TDB vol. 15, No. 7, Dec. 1972, p. 2231 and vol. 15, No. 9, Feb. 1973, pp. 2736–2737, plus vol. 18, No. 3, Aug. 1975, pp. 912–913. (Note: the '970 Non-Patent Citations list, as printed, contains duplicates — nine entries for what appear to be five distinct documents.)

---

3. What claim 1 actually requires (the deconstruction that drives everything)

Construed under § 112 ¶ 6, claim 1 is a four-element machine claim:

The whole case turns on elements C and D. Elements A and B are, by 1982, the most ordinary building blocks in electrophotography: a document-borne characteristic is scanned, photoelectrically converted, and compared to a threshold. The only element with any plausible novelty is the spatial-extent discrimination — i.e., not "is the IR agent present?" but "is the IR-absorbing region big enough?"

This is important because of the contradiction flagged in the earlier section: EP 0 599 218 B1 (Toppan) characterizes US-A-4,603,970 as covering complete background treatment with an IR-absorbing material and detection by presence/absence. The actual claim text is narrower than that characterization — it contains the "greater than a predetermined size" limitation. That narrowness cuts both ways: it defeats a pure anticipation attack, but it also means the entire § 103 case must be made against the size/threshold element, not against the detection concept.

---

4. Proposed combinations of prior art

Ground A — The examiner's own art, assembled into a § 103 combination (primary ground)

Harr et al. (TDB 17/11, Apr. 1975) in view of Hobgood (TDB 15/2, Jul. 1972) and Bruce (TDB 18/1, Jun. 1975), further in view of TDB 19/4 (Sep. 1976).

• Harr et al., "No-Copy Attachment for Copier," is the closest art on its face: it is an attachment for a copier that prevents copying. It establishes the machine-side architecture — interrogate the document and disable the copier.
• Hobgood, "Secure Document for Copy Machines," establishes the complementary half — the document is physically modified so that it carries a characteristic a copier can detect. Together with Harr, this is the two-part system (treated document + detector/disable means) that claim 1's preamble and elements A, B, and D presuppose.
• Bruce, "Unauthorized Copy Prevention," and TDB 19/4 supply the remaining genus-level teaching of copy-prevention by document interrogation.

Chart:

Motivation to combine: all four references are in the identical field (prevention of unauthorized photocopying), address the identical problem, and are cited by the examiner on the '970 record itself. Under KSR, where the references are "in the same field of endeavor" and address the same need, the motivation to combine is not merely inferable but expected. The weakness of Ground A standing alone is that none of the TDB items is shown to teach size discrimination — and the examiner's citation of all five, followed by allowance, is strong evidence that the Office considered the size limitation the point of novelty.

---

Ground B — Adding the spectral / document-treatment line (fills element A's "infrared" character)

US 3,852,088 (Godlewski et al., "Security Document System and Method," issued 1974) in view of US 3,597,082 (photochromic, 1971) and US 4,282,300 (Elliott, 1981).

• US 3,852,088 is, I assess with high confidence, the "masking background / specific ink" method the '970 BACKGROUND describes without naming it. Its text expressly states that the masking background "absorbs light to a greater degree than the blue ink in all classes of copiers operating predominantly in the blue, green, blue-green or infrared regions," and that "to all these copiers, the background is similar to black." That is a direct teaching that a document background can be engineered to be differentially absorptive in the infrared band — precisely the physical precondition for claim 1's element A.
• US 4,282,300 (Elliott, 1981) teaches restricting a copier's illumination to the emission spectrum absorbed by the image, i.e., spectral engineering of the optical/detection channel. It supplies the reason a skilled artisan would select a narrow band (here, near-IR) to maximize the difference between the treated background and ordinary print.
• US 3,597,082 supplies photochromic/altered-reflectance document treatment and is a patentee admission (cited in the '970 BACKGROUND as the known state of the art).

Motivation to combine: the '970 specification itself supplies it. The BACKGROUND states that the prior document-side methods (photochromic paper; specific ink on a specific background) "can be inhibited only when specific types of copying machines are employed; they cannot prevent copying using a so-called white-light copier in which light-sensitive materials coated with zinc oxide are used," and that "it is therefore desired to develop an apparatus which is not seriously influenced by the color of images." That is the patentee's own statement of the problem and thus an admissible articulation of the motivation to move to a copier-side, spectrally-selective detector. Under KSR, a patentee's recitation of the problem in the specification is a legitimate source of the motivation to combine.

---

Ground C — The "comparison means" (element B) and the "adapt a conventional copier" teaching

US 3,831,007 (Braun) in view of any of the copier-control references (e.g., US 4,304,486).

I retrieved the operative text of Braun, which is unusually on-point for element B and for the general enabling question:

• It describes how "convenience office copiers of conventional types may readily be adapted to sense and respond to detection of the delta-distance bar code pattern on document 10," and shows a copier in which "each phototransistor… is driven into conduction by light… Thus, there is provided… a series of signals which correspond to the light and dark areas of the background pattern (as well as other indicia) on document 10."
• It then shows compare circuitry that "compares the parallel outputs from each shift register with a static four-bit predetermined binary code word."

This is materially damaging to claims 1 and 4:

• It teaches scanning a document in a copier with photosensors and comparing the sensor output against a predetermined reference (element B) — and, because the code is a delta-distance code, it teaches that the spatial extent/interval of scanned features carries information that can be discriminated.
• Claim 4 ("said scanning means provides an output… upon encountering said infrared ray-absorbing agent or upon encountering information on said manuscript sheet") reads almost verbatim onto Braun's statement that the phototransistor signals "correspond to the light and dark areas of the background pattern (as well as other indicia) on document 10." Braun does not distinguish the security pattern from the ordinary indicia — which is exactly claim 4's limitation.

Motivation to combine: Braun solves a different end (bar-code reading, not copy inhibition), but the motivation is supplied by Grounds A and B: once one accepts machine-side interrogation, one turns to the existing copier-scanning and compare-circuitry art to implement it, because that art already sits inside the copier and requires no new scanning hardware — which is also the rationale for claim 3.

---

Ground D — The "make unreadable" branch (claims 6, 7, 8)

Claims 7 and 8 require only that the disabling means be (7) turning off the charge-transfer corotron discharge or (8) turning off the exposure lamp. Both are the canonical ways to defeat xerographic imaging and are conventional copier controls; US 4,304,486 (1981) is a copier control-system reference showing routine regulation of charging/exposure, and the '970 specification itself recites these two options in a single sentence, without any assertion that either is novel or non-obvious. The forward-cited Xerox US 4,791,449 (1988) confirms the same technique ("cut off power to exposure lamp 108") was the industry's standard disabling mechanism — though, again, it is not itself prior art.

Motivation: once the decision to disable is made, killing the exposure lamp or the transfer corotron is the cheapest available disabling step, and both are illustrated side by side in the specification as equivalents. There is no non-obviousness to be found in choosing one of two conventional copier controls.

---

5. Dependent claims

Because claims 2 and 5–8 all depend from claim 1, a successful § 103 attack on claim 1 collapses the entire claim set — the classic combination-patent posture.

---

6. Why a PHOSITA would have combined these — the KSR synthesis

The § 103 case is strongest when the gap (size discrimination) is closed not by a reference but by routine engineering, which KSR expressly permits.

1. The false-positive problem is inherent and self-evident. Any presence/absence detector for an IR-absorbing agent will also fire on ordinary printed image areas, because ordinary black ink and the disclosed IR-absorbing agents both exhibit reduced reflectance. The '970 specification concedes exactly this: with a usual manuscript, "the output voltage is low at image areas" (curve C of FIG. 5). A detector that triggers on every dark area would block copying of ordinary documents — a commercially fatal false-positive rate. Any competent artisan implementing Harr's "No-Copy Attachment" would therefore necessarily add some discrimination criterion.

2. The claimed criterion is the most predictable one available. Given a line-scanning sensor, the only way to discriminate a large region from a small one in real time is to measure how long the below-threshold condition persists. The patent's own arithmetic (T derived from scan speed and manuscript size) shows there is no invention in the conversion; it is a unit conversion. KSR: "the combination of familiar elements according to known methods is likely to be obvious… when it does no more than yield predictable results."

3. The specification states the motivation itself. In the "advantages" paragraph, the patentee says the apparatus is advantageous because "by optionally setting the reference voltage VREF and the reference time T, erroneous detection can be avoided." That sentence is the motivation to combine, in the patentee's own words. An admission in the specification of why the threshold is desirable is powerful § 103 evidence.

4. The field was under intense, documented pressure. The BACKGROUND begins: "With the development and spread of office copiers or copying machines and facsimile, it has been strongly desired to provide a practical means to inhibit the copying of confidential documents or copyrighted materials." That is an admitted, long-felt, widely-recognized need.

5. Contemporaneous convergence. Within months of '970 issuing, Xerox (US 4,678,322; US 4,791,449) and Eastman Kodak (US 4,739,377) independently filed on the same three-part architecture — treated document, in-copier detector, disabled imaging cycle — using a different physics (phosphorescence). Under KSR, "if a technique has been used to improve one device, and a person of ordinary skill in the art would recognize that it would improve similar devices in the same way, using the technique is obvious." The 1986 convergence shows the architecture was the obvious solution; the fact that Xerox and Kodak chose different marker physics is evidence that the detection concept was well-trodden, not that '970 was inventive.

6. The specification disparages the prior art on grounds that cut against it. The BACKGROUND dismisses US 3,597,082 and the specific-ink method because they are machine-specific. But the same paragraph shows the artisan already possessed the conceptual tools — differential document reflectance in a selected spectral band — to which only a detector and a comparator needed to be added.

---

7. Where the § 103 challenge is weak — the patentee's best rebuttals

I should be candid about the limits of the above.

1. The size element is a genuine gap in the cited art. On the record I can reach, none of the five examiner-cited IBM TDB items, nor US 3,597,082, is shown to teach or suggest discriminating confidential documents by the extent of the absorbing region. Ground A therefore depends entirely on the "routine engineering / predictable use" branch of KSR, not on an express teaching, suggestion, or motivation. A patentee can argue that the references were directed to binary presence/absence detection and that no reference "recognized" the false-positive problem the size test solves. The counter (that the problem is inherent and conceded in the spec) is strong, but it is a factual dispute an examiner or court must resolve — it is not a slam dunk.

2. The "recognition of the problem" argument is contestable. KSR permits an obviousness finding even where the problem was newly recognized — but a fact-finder may nonetheless credit evidence that the art's binary-detection paradigm discouraged adding a size discriminator.

3. Claim 3's "only a single scanning means" is arguably structural. Although eliminating a duplicate scanner looks like an obvious economy, the claim as drafted is a negative limitation on the apparatus, and it could be argued that the cited art merely permits sharing without teaching the single-scanner apparatus.

4. Citations ≠ disclosures. The strongest technically relevant references I identified (US 3,852,088, US 3,831,007) came from outside the '970 page's own Prior Art section (see § 8 caveat). If they are not part of the asserted art, Grounds B and C weaken considerably, and the case reduces to Ground A alone.

5. Secondary considerations. The patentee would point to the commercial success of the Fuji Xerox FX-series machines named in Examples 1–2 (FX-4600, FX-2830), the nexus being that these are copiers adapted per the invention. This is weak here: the specification's own examples describe making manuscript sheets for use with pre-existing Xerox copiers, and the "advantage" recited is a design choice (adjustable VREF/T) rather than an unexpected result. Still, a well-documented commercial-success record with a genuine nexus would have to be weighed.

6. No live controversy. Per the previously generated Litigation section, the patent lapsed for failure to pay maintenance fees effective 1998-08-05 and was never, as far as I could determine, litigated. There is no actual § 103 contest to resolve; this is a retrospective analysis.

---

8. Confidence and sourcing caveats

• The five IBM TDB items and US 3,597,082 are squarely on the '970 record and predate the 1982 priority — high confidence.
• I could not retrieve the full text of the five IBM TDB items. They are not digitally indexed in a form my searches reached. My characterization of "Harr et al." / "Hobgood" / "Bruce" rests on (a) their full titles, recovered verbatim from the citation lists of US 4,791,449 and US 4,908,873, and (b) the examiner's citation of them on the '970 record. I do not know what detection physics, circuitry, or document treatment each discloses. Any § 103 ground resting on their content must be verified against the TDB articles themselves before it is asserted. If, for example, Harr et al. already describes a threshold or duration test, Ground A becomes far stronger; if it discloses only presence/absence and expressly warns against false triggers, it becomes an anticipation/teaching-away problem.
• US 3,852,088 — I identified it as the "masking background with a specific ink" method described (unnumbered) in the '970 BACKGROUND on the strength of a search hit showing its text on masking backgrounds and infrared-region copiers. The identification is inferred, not verified by a direct reading of the '970's citation list (which does not list it).
• US 3,831,007 (Braun) — I retrieved operative text (compare circuitry, "predetermined binary code word," retrofitting conventional office copiers). I did not verify its issue date from a primary record; the "similar documents" search surfaced it without a date, so treat "1974" as approximate pending verification.
• The forward-citation correction in § 2 is the one finding I would stake the most on: US 4,678,322 / US 4,739,377 / US 4,791,449 / US 4,956,666 / US 4,980,719 / EP 0 599 218 / US 5,622,798 all have 1986-or-later priority dates and cannot be used as § 103 art against a 1982-priority patent. If an earlier-generated section or a downstream document treats them as prior art, that is an error to correct.

---

9. Bottom line

Claims 1, 2, 4, 5, 6, 7, and 8 are vulnerable to a § 103 rejection, and claim 3 is moderately vulnerable — but the rejection stands or falls on a single limitation.

• Elements A (scanning), B (comparison to a reference level), and D's "prevent copying" function were all squarely in the art by 1982, taught by the examiner-cited IBM TDB copy-prevention items (Harr et al.; Hobgood; Bruce), reinforced by US 3,831,007's teaching of document scanning with photosensors and compare circuitry in a conventional office copier, and reinforced on the spectral side by US 3,852,088's teaching that a document background can be made differentially absorptive in the infrared.
• The only meaningful novelty is elements C/D — the "greater than a predetermined size" test. The best § 103 attack on that element is not a reference but the KSR "predictable use" argument:
  1. any presence/absence IR detector necessarily false-triggers on ordinary dark image areas (the specification concedes this);
  2. the only real-time way to distinguish a large absorbing region from a small one with a line scanner is to measure the duration of the below-threshold signal;
  3. the patentee's own specification supplies the motivation ("by optionally setting the reference voltage VREF and the reference time T, erroneous detection can be avoided");
  4. and the field's documented pressure, plus the 1986 convergence of Xerox and Kodak on the identical architecture, confirm the combination was the expected path.
• The soft spot is purely evidentiary: I have not located prior art that expressly teaches spatial-extent discrimination of an IR-absorbing region, and the examiner's allowance over five TDB items implies the Office treated the size limitation as the point of novelty. A challenger who finds a reference teaching extent/duration discrimination (or who successfully frames it as routine engineering under KSR) invalidates the whole claim set via claim 1; a challenger who does not will likely fail, because all eight claims rise or fall with that one element.