US 4653899

I'll search for this specific patent number across the requested sources.

The generic searches returned noise. Let me refine with more targeted queries.

Found the patent record. Now let me check for any litigation/CAFC docket activity and verify against USPTO records.

Summary — US 4,653,899

Sourcing note: My searches did surface the patent record itself (Google Patents, Justia, and a Google Patents classification sitemap entry), but the generic queries I ran did not return results from the USPTO Patent Public Search / PatentCenter interface or from any Court of Appeals for the Federal Circuit (CAFC) 2026 docket database. I ran out of search steps before I could run targeted docket queries. Treat everything in the "litigation" section below as absence of evidence, not evidence of absence.

Bibliographic data (well-corroborated)

The issue date and title are independently corroborated by Google Patents and by a Google Patents sitemap listing for the G03G 15 class ("03/31/1987 | US4653899 Image forming apparatus with area selection and preservation functions"). The Justia page (patents.justia.com/patent/4653899) returned matching specification text, confirming the number resolves to this document and not a similar one.

Abstract

The apparatus includes an original table, a transmission light emitting section, an erasure area specifying section for shifting the emitted light to an unnecessary portion of the original to specify a portion to be erased, a memory, a first controller for storing position data of the portion to be erased in the memory, a data hold request section, a data holding section having a holding medium that can repeatedly read out written data, a second controller for reading the position data out of the memory to hold it in the data holding section on request, a held-data readout request section, a third controller for reading the held position data back into the memory on request, an original scanning section, an image forming section, an image erasing section for selectively erasing an image to be formed, and a fourth controller for reading the stored position data during an image forming operation and supplying it to the image erasing section.

Claims — plain-language overview

The patent has 7 claims; claim 1 is the sole independent claim.

Claim 1 — the independent claim. This is a "mask/erase-by-editing" copier, and in plain terms it covers the combination of:

1. An original table (glass platen) for a light-transmitting original.
2. A transmission light source mounted to move along the platen, shining light through the original (so the operator sees a spot through the sheet — this works for thick originals such as a postcard).
3. An erasure-area specifying mechanism that moves that light spot to the unwanted part of the original to designate what should be erased.
4. A temporary data store (memory) plus a first controller that writes the position data for the designated erasure area into it.
5. A data-hold request mechanism (a user key) and a removable/repeatable data-holding medium (elsewhere described as a "quick disk" magnetic medium).
6. A second controller that copies the stored position data from temporary memory out to the holding medium on request.
7. A held-data readout request mechanism and a third controller that pulls the position data back off the holding medium into temporary memory on request — i.e., a save/recall round trip.
8. An original scanning optical system that moves along the platen to scan the original.
9. An image forming section that forms and develops a reflected-light image onto an image-receiving medium (the electrophotographic drum/paper path).
10. An image erasing section that selectively erases parts of the image being formed.
11. A fourth controller that reads the position data (from either the first or third controller's path, directly or via the holding medium) during the image-forming operation and feeds it to the erasing section.

The practical gist: erase a specified region of a copy, and save/reload that region definition so the same edit can be repeated.

Dependent claims (paraphrased):

• Claim 2 (dep. 1): the transmission light source is a light-emitting element plus a lens, movable perpendicular to the scan direction, forming spot light.
• Claim 3 (dep. 1): the erasure-area specifying mechanism includes means for calculating the position data corresponding to the specified area.
• Claim 4 (dep. 1): the erasing section is a linear array of light-emitting elements facing the image forming section.
• Claim 5 (dep. 4): those array elements are positioned so as to be selectively illuminated during focusing — corresponding to the disclosed array placed between the charger and the exposure region (latent image selectively discharged before exposure).
• Claim 6 (dep. 4): the array elements are positioned so as to be selectively illuminated during development — corresponding to the disclosed alternative placement between the exposure region and the developing unit.
• Claim 7 (dep. 1): the data holding section is a "quick disk" device with a rotatable magnetic recording medium, detachably loaded as the holding medium.

Litigation / CAFC 2026 docket

• No CAFC 2026 docket activity for US 4,653,899 was found in my searches. I was unable to reach a live CAFC docket query before exhausting my search budget.
• Corroborating circumstantial evidence points against current litigation: the patent's recorded legal status is "Expired – Fee Related," with anticipated expiration 2006-02-04. Even at its maximum 20-year term from the 1986 filing date, protection would have lapsed by 2006, so an appeal on the merits of this patent reaching the Federal Circuit in 2026 would be unusual (though ancillary or historical matters are not impossible).
• Forward citations are ordinary art citations from 1985–2005 (mostly Toshiba, Minolta, Xerox, Sanyo) and do not indicate litigation.

Explicit uncertainties

1. I could not confirm the bibliographic record against the USPTO Patent Public Search / PatentCenter directly; my confirmation rests on Google Patents, Justia, and a Google Patents sitemap entry.
2. I could not verify maintenance-fee lapse dates or any reexamination/reexamination-certificate history. One search result surfaced a Google Patents-hosted PDF path containing "RED_FLAGS_Oct.2007_" for this patent number; I do not have authoritative information on what that filename signifies, so I am not characterizing it.
3. Any statement about CAFC 2026 docket status is unverified — I found nothing, but did not complete a definitive docket search.

I'll run targeted litigation searches for this specific patent number.

Let me run more targeted litigation-database queries.

Let me check a few final angles — Toshiba-side assertions and PTAB challenges.

Litigation search results for US Patent No. 4,653,899

Bottom line: No litigation involving US 4,653,899 was found. No district court case, no CAFC appeal, and no PTAB (IPR/PGR) proceeding naming this patent could be identified.

I want to be explicit about the limits of that statement, because the task instructions asked for plaintiff/defendant/jurisdiction/case number/filing date/status for each case — and I have no such case to report. This is an absence of evidence, not a certified negative. I was unable to reach a live, patent-number-indexed query on Unified Patents' litigation portal (portal.unifiedpatents.com/litigation/caselist) or a Docket Alarm / PACER party-and-patent search before exhausting my search budget. What follows is what the searches actually returned.

Database-by-database result

Important: unrelated documents bearing the string "4653899"

Per the strict rule to interpret identifiers literally and not auto-correct, I note these collisions and explicitly decline to attribute them to US 4,653,899:

• JP Patent No. 4653899 (JP 4653899 B2) — a Japanese patent, "日よけ布およびその製造方法" (awning cloth and method of manufacture), Schmitz-Werke GmbH & Co. KG, application JP 2001-107258, registered 2010-12-24 / published 2011-03-16. This is a different, unrelated patent in a different jurisdiction and a different technology. It is not US 4,653,899.
• A Brazilian JUCIS-RS filing number "4653899" for Drogarias Bitello Ltda - ME.
• A UK Companies House gazette item numbered "(4653899)" (Shantinagar Ltd / Westwood Leisure matters).
• Note the near-miss "(4653898)" in the Edinburgh Gazette — a still different number.

None of these are litigation, and none involve the patent in question.

Consistency with the earlier section — no contradiction, but one open item

This is consistent with, and reinforces, the earlier "Patent summary" section, which reached the same conclusion and flagged the same caveat. Two points carry forward:

1. The patent's legal status undercuts any plausible 2026 litigation. US 4,653,899 is recorded as "Expired – Fee Related," with anticipated expiration 2006-02-04. Any infringement action would necessarily concern pre-2006 conduct, and any appeal reaching the Federal Circuit in 2026 would be extraordinary.
2. Unresolved item, still unresolved: the earlier section noted a Google Patents-hosted PDF whose URL path contains RED_FLAGS_Oct.2007_ for this patent number (patents.google.com/patent/US4653899A/RED_FLAGS_Oct.2007_.pdf). My search reproduced that same URL, and its retrieved content is simply the patent's own specification text. I still do not have authoritative information on what the filename token signifies, and I am not characterizing it as evidence of a reexamination, a disclaimer, a certificate of correction, or litigation. If it has legal significance, it would need to be confirmed directly against the USPTO Patent Public Search / PatentCenter file wrapper.

Recommended next steps for a definitive negative

If a legally defensible "no litigation" conclusion is needed, a patent-number-indexed search should be run directly against: (a) PACER / CourtListener RECAP using the exact patent number as a docket-text query; (b) USPTO PTAB proceedings search; (c) USPTO Patent Public Search file wrapper for this patent (to resolve the RED_FLAGS item); and (d) Unified Patents' litigation portal with case-insensitive exact-string matching. A negative result across those four is what would make the conclusion credible. Until then: no known litigation.

I'll verify the structured "no proceedings" finding with targeted searches before writing up.

Proceedings overview

Total AIA trial proceedings on file for US 4,653,899: 0. The structured USPTO Open Data Portal (ODP) block supplied with this prompt returns no AIA trial proceedings — no IPRs, no PGRs, no CBMs — against this patent, which means there is nothing in the "active / claims invalidated / claims sustained / settled / institution denied" buckets to report. My independent web searches for the patent number in combination with PTAB/IPR/PGR/CBM terms surfaced no proceeding, no institution decision, and no Final Written Decision; the only hits were unrelated matters (a different Japanese patent numbered JP 4653899, an unrelated ITC/FRN document referencing a different '899 patent, and Toshiba-side litigation over other patents entirely).

The bottom line for a defendant: there is no PTAB record to lean on. No claim of US 4,653,899 has been canceled or even challenged at the Board, and there is no estoppel-creating petitioner history. Critically, though, the patent's own record states it expired 2006-02-04 ("Expired – Fee Related," anticipated expiration 2006-02-04), so the practical defensive question is not "how do I kill it at the PTAB" but "is this patent even enforceable against my client's conduct today." A demand letter citing US 4,653,899 in 2026 requires an immediate statute-of-limitations and damages-window analysis before any merits work.

Per-proceeding sections

None apply. There are no proceedings to render in the ### {PROCEEDING_NUMBER} — {Petitioner} v. {Patent Owner} format. Per the constraint against fabricating proceeding numbers, I am not creating placeholder entries.

Verification trail and its limits

• Canonical source (authoritative for this task): the structured "PTAB proceedings on file" block — result: empty / no AIA trial proceedings.
• Corroborating searches run: queries combining "4653899" / "4,653,899" with IPR, PTAB, inter partes review, and institution decision. All returned zero PTAB records for this patent.
• What I could not do: I could not open the PTAB E2E / PTAB Center proceeding search directly, and one attempted query was cut off by a search-step limit. I also did not reach the CAFC docket or CourtListener. So the zero count rests on (a) the authoritative ODP block and (b) negative web results — strong, but not a line-by-line E2E confirmation. I am flagging this rather than asserting E2E-level certainty.

One loose thread worth naming

My searches again surfaced a Google Patents-hosted PDF path keyed to this patent: patents.google.com/patent/US4653899A/RED_FLAGS_Oct.2007_.pdf. This was already noted as unresolved in the earlier bibliographic section, and I still cannot characterize what it is or who produced it. It is not a PTAB document and carries no evidentiary weight on the question at hand. Do not treat it as a proceeding.

Strategic summary

Claim status: all claims UNTESTED. Claims 1–7 of US 4,653,899 stand exactly as issued on 1987-03-31 — claim 1 (independent) plus dependent claims 2–7. Nothing has been canceled, narrowed, disclaimed, or confirmed by the Board. There is no reexamination or reissue history that my searches revealed either (though, as noted in the earlier section, I could not conclusively verify reexamination/reexam-certificate history against USPTO sources). This is a pre-AIA, never-challenged patent — meaning any invalidity position you take will be litigated from scratch, with Phillips claim construction and no PTAB-created record to borrow.

Estoppel landscape: 35 U.S.C. § 315(e)(2) is inapplicable — estoppel is triggered only by a petitioner who obtains a Final Written Decision, and there has never been a petitioner. No private party is estopped from raising any ground, and no ground has been "reasonably could have raised"-consumed. Practically, that cuts both ways: you have maximum freedom of art, but you also have no early, cheap kill mechanism already paid for by someone else and no roadmap from a prior panel's claim constructions of terms like "transmission light emitting means" (a § 112 ¶ 6 / means-plus-function term that any litigator should expect to fight over) or "associated data holding means."

Pattern signals: none available. There is no repeat-petitioner pattern, no patent-owner appeal history at the Federal Circuit, and no evidence of a defensive aggregator (Unified Patents, RPX, etc.) having taken this patent on. The reason is almost certainly economic and temporal rather than legal: the patent expired 2006-02-04, which means (i) no ongoing royalty base, (ii) damages capped by the six-year § 286 lookback into a window that has been closed for roughly two decades, and (iii) no rational economic basis for an IPR petition, which is why the proceeding count is zero. The absence of IPRs here is not a signal that the patent was tested and survived — it is a signal that nobody had a financial reason to file.

Recommended next steps

If you are a defendant and a demand letter cites US 4,653,899:

1. Check the expiration math first, before anything else. The Google Patents record for US 4,653,899 lists legal status "Expired – Fee Related" with anticipated expiration 2006-02-04. Even setting that aside, a patent with a 1986-02-04 filing date and 1985-02-06 JP priority would reach the end of its enforceable term no later than 2006. Any infringement theory must therefore be confined to conduct before that date. If the demand targets present-day products or conduct, the appropriate response is a short, dated letter pointing to the expiration and demanding the asserted acts, products, and date ranges — not an invalidity study.
2. Do not expect a PTAB shortcut, and do not budget for one. There is no PTAB record, no FWD to link to, and no invalidated claim to quote back at opposing counsel. There is no PTAB E2E or CourtListener opinion to cite in this matter because none exists. If you nevertheless need an AIA route (e.g., to obtain a Phillips construction or a claim-cancellation ruling for a residual pre-2006 damages claim), an IPR petition against an expired patent is legally available but rarely economical — confirm with counsel whether the remaining damages exposure justifies the filing fees, expert costs, and the statutory one-year trial clock that would run from institution.
3. Prefer a § 285 / Rule 11 posture if the demand is meritless on its face. Where the demand letter fails to account for a patent that has been expired since 2006-02-04, or where it relies on expired claims to extract a license, that is a fact pattern courts have treated as fee-shifting material. Preserve the demand letter and any pre-suit communications with that in mind.
4. Verify against USPTO primary sources what I could not. If this patent matters commercially, pull the full file wrapper from USPTO PatentCenter (application 06/825,876) to confirm (a) maintenance-fee lapse and expiration, (b) absence of any reexamination or reissue certificate, and (c) any assignment history beyond the recorded 1986-02-04 assignment of Watanabe → Kabushiki Kaisha Toshiba. Then run the proceeding search directly at PTAB E2E / PTAB Center to independently confirm the zero count that the ODP block and my searches indicate.
5. Identify the real patent owner before doing any of the above work. The record shows Original Assignee Toshiba Corp. and the 1986-02-04 assignment to Kabushiki Kaisha Toshiba. If the demand comes from a non-practicing entity, verify the chain of title from Toshiba to the asserting party — a broken or unrecorded chain changes the calculus on standing and on who, if anyone, can enforce a 2006-expired patent.

If you are trying to assess whether the patent has been "hardened" or "weakened": it is neither. It is untested and expired. Do not describe it as surviving IPRs, and do not describe any claim as canceled. On the record I can verify from the ODP block and web sources, all seven claims stand as issued, and no AIA trial proceeding has ever been filed against them.

I'll query the assignment record and related sources now.

Sourcing caveat (read first)

I could not reach either USPTO assignment interface in this session. Targeted queries against assignmentcenter.uspto.gov, legacy-assignments.uspto.gov, and assignment.uspto.gov returned no assignment instrument for US 4,653,899. The one assignment event I can report comes from the Google Patents "Legal Events" record reproduced in the authoritative full text supplied for this patent (which is itself sourced from USPTO assignment data), and that record does not include reel/frame numbers or a correspondent name.

So: I can confirm that at least one assignment was recorded, but I cannot report reel/frame or correspondent for it. I will not invent those numbers. Treat the "Assignment timeline" section below as one verified event plus an explicit gap.

Inventors

• Sole inventor. No co-inventor is named on the front page, in the specification, or in the Google Patents legal-events record, which lists exactly one assignor: "Assignors: WATANABE, JUNJI."
• Departure pattern: Not determinable. A single-inventor, single-assignor, employer-assignment case from 1986 Toshiba gives no signal about whether the inventor left within 12 months of filing, and I found no employment or personnel record. I am not going to read anything into the absence of data — a sole inventor assigning to his employer under an employee-obligation clause is the most ordinary fact pattern in patent prosecution, not a precursor to a fire-sale.
• No successor inventor entities (no second-generation filings by this inventor as assignor) surfaced in the forward-citation set, which is comprised of ordinary art citations from Toshiba, Minolta, Xerox, and Sanyo.

Original assignee

Kabushiki Kaisha Toshiba (Toshiba Corp.), a corporation of Japan, 1-1 Shibaura 1-chome, Minato-ku, Tokyo.

• Primary line of business: Diversified electrical/electronic manufacturing — heavy electrical equipment, semiconductors, consumer electronics, and office equipment including electrophotographic copiers. The patent itself is directed to the copier line (G03G 15/60, G03G 15/605, G03G 15/36), and the specification describes a full production-style machine: optical scanning carriages, a photosensitive drum, chargers, a developing unit, cassette feeds, a fixing roller pair, and an LED erasure array driven by a rank-by-rank memory readout.
• Did Toshiba ship a product embodying the claims? Unclear — I could not confirm this. Toshiba was unquestionably a major copier OEM in this period, and the disclosed "quick disk" (a removable magnetic diskette, claimed in claim 7) was a real commodity removable medium used in Japanese word processors and office machines of the mid-1980s. But I found no evidence tying a specific Toshiba copier model to the claimed combination — moving transmission spot light through the original to designate an erasure area, plus save/recall of that area data to a removable magnetic medium. Do not treat the OEM's general copier business as proof of embodiment.
• Current status: Toshiba Corp. was delisted and taken private in 2023 by a Japan Industrial Partners-led consortium (TBJH Inc.) — a restructuring, not a liquidation or Chapter 7/11 of the assignee. Distinguish this carefully: Toshiba's 2017 Chapter 11 filing was Westinghouse Electric, a U.S. subsidiary, in the nuclear business, unrelated to this patent. The office-equipment business has long been carried by the separately listed Toshiba Tec Corporation. Whether US 4,653,899's expired rights sit with Toshiba Corp. or Toshiba Tec today is moot (see next line) and I did not verify it.
• Legal status: Expired – Fee Related, anticipated expiration 2006-02-04.

Assignment timeline

Recorded 1986-02-04 — Reel/Frame: NOT RETRIEVED

• Executed: date not shown in the available record (necessarily on or before 1986-02-04; the U.S. application 06/825,876 was filed the same day)
• Conveyance: Assignment ("ASSIGNMENT OF ASSIGNORS INTEREST")
• Assignor: Junji Watanabe
• Assignee: KABUSHIKI KAISHA TOSHIBA, A CORP. OF JAPAN
• Correspondent: Not retrieved. This is the field most useful for your NPE-family linkage, and its absence is the single biggest hole in this report. Google Patents' legal-events feed does not carry the correspondent field; only the Assignment Center record (or the recorded instrument PDF on legacy-assignments.uspto.gov) does.
• Context: Ordinary employee-inventor-to-employer assignment executed as part of the original filing, not an acquisition, reorg, securitization, or transfer-to-asserter.

No further assignment events are recorded. Specifically, there is no evidence of:

• any post-issuance transfer to an IP holding vehicle,
• any security agreement or release,
• any merger or change-of-name record (notable, given Toshiba's subsequent corporate reorganizations — but with the patent expiring 2006-02-04, there was no reason to record one),
• any license or transfer to an NPE or aggregator.

This absence is itself the finding: the patent appears to have remained with Toshiba from filing until expiration. Consistent with that, its full 20-year term ran out in 2006 (anticipated expiration recorded 2006-02-04), leaving nothing to sell.

One loose end, flagged not characterized

A Google Patents-hosted file path exists under this patent: https://patents.google.com/patent/US4653899A/RED_FLAGS_Oct.2007_.pdf. My search this session returned the patent's own specification text from that URL, which indicates it is a Google-hosted copy/derivative of the patent document rather than a USPTO assignment instrument. I cannot confirm what the "RED_FLAGS" filename signifies and I am expressly not characterizing it. It is not an assignment record and does not appear in the timeline.

Timeline diagram

timeline
    title Ownership of US 4653899
    1985 : Invented by Junji Watanabe
         : JP priority application 60-21081 filed
    1986 : US application 825876 filed
         : Assigned to Kabushiki Kaisha Toshiba
    1987 : US 4653899 issued to Toshiba
    2006 : Patent expired

NPE / troll-pattern signals

1. Shell-entity transfer — not present. The only assignment of record moves the application from an individual employee-inventor into an operating manufacturer, the inverse of the shell-entity pattern. No "IP / Holdings / Licensing / Ventures" suffix appears anywhere in the chain, and no Delaware or Texas single-purpose LLC is named.

2. Known asserter in the chain — not present. Neither assignor nor assignee matches any entity on the public NPE lists you supplied (Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities). The single recorded assignee is Toshiba. Note the forward-citation set — US4707110 (Toshiba), US4724464 (Mita), US4734789 (Xerox), US4742373 / US4806976 / US4806978 (Minolta), US5030990 (Sanyo) — is a set of operating competitors citing art, which is the normal profile for a 1980s editing-copier patent, not an assertion-driven citation cluster.

3. Repeat correspondent across the chain — unclear, and cannot be assessed. There is only one recorded link, so "recurrence" is undefined; and critically, the correspondent field was not retrievable for that one link. I have no attorney name or firm of record to run against Unified Patents / RPX / Patent Progress lists. This is a genuine data gap, not a negative finding.

4. Cascading transfers — not present. One transfer, in 1986, at filing. There is no chained-LLC sequence, and therefore no shared-address or common-principal pattern to evaluate. The 20-year gap between the 1986 recording and the 2006-02-04 expiration is the opposite of a <24-month cascade.

5. Pre-litigation transfer — not present. No infringement suit naming this patent was found in either this session or the prior one, so there is no litigation date against which a transfer could fall within 6 months. The only transfer precedes issuance by 14 months.

6. Bankruptcy fire-sale — not present. Toshiba has undergone significant restructuring, but this patent expired in 2006, well before the notable Toshiba distress events (the 2017 Westinghouse Electric Chapter 11, and the 2023 take-private). There is no record of this patent being sold in any proceeding, and there was no unexpired term left to sell. I explicitly decline to infer a fire-sale from Toshiba's general corporate history.

7. Privateering — not present. No NPE was ever put in the chain, so there was no vehicle through which Toshiba could assert this patent against competitors off-balance-sheet.

8. Defensive aggregator (anti-NPE) — not present. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. The patent did not need neutralizing — it expired.

Verdict

Insufficient data — on the strictest reading of your rubric, the record contains only the original assignment (employee-to-employer, recorded 1986-02-04, assignor Junji Watanabe, assignee Kabushiki Kaisha Toshiba) and nothing post-issuance. That is a category match for "Insufficient data," and it is also the substantively correct call: none of the eight NPE signals is present, six are affirmatively negated, and two (repeat correspondent, cascading transfers) are unassessable solely because the reel/frame and correspondent fields could not be retrieved from a USPTO assignment interface this session.

The single most decisive fact is the 2006-02-04 anticipated expiration on an "Expired – Fee Related" record: any transfer chain built to enable assertion would have had to occur while the patent was still enforceable, and no such transfer exists in the record. There is no basis to call this an NPE patent, and no basis to call it a defensive-aggregator kill either — it simply ran its course inside Toshiba.

Verification link: https://assignmentcenter.uspto.gov/ (search patent number 4653899) and the legacy mirror at https://assignment.uspto.gov/patent/index.html — retrieving reel/frame and correspondent from either is the one outstanding action that would close this report's only material gap, and it is a retrieval I could not complete here rather than a search that came back empty.

I'll research this systematically, starting with the patent record itself and its cited references.

The patent record is confirmed. Now let me pull details on each cited reference.

Sourcing note and a flagged contradiction

Date discrepancy (flagged as instructed): the task header says April 26, 2026, while the operating context for this session states today's date is 2026-09-30. I cannot reconcile these; I note it rather than picking one silently. Nothing in the analysis below turns on which is correct.

Database sourcing: I searched for the number 4653899 literally and did not auto-correct it. My search budget was exhausted before I could run a query directly inside the USPTO Patent Public Search / PatentCenter interface, so the record I relied on is the Google Patents publication (which mirrors the USPTO printed patent) plus corroborating third-party mirrors. I am therefore reporting the references cited on the face of US 4,653,899 as reproduced in the authoritative full text supplied to me, cross-checked against Google Patents for the individual references. I am not claiming to have independently re-pulled the USPTO "References Cited" field.

One earlier uncertainty is now resolved: the previously-flagged URL path https://patents.google.com/patent/US4653899A/RED_FLAGS_Oct.2007_.pdf is the full-text PDF of this very patent — my search returned that URL with the patent's own specification and claims as its content. It is a Google bulk-scan filename, not a litigation or "red flag" indicator for this patent. The earlier section was right to leave it uncharacterized; it can now be dismissed as a false alarm.

The citation set

The front-page "References Cited" block of US 4,653,899 lists seven U.S. patent references. The text I was given is truncated — it reproduces only the first five (it cuts off mid-sentence inside the fifth, US 4,215,929, at "…capable of controlling an electrostatic image for"). I have no reliable content for references 6 and 7 and will not invent them. If a complete anticipation analysis is required, the remaining two must be pulled from the USPTO full-text record.

The five retrievable citations, verbatim as listed:

Also on the face of the patent (family, not prior art): the JP priority JP 60-21081 / JPS61180265A (1985-02-06) and the German counterpart DE 3603755 A1. The 25 "Cited By" entries are later art citing this patent and are therefore not §102 art against it.

Per-reference §102 assessment

Governing law caveat: with a 1985-02-06 priority date and 1986-02-04 filing, this patent is governed by pre-AIA 35 U.S.C. § 102. Anticipation requires a single reference disclosing every element of a claim, arranged as in the claim ("arranged as in the claim"). None of the five references below meets that bar for claim 1.

1. US 3,827,799 (Ricoh)

Disclosure. A pair of charge-preventing plates sit at the transverse ends of the corona charging device and are driven by indicators slid manually along graduated edges of the original glass. A separate light-intercepting plate carries a lamp 16 aimed at a marginal photoreceptor strip "so that the charge carried by such marginal area can be removed." So it does teach (a) an original table, (l) a photoreceptor image-forming system, and (m) a selectively energizable discharge lamp acting as image-erasing means, plus manual setting of the region.

§102 reach. No anticipation of claim 1. The area-defining input is a manual mechanical indicator, not "transmission light emitting means … for emitting transmission light through the original," and there is no (d)–(j) data-store / data-hold / read-back architecture and no (n) fourth control means reading stored position data during image formation. Claim 4 (a "plurality of light-emitting elements linearly arranged") is also not met: US 3,827,799 uses one discrete lamp plus mechanical plates, not a linear emitter array. Best characterizable as §103 background for the "erase the margin by discharging the drum" concept.

2. US 3,997,873 (Manitou Systems)

Disclosure. A photocopier accounting network. Each copier terminal reads a punched badge (a portable information medium), links to a central recorder, and the recorder records terminal/user/count data on punched paper tape and verifies badge validity (claims 8–11; claim 11 recites "said associated devices are copying machines").

§102 reach. No anticipation of any claim. It is cited, in my reading, as general evidence that copiers of the era were known to write data to, and repeatedly read data from, a portable/removable medium — tangential support for the "data holding… which can repeatedly read out the data held" limb of claim 1. But it stores usage counts, not erasure-area position data; it has no image-erasing means; and it is not a "rotatable magnetic recording medium." It therefore cannot anticipate claim 1 and specifically cannot anticipate claim 7 (punched tape/badge ≠ "quick disk device including a rotatable magnetic recording medium"). §103 relevance only, and even that is thin.

3. US 4,045,218 (Xerox / McVeigh)

Disclosure. This is the most substantively relevant of the five. Per the abstract: after exposure, "the non-selected portions of the resulting latent image are erased prior to developing the image with a given colored toner," using a selector (FIG. 2, "color highlighting selector"); a second pass erases the selected portions and develops them in a second color.

§102 reach. Strong as to discrete elements of claim 1 — it plainly discloses (a) an original table/support, (l) an electrophotographic image-forming means, and (m) selective image-erasing means operating on the latent image before development — and it is direct evidence for claim 6's concept of erasure at the development stage. But it lacks (b)/(c) the through-original transmission light spot used by the operator to designate the region, and lacks the entire (d)–(j) temporary-store / hold-to-removable-medium / read-back chain and (n) the fourth control means. No claim 1 anticipation; strong §103 art for the erasing-means and claim-6 limitations. (I did not verify the selector's physical implementation — whether it is an array or a mechanical shutter — so I flag that a claim 4 "linear array of light-emitting elements" reading is unconfirmed.)

4. US 4,157,822 (IBM)

Disclosure. A recirculating document feeder with unloading — sheet-transport hardware. In my searches I obtained only its bibliographic entry; I did not retrieve its specification.

§102 reach. No anticipation of any claim. It bears on neither region selection nor data storage nor erasure. It reads as a peripheral "copying-machine environment" citation (possibly cited for the platen/original-handling context). I state this on limited evidence and flag the incompleteness.

5. US 4,215,929 (Canon)

Disclosure. The retrieved text confirms it addresses precisely the problem this patent addresses: selecting only the required region of an original, and it enumerates the then-known techniques — "(i) In the electric charging step, the unnecessary image region … is prevented from being charged. (ii) In the developing step, the unnecessary image region … is prevented from being developed. (iii) In the image transfer step, the unnecessary image region … is prevented from being transferred." It uses "masking means including masking members" moved relative to a "control light" irradiating region, and contemplates selecting the necessary region "from an image original projected on a monitoring device." Its stated objects even include choosing between retaining the surrounding region versus the enclosed region — i.e., the same inside/outside mask polarity pair used in the present patent's keys 30c / 30p.

§102 reach. This is the closest art on the erasing/selection side and is the best candidate to be the examiner's principal reference for the characterising concept. Even so, it does not anticipate claim 1: the region is set by masking members/control light rather than by a movable transmission light source shining through the original, and there is no data storage, no removable holding medium, and no read-back control chain. Its "background art" discussion is squarely §103 territory for claims 4, 5 and 6 (erasure effected at charge/focus stage vs. development stage), which is exactly the two-placement distinction the present patent draws in FIGS. 21A/21B.

6 and 7. Not retrievable

The front-page citation list contains two further U.S. references that the supplied text truncates before displaying. I have no citation details for them and decline to guess. Any of them could in principle be more on-point than the five above, so this gap is material.

Bottom line

• No single one of the retrieved citations anticipates claim 1 under pre-AIA § 102. Claim 1's novelty/priority over this art, if any, resides in the combination of (i) operator designation of the erase region by a movable spot of light transmitted through the original and (ii) the temporary-memory ⇄ removable holding-medium ⇄ read-back control architecture feeding (iii) the during-image-formation erasure controller.
• Best prior art, ranked by relevance: US 4,215,929 (Canon) and US 4,045,218 (Xerox) — both are §103 art against the erasing-means limitations, and US 4,215,929's explicit enumeration of charge-stage vs. development-stage suppression is the nearest thing to an anticipation argument against dependent claims 4/5/6. US 3,827,799 (Ricoh) is §103 art for margin-erasure. US 3,997,873 (Manitou) is weak §103 art for portable copier data media. US 4,157,822 (IBM) is peripheral.
• Nothing here anticipates claim 7; no retrieved reference discloses a detachable rotatable magnetic medium ("quick disk") as the holding medium.

Explicit residual uncertainties: (1) I did not query the USPTO PatentCenter directly — the record rests on the supplied full text plus Google Patents; (2) two of the seven cited references are missing from the supplied text and are unanalysed; (3) I did not retrieve the US 4,157,822 specification or the US 4,045,218 selector drawings, so the claim-4 "linear array" point is unconfirmed for those two; (4) I did not complete a CAFC/litigation check in this pass, and the patent's recorded status is Expired – Fee Related (anticipated expiration 2006-02-04), which is inconsistent with live 2026 merits litigation.

I'll gather details on the cited prior-art references before building the § 103 analysis.

Good detail on the first two citations. Now let me retrieve the remaining cited references and confirm the full citation list.

§ 103 Obviousness Analysis — US 4,653,899 (Watanabe / Toshiba)

Note on scope and on contradictions. This analysis builds on, and does not restate, the claim chart and bibliographic work already produced. Two housekeeping flags before the substance:

1. Date discrepancy: the operating rules state today is 2026‑09‑30, while the task header says April 26, 2026. I have proceeded on the substance (immaterial to the analysis, since the patent expired 2006‑02‑04), but flagging it per the instruction to surface contradictions.
2. Prior-art corpus used: per instruction, I used the "Prior Art" portions of the US 4,653,899 page — i.e. the "Citations (7)" list. That list is truncated in the supplied page text: only five entries are fully visible (US 3,827,799; US 3,997,873; US 4,045,218; US 4,157,822; US 4,215,929), with the fifth cutting off mid‑abstract. References 6 and 7 of the seven are not identifiable from the material provided, and I could not retrieve them before exhausting my search budget. Every combination below therefore rests on the five confirmed references. I also note that the 25 "Cited By" items and the family members DE 3603755 A1 / JPS 61‑180265 A are not available as prior art (see § 5).

1. Governing legal framework

• Pre‑AIA § 103 applies (priority 1985‑02‑06; filed 1986‑02‑04; issued 1987‑03‑31). The hypothetical person having ordinary skill in the art ("PHOSITA") is a copier-systems engineer of the mid‑1980s with working knowledge of xerographic process stations, optical scan carriages, platen accessories, LED erase arrays, and microcomputer-controlled copier panels.
• Analysis under Graham v. John Deere (scope/content of prior art; differences; PHOSITA level; secondary considerations), with the motivation inquiry governed by KSR Int'l v. Teleflex: a combination is obvious if it flows from "known methods… according to known methods," from "a finite number of identified, predictable solutions," or from "market demand" rather than from hindsight.
• All claims are in means‑plus‑function form ("…control means for…"). Under pre‑AIA § 112 ¶ 6 the "means" recitations are construed to cover the disclosed structures and equivalents; that breadth makes the claims correspondingly easier to read onto combinations of old copier elements.

2. The confirmed prior art of record and what each actually teaches

3. Differences between claim 1 and the prior art

Mapping claim 1's elements (labels A–K per the earlier chart) against the confirmed art:

The bottom line: four of the five confirmed references are copier‑editing/area‑control art (Ricoh, Canon, Xerox) or copier‑data‑handling art (IBM, Manitou). The only textual gap on claim 1 is element B — a spot light transmitted through the original as the designation mechanism. Everything else is a question of combination, not of missing subject matter.

4. Combinations that would render the claims obvious

Combination I — US 4,215,929 (primary) + US 3,827,799 + US 4,045,218 → anticipates the substance of "select an area, erase the latent image in that area"

• US 4,215,929 already teaches the core inventive concept of claim 1: designating an arbitrary, continuously settable region of the original (by control‑moved masks), and then eliminating charge only in the unnecessary image region before development, using a light source that may be "a linear light source."
• US 3,827,799 supplies the platen‑referenced, manually operated designation of which portions are to be imaged/erased (indicators slid along the glass edge, mechanically linked to the charge‑preventing/illuminating hardware), and confirms that removing charge by light is an established erase mechanism.
• US 4,045,218 confirms the "erase the non‑selected portion, then develop the selected portion" sequence, and expressly motivates it commercially ("highly useful for emphasizing selected paragraphs or portions of a letter").

Motivation (KSR‑compliant): all three address the same pre‑existing problem — preventing reproduction of unwanted portions of an original — and all three solve it inside the same xerographic process stations. Combining a mask‑type area selector with platen‑referenced designation means and a pre‑development latent‑image erase was the predictable, finite set of known solutions to that problem. No new mechanism, no change in principle of operation.

Result: claims 1 (except element B), 4, 5, and 6 are strongly suggested.

• Claim 4 ("plurality of light‑emitting elements linearly arranged … opposing said image forming means") is essentially read on by US 4,215,929's "spot light sources sufficient in number to uniformly irradiate … or a linear light source" and by US 3,827,799's erase lamp.
• Claim 5 ("selective light emission … during focusing") reads on the pre‑exposure/at‑exposure placement (US 4,215,929's control light in frame 15 alongside the exposure slit; US 3,827,799's marginal‑erase lamp before exposure).
• Claim 6 ("…during development") reads on the post‑exposure/pre‑development placement of US 4,045,218 and US 4,215,929.
  The two placements are the only two positions in the disclosed process for a latent‑image eraser; trying both is a design choice between two known options (KSR: "a finite number of identified, predictable solutions").

Combination II — Combination I + US 4,157,822 → supplies the "hold / re‑read / repeat" round trip

• US 4,157,822 discloses a machine‑readable medium (header card) carrying job‑defining data — including copy count and reduction size — read by a code reader to "program[] the machine for the subsequent job," in a machine that recirculates the same original to make many copies. That is the functional equivalent of claim 1's elements E, F, and G (hold request, write to medium, read‑back request, restore to memory), applied to copier setup data.
• Motivation: the specification itself supplies it. The patent states as an admitted fact that "quick disk Qd stores display data specifying each operation of the copying machine, control data for controlling the operation thereof, and the like." Once the machine already reads operating/control data off a removable medium, extending that same medium to carry the coordinates of the area to be erased — so an identical edit can be repeated for a run of copies — is an obvious substitution of one class of copier data for another on an already‑present data path. This is the strongest single obviousness argument against claim 1's storage/recall sub‑combination, because the patentee's own disclosure concedes the medium‑driven‑control architecture.

Combination III — Combination II + US 3,997,873 → supplies the recording/read‑back of per‑machine data

• US 3,997,873 shows a per‑copier terminal with memory that records data and reads it back (memories 30/38/34; data verified; recorded centrally). It corroborates that, as of 1976, storing copier‑operation data on a re‑readable medium and reading it back to control the copier was routine.
• Motivation: accounting/administrative reuse of a portable record (fewer costly hardware duplications, portability between machines) — the same market-driven rationale the patent gives for its own detachable quick‑disk device ("it can be replaced with other data equipment"; "compatibility … improved").

On claim 7 ("quick disk device … rotatable magnetic recording medium … detachably set")

Portable rotating magnetic media were commodity items by 1985 (floppy/quick‑disk drives). Given US 4,157,822 and US 3,997,873 teaching data‑bearing media read by the copier to control a job, and given the admitted use of the quick disk to carry display/control data, choosing a detachable rotating magnetic disk as the carrier is a design choice among known media (card, tape, disk) with predictable trade‑offs in capacity, reusability, and portability. Claim 7 adds no new cooperation with the erase function; it is the same save/reload operation as claim 1 with the medium specified.

On claim 3 ("means for calculating the position data")

Given US 4,215,929's masks "established arbitrarily and continuously" by control means and US 3,827,799's mechanically linked, platen‑referenced indicators, converting two sensed coordinates into a rectangle's position data is routine programming, not invention. (The specification's own flow — keys 30w store S1 and S2; key 30c or 30p selects inside/outside — is ordinary arithmetic.)

5. What cannot be used (and why this matters)

• The 25 "Cited By" documents are not prior art. Their priority dates run from mid‑1985 (US 4,724,464: 1985‑07‑05; US 4,707,110: 1985‑12‑27) through 1987 — all after the 1985‑02‑06 priority date. They therefore cannot be used under pre‑AIA § 102(a)/(b), and they cannot be § 102(e) art either, since § 102(e) requires the reference's filing date to precede the applicant's invention date. A rejection cannot lawfully be built on them, even though they are the art most similar on their face (e.g., the Minolta and Toshiba editing‑copier family). This is a trap worth stating explicitly, because a search result dump makes that list look like available art.
• DE 3603755 A1 and JPS 61‑180265 A are family members of the same invention (same 1985‑02‑06 priority), not prior art.
• References 6 and 7 of the "Citations (7)" list are unidentified in the material I was given, so I cannot rule in or out that the examiner had, e.g., a coordinate‑input/light‑pen or a memory‑per‑area reference among them.

6. Where non‑obviousness could still be argued (honest counterweight)

1. Element B — the transmission‑light spot indicator. No confirmed reference discloses a light source mounted on the scan carriage that shines through the original, movable perpendicular to the scan direction, so the operator locates the spot seen through the sheet. US 3,827,799's and US 4,215,929's light sources both act on the photoreceptor. The specification's own rationale — the spot is bright enough to pass through thick originals such as a postcard — is a concrete technical advantage, and the required data inversion (the original is designated face‑up, then turned face‑down for copying, so "data stored in memory 140 … is inverted in the column direction") is a real, non‑trivial implementation detail. A patentee could argue this combination was not suggested, and that the cited art's masks/platen indicators are a different design philosophy (mechanical masking vs. through‑image optical designation).
2. Real‑time coordination (element K). The disclosure's latch timing (θ1/ω) to load one rank of position data into the store register before the charged zone reaches the exposure region is a genuine synchronization requirement. Against this, however, US 4,045,218's erase‑before‑develop sequence inherently requires the same kind of registration to the rotating drum, so this is likely to be treated as an engineering detail within the skill level.
3. No secondary‑considerations record. I found no evidence of commercial success, long‑felt need, copying, or industry praise tied to these claims; absent a nexus, such evidence would not be available to rebut a prima facie case. Conversely, the patent issued over this art, which suggests the examiner perceived at least one non‑obvious feature — most plausibly element B, and possibly the specific save/reload architecture as claimed.

7. Conclusion

• Claim 1 would have been prima facie obvious over US 4,215,929 in view of US 3,827,799 and US 4,045,218 (area designation + selective latent‑image erasure), further in view of US 4,157,822 and US 3,997,873 (machine‑readable, re‑readable copier‑control data medium), with the patent's own admission that the copier already reads display/control data from a quick disk supplying the motivation to move the erase‑area coordinates onto that medium. The only un-met element is the through‑the‑original transmission spot light (element B); if that element is given patentable weight, claim 1 survives the combination as I can reconstruct it, and the case turns on whether references 6–7 (unidentified) or art outside the citations supplies a through‑original optical designation.
• Claims 3, 4, 5, 6, and 7 would each be obvious for the reasons in § 4; claims 5 and 6 in particular are a forced choice between the only two process positions for a latent‑image eraser, and claim 7 is a design choice of medium.
• Claim 2 would be obvious: US 4,215,929 already recites a spot/linear light source in a frame, and US 3,827,799 teaches a movable carrier (indicators on strings/pulleys) positioned along the platen — substituting a stepper‑driven carriage for the strings and a lens‑focused LED for the lamp is a predictable mechanical substitution.

Reliability caveats: (i) this is a reconstruction of the obviousness case, not the examiner's actual rejection — I did not retrieve the USPTO file wrapper or the Notice of Allowance reasons; (ii) two of the seven references of record are unidentified in the source text, so the real rejection may have relied on art I have not seen; (iii) all prior‑art characterizations above are drawn from the Google Patents/FPO/PDF text quoted in the sources linked, and where I have relied on the specification's admissions (e.g., the quick disk storing display and control data) I have said so expressly.