US 5426282

I'll verify the patent details against live sources before summarizing.

Search results — US 5,426,282

I searched for the exact number (no fuzzy matching) and did not return any other patent numbers. Below is what the authoritative full text (Google Patents, fetched 2026‑09‑24) and the corroborating sources say, distinguished from what I could not verify.

1. Bibliographic data

Uncertainty flags:

• The assignee record is unreliable as stated. Google itself disclaims accuracy, and pre‑AIA patents of this era often changed hands without appearing in the patent face. I could not confirm ownership from an assignment database (USPTO Assignment) in this session.
• I found no independent confirmation of the inventor's then‑current employer for this filing. Note only that Humble is a named inventor on the cited Checkrobot patent US4676343A, and the specification extensively discusses CheckRobot, Inc.'s US4964053 as the closest prior approach — but that does not establish assignee status for the '282.

2. Abstract (as granted)

"A retail checkout station for purchaser self-checkout is coupled to a service terminal with a video monitor directed at the checkout. The service terminal preferably services a number of checkouts. Product codes can be downloaded from a remote location to a checkout presented with a product lacking a product code barcode label, such as an item of bulk produce. A product database having product price information stored in a memory is normally indexed to product codes which are scanned in, and also is at least partly indexed by product identity information. The checkout station can have an electronic scale for determining the weight of a product. A video image of an uncoded product in the security zone is captured by a camera and displayed on a video monitor which is viewable by store personnel. A keyboard associated with the video monitor permits the store personnel to enter a product code corresponding to the product displayed on the video monitor. The service terminal function can be met by cashiers at idle checkout lanes or payment stations. The data processor accesses product price information from memory, and if necessary calculates a purchase price of the product by weight. The purchase price is added to the price of other consumer purchases for payment by the purchaser."

3. Plain-language overview of the independent claims

There are 10 claims; only claims 1 and 8 are independent (2–7 depend from 1/2/5; 9–10 depend from 8).

Claim 1 — remote video verification at the payment station (the core idea).
A customer-operated self-checkout system with a scanner, data processor, and memory that links scanned codes to prices and accumulates a transaction total. Its distinguishing elements:

• a video camera at the scanning checkout station that records and stores in memory a video image of products passing the scanner;
• at least one display at a payment station that is physically separate from the scanning checkout station, which retrieves and shows that stored image at the payment station;
• data input means (keyboard-type) coupled to the processor, letting staff enter an identity code for a product shown in the image that lacks a product code;
• crucially, the code may be entered remotely, after scanning is complete but before payment, and stands in for a scanned code in the running transaction total.

Plain English: the machine photographs the un‑barcoded item, and the human who identifies it can be sitting at a different station entirely — even after the customer has finished scanning.

Claim 8 — the bulk-produce self-checkout apparatus with weigh‑stop, camera, and payment-station lookup.
A system for self-checkout of products picked from bulk, comprising:

• a product database of price information;
• a checkout station with a security zone, a conveyor, and weighing means;
• a sensor that detects the product and temporarily stops the conveyor with the product on the scale;
• a camera capturing a video image in the security zone;
• memory storing the image for later retrieval;
• a video monitor at a payment station displaying the image to store personnel after the purchaser has exited the checkout station and entered the payment station;
• keyboard means for staff to enter the product code;
• a processor computing the purchase price from the identification code, the measured weight, and stored price data;
• means for transmitting the purchase price to the payment station.

Plain English: weigh-and-photograph at the lane, identify-and-price later at the register — the "offline/deferred" variant of the same concept.

Dependent claims, briefly:

• 2 — multiple networked checkout stations; the processor switchably routes any lane's stored image to the display and associates remotely entered codes back to the correct lane's transaction.
• 3 — multiple payment stations; the stored image can be displayed at any of them.
• 4 — the image is digitized before storage.
• 5 — customer-operable input at the lane for product identity/quantity, with remote staff override when what the customer entered does not match the displayed image.
• 6 — the customer input is a keyboard, a touch-sensitive screen, and/or voice-operated input.
• 7 — the customer input is a touch screen displaying product images for selection.
• 9 — the keyboard also accepts a unit count, and the processor uses that count in the price calculation.
• 10 — store personnel can selectively retrieve images from memory.

4. Notable context from the specification

The patent frames itself as solving an icon-selection problem: it criticizes the icon approach of US4964053 (described as application Ser. No. 07/185,167, abandoned, continued as Ser. No. 07/420,685, "owned by CheckRobot, Inc.") as limited by screen size and by visually similar items (specifically identifying "apples vs. plums, red grapes vs. concords"). It also describes three intervention modes — "real time," "off line," and cashier processing at the payment station — plus voice recognition (microphone 74 / recognizer 76) and, as optional prior-art-adjacent alternatives, pattern-recognition image processing.

5. Litigation / CAFC 2026 docket check

No CAFC 2026 docket entry, district court case, or IPR/PGR involving US 5,426,282 was located. I want to be explicit about the limits of that statement:

• My searches were general web searches, not direct queries of PACER, the CAFC docket, or the USPTO Patent Trial and Appeal Board API. A null result from web search is not conclusive of a null docket.
• The absence is nonetheless consistent with the record: the patent is marked "Expired – Fee Related" with an anticipated expiration of 1993‑08‑05 + 20 years = 2013‑08‑05. Even ignoring the maintenance-fee lapse, the statutory term had run before 2026, so it is not an enforceable basis for a 2026 infringement suit.
• Do not confuse this patent with unrelated matters whose search hits contained the word "Humble" (e.g., Kinnear‑Weed Corp. v. Humble Oil & Refining Co., a Fifth Circuit patent case with no connection to this patent).

6. Summary of what I could not verify

1. Any assignment/ownership chain beyond the grant face data.
2. Whether any litigation ever occurred during the enforceable life (1995–2013).
3. The primary examiner and any reexamination history for the '282 itself.
4. The exact forward-citation count (the given source self‑contradicts: 140 vs. 227).

If you need any of those confirmed, they would require a direct USPTO Patent Center / Assignment search and a PACER or CAFC docket query, which I could not perform with the tools available here.

I'll search for litigation involving US patent 5426282 across patent litigation databases.

Let me search more targeted litigation databases.

Let me try more specific litigation search sources.

Findings: Litigation involving US Patent 5,426,282

Based on my searches of patent litigation databases and related sources, I was unable to identify any litigation (district court infringement suits, ITC investigations, or Patent Trial and Appeal Board proceedings) in which US Patent 5,426,282 was asserted or challenged.

I want to be explicit that this is a negative result from a limited search, not a definitive confirmation that no litigation exists.

What I checked

• The Unified Patents patent portal entry for the patent — https://portal.unifiedpatents.com/patents/patent/US-5426282-A — which contains bibliographic data, prior art, and "Referenced By" citations, but the retrieved content did not surface any litigation record for this patent.
• Google Patents entry — https://patents.google.com/patent/[US5426282A](/patent/US5426282A)/en — which lists cited art, citing art, and family members, but no "Litigation" or "Cases" section (Google Patents surfaces those only when it has litigation data).
• General searches for the patent number, the title ("System for self-checkout of bulk produce items"), and the inventor (David R. Humble) in combination with litigation terms.

The only "case"-style matches for the string "5426282" that my searches returned were unrelated (e.g., an unrelated Brazilian corporate-registration number in a JUCE Ceará filing) and are not patent litigation. I am flagging these solely to note that I did not treat them as relevant.

Relevant context about the patent

• Patent: US 5,426,282 A, "System for self-checkout of bulk produce items"
• Inventor / original assignee: David R. Humble (individual; no corporate assignee listed on the face of the patent)
• Application: US 08/102,763; Filed: 1993-08-05; Granted: 1995-06-20
• Legal status: Expired – Fee Related; anticipated expiration 2013-08-05

This status matters for the practical question: the patent's 20-year term ran from the 1993-08-05 filing date, so any enforceable term would have lapsed around August 2013. That window, combined with the individual-ownership profile (rather than a known NPE/assertion entity), is consistent with — though not proof of — an absence of infringement litigation. The patent does appear frequently as prior art in later patents and in reexamination/IPR-type contexts where it is cited against other parties (e.g., it is cited in the WO2010000986A1 and WO2000026875A1 search reports, and referenced within various post-1995 patents). That is prior-art citation, not litigation involving the '282 patent itself.

Important caveats

1. Coverage limits. Unified Patents' portal, Google Patents, and general web results are not exhaustive litigation databases. Comprehensive litigation records live in Docket Navigator, Lex Machina, RPX Insight, and PACER, which I could not fully query here.
2. Timing. Because this patent predates the widespread electronic docketing era and expired in 2013, any early-1990s/2000s case may be poorly indexed in the sources I could reach.
3. Terminology risk. Cases could identify the patent by its title or by the parties' names rather than the number, which would not be caught by number-based searches.

Recommendation

Before treating "no litigation" as final, I would verify against at least one subscription litigation database (Lex Machina or Docket Navigator) and a PACER search for civil cases citing "5,426,282," plus a manual check of the USPTO Patent Trial and Appeal Board (PTAB) and the patent's USPTO PatentCenter/Assignment records for any recorded license, assignment, or suit notices. If you'd like, I can draft the exact PACER and Docket Navigator query strings to run.

Bottom line: I found no verifiable litigation involving US 5,426,282, so I cannot provide plaintiff(s), defendant(s), jurisdiction, case number, filing date, or outcome without fabricating them — which I will not do. The correct statement on the current record is that no litigation is known.

I'll verify the PTAB record against the structured data and run independent searches to catch anything the ODP ingest might have missed.

Let me run a few more targeted checks, including the near-miss patent numbers my first search surfaced (which are easy to confuse with the '282).

Let me run two final targeted checks to rule out CBM/PGR activity and any docket-indexed trial I might have missed.

I ran the canonical structured source plus independent web checks. Here is the result.

Proceedings overview

Total AIA trial proceedings on US 5,426,282: 0. The USPTO Open Data Portal (the canonical source in the "PTAB proceedings on file" block) returns no IPR, no PGR, and no CBM for this patent, and independent web searches surfaced nothing that contradicts that ingest. Breakdown: 0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denials — the patent has never been through an AIA trial.

Defensive posture for a defendant: the record is empty, and that emptiness cuts against — not for — a validity defense. There is no FWD here to lean on, no claim you can point to as already canceled, and no estoppel to exploit. If you are asserted against, you cannot say "claim 1 is dead." You would be the first to test these claims, which means you carry the full burden of building an invalidity case from scratch (or filing the first petition) rather than importing a prior outcome.

Two structural facts are worth stating up front because they constrain the options:

• PGR is unavailable as a matter of law. The '282 was filed 1993-08-05 — pre-AIA. Post-grant review requires an effective filing date on or after 2016-03-16 (the AIA first-inventor-to-file date, 2013-03-16). So no PGR path exists.
• CBM is no longer available. CBM's transitional program sunset on 2020-09-16, and in any event no CBM was ever filed against this patent.
• The patent's term ran out. Anticipated expiration was 2013-08-05 (Google Patents; consistent with the 20-year term from the 1993-08-05 filing), and it is marked Expired – Fee Related. An expired patent can still technically be the subject of an IPR, but the practical incentive is near zero when there is no live infringement exposure.

Proceedings

None. I will not list a proceeding that does not exist, and I will not invent a number. Per the constraints on this task, the correct output is that there is no AIA trial activity on file for US 5,426,282.

Because "no PTAB activity" is the answer, the more useful thing I can give you is the set of near-miss records that look like this patent but are not it — these are the exact confusions that could mislead someone researching this patent.

Near-miss 1 — US 5,245,282 (NOT this patent)

• What it is: "Three-Dimensional Magnetic Resonance Imaging," owned by the University of Virginia Patent Foundation.
• Why it's confused: The number transposes two digits versus US 5,426,282. It was litigated in University of Virginia Patent Foundation v. General Electric Co., No. 3:2008-cv-00025 (W.D. Va.) — https://www.courtlistener.com/opinion/[2471266](/patent/2471266)/university-of-virginia-patent-foundation-v-general-electric-co/ — and was the subject of an ex parte reexamination (a reexam certificate issued 2010-05-04).
• Not relevant as precedent for the '282: It is a different patent, a different technology (MRI pulse sequences), and the proceeding was an examiner-side reexam, not an AIA trial. It is an invention-rights/intervening-rights case (§§ 252, 307(b)), not a PTAB case. Do not cite it as PTAB precedent for this patent.

Near-miss 2 — US 6,245,282 (NOT this patent)

• What it is: A different patent that was the subject of IPR2017-00633, where the patent owner filed a preliminary response arguing discretionary denial under § 325(d).
• Why it's confused: It shares the "282" suffix. Source document: https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1479434](/patent/1479434)/download-documents
• Not relevant: Different patent number, different owner, different art.

Near-miss 3 — the "Humble" reference in IPR2013-00164 (NOT this patent)

• In Zynga Inc. v. Personalized Media Communications, LLC, IPR2013-00164 (US 7,797,717), the petitioner's reply repeatedly refers to a "Humble" prior-art reference describing a "product checkout system that generates a promotional message in response to a scanner signal." See https://www.docketalarm.com/cases/PTAB/IPR2013-00164/... .
• Why it's confused: The name "Humble" matches the inventor of our patent (David R. Humble). But that proceeding is about US 7,797,717, and the "Humble" cited there is a prior-art checkout document used as evidence, not US 5,426,282 as the patent under review. Do not treat IPR2013-00164 as a proceeding "on" the '282.

Near-miss 4 — the "5426282" string in non-patent data

• A web hit for the bare string "5426282" is a Brazilian corporate registration number in a JUCE Ceará filing (https://www.jucec.ce.gov.br/...). Completely unrelated. I flag it only so it is not mistaken for a docketed case.

There is one further genuine appearance of the '282 that is not a PTAB proceeding but is worth logging: US 5,426,282 is cited as prior art in later patents (e.g., it appears on the face of IBM's US 7,118,026 B2, "Apparatus, method, and system for positively identifying an item," listing "5,426,282 A * 6/1995 Humble ...... 235/383"). That is forward citation, not an attack on the '282. It is often misread in citation databases as "activity," but it confers no defensive value.

Strategic summary

Claim status. There is no PTAB-derived claim status to report. Nothing on US 5,426,282 has been canceled, sustained, or even tested at the Board. All ten claims — independent claims 1 and 8, plus dependents 2–7 and 9–10 — stand as granted. Anyone describing claims of this patent as "invalidated" is describing a different patent (most likely the 5,245,282 or 6,245,282 near-misses above).

Estoppel landscape. There is no § 315(e)(2) estoppel anywhere on this patent, because estoppel attaches only to claims addressed in a final written decision and no FWD exists. Practically, this means a defendant facing assertion has the full universe of prior-art grounds still available — § 102 and § 103 on patents and printed publications for an IPR, and every validity ground in district court. Nothing has been spent, and no petitioner has burned any art. The flip side is the one I noted above: you get no benefit from someone else's work, and you would be first.

Pattern signals. There is no pattern to detect. No petitioner has filed even one IPR against this patent, let alone multiple. There is no defensive aggregator (no Unified Patents, RPX, or similar) in the chain for this patent — I specifically checked the Unified Patents portal entry (https://portal.unifiedpatents.com/patents/patent/US-5426282-A), which shows bibliographic data, prior art, and a "Referenced By" list but no litigation or PTAB record. There is no patent-owner appeal activity, for the obvious reason that the owner never had a PTAB outcome to appeal. And the patent's own posture — individual ownership, no corporate assignee, expired since 2013, an empty assertion history — is consistent with a patent that was never commercialized as an assertion asset. That is a signal about risk, not a validity finding: it reduces the likelihood of a high-stakes assertion campaign, but it does nothing to strengthen or weaken the claims.

Recommended next steps

1. Treat "no PTAB activity" as the operative finding, but verify at the source. The ODP structured block is the canonical list, and web search corroborates it. If you need a belt-and-suspenders confirmation, query PTAB E2E / PTAB Center by the patent number and by "Humble" as patent owner, and check the USPTO Patent Center for the '282's reexamination history — the near-miss cases above show how easily a different "282" patent's reexam or IPR can be misattributed.

2. Confirm the enforceability posture before building any defense. The patent is marked Expired – Fee Related with an anticipated expiration of 2013-08-05. If a demand letter or complaint cites this patent, the first question is not validity — it is whether any enforceable term remains at all. On the current record, the answer is no; the statutory term ran out roughly a decade ago. That is a threshold, case-dispositive point that costs less to establish than an IPR.

3. If you nevertheless need an invalidity position, you are starting from zero — and that is actually favorable on economics. With no FWD and no estoppel, nothing is foreclosed. The five references cited on the face of the '282 are the natural starting point: GB 2161631 A, US 4,676,343 A, US 4,792,018 A, US 4,964,053 A (all Checkrobot Inc.), and US 5,083,638 A (Howard Schneider). Note the specification's own framing: it distinguishes the icon-selection approach of US 4,964,053 and argues the screen-size/similar-appearance limitation as the problem the '282 solves. That admission is itself worth mining for a § 103 obviousness theory, since the inventor characterizes the move to remote human video verification as a solution to a recognized limitation of a known system.

4. Do not claim a FWD disposition you cannot link. Because there is no FWD, there is no disposition to quote and no PTAB E2E or CourtListener opinion to cite for this patent. Any statement to a court or client that these claims were "upheld" or "canceled" at the Board would be unsupported — say instead that the patent has never been challenged in an AIA trial.

Bottom line for a defendant: US 5,426,282 has no PTAB history of any kind — no IPR, PGR, or CBM, and therefore no claim cancellations and no estoppel to leverage. The patent's expired status is the dominant fact and should be your first line of inquiry. If that fails, you retain the complete set of invalidity grounds with no § 315(e)(2) constraints, but you will be the first to develop them.

I'll research the assignment record for US 5,426,282 before drafting. Let me run targeted searches.

Let me dig into the assignment records and the inventor's corporate history.

Ownership & Assignment Analysis — US 5,426,282

Scope note up front. I could not execute the interactive query against USPTO Patent Assignment Search (assignmentcenter.uspto.gov / assignment.uspto.gov) with the tools available in this session — both are JavaScript applications that my search interface cannot drive. What follows is therefore built from the corroborating public record (Google Patents legal/assignee fields, Unified Patents portal, Justia/FPO assignee displays, the EPO/DOCDB family table, and the inventor's public biography), and I mark every inference as such. No reel/frame numbers are cited because I found none — and I am not going to invent any.

Inventors

Context and unusual patterns:

• Sole inventor, no co-inventors, no corporate assignee. For a 1993 supermarket-equipment filing by a career founder, this is the single most distinctive fact about the '282. Compare the Checkrobot, Inc. patents in the same field cited as prior art against it — US 4,676,343 (Humble, Gentzler +1, assigned to Checkrobot Inc.), US 4,792,018, US 4,964,053 — every one carries a corporate (73) assignee. The '282 carries none.
• Same-inventor "improvement filed personally" pattern. The specification's own target is US 4,964,053, "Self-checkout of produce items … owned by CheckRobot, Inc." Espacenet's cited-documents rendering of the sibling US 7,034,679 file lists Humble, David R. as inventor on US 4,964,053 as well. If that reading is correct, Humble is critiquing and improving on his own earlier Checkrobot-assigned work — but filing the improvement outside Checkrobot. I flag this as medium confidence; the Espacenet field alignment for that entry was ambiguous in the retrieved text.
• Same-day sibling filing. WO 1995/004491 A1, "Integrated automated retail checkout terminal," inventor Humble, David Raymond, carries the same 1993-08-05 priority date as the '282 (issued in the US as US 5,494,136, also Humble). So two Humble applications were filed the same day in his own name. Note the oddity: the EPO/DOCDB family table reproduced in the WO 2002063581A3 search report lists US 5,426,282's family member column as "NONE" — i.e., the '282 is recorded without the sibling it was plainly filed alongside.
• No "inventors departed within 12 months" signal to evaluate — with a single sole inventor and no corporate assignee, that fire-sale-precursor pattern is structurally inapplicable.
• Inventor deceased. Humble's obituary (Sun-Sentinel/Legacy.com, Fort Lauderdale) records his death on 2022-10-12, and describes him as having founded CheckRobot, Inc. (1985), Advanced Promotion Technologies Inc. (1987) and eDiets (1996), and as part of the original management team of Sensormatic Electronics. Any residual rights would have passed to his estate/heirs; no assignment to an estate or personal representative appears in any source I retrieved.

Original assignee

None — "Individual."

• The issued patent carries no (73) assignee. Google Patents records both Original Assignee and Current Assignee as "Individual" (with Google's own disclaimer that assignee listings "may be inaccurate"). Justia and FreePatentsOnline display the '282's reference row simply as "5426282 | June 20, 1995 | Humble" — with no assignee string, whereas neighbouring rows on those same pages do show assignees. Unified Patents lists an "Assignees:" field but did not surface a corporate name in the retrieved content.
• Did they ship a product embodying the claims? There is no operating assignee entity, so there is no product line and no corporate revenue associated with this patent. The named owner was a professional inventor/founder, not an operating company holding a product.
• Primary line of business of the owner: serial technology founder (retail checkout automation, electronic marketing, and later consumer internet). Note that the field was commercialised by others — Checkrobot/NCR/Optimal-Fujitsu-style self-checkout lanes — but there is no evidence the '282 itself was ever practised by, licensed through, or asserted by a Humble-controlled entity.
• Current status: individual owner (natural person), deceased 2022-10-12. The patent itself is Expired – Fee Related, nominal 20-year term running from the 1993-08-05 filing to 2013-08-05.

Assignment timeline

Plainly stated: I found no recorded assignment of US 5,426,282 — from any assignor to any assignee, at any time, in any source I could retrieve.

There is no reel/frame to list, no conveyance type, no correspondent, and consequently no correspondent-recurrence analysis is possible (there is nobody of record on an assignment document to recur).

Why the absence is meaningful rather than an artifact — with a caveat:

• The USPTO Assignment Search database covers all recorded patent assignments back to August 1980 (per USPTO's own Assignment Database documentation). A 1993-executed assignment, if recorded, would be in scope of the database. So this is not a pre-electronic-records gap.
• Every secondary aggregator that derives current ownership from the recorded-assignment feed (Google Patents, Unified Patents, Justia) shows the same thing: no assignee other than the individual Humble, no security agreement, no release, no change of name.
• Caveat (this is the honest limit): a recorded assignment can be missed by aggregators if the recording was late, corrected, or indexed only by application number rather than patent number. I therefore cannot elevate this to "confirmed zero." It is "no assignment found, and all available corroboration agrees."

Added observation — the fee lapse is itself evidence: the status "Expired – Fee Related" is Google's classification for a patent that lapsed for non-payment of a maintenance fee, not one that ran its full term. The '282 issued 1995-06-20, so its 3.5-, 7.5- and 11.5-year fee windows fell at roughly 1998-12, 2002-12 and 2006-12. I could not determine from the sources retrieved which window was missed — that requires the USPTO maintenance-fee event history in Patent Center. But the direction is unambiguous: a patent that lapses for non-payment before the end of its term has not been maintained as an assertion asset. Note the internal tension in the source text — Google simultaneously prints "Expired – Fee Related" and "anticipated expiration 2013-08-05"; the latter is a nominal 20-year calculation, not a finding that the patent stayed alive to 2013. Flagging this as a source-level inconsistency, consistent with the tension already noted in the prior "Patent summary" section.

Per the task instruction, this section is where a "no records" finding would normally terminate the enumeration. I have continued into the diagram and signals below because the requested verdict depends on them, and because the absence of records is itself the finding here.

Timeline diagram

timeline
    title Ownership record of US 5426282
    1993 : Filed by David R Humble as individual
         : Same day sibling filing by Humble
    1995 : Issued with no corporate assignee
    2013 : Nominal 20 year term would have ended
    2022 : Inventor David R Humble dies in Florida

There is deliberately no assignment event to plot — that is the result.

NPE / troll-pattern signals

Inverse signal observed (not one of the eight, but material): the maintenance-fee lapse indicated by "Expired – Fee Related" is affirmative evidence of non-assertion. An entity holding the patent for enforcement keeps the fees current; an individual who let it lapse did not.

Verdict

Insufficient data — applying the rubric's own definition ("no records, or only the original assignment").

Justification. There is no assignment record to analyse: no reel/frame, no conveyance, no assignee other than the sole inventor David R. Humble, and no correspondent to test for the repeat-attorney signal. Because the USPTO assignment database reaches back to August 1980, a 1993-executed assignment would be expected to appear if recorded — and none does in Google Patents, Unified Patents, Justia or FPO — but I could not run the authoritative Patent Assignment Search query myself, so I cannot certify a true null.

Practical read, stated separately from the verdict: every affirmative NPE indicator is absent, and two facts point away from assertion — the patent lapsed for failure to pay a maintenance fee ("Expired – Fee Related") and the individual owner died in 2022, with no recorded transfer to an estate or successor. On this record the patent was never weaponised; the "Insufficient data" verdict reflects an empty assignment record, not any suspicion of concealed transfers.

Verification queries to run against the authoritative source (I could not execute these):

1. https://assignment.uspto.gov/patent/index.html → Patent Number field: 5426282 (also try 5,426,282 and Application Number 08/102,763).
2. Same interface → Assignor Humble, David R. and Assignee Humble (wildcard), to catch any recording indexed only by name.
3. Google Patents legal events tab for US5426282A — https://patents.google.com/patent/US5426282A/en — to list any recorded assignment, fee-payment, or lapse event.
4. USPTO Patent Center maintenance-fee event history, to fix the actual lapse date (which fee window was missed).
5. USPTO Patent Assignment records for the same-day sibling US 5,494,136 / WO 1995/004491 (Humble, David Raymond), which may show a corporate assignee and clarify whether Humble retained rights personally or in fact held them through an entity for the 1993 filings.

Bottom line: US 5,426,282 was owned by David R. Humble personally from filing through expiration; no assignment of any kind is recorded; no NPE or defensive-aggregator chain exists; the patent lapsed for non-payment and its inventor died in 2022.

I'll verify the patent through USPTO sources and pull the details of each cited reference.

Let me confirm the two remaining cited references.

USPTO lookup — US 5,426,282

Searched the patent number exactly (5426282); no other numbers were substituted or auto-corrected. The record is corroborated by the Google Patents full text (fetched 2026‑09‑24), the USPTO‑derived page at https://uspto.report/patent/grant/[5426282](/patent/5426282), and the Unified Patents record at https://portal.unifiedpatents.com/patents/patent/US-5426282-A. Bibliographic data (title, inventor David R. Humble, app 08/102,763, filed & priority 1993‑08‑05, granted 1995‑06‑20, individual assignee, expired) matches the earlier sections — I won't repeat those tables.

One point of housekeeping that matters for the prior-art list: the authoritative full text shows two citation tables that disagree in count — a "Citations (4)" table (GB2161631A, US4676343A, US4964053A, US5083638A) and a "Patent Citations (5)" table that adds US4792018A. The complete set of examiner-cited patent references is therefore five, and all five are analyzed below. Non‑patent literature cited: 0 (Unified Patents).

The five cited references

All five predate the 1993‑08‑05 filing by more than a year, so each is available at minimum under pre‑AIA 35 U.S.C. § 102(b) (printed publication/patent more than one year before the U.S. filing date). Two of them (US 4,676,343 and US 4,792,018) name David R. Humble as a joint inventor — i.e., the '282 inventor's own earlier patents. Under § 102(b), an inventor's own patent that issued more than one year before the later filing date is still prior art against the later claims.

1. GB 2161631 A

• Full citation: GB 2161631 A, "System for security processing of retailed articles," Checkrobot Inc. (UK). Priority 1984‑07‑09; published 1986‑01‑15. https://patents.google.com/patent/GB2161631A/en
• Description: UK counterpart of the Checkrobot security-checkout family. Discloses supermarket self-checkout with a laser UPC reader (10), entry/exit conveyors (12/13), a secured zone or "security tunnel" (15), light curtains as an "entrance sentry," weight measurement/comparison against stored values, an interactive display screen (11), and a cashier station shared by a group of counters with a master monitor screen, keyboard, cash drawer and receipt printer. Version 1 compares article shape; later versions add a second in-tunnel code reader and EAS detection.
• § 102 anticipation assessment: Does not anticipate any claim of the '282. It supplies the self-checkout/scanner/processor/security-zone environment but discloses no video camera and no stored video image displayed remotely — the limitation that defines claim 1(d)–(f) and claim 8's camera/monitor elements. At most it is § 103 background art for the preamble of claims 1 and 8.

2. US 4,676,343 A

• Full citation: US 4,676,343 A, "Self-service distribution system," Checkrobot Inc.; inventors David R. Humble, David L. Gentzler, Steven J. Tilidetzke. Filed 1984‑07‑09; issued 1987‑06‑30. 32 claims, 11 figures. https://patents.google.com/patent/US4676343A/en ; PDF: https://patentimages.storage.googleapis.com/48/e7/90/f1ba03beb25762/US4676343.pdf
• Description: The foundational Checkrobot operator-unassisted checkout. Laser scanner reads UPC; article placed on entry conveyor, passes an entry light curtain into a security tunnel; a scale measures weight and the processor compares measured vs. stored expected weight; conveyor stops/reverses on mismatch; a touch-activated screen prompts the customer; on completion the customer receives a printed itemized list taken to a cashier for payment and a final receipt. EAS optional.
• § 102 anticipation assessment: Does not anticipate any claim. It discloses the "product code scanner coupled to a data processor and a memory … to accumulate a transaction total" limitations of claim 1(b), the conveyor/security-zone/scale architecture of claim 8, and a multi-lane arrangement relevant to the preamble of claim 2 — but it has no camera, no stored image, and no remote/payment-station image display or remote code entry. Claim 2's "switchably couple the stored video image … to the display" is absent. § 103 material, not § 102.

3. US 4,792,018 A

• Full citation: US 4,792,018 A, "System for security processing of retailed articles," Checkrobot Inc.; inventors David R. Humble, David L. Gentzler, Steven J. Tilidetzke. Continuation‑in‑part of Ser. No. 628,913 filed 1984‑07‑09; issued 1988‑12‑20. https://patents.google.com/patent/US4792018A/en
• Description: CIP that hardens the '343 system against customer fraud: article shape detection via the entry light curtain cross‑correlated with UPC; repeat UPC reading inside the tunnel and comparison to the scan; and EAS‑tag logic (whether a given UPC should carry a tag). Same scanner/conveyor/tunnel/scale/interactive-display/cashier‑receipt architecture.
• § 102 anticipation assessment: Does not anticipate any claim. Adds security features, not remote video identification. Same gap as the '343: no camera, no stored image, no payment‑station image display or remote code entry. Note this reference appears only in the five‑row "Patent Citations" table, not the four‑row table — a source inconsistency worth flagging, but it is a genuine member of the cited set.

4. US 4,964,053 A — the reference the '282 specification itself targets

• Full citation: US 4,964,053 A, "Self-checkout of produce items," Checkrobot, Inc.; filed 1988‑04‑22; issued 1990‑10‑16. (Named inventor field is Humble — the PDF header reads "United States Patent [19] Humble.") https://patents.google.com/patent/US4964053/en
• Description: Adds unlabeled-produce handling to the Checkrobot self-checkout. A touch-actuable video screen presents hierarchical produce icons (fruit → vegetables → tomatoes → beefsteak/cherry/sauce); the customer selects an icon, the article is conveyed to a weigh scale, and the processor computes price from the selected icon, stored price data and measured weight (touch‑screen/icon selection, weight‑range security checks). It also discloses a cashier station with a master monitor screen and keyboard at which the cashier can add items "that could not be handled automatically by the counter … almost certainly all produce items."
• § 102 anticipation assessment: Does not anticipate any claim standing alone, because it lacks the camera‑image‑to‑remote‑display element. It is nonetheless the semantically closest Checkrobot reference and would supply most non‑camera elements:
  • Claim 1 preamble + 1(a)–(b): self‑checkout of unlabeled products; scanner/processor/memory/total.
  • Claim 5, 6, 7: customer‑operable input (touch screen showing product images for selection) — but these depend from claim 1, so they cannot be anticipated without claim 1 being anticipated.
  • Claim 8 preamble + database, security zone, conveyor, scale, stop‑the‑product sensor, processor‑computes‑price limitations — but not the camera, stored‑image, video‑monitor‑at‑payment‑station, or keyboard‑at‑payment‑station limitations.
    This reference is exactly what the '282 specification characterizes as the limited-icon approach it improves upon ("limited by the size of the display screen and the limited number of produce icons" — e.g., "apples vs. plums, red grapes vs. concords").

5. US 5,083,638 A — the strongest § 102 candidate

• Full citation: US 5,083,638 A, "Automated point-of-sale machine," Howard Schneider (individual; not Checkrobot). Priority/filing 1990‑09‑18; issued 1992‑01‑28.
• Description: Multiple "robot modules" (checkout stations), each with a bar‑code scanner and a packing scale, plus a video camera; a supervisor module with a set of video monitors where a supervisory employee visually identifies produce or non‑validly‑coded items from the camera image and grants "video approval," with a produce area (optionally with its own scale) where unlabeled items are deposited for viewing. A later application (EP 0 498 311 A1) expressly describes the '638 as having "a set of video and camera monitors at the supervisor module" and quotes the problem that "the supervisory employee [must] distinguish produce items placed into the bag on the packing scale for identification."
• § 102 anticipation assessment: This is the only cited reference that discloses the core distinguishing feature of the '282 — a camera image of an un‑coded product viewed by store personnel who then enter/approve the item identity. It therefore is the reference potentially anticipating claims 1 and 8, provided that the following limitations are also met on the face of the '638:
  • claim 1(d)/(f): the display is at a "payment station apart from the scanning checkout station," and the code is entered "after completion of scanning and prior to payment";
  • claim 8: a "video monitor means disposed at a payment station" showing the image "after the purchaser has exited the checkout station and entered said payment station," plus keyboard means at that monitor and "means for transmitting the purchase price to [the] payment station."
    Whether the supervisor module of the '638 is a "payment station" in the claim sense (vs. a monitoring/approval station) and whether its approval timing maps onto the claim's post‑scan/pre‑payment window are the points on which anticipation would rise or fall. Because I could not retrieve and read the full '638 specification in this session, I am flagging this as a candidate, not a confirmed § 102 hit. If the '638 lacks either the payment‑station location or the post‑scan/pre‑payment timing, it falls back to § 103 art in combination with the Checkrobot family, rather than anticipatory art. This is the reference I would read in full first.

Mapping summary (claims potentially anticipated under § 102)

Important caveats and what I could not verify

1. These are "references considered," not a finding of anticipation. All five appear in the granted patent's citation list, and the '282 issued with all ten claims intact over them. That means the examiner did not ultimately sustain a § 102 or § 103 rejection, or any rejection was overcome. I do not have the file wrapper / Office actions in this session, so I cannot tell you which of these, if any, generated an actual rejection or whether the claims were amended to avoid one. Treat the table above as potential § 102 relevance, not as a record of what happened at the USPTO.
2. Anticipation is claim‑as‑a‑whole. Under § 102 a single reference must disclose every limitation. Because claims 2–7 and 9–10 are dependent (and therefore incorporate all limitations of claim 1 or 8), none of them can be anticipated by a reference that fails to anticipate claim 1 or 8 — even though US 4,964,053 arguably supplies the touch‑screen/icon and customer‑input elements of claims 5–7.
3. Unverified item: I could not retrieve the full text of US 5,083,638 in this session (I was working from the citation metadata and the later EP 0 498 311 A1 description of "Schneider '638"). The description of its camera/supervisor‑module architecture is therefore second‑hand. Before relying on the § 102 candidate status of claims 1/8, read the '638 specification directly (https://patents.google.com/patent/US5083638A/en) and confirm the "payment station" location and the post‑scan/pre‑payment timing.
4. Direct USPTO systems were not queried. I used Google Patents' rendering of USPTO data and the USPTO‑derived uspto.report page; I did not query Patent Center/Patent Public Search or the Assignment database directly, so I cannot confirm examiner identity, examiner's reasons‑for‑allowance, or any reexamination history of the '282.
5. A note on the disclosure's own prior‑art framing: the '282 body text does not mention Schneider '638 by name — it names only the Checkrobot '053 (as Ser. Nos. 07/185,167 / 07/420,685) and the general icon approach. That asymmetry (the applicant discusses the Checkrobot art but not Schneider) is worth noting, but assigning any legal significance to it would be speculation.

If you want, I can next pull the full text of US 5,083,638 and do a limitation‑by‑limitation § 102 chart against claims 1 and 8, and separately chart US 4,964,053 against claim 8's apparatus elements.

I'll ground the analysis in the actual content of the five references listed in the Prior Art section before building the §103 combinations.

Obviousness Analysis — US 5,426,282 under 35 U.S.C. § 103

Scope note: This analysis uses only the five references in the "Prior Art" section of the record you supplied — the Patent Citations (5) table: GB2161631A; US4676343A; US4792018A; US4964053A; US5083638A. (The page also contains a "Citations (4)" table that omits US4792018A — a self-contradiction I am flagging and resolving in favor of the 5-reference set, since '018 appears in both the "Patent Citations (5)" table and the machine-readable citation data.) All five issued/published more than one year before the 1993‑08‑05 filing date (the most recent, US5083638A, issued 1992‑01‑28), so all qualify as pre‑AIA § 102(b) art usable in a § 103 combination. Note that § 103(c) does not remove them, even though US4964053A shares the inventor (Humble) with the '282 — § 103(c) only reaches § 102(e)/(f)/(g) art, not § 102(b) patents/printed publications.

Caveat on sources: I verified the disclosure content of the five references from the fetched full texts listed below. Where I rely on a reference's own text, I quote it. Where I infer, I say so.

1. The five references, as verified

2. Level of ordinary skill (POSITA)

A POSITA here is an engineer with ~2–3 years in retail point-of-sale/self-checkout design, familiar with: UPC scanning, conveyor/light-curtain security tunnels, legal-for-trade scales, PLU lookup tables, video/CCTV, and store-wide POS networks (a controller serving multiple lanes and a cashier station). All the asserted combinations are within that person's ordinary design repertoire.

3. Claim 1 — element mapping (primary combination: '053 + '638, secondarily '343/'018/GB'631)

Literal-text note: the granted claim contains typographical errors ("stored in memory a video image," "can bee entered") and the specification contains others ("throught," "six naval oranges"). Per instruction I have not corrected them; they do not affect the elements' scope but are worth noting for any claim-construction dispute.

Why a POSITA would combine '638 with '053: both address the same problem — non‑UPC produce in an operator-unattended lane — and '638's own front matter lists a Checkrobot patent (US4,792,018) among its references, evidencing that the art was treating these as a single line of development. '638 expressly frames its advance as improving handling of "non-coded products." Substituting a transmitted photograph of the actual item for (or as a verification of) a self-selected icon is a known technique applied to a known problem, with a predictable benefit — the '282's own specification concedes the motivating defect of the icon approach ("apples vs. plums, red grapes vs. concords").

4. Claim 8 — element mapping

Claim 8 is the "weigh-and-photograph now, price later at the payment station" variant. Every element is supplied by the same cluster:

• Product database / price information — '053 (price data store 24); '018 (CPU 44 with price storage).
• Checkout station with a security zone — '018 secured zone 15; GB'631 tunnel 15.
• Conveyor means — '343, '018 (entry/exit conveyors 12/13).
• Weighing means disposed for determining a weight of a product in the security zone — '018 weight sensor 43 beneath entry conveyor 12; '053 weigh scale 22.
• Sensor means for detecting the product… and temporarily stopping the conveyor… when the product is disposed on the weighing means — '018/GB'631 entrance sentry light curtain 33 + controller for selective conveyor movement (accept/reject), i.e., stopping/reversing the belt on a sensed event.
• Camera means… capturing a video image of the product in the security zone / memory means storing said image for later retrieval — '5083638 camera and image signal, with storage being the conventional handling step.
• "video monitor means disposed at a payment station… after the purchaser has exited the checkout station and entered said payment station" — '638's supervisor module (customer walks there to pay) and '053's cashier register station serving a group of counters, with master monitor screen 22 and keyboard 23.
• Keyboard means… to enter a product code — '638; '053.
• Processor means operating on identification, weight, and price information — '053 (weight × stored price per unit weight); '018 CPU.
• Means for transmitting the purchase price to the payment station — '343/'018 ("issues a receipt to the customer for use in payment at a cashing station"), '053 (final sales slip printer and cashier station display).

5. Dependent claims 2–7, 9, 10

• Claim 2 (multiple lanes networked; processor switchably couples a lane's stored image to the display and associates remotely entered codes to individual lanes' transactions): '053 expressly discloses this architecture — one cashier station serves a group of counters, and each counter's coded identifier plus its subtotal appear on the master monitor screen, so the cashier's entries are associated with the correct lane. '638's supervisor-module/supervisory-employee relationship supplies the multi-lane supervision concept. Very strong.
• Claim 3 (multiple payment stations; image displayable at any): a mere duplication of known elements ('053's "each group having two or more counters" with a cashier station per group; '053 states the central processor "may form part of a local, network or remote computer").
• Claim 4 (image digitized before storage): design choice/known technique with predictable result; the '282's own specification treats digitization as routine.
• Claim 5 (customer input at the lane; remote staff code entry when the customer's entry "do[es] not match the video image"): essentially disclosed — '053's comparison of actual weight to a predetermined range "for detecting customer selection of incorrect icons," and '638's presentation of "an image of the bag's new contents along with the customer's identification… to the supervisory employee for verification." Very strong.
• Claim 6 (keyboard, touch screen, and/or voice input — note the "at least one of" transitional language): '053's touch-actuable screen alone satisfies the claim; '638's supervisor-module/user interfaces and its bidirectional intercom supply the rest. Voice input is a permissive alternative that need not be met to infringe.
• Claim 7 (touch screen displaying product images for selection): directly disclosed in '053 (icon store 25 + touch-actuable video terminal 11; '053 claims 3, 6, 7).
• Claim 9 (keyboard also enters unit count; processor uses it): '638 — the experienced-customer menu lets the customer "identify the product and optionally identify its quantity"; '053 contemplates piece-priced produce.
• Claim 10 (store personnel can selectively retrieve images from memory): weakest element. '053's master monitor screen (cashier retrieves a lane's data on demand) and '638's supervisory review support it, but neither reference expressly describes on-demand retrieval of an archived image keyed to a past transaction. This claim would depend on treating "selective retrieval" as the conventional capacity of any image database — arguable, not airtight.

6. KSR rationales to plead

1. Known elements, known method, predictable result — camera + remote monitor + keyboard (in '638) plus the multi-lane cashier station with subtotal reconciliation (in '053/'343/'018/'GB'631).
2. Simple substitution of one known element for another — item image reviewed by a human in place of the icon the customer tapped, both used for the same purpose (identifying non-coded produce).
3. Known technique to improve similar devices in the same way — deferred, off-line completion of a self-checkout transaction at a cashier station was already standard in the Checkrobot family.
4. Design incentive / market pressure — the field was explicitly driven by labor cost ("a limited number of operators" serving many lanes; '638's stated goal of requiring only ~a second of supervisory attention).
5. Express teaching in the art — '638 literally instructs sending the image of an un-coded item to a supervisor display whose keyboard is used to key in the product number, and locates payment at that same module.
6. Reasonable expectation of success — no new mechanism is required; all signals (video, weight, PLU) and the network were already in the references.

7. Where the § 103 case is weakest (be honest)

• The storage/retrieval limitation is the crux. Claim 1 requires the image be "record[ed] and stored in memory" and later retrieved at the payment station; claim 8 requires storage "for later retrieval" and display "after the purchaser has exited the checkout station." '638 discloses contemporaneous image transmission, not archival storage and later retrieval. A challenger must argue that storing the camera signal in memory is conventional (supported by claim 4's own framing and by the specification's statement that the image is "either stored temporarily in a memory 54 or transmitted immediately to a video monitor 58"). Counter-risk for the patentee: that very "either/or" sentence in the '282 shows the inventor regarded storage vs. immediate transmission as alternatives, which weakens the storage limitation as a point of novelty.
• Nothing in the cited set expressly keys a stored item image to a lane/transaction record. Claim 2's "switchably… associate" element is supported by '053's lane-ID/subtotal-on-master-monitor teaching, but the image-keying link is an inference.
• Claim 10 rests on the thinnest support (see above).
• Procedural posture: the examiner had all five references and still allowed the claims (granted 1995‑06‑20), so a challenger must show the Office failed to appreciate a teaching — most plausibly by pointing to '638's beginner-customer "no bar code" image-to-supervisor flow, which on its face reads on claim 1's remote-code-entry element.

8. Secondary considerations

No objective evidence (commercial success, long-felt need, industry praise, copying, licensing) is of record in the materials I have. The '282's own specification is an admission that the icon-selection approach was known and that the problem was its accuracy limits — which cuts against the patentee on nexus arguments. I could not verify any commercial or licensing activity; treat any such assertion as unverified.

9. Bottom line

• Claims 1, 5, 8, and 9 are the strongest candidates for invalidation via '5083638 (Schneider) in view of '4964053 (Humble/Checkrobot), further in view of '4676343/'4792018/GB2161631 — the references supply every element, and '638 provides the express teaching of sending an un-coded product's image to a separate, pay-at module where an employee keys the product code. Confidence: moderate-to-high, contingent on resolving the storage/retrieval limitation as a conventional design choice.
• Claims 7 and 2/4 are the most vulnerable to a standalone or near-standalone attack on '4964053 (icons on a touch screen; single cashier station with master monitor and lane-ID subtotals) plus routine digitization. Confidence: high.
• Claim 10 is the least vulnerable on the cited record. Confidence: low.
• One alternative theory the record does not address but which is worth a separate check: whether '282 and '4964053/'4676343 were commonly owned (Checkrobot vs. "Individual") at the relevant times, which would open a nonstatutory double-patenting theory independent of § 103. I could not verify any assignment chain — the earlier sections already flagged that the assignee data for '282 is unreliable.

Identifiers were interpreted literally throughout; I did not auto-correct the patent number, the claim typos, or the citation-table inconsistency.