US 7657333

I'll search for this specific patent number and any 2026 litigation activity.

Let me check specifically for any 2026 court activity and confirm the claim set.

US 7,657,333 B2 — Analyst Summary

Caution up front: my searches surfaced no 2026 CAFC docket activity and no litigation or PTAB proceeding involving US 7,657,333. I could not run a direct USPTO PatentCenter / PACER / CAFC CM-ECF query in this session, so treat the litigation finding as "nothing found," not as certified absence. One search result mentioned a "'733 patent" in an Ameranth CBM proceeding — that is a different patent (Ameranth's own '733), not US 7,657,333. I did not conflate the two.

Bibliographic data (from the patent document provided)

Abstract (verbatim)

Systems and methods that vary multiple data sampling rates, to collect sets of data with different levels of granularity for an industrial system. The data for such industrial system includes sets of data from the "internal" data stream(s) (e.g., history data collected from an industrial unit) and sets of data from an "external" (e.g., traffic data on network services) data stream(s), based in part on the criticality/importance criteria assigned to each collection stage. Each set of data can be assigned its own unique data collection rate. For example, a higher sample rate can be employed when collecting data from the network during an operation stage that is deemed more critical (e.g., dynamic attribution of predetermined importance factors) than the rest of the operation.

Independent claims in plain language

Claim 1 — system. An industrial automation system with two cooperating parts: (a) a coordination component that synchronizes multiple data streams using a common timestamp or sequence-count generator, where each stream has its own data sampling rate defining how fast data is pulled from it; and (b) a sample rate adjustment component that changes the sampling rate of at least one stream based on either a predetermined criterion or a predicted required sampling rate. Essentially: keep streams time-aligned, then selectively re-rate them.

Claim 13 — method. A method of collecting data in an industrial plant: collect data from multiple internal and external streams; maintain a sequence relationship among them using a common time-stamp generator or common sequence-count generator; and adjust the collection rate for at least one of those streams based on a predetermined criterion. This is the method counterpart to claim 1, adding the internal/external distinction explicitly.

Claim 20 — means-plus-function system. Same functional scope as claim 1, drafted as "means for synchronizing…" plus "means for adjusting…" the sampling rate based on a predetermined criterion or predicted required rate.

Notable dependent claims: internal = device data / external = network traffic (cl. 2); recognition component for event-cause trends (cl. 3); "weaving" streams together (cl. 4); estimation component using an analytical model to predict the rate (cl. 5); fault detection or granularity level as the criterion (cl. 6); matching component subscribing industrial zones to triggering events (cl. 7); zone-by-zone rate variation (cl. 8); AI component inferring when/where/how to vary the rate, with probabilistic/statistical model, prior data, and user-feedback adjustment (cl. 9–11); rate increase on fault criteria (cl. 12); function blocks and triggering events tied to granularity levels (cl. 14–15); user-feedback rate adjustment (cl. 16).

Analytical observations

1. Title vs. claim scope mismatch. The title stresses "anomaly detection," but the claimed triggers are broader — "a predetermined criteria" or "a predicted required sampling rate," with fault detection appearing only as a dependent claim (6, 12, 18). Anomaly detection is the motivational use case, not a claim limitation.
2. The inventive hook is synchronization + variable rate, not data collection per se. Collected history data and network-traffic data were known; the asserted novelty is coordinating streams on a shared time-stamp/sequence-count basis (IEEE 1588 is mentioned in the spec) and then varying per-stream granularity.
3. Broadest claim (1) is relatively thin. It requires only two functional components; the "predetermined criteria OR predicted rate" disjunction is permissive, which could matter in any invalidity/infringement analysis.
4. Spec detail not in the claims. The embedded network traffic analyzer (Fig. 3, ASIC implementation, filter components), the unified repository/selective-decay concept, and the multi-module routing of rate adjustment (Fig. 9b) are described but largely unclaimed.

Uncertainty flags

• The bibliographic data above comes from the patent document supplied to me and secondary aggregators (Google Patents, Justia, FreePatentsOnline, Espacenet, RPX). I did not independently authenticate it against USPTO PatentCenter; assignee/status fields are stated by Google Patents to be assumptions.
• Litigation/PTAB/CAFC status for this specific number is unverified. My searches returned nothing on point; absent a direct docket pull, I cannot certify that no 2026 proceeding exists.
• The Ameranth "'733" references in search results are a different patent — disregard them for this number.

I'll search for litigation involving US patent 7657333 specifically.

The initial results are dominated by different patents that merely share the "733" suffix (Ameranth's 6,982,733 and Headwater's 8,406,733). Let me run more targeted searches to isolate US 7657333 itself.

Search Results: Litigation Involving U.S. Patent No. 7,657,333

No litigation involving U.S. Patent No. 7,657,333 was found.

What I searched

I ran multiple targeted queries (Google Patents, USPTO/ptacts, CourtListener, Unified Patents-style aggregators, and general litigation coverage) for the literal number "7657333" / "7,657,333" combined with litigation terms (lawsuit, complaint, "case no," district court, infringement).

Key finding — do not confuse this patent with look‑alike numbers

Nearly all search hits containing the string "733" refer to different patents that merely share a numeric suffix. Per your strict rule, I am interpreting these literally and flagging them as not US 7,657,333:

The Ameranth '733 (6,982,733) and Headwater '733 (8,406,733) are unrelated to your target and involve entirely different technologies (wireless/restaurant ordering and wireless-device policy management, respectively). I am not reporting them as litigation against US 7,657,333.

About US 7,657,333

• Title: "Adjustment of data collection rate based on anomaly detection"
• Patent no.: US 7,657,333 B2
• Application no.: 11/862,983; filed Sept. 27, 2007; granted Feb. 2, 2010
• Assignee: Rockwell Automation Technologies, Inc. (Mayfield Heights, OH)
• Inventors: Bradford, Siorek, Gach, Balewski, Kretschmann, Harris, Hall, Rischar
• Status (per Google Patents): Active; adjusted expiration listed as 2028‑02‑22

I found no docket entries, complaints, ITC § 337 investigations, IPR/PGR petitions, or CAFC/PACER records naming US 7,657,333 as a patent-in-suit. Notably, Rockwell's known enforcement activity (e.g., the Solaia '318 disputes, Rockwell v. Beckhoff, MagneMotion v. Rockwell, and ITC Inv. No. 337‑TA‑1020) involves different patents — none of which is the '333 patent.

Confidence / caveat

• High confidence that the widely-circulated "'733 patent" litigation stories are about 6,982,733 or 8,406,733, not 7,657,333 — the numbers were verified against the complaint/PTAB text.
• Moderate confidence that no litigation exists at all for 7,657,333. My searches surfaced no positive match, which is consistent with an older (2010) industrial-automation patent that has not been asserted. However, I could not verify against a paid/comprehensive docket database (e.g., a full Docket Navigator or Lex Machina run), and free litigation portals can lag or omit smaller cases.

If you need a definitive negative, the authoritative checks would be: (1) a Lex Machina / Docket Navigator query filtered to patent number 7,657,333; (2) PACER/RECAP full‑text search on CourtListener for "7,657,333"; and (3) the USPTO Patent Center "Litigation" tab for that patent. I did not have authenticated access to those proprietary dockets in this session, so I am expressly stating I cannot rule out a case that isn't indexed by the free sources searched.

I'll start by verifying the canonical PTAB list against web sources, then check for any recent filings the ODP ingest may have missed.

Let me run a couple more targeted checks before concluding.

Proceedings overview

Total AIA trial proceedings on U.S. Patent No. 7,657,333: 0. The USPTO Open Data Portal (ODP) API returns no AIA trial proceedings for this patent as of the most recent ingest, and my independent web searches for IPR / PGR / CBM petitions, institution decisions, Final Written Decisions, and Federal Circuit appeals naming this patent returned nothing on point. Breakdown by status is therefore trivially: 0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denials.

Bottom-line defensive posture for a defendant: the patent is untested and unhardened rather than "survived two IPRs." No petitioner has ever put claims 1–20 at risk before the Board, so there is no estoppel (§ 315(e)(2)) and no Board claim construction or validity ruling to lean on — but equally, there is no neutralizing FWD you can wave at opposing counsel. A defendant facing a demand letter today is litigating against a fully intact 20-claim patent whose validity has never been adjudicated administratively.

---

No proceedings to report

There are no proceeding entries to list. Per the operating constraint do not invent proceeding numbers, I am not fabricating placeholder ### blocks. The correct output for this patent is a null set.

For completeness, here is what I checked and what I ruled out:

Look-alikes I excluded (do not confuse these with US 7,657,333)

• Ameranth's "the '733 patent" = U.S. Patent No. 6,982,733 (wireless restaurant ordering), litigated in Ameranth v. OpenTable and subject to a CBM proceeding. Different patent, different owner, different technology.
• Headwater Research's "the '733 Patent" = U.S. Patent No. 8,406,733, asserted in Headwater Research LLC v. [Samsung Electronics Co.](/litigations/by-defendant/Samsung%20Electronics%20Co.), No. 2:23-cv-00103 (E.D. Tex.). Different patent.
• IPR2015-001733 — this is a proceeding number, not a patent number; it concerns U.S. Patent No. 7,976,838. Its trailing digits invite confusion with 7,657,333; they are unrelated.
• Rockwell portfolio-adjacent patents — US 7,933,666 ("Adjustable data collection rate for embedded historians") and US 7,908,360 ("Correlation of non-times series events in industrial systems") share inventors and subject matter with the '333 patent. My searches surfaced no PTAB proceedings against those either, but I did not exhaustively verify them, and they are not the patent at issue here.

---

Strategic summary

Claim status: everything is UNTESTED. All 20 claims remain in force as issued: independent claims 1 (system), 13 (method), and 20 (means-plus-function), plus dependents 2–12, 14–19. Not one claim has been canceled, disclaimed, disclaimed-by-amendment, or held unpatentable — because no one has ever filed. Contrast this with the "hardened patent" scenario: there is no FWD to cite, no Board construction of "coordination component," "common timestamp or sequence count generator," or "predetermined criteria," and no IPR-driven narrowing to argue the patent owner has conceded. The claim text I confirmed from the published record (Justia, Espacenet, uspto.report) matches the grant, and per the earlier analysis in this file the "anomaly detection" hook lives in the title and specification, not in the independent claims — a point that cuts in your favor on § 112 written-description/enablement theories but does nothing for you until someone actually runs an IPR.

Estoppel landscape: clean slate. Section 315(e)(2) estoppel attaches only to a petitioner that obtains an instituted IPR and only as to grounds raised or that reasonably could have been raised. With zero instituted proceedings, no party is estopped, and every prior-art ground remains available — including the references already of record (Dombrosky 5,099,449; Murphy 5,764,927; Shaw 5,815,689; Duan 6,728,584; Siorek 6,732,287; Zocchi 2004/0174818; Thong 2006/0200743; Raymond 2007/0260410; Angrilli 2008/0107336; Ridel 2008/0205499). Note that several of these were cited on the face of the patent, which raises a § 325(d) risk if you petition on them alone — the Board routinely exercises discretion to deny where the same art and arguments were previously before the Office. New art is the better path.

Filing-window reality check. This is a pre-AIA patent (filed 2007-09-27, issued 2010-02-02), which closes two doors:

• PGR is unavailable — post-grant review reaches only first-inventor-to-file patents issued from applications filed on or after 2013-03-16.
• CBM is unavailable — the transitional program sunset on 2020-09-16, and independently, an industrial-automation data-collection patent would almost certainly qualify as a "technological invention" excluded from CBM eligibility under AIA § 18(d)(2).

That leaves IPR as the only live vehicle, and it remains available to a defendant who acquires standing by being served with a complaint — subject to the § 315(b) one-year bar running from service. Absent service, there is no § 315(b) clock and no standing problem for a petitioner filing defensively.

Pattern signals. None to report: no repeat petitioner, no serial filings, no joinder activity, no defensive aggregator (Unified Patents or similar) in the chain. That is consistent with the earlier finding in this file that the patent has never been asserted — and Rockwell's known enforcement campaigns (the ITC § 337 action against Advantech and 3S-Smart over a different eight-patent set including the '226, '817, '960, '225, '704, '196, '585, and '800 patents; the Solaia '318 disputes; Rockwell v. Beckhoff) have never pulled the '333 patent in. Well-asserted patents attract IPRs; this one has not been asserted, so the absence of IPRs is expected rather than anomalous.

---

Recommended next steps

If you are the defendant and have been served: the absence of any PTAB record means you are free to file the first IPR against this patent. Practical checklist:

1. Confirm the patent is live before spending. Google Patents' legal-events record for US7657333B2 shows a maintenance-fee payment at the 12th year, large entity, and an adjusted expiration of 2028-02-22. That means the fee was paid and the patent is in force, so there is real runway to litigate — but verify current fee status on USPTO Patent Center rather than relying on the aggregator.
2. Pull the prosecution history for § 325(d) exposure. The file history (Attorney Docket via Turocy & Watson LLP; Primary Examiner Ramesh B. Patel) will tell you which references were considered and argued. Art that was squarely before the Examiner invites discretionary denial; art that was not is your strongest ground.
3. Build the § 103 case around the synchronization limitation, not "data collection." The earlier analysis in this file is right that claim 1's inventive hook is the combination of (a) stream synchronization via a common timestamp or sequence count generator and (b) per-stream variable sampling rate. The specification expressly points to IEEE 1588 synchronization and to embedded historians plus an embedded network traffic analyzer sharing a common timestamp/sequence counter (370). Prior art showing synchronized multi-stream capture with independent per-stream sample rates — e.g., synchronized data-acquisition or network-capture systems — is where the § 103 argument should center.
4. Weigh the redundancy of claims 2, 13, and 20. Claim 13 is the method counterpart to claim 1 with the internal/external distinction made explicit; claim 20 is the means-plus-function clone of claim 1 (which will be construed under § 112(f) against the corresponding structures in the spec). A well-constructed petition should address all three independents, and SAS v. Iancu means the Board will institute on all challenged claims or none.
5. If instead you are the patent owner: treat the absence of IPR activity as durable goodwill you should not squander. A first assertion of this patent will likely draw a first-ever IPR petition within a year of service, and a null PTAB record means there is no prior Board endorsement of any claim to point to in a Fintiv-style discretionary-denial argument. Any institution decision deadline, oral hearing, and FWD (statutory one year from institution, extendable to 18 months for good cause) would all start from zero.

Live monitoring links (for the "nothing found" conclusion to be re-verified before any filing or response deadline):

• USPTO PTAB E2E / Patent Trial and Appeal Board End-to-End: https://e2e.uspto.gov (search by patent number 7657333)
• USPTO PTAB Decisions: https://www.uspto.gov/patents/ptab/decisions
• CourtListener RECAP full-text and CAFC docket search: https://www.courtlistener.com
• Google Patents record for this patent: https://patents.google.com/patent/US7657333/en

Uncertainty flags

• The null result is well-supported but not certified. The ODP ingest is the canonical source and it is empty; my web searches independently confirm it. However, I could not run an authenticated Docket Navigator or Lex Machina query, and free portals can lag newly filed petitions by weeks. Before relying on "no IPR exists" in a filing, re-run a PTAB E2E search by patent number on the day you rely on it.
• I did not locate any proceeding, so I make no assertions about judge panels, petition grounds, institution reasoning, FWD claim-level dispositions, settlement terms, or appeals. None exist to describe. Any such detail in this output would be fabricated.
• Pre-existing uncertainty from the earlier sections stands: the earlier summary noted the patent's "Active / 2028-02-22 adjusted expiration" status is stated by Google Patents to be an assumption, and that no litigation has been identified for this patent. The "no IPR" finding here is consistent with, and reinforces, the "never asserted" finding there. No contradiction between sections.

I'll research the assignment record for US 7,657,333, starting with the USPTO Assignment Center and cross-referencing sources.

Let me search more specifically for the assignment record details and cross-reference sources.

Let me try to access the specific assignment records and Patent Center data.

Inventors

Eight named inventors, all appearing (by residence and by the assignee relationship disclosed on the face of the patent and its EP family member EP2042960A/B) to have been Rockwell Automation employees at the time of filing. All list Ohio residences clustered around Rockwell's Mayfield Heights, OH headquarters:

Pattern notes:

• This is a classic large-company inventor team, not a lone-inventor/individual-owner patent. The residence addresses are corroborated by the EP family publication (EP2053786A3 / EP2042960 family), which lists the same inventors with the same Ohio addresses under applicant Rockwell Automation Technologies, Inc., Mayfield Heights, OH 44124.
• No inventor-departure signal. There is no recorded post-filing assignment from any individual inventor, and no indication any inventor took a personal ownership interest that was later separately conveyed. The rights flowed to Rockwell at/or near filing (recorded 2007-12-05), the standard corporate-assignment pattern.
• Family-cluster signal. The same core team (Bradford, Siorek, Gach, Kretschmann, Harris, Hall, Rischar, Balewski) filed a cluster of related applications on consecutive days in late September 2007 — e.g., application 11/863,449 ("Visualization of Non-Time Series Events," pub. 2009/0089xxx, Sept. 28, 2007) and 11/863,510 (issued as US 7,908,360, "Correlation of Non-Times Series Events in Industrial Systems") — all assigned to Rockwell. That is a productive internal IP harvest, not a fire-sale precursor.
• ⚠️ Data limitation: I could not retrieve each inventor's employment contract or a record of any of them leaving Rockwell within 12 months of filing. No such departure is evidenced in any source I could reach; absence of a departure record is not proof of continued employment.

---

Original assignee

Rockwell Automation Technologies, Inc. (a Rockwell Automation holding/IP subsidiary), Mayfield Heights, OH; original and sole assignee on the issued patent.

• Primary line of business: industrial automation and information — programmable logic controllers (Logix/ControlLogix, PLC‑5, SLC‑500, MicroLogix named in the spec itself), I/O, drives, motion, safety, and industrial software (FactoryTalk; embedded historians and network interfaces). Parent Rockwell Automation, Inc. (NYSE: ROK) is a large, publicly traded, currently operating company; no bankruptcy, receivership, or dissolution event. (General/sector knowledge, plus the assignee's own 10‑K/8‑K disclosure practice; I did not pull a specific filing in this session.)
• Did they ship a product embodying the claims? Highly likely, and consistent with the spec's own internal references to embedded historians, traffic analyzers, ControlNet/DeviceNet/Ethernet interfaces, and the backplane/module architecture — all Rockwell product categories. This is a product-implemented industrial-automation patent, not a paper patent. (I am stating this as a well-grounded inference from the specification; I did not verify a specific SKU.)
• Current status: Operating; assignee/owner of record unchanged as of the patent's Google Patents legal‑events record fetched 2026‑09‑29.

---

Assignment timeline

I attempted to query the USPTO Assignment Center (assignmentcenter.uspto.gov) and the older assignment.uspto.gov index directly, plus cross-referenced Google Patents legal events. I was not able to load live Assignment Center records in this session, so I cannot cite the reel/frame number for the one assignment that exists. I will not fabricate one.

The authoritative record I do have — the patent's own legal‑events data — shows exactly one recorded assignment in the chain:

• Executed on or about filing (date not specified in the record) / recorded 2007-12-05 — Reel NNNNNN/NNNN (⚠️ not retrieved)
  • Conveyance: Assignment — "ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)"
  • Assignor: All eight named inventors — Harris, Kendal R.; Hall, Kenwood H.; Rischar, Charles Martin; Balewski, Mark Joseph; Bradford, Jonathan D.; Gach, Martin George; Siorek, Timothy; Kretschmann, Robert J.
  • Assignee: ROCKWELL AUTOMATION TECHNOLOGIES, INC.
  • Correspondent: not determinable from the sources available to me. The USPTO Assignment Center "correspondent" field was not retrievable, and the patent text supplied to me does not show the assignment-transmittal attorney. ⚠️ I am flagging this explicitly rather than guessing — the correspondent field is the single most useful NPE tell per the task brief, and here it is simply unavailable. (For the record, the prosecution correspondence/attorney-of-record field is also not present in the text I was given.)
  • Context: Internal origination / standard employee-to-employer assignment at filing of a corporate-funded invention. Not an acquisition, not a fire-sale, not a reorg, not a securitization, not a transfer to an asserter.

No other assignments exist in the public record for this patent. Specifically, there is:

• No post-issuance assignment (nothing after Feb. 2, 2010).
• No security agreement, merger, change-of-name, license-and-record, or release recorded against 11/862,983 / 7,657,333.
• No assignment associated with the EP or CN family members that touches the US patent.

(Foreign-family context, not assignments: Google Patents also lists "2008-09-25 Priority to EP08165117.6A" and "2008-09-27 Priority to CN2008201814878U." These are priority filings of family members, not transfers of the US patent, and I am not reporting them as assignments.)

Cross-checks that produced nothing:

• Google Patents legal events for US 7,657,333: only the 2007‑12‑05 assignment, the 2010‑02‑02 grant, the EP/CN priority entries, and the adjusted-expiration entry.
• Prior-session litigation sweep (referenced in the Patent/Litigation summaries provided to me): no IPR/PGR/CBM, no ITC § 337, no district-court or CAFC case naming US 7,657,333. A "pre-litigation transfer" signal therefore has no triggering suit to attach to.
• No RPX/Unified Patents asserter-directory entry surfaced for this number.

---

Timeline diagram

timeline
    title Ownership of US 7657333
    2007 : Application filed 27 Sep
         : Inventors assign to Rockwell Automation Technologies
         : Assignment recorded 5 Dec
    2010 : Patent issued 2 Feb
    2028 : Adjusted expiration 22 Feb

---

NPE / troll-pattern signals

Score: 0 of 8 NPE signals present (1 not assessable due to missing correspondent data).

---

Verdict

Insufficient data — in the task's own terms, "only the original assignment."

Justification: the public record for US 7,657,333 contains exactly one ownership transaction — the inventors → Rockwell Automation Technologies, Inc. assignment recorded 2007‑12‑05 — and nothing after the 2010‑02‑02 grant. The chain never left an operating industrial-automation manufacturer, so every NPE signal (shell transfer, known asserter, cascading transfers, pre-litigation transfer, fire-sale, privateering) is affirmatively not present, which is the opposite of a troll pattern; I cannot return an "NPE — high/moderate" verdict, and I cannot return "operating-company assertion" because, per the prior litigation sweep, Rockwell has asserted no suit on this patent. Two caveats keep this from being a stronger "clean ownership" statement: (a) I could not retrieve the reel/frame or the correspondent of record from USPTO Assignment Center in this session, so I am expressly declining to state either rather than invent them; and (b) the "no litigation" finding rests on the earlier search-based sweep, not a certified docket pull.

Verify at: https://assignmentcenter.uspto.gov/ (search patent number 7,657,333; also mirrored at https://assignment.uspto.gov/patent/index.html). Recommended definitive cross-checks: USPTO Patent Center "Assignment/Litigation" tab for 7,657,333, and a Docket Navigator/Lex Machina query filtered to patent number 7,657,333.

---

Cross-reference note on the earlier sections

No contradictions found. The Patent Summary's "assignee recorded Dec. 5, 2007" and the Litigation Summary's "no litigation / no IPR / no ITC naming US 7,657,333" are both consistent with everything above. I did not disturb the caution against confusing this patent with Ameranth's 6,982,733 or Headwater's 8,406,733 — those remain unrelated. The one thing I must add to those earlier sections is the missing reel/frame and missing correspondent for the single 2007 assignment, which the prior sections did not supply and which I could not obtain here.

I'll search for the official USPTO record and the prior-art citations associated with US 7,657,333.

Let me pull the "Patent Citations" and "Cited By" sections directly from the patent record.

Let me try to retrieve the actual citation lists from the patent page and uspto.report.

USPTO Search — US 7,657,333: Prior-Art Citation Analysis

Bottom line up front (read this first)

I could not retrieve the authoritative prosecution citation list ("References Cited" / PTO‑892 / IDS) for US 7,657,333 from the sources available to me in this session, and I am not going to fabricate patent numbers and attribute them as the cited prior art. That would violate the "do not fabricate" operating rule and would corrupt any downstream § 102 analysis.

Concretely:

• The authoritative patent text supplied to me does not contain the citation tables. The scrape of the Google Patents page includes the bibliographic header, classifications, abstract, description, and the beginning of the claims — but the "Patent Citations" and "Cited By" tables were stripped out of the provided text.
• My targeted searches for "7657333 cited references," "References Cited," and citation-list queries returned citation lists belonging to other patents (e.g., US 9,806,968, US 10,943,471, US 11,418,487, Keysight IPR exhibits), not the citation list for US 7,657,333. None of those is the '333 record, and I am not conflating them.

So the honest status of the specific request ("look at each patent citation for 7657333, give full citation, date, description, and claim(s) anticipated") is: the citation set itself is unverified/not obtained. Below is what I can ground, plus how to close the gap definitively.

What the search did confirm (ground truth)

Primary source for the record: https://patents.google.com/patent/US7657333/en (Google Patents; accessed 2026‑09‑29).

Cross-citation evidence I did observe (flagged as unverified)

These appeared as snippets on pages that reference US 7,657,333. I am reporting them only as observed search artifacts with uncertain citation direction — not as confirmed prior art, because I could not open the citation tables directly:

• CA 2408901 A1 — "System for acquiring data from a facility and method." The Google Patents page for CA2408901 lists US7657333B2 (2007‑09‑27 / 2010‑02‑02, Rockwell) in a reference table. If this is a "Cited By" row on the CA page, it implies US 7,657,333 cites CA2408901 — which would make CA2408901 a candidate backward citation (earlier data-acquisition-from-a-facility art) worth checking. I could not confirm direction. (https://patents.google.com/patent/CA2408901A1/en)
• US 2017/0220692 A1 — "User-Trained Searching Application System and Method." Its page lists US7657333B2 in a reference table. This is a later publication (2017), so it is almost certainly a forward citation (it cites the '333) and is not prior art to the '333.
• Related-family note carried over from the earlier section: US 7,933,666 B2, "Adjustable data collection rate for embedded historians" (Campbell et al., Rockwell Automation). This overlaps the '333 subject matter, but its date is later than the '333 priority date, so it is not § 102 prior art to the '333 — it is, at most, evidence of the same corporate disclosure stream.

To be explicit: none of the above has been verified as a reference cited on the face of US 7,657,333, and I am not asserting it is.

What a § 102 analysis of the '333 would actually require

Because I don't have the real cited references, here is the test each candidate reference must meet — so the analysis can be completed the moment the true citation list is pulled. Recall from the patent-summary section that independent claims are 1 (system), 13 (method), 20 (means-plus-function), and that the asserted novelty is synchronization of multiple streams on a common timestamp/sequence-count + per-stream variable sampling rate, not data collection per se.

For a reference to anticipate claim 1 under § 102, it must disclose, in a single reference:

• a coordination component that synchronizes a plurality of data streams using a common timestamp generator or common sequence-count generator; and
• each stream having an associated data sampling rate defining how fast data is collected from it; and
• a sample rate adjustment component that adjusts the sampling rate of at least one stream based on (a) a predetermined criterion OR (b) a predicted required sampling rate.

Claim 13 (§ 102): a single reference must disclose collecting data from multiple internal and external streams, maintaining a sequence relationship via a common time-stamp/sequence-count generator, and adjusting the collection rate of at least one stream based on a predetermined criterion.

Claim 20 (§ 102 / § 112(f)): the same functional elements, recited as "means for synchronizing…" + "means for adjusting…" — construed to cover the corresponding structure in the spec.

Dependent-claim anticipation targets (each needs its own reference or combination): cl. 2 (internal = device data, external = network traffic); cl. 3 (recognition component); cl. 4 (weaving streams); cl. 5 (estimation component / analytical model); cl. 6 & 12 (fault-detection or granularity criterion; rate increase on fault); cl. 7–8 (matching component; zone-based rate variation); cl. 9–11 (AI component, probabilistic/statistical model, prior data, user feedback); cl. 14–15 (function blocks + triggering events tied to granularity); cl. 16 (user-feedback rate adjustment).

Practical note for whoever completes the § 102 pass: the hard limitation is the synchronized, shared-timestamp/sequence-count coordination of streams with independently variable rates. Network traffic analyzers (e.g., the references surfaced in unrelated searches) and historians that log independently will not, standing alone, anticipate claim 1 or 13 — they must be shown to tag both streams with a common time base (the spec points to IEEE 1588) and to vary per-stream rate. That is the crux on which any § 102 challenge to the independent claims will stand or fall.

How to obtain the definitive citation set (authoritative, not inferred)

1. USPTO PatentCenter — application 11/862,983, "References Cited" tab, and the PTO‑892 (Notice of References Cited) / IDS documents in the file wrapper, plus the examiner's search notes. This is the controlling record. (https://patentcenter.uspto.gov)
2. USPTO Public Pair / Patent Full‑Text (PatFT) for US 7,657,333 — the "Referenced By" and "References Cited" fields.
3. Google Patents US7657333B2 — the "Patent Citations," "Non-Patent Citations," and "Cited By" tables (present on the live page; absent from the scrape I was given).
4. Espacenet US7657333B2 (original document) — "Cited documents" and "Citing documents" lists.
5. For litigation/§ 102 challenges specifically: Lex Machina / Docket Navigator and PTAB (ptacts.uspto.gov) filtered to patent number 7,657,333 (per the litigation section, nothing on point was found — treat as "nothing found," not certified absence).

Contradiction / consistency flags

• Date inconsistency: the previously generated sections were produced under "Today's date 2026-09-29," while this task header states "Current Date: April 26, 2026." I flag this rather than silently resolving it; it does not change any bibliographic fact above.
• Do not auto-correct: all identifiers above are reproduced literally as found. The "'733" strings attached to Ameranth (US 6,982,733) and Headwater (US 8,406,733) in search results remain distinct patents and are excluded, consistent with the litigation section.
• Limitation carried forward: I still cannot certify the prosecution citation list, so any statement that a particular reference "anticipates claim X" would be unsupported until items 1–4 above are pulled. I am stating that explicitly rather than filling the gap with plausible-looking patent numbers.

I'll ground this analysis by pulling the actual prior-art citations and family references for this patent, since the fetched page text gave me only prior-art keywords and background discussion rather than a citation table.

I have the EPO search report (the examiner's own prior-art citations for this family) and several sibling references. Let me pin down the two references the EPO actually relied on with an "XY"/"Y" relevance rating, plus two candidate third-party references.

Obviousness Analysis — U.S. Patent No. 7,657,333 B2 (35 U.S.C. § 103)

Analyst note on scope and evidence. Before the analysis, one evidence-integrity flag that matters for everything below.

1. The "Prior Art section" of the supplied page is incomplete. The fetched US7657333B2 Google Patents text contains only the machine-generated prior-art keywords (data, industrial, automation system, industrial automation, component) and the family/"Other versions" links. The substantive "Patent Citations" and "Cited By" tables were stripped out in the fetch. Critically, one search result displayed a "Patent Citations (32)" table that belongs to U.S. 2017/0220692 A1 ("User-Trained Searching Application"), not to the '333 — the fetch interleaved two documents' citation tables. I have not attributed those 32 references (AOL, Google, Kofax, etc.) to the '333.

2. What I therefore relied on for prior art. Because the page's citation table was missing, I reconstructed the prior-art set from two on-point sources: (a) the EPO/Global Patent Index search reports for the two EP family members, which mirror this family and show the examiner's own relevance ratings; and (b) the citation lists exposed in search results for the sibling patents. This is stated transparently because the EPO search report is the single most probative "prior art section" item available for this patent.

3. A contradiction with the earlier-generated summary. The earlier summary said the embedded network traffic analyzer (NTA) was "largely unclaimed." The pre-grant publication's claims (as rendered by Espacenet for US 2009/0089558 A1) appear to recite the NTA in dependent claims 6 and 8. Those two renderings cannot both be right, and the Espacenet claim rendering is internally inconsistent (its claim 8 refers to an "NTA" that is never introduced in its stated parent claim 5). I treat this as an unverified prosecution-history signal, not fact — see §IX.

---

I. Governing law

The application was filed 27 Sept 2007, so pre-AIA § 103 applies throughout.

• Graham v. John Deere Co., 383 U.S. 1 (1966): scope/content of the prior art; differences between prior art and claims; PHOSITA level; secondary considerations.
• KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007): the TSM test is not the exclusive test; a combination is obvious where the improvement is a predictable use of prior art elements according to known methods, is a simple substitution of one known element for another, or is a use of a known technique to improve a similar device in the same way. Broad, enumerated rationales, each requiring only an "articulated reasoning with a rational underpinning."
• MPEP § 2143 exemplary rationales (A)–(G); MPEP § 2129 (applicant's own specification statements are usable as prior-art admissions).
• Pre-AIA 35 U.S.C. § 103(c)(1) — this is the pivot of the whole analysis (§IV.B).

II. PHOSITA and claim construction

PHOSITA: a B.S. in electrical/computer engineering or comparable, plus ~2–4 years in industrial automation, including exposure to controller backplanes, historians/event loggers, and industrial network analysis (ControlNet/DeviceNet/Ethernet). The claimed subject matter mixes three well-trodden engineering sub-fields — historians, network analyzers, and time-synchronization — so this is a combination of known arts, not a new art.

Key limitations construed:

• "coordination component that synchronizes multiple data streams … according to a common timestamp or sequence count generator" — a shared time/frame-of-reference generator tagging records from ≥2 streams so they can be interleaved. Notably, the '333 specification itself concedes IEEE 1588 synchronization as known ("provided that the timestamp/sequence count generators are synchronized (e.g. via IEEE 1588)").
• "each of the multiple data streams has a data sampling rate that defines a rate at which data is collected" — per-stream rate; the specification says any rate, including zero ("filtered out … by setting their collection rates to zero").
• "varies the data sampling rate … based upon at least one of a predetermined criteria or a predicted required sampling rate" — a disjunctive, permissive trigger. This is the broadest and most vulnerable element.
• Claim 20 — § 112 ¶ 6 means-plus-function; scope is the corresponding structure (the coordination component / sample rate adjustment component implementations) plus equivalents.

III. Prior-art inventory

On the page's own prior-art keywords. Google's extracted keywords for this patent are data, industrial, automation system, industrial automation, component. The patent's claim vocabulary is dominated by generic industrial-automation nouns and the two coined functional labels ("coordination component," "sample rate adjustment component"). That is the linguistic signature of a combination-of-known-elements claim rather than a claim to a new physical mechanism — relevant to the KSR analysis but not, standing alone, dispositive.

IV. The two structural facts that drive the outcome

A. The specification concedes most of the elements (MPEP § 2129)

The '333 specification expressly characterizes as known: network traffic analyzers and that they are "controlled to begin and/or end recording based on the presence of certain conditions"; historians that "communicate with controllers directly through the backplane"; backplane historian/recorder modules; and IEEE 1588 synchronization. It further concedes that "timing relationships, sequence counting, and the like" are "available relationships." These admissions supply, as prior art, the synchronization half and the conditional-capture half of claims 1/13/20. What remains genuinely claimed is only the conjunctive architecture: multiple streams, each with an independently variable rate, kept aligned on one common counter.

B. § 103(c) removes the closest art (this is the decisive point)

Every reference in family F is Rockwell-owned and qualifies only under §102(e) (all published in May 2008 or later, i.e., after the 27 Sept 2007 filing, so no §102(b) statutory-bar hook). Pre-AIA § 103(c)(1) provides that subject matter developed by another person that qualifies as prior art only under §102(e)/(f)/(g) "shall not preclude patentability" under §103 where the subject matter and the claimed invention were commonly owned at the time of invention. Rockwell Automation Technologies, Inc. is the assignee of both the '333 and the Campbell/Baier family.

Consequence: the examiner/applicant can prevail on § 103(c) to disqualify US 7,933,666 — whose claim 1 ("rate adjustment component configured to vary the sampling rate … based on a stage of the industrial process … a first sampling rate during the normal operation stage … a second sampling rate during the critical stage") is substantively the rate-adjustment half of the '333 — from the obviousness combination. The same disqualification removes US 2008/0114474 from the function-block/triggering combination for claims 7, 14 and 15.

This is a defense the patent holder can raise, not a defect in the patent. Any serious § 103 challenge to the '333 must therefore be built entirely from third-party art — refs A, B, C, D, E — which is exactly what the EPO search report did.

V. Element charts and the candidate combinations

Combination 1 — Ito + Guo (the EPO's own combination)

The EPO's search report for EP 2042960 (the EP member of this family, priority EP08165117.6A) cites:

• [XY] US 2003/0154056 A1 (Ito et al.)
• [Y] US 2004/0186685 A1 (Guo et al.)
• [A] US 2007/0025485 A1 (Caselli et al.)

In EPO practice, X = particularly relevant taken alone; Y = particularly relevant combined with another document; A = background. A claim carrying both an XY and a Y citation is the examiner saying, in effect, Ito + Guo collectively render this claim obvious.

Honest limitation: I could not retrieve the full text of Ito or Guo in this session, so I will not quote their disclosures. What can be said with confidence is the force of the citation: the EPO applied this pairing against claims directed to synchronizing multiple data streams and varying a data sampling rate. Ito carries the "X" — i.e., the EPO regarded it as teaching the core of the then-pending claim 1; Guo supplies the "Y" — the supplemental teaching the examiner needed to close the gap. To convert this into a written § 103 rejection, a challenger must locate in Ito the multiple-data-stream/per-stream-collection-rate architecture, and in Guo the timestamp/sequence-based correlation (or vice versa).

Combination 2 — US 7,558,687 (primary) + Ito or Guo (secondary) + the § 2129 admissions

This is the combination I can document end-to-end from retrieved text.

Why the combination would be motivated (KSR/MPEP 2143):

1. Same field, same problem, same purpose (MPEP 2143(A)). Internal (historian) and external (traffic-analyzer) data are both diagnostic streams used to troubleshoot the same industrial network. Correlating them is the recognized goal; the '333 background states the problem itself ("such two information streams are not synchronized together, and their collection does not depend on criticality of collection stage").
2. Design incentive / market force (KSR). Storage and network bandwidth are finite; sampling everything at maximum resolution is infeasible. A POSITA facing that constraint would predictably (a) synchronize streams on a shared clock so records interleave, and (b) raise resolution only where it matters. Ref. D makes the bandwidth constraint and the fault-triggered increase explicit, supplying the "rational underpinning."
3. Known technique to improve a similar device (MPEP 2143(C)). Tagging event logs with a common time-stamp/sequence-count is the textbook way to correlate two asynchronously collected logs; applying IEEE 1588 to the historian/traffic-analyzer pair is the canonical application.
4. Reasonable expectation of success. Each element operates as it does in its prior use; the combination requires no re-design of the underlying collection hardware.

Combination 3 — dependent claims 5 and 9–11 (estimation / AI / statistical model / feedback)

Ref. E (US 7,647,131 B1) is close to a one-to-one match: probabilistic analysis correlating historical data to predict failure → automatically determine the sampling rate; plus data mining for trends/patterns/correlations; aggregation across a plurality of controllers; audit logging; user security control over manually altering the sampling rate. That squarely meets claim 5 (estimation component predicting a required rate for a stage), claim 9 (AI component), claim 10 (statistical model based on prior data), and claim 11 (feedback to adjust the model). The '333's own admission that "various artificial intelligence systems/methodologies can also be employed" independently supplies claim 9.
Caveat: if E is Rockwell-owned and § 102(e)-only, § 103(c) disqualifies it too. I could not verify ref. E's assignee; this must be checked.

Combination 4 — dependent claims 7, 8 and 14, 15 (zones, triggering events, function blocks)

Rockwell's US 2008/0114474 A1 (family F) is the near-perfect anticipatory map: an analysis component that determines the function blocks of the control algorithm, a triggering component that starts data capture "at onset of activities relating to a function block" and stops "upon completion," and a feedback component that adjusts "frequency of data collection, granularity." If that reference is available, claims 7, 8, 14 and 15 fall quickly. But it is the paradigm case for § 103(c) — same assignee, § 102(e)-only, published 2008-05-15. For a § 103 challenge it is likely unusable; for a § 102 anticipation challenge the analysis is different and it remains available. A challenger would instead substitute generic IEC 61131-3 function-block / event-triggered-logging art, which the '333 itself treats as conventional.

Combination 5 — claim 12 (rate increase on fault)

US 7,558,687 states the limitation almost verbatim: raising a portion of the sampling rates "in response to detecting at least one fault or unexpected event." Alone, this makes claim 12's added subject matter very weak.

Combination 6 — claim 20

Means-plus-function mirror of claim 1; rises or falls with Combination 1 or 2. The disclosed corresponding structures are generic (a shared clock/counter generator; a rate-selecting controller), so there is no structural escape from the § 103 analysis.

VI. Why the claims may nonetheless survive

A balanced § 103 opinion must give the patent its due. Four points cut against obviousness:

1. § 103(c) is a statutory shield over the closest art (§IV.B). Strip out family F and the third-party art must independently supply both synchronization and per-stream variable rate in a single coherent architecture — a materially harder showing.
2. The EPO's XY objection did not prevail. The search report is an interim opinion. Google Patents lists EP 2042960B1 as a granted family member, which means the EPO ultimately allowed the case despite the Ito/Guo citation. That is probative — though not dispositive in the U.S., where the prior-art set and § 103(c) posture differ — evidence that a skilled examiner who saw the closest combination still did not consider the claims unpatentable.
3. The specific conjunctive claim-1 architecture is arguably not disclosed in one place. The third-party art teaches variable sampling rate on anomaly (D, E) and time-correlation of streams (B), but the requirement that each of multiple streams carries its own adjustable rate while all are held on one common timestamp/sequence-count generator is a configuration, not a mere aggregation. Under KSR this is still a strong candidate for obviousness, but it is not a slam dunk absent a reference that ties the two halves together.
4. No § 103 rejection appears to have issued. The patent granted with 20 claims and no apparent terminal disclaimer in the record I retrieved, and the EPO search report's "[A]" treatment of Caselli suggests the art was thin. Presumption of validity under 35 U.S.C. § 282 attaches; a challenger bears the burden of clear and convincing evidence at trial, or a preponderance in an IPR/PGR (which, being a post-grant challenge to a pre-AIA patent, would be governed by pre-AIA § 103 and — importantly — § 103(c) is not available as a defense in an IPR if the challenger is a third party; but common ownership of the prior art to the patent owner still matters if the petitioner relies on the patent owner's own § 102(e) art... which a petitioner generally cannot do for § 103 in view of 103(c)'s codified policy). Practically: a petitioner cannot easily use Rockwell's own '666/'474 family against Rockwell's '333 in a § 103 theory. That is a genuinely strong, and non-obvious, defensive position.

VII. Secondary considerations (35 U.S.C. § 282; Graham factor 4)

The following are potential objective indicia, flagged as unverified:

• Commercial success — Rockwell's FactoryTalk Historian ME line (1756-HIST modules) embodies embedded-historian technology. Nexus problem: that product more plausibly embodies the '666 claims (adjustable rate for embedded historians) than the '333's synchronization-of-external-traffic-with-internal-history claim. A nexus argument would have to tie the product to the claimed coordination function to be probative. I could not verify any nexus in this session.
• Long-felt but unmet need — the '333 is explicit that the two information streams "are collected independently and analyzed separately," which courts sometimes read as a recognized unresolved problem. This cuts for the patent, but is equally consistent with the solution being an obvious next step.
• Industry praise / licensing — no evidence found.

VIII. Bottom line

• The most legally coherent § 103 onslaught is: US 7,558,687 B1 + US 2003/0154056 A1 (or US 2004/0186685 A1), with the '333's own specification admissions supplying the historian/traffic-analyzer/sequence-count background. It attacks the rate-adjustment prong (D, near-verbatim on claim 12 and on the "predetermined criteria" disjunct) and the synchronization prong (B/Ito plus the applicant's own IEEE 1588 admission).
• A secondary attack on the dependent claims runs through US 7,647,131 B1 (claims 5, 9–11) and generic function-block/event-triggered-logging art (claims 7, 8, 14, 15).
• But the combination is not clean, and the patent has a real defense: pre-AIA § 103(c) disqualifies the entire Rockwell sibling family, which contains the one reference (US 7,933,666 B2) that most closely reads on the rate-adjustment element. A challenger using only third-party art must bridge a genuine gap — the per-stream variable rate bound to a common counter — and the EPO, which saw the best third-party combination, nonetheless allowed the EP counterpart.
• My estimate: independent claims 1, 13 and 20 present a moderate obviousness risk on the Ito/Guo + US 7,558,687 theory; dependent claims 12, 5, 9, 10, 11 present a higher risk (near-verbatim disclosure in D and E); claims 7, 8, 14, 15 present a low-to-moderate risk once § 103(c) is invoked. None of these is a certainty, and the D/E references' ownership status is the swing variable.

IX. Confidence and verification checklist

Verified this session: EP family search-report citations and their XY/Y/A ratings; the full text of US 7,558,687's key claims; the full claims of US 7,647,131; the family-F sibling disclosures and dates; the '333 claims and specification admissions.

Not verified — do not rely on without checking:

1. Full disclosures of US 2003/0154056 A1 (Ito) and US 2004/0186685 A1 (Guo) — I could not retrieve their text, so Combination 1 rests on the EPO's relevance rating alone.
2. Assignee/ownership of US 7,558,687 (D) and US 7,647,131 (E) — if either is Rockwell-owned and § 102(e)-only, § 103(c) removes it and Combinations 2 and 3 lose their primary references. This is the single highest-value check.
3. Filing date of US 7,647,131 — needed to confirm its § 102(e) availability date.
4. The '333 file wrapper (Patent Public Search / PatentCenter) — to see whether the examiner made any § 103 rejection, what art was cited, and whether the claim set was amended from the broader published version. The apparent gap between the published claims and issued claims, and the internal inconsistency in the Espacenet rendering, should be resolved against the file wrapper, not against aggregator text.
5. EP 2042960B1 grant date and file history — to confirm the EPO dropped the Ito/Guo objection and on what basis.
6. Re-run for any 2026 IPR/PGR — the earlier-generated summary and litigation section both found no proceeding, and this analysis adds no contrary evidence, but neither can certify a negative without a Docket Navigator/Lex Machina pull.