US 7591223

I'll search for information on this patent number, including any CAFC 2026 docket activity.

Search scope and caveats

My tools here are general web searches, not a direct query of the USPTO PatentCenter/PAIR system or the CAFC docket. So the following is grounded in (a) the authoritative full text of US 7,591,223 supplied to me and (b) Google Patents / FreePatentsOnline / Justia records for that exact number. I found no CAFC 2026 docket entry, appeal, or IPR appeal involving US 7,591,223, and I am not able to state affirmatively that one does not exist — I simply have no authoritative source showing one. Note that my searches on "Ferag" + litigation surfaced only older, unrelated decisions (e.g., Ferag AG v. Grapha-Holding AG, Ferag AG v. Quipp, Inc., 45 F.3d 1562), which do not involve the '223 patent. I did not return results for similar numbers.

Bibliographic data for US 7,591,223 (interpreted literally)

Google Patents flags "Family has litigation" (Darts-ip family ID 34935098), which indicates litigation somewhere in the family at some point — but I could not verify any specific case, and specifically none tied to a 2026 CAFC docket.

Abstract (verbatim)

"A device and method for transport of flexible, two dimensional products and performance of associated production processing of those products, by which the flexible, two-dimensional products are serially conveyed to a workstation and are conveyed away from the workstation again by the same gripper transporter, which enables the workstation to be selectively removed from the production process being performed along the transport path of the printed products, and to be selectively re-introduced again the transport path, in a flexible manner."

Plain-language overview of each independent claim

Claim 1 — Method (relative repositioning of station vs. transport path).
Continuously convey flexible flat products along a transport path in a conveying direction. Provide at least one connecting station whose at least one connecting element rotates around the station's axis along a circle. The station and the transport path can assume two relative positions:

• First position: the transport path is tangential to that circle, so the rotating connecting element can process the products on the path;
• Second position: the transport path is at a distance from the circle, so products pass the station unprocessed.

The method then includes continuously conveying the products and changing between the two positions — to stop processing (first → second) or to begin processing (second → first). In essence: the processing station can be "switched out" and back in without rebuilding the transport path.

Claim 1's dependents: 2 (change made via controllable diverters in the path); 3 (change made by moving the station's axis relative to the path); 4 (multiple workstations in serial or parallel arrangement); 5 (a gripper conveys one product to a workstation and another product away from it).

Claim 6 — Device (relative repositioning, apparatus form of claim 1).
A transporter continuously conveys the flexible products along a transport path to at least one connecting station. The connecting station has at least one connecting element rotating around the station axis on a circle. The station and path are arrangeable so that (a) in a first position the path is tangential to the circle for processing, and (b) in a second position the path is spaced from the circle so products pass unprocessed. A changing means shifts between the two positions to stop or begin processing.

• Dependent claim 7: the changing means operates electrically and/or hydraulically and/or pneumatically.

Claim 8 — Method (switching via diverters).
Same two-position tangential-vs-spaced framework as claim 1, but the change between first and second positions (stopping/starting processing) is performed by means of controllable diverters located in the transport path (i.e., routing products onto a different line rather than moving the station). Dependents 9–10 mirror claim 4 and 5 (multiple workstations; gripper carrying one product to and another away from a workstation).

Claim 11 — Device (parallel lines / diverter-based).
A transporter continuously conveys the products; at least one diverter in the transport path divides it into at least two parallel lines; at least one connecting station has a connecting element rotating on a circle. The diverter is switchable so that:

• products are continuously conveyed along one parallel line where the station/path relationship puts the transport path tangential to the circle (first position, processing occurs), or
• along another parallel line where the path is at a distance from the circle (second position, products pass unprocessed).

Dependents 12–13 add multiple workstations (serial or parallel) and the gripper that conveys one product to and another away from a workstation.

Claim 14 — Method (switching by moving the station axis).
Same tangential-vs-spaced two-position framework as claim 1, but the change between positions is accomplished specifically by moving the axis of the connecting station relative to the transport path (first → second to stop processing, or second → first to begin). Dependents 15–16 mirror multiple workstations and the gripper exchange feature.

Claim 17 — Device (movable station axis).
A transporter continuously conveys products along a transport path; at least one connecting station has at least one connecting element rotating around a station axis on a circle. The axis is movable: to a first position in which the station and path are positioned so the path is tangential to the circle (processing), and to a second position allowing products on the path to pass the station unprocessed. Dependents 18–19 add multiple serially/parallel-arranged workstations and the gripper exchange feature.

Technical context (from the specification)

The commercial problem addressed is inflexible fixed bindery/print-finishing lines: adding or skipping a work step previously required constructing a new transport path or spanning a removed workstation's footprint. The invention keeps the products on the same gripper transporter into and out of each workstation and makes the station removable (lowered, lifted, pivoted — pneumatically, hydraulically, or electrically) or routable-around (diverters/points), with a cover protecting products passing an idle station. Workstations contemplated include stapling/stitching (including rotary stapler/stitcher 38 with heads 40 on carrier disk 39), gluing/ adhesive binding, ultrasonic connection, insertion (with reference to EP 1254857 and its FIG. 12), collector/insert drums, and cutting stations. Grippers may hold products at the fold or at the open end, and may stay closed through processing (e.g., stapling) or deposit products onto support elements 34 (saddles/benders) and re-grasp them downstream. Cited prior art includes EP 0762950 B1, GB 5861, WO 02/36474, EP 0606555, EP 0691215, EP 1418146 A2 / US 2004/0089991 A1, EP 0390733/EP 0390734, EP 1254857, and others.

Uncertainty notes

• Litigation: I cannot confirm or describe any 2026 CAFC docket activity for US 7,591,223; none appeared in my searches. The only family-litigation signal is the generic Google Patents/Darts-ip flag, which does not identify a case.
• Fees/status: Google Patents lists an adjusted expiration of 2026-01-02 and status "Expired – Lifetime," while the Legal Events show maintenance fees paid through the 12th year (2020-09-28). If you need the definitive current status, verify in USPTO PatentCenter/PAIR directly.
• Inventor name spelling: the text uses "Heinz Möckli" while the assignment record renders it "MOCKLI, HEINZ"; I have not normalized either form.

I'll search litigation databases for this specific patent number.

Let me search additional litigation-specific sources.

Litigation search results for US 7,591,223

Search sources checked. I ran targeted queries against general web index that surfaces the litigation-aggregator and court-opinion sources you named — including Unified Patents' litigation portal (portal.unifiedpatents.com/litigation/caselist), CourtListener, PACER-adjacent dockets, Justia Patents, Google Patents, and the CAFC opinions/orders site — using the exact string 7591223 / 7,591,223 and Ferag-specific terms. I did not have a direct authenticated API query into PACER or Docket Alarm, so this is web-search-grounded rather than a docket-database export.

Finding: no litigation located

I found no U.S. litigation — no district-court action, no CAFC appeal, no ITC Section 337 investigation, and no PTAB proceeding (IPR/PGR/CBM) — involving US 7,591,223. Every hit that contained a similar-looking number resolved to a different patent, defectively, or to a case with no connection to the '223 patent:

The one dispute signal in the family (not U.S. litigation)

The European counterpart EP 1 588 971 B2 (same Ferag AG family; priority CH 707/2004) was opposed at the EPO — the EPO publication record I retrieved lists cited opponent prior art (e.g., EP 1 310 444 A1, Grapa Holding AG; WO 96/34724 A1; EP 0 662 440 A1; EP 0 686 463 A1; EP 1 254 857 A2) and shows the patent surviving as B2 (published 2019-06-19). That is an EPO opposition, not U.S. litigation, and it does not name a U.S. case number, court, or party.

Google Patents' "Family has litigation" flag — unresolved

Google Patents displays a "Family has litigation" banner with a Darts-ip family ID 34935098. That flag is generic and does not identify a case. I could not retrieve any Darts-ip case record, complaint, docket entry, or opinion corresponding to it. It may reflect non-U.S. proceedings or a since-closed matter, but I cannot state that affirmatively. I am not able to confirm any active litigation, and I cannot confirm that none exists — I simply have no authoritative source showing one.

Bottom line

Based only on the searches available to me, no litigation involving US 7,591,223 is identified: no plaintiff, no defendant, no jurisdiction, no case number, no filing date, and no outcome can be reported, because no case was located. The only litigation-adjacent signals are (a) the unexplained generic Google Patents/Darts-ip "family has litigation" flag, and (b) an EPO opposition to the related European patent.

Two flags worth noting:

1. Date discrepancy. Your prompt sets "current date: April 26, 2026," while my environment date indicates September 30, 2026. I have not assumed either date is authoritative; if you need a docket as of a specific date, please confirm which.
2. Verification limitation. To definitively rule litigation in or out, this should be run against an authenticated PACER / Docket Alarm / Lex Machina query for the '223 patent and Ferag AG as assignee/plaintiff, plus a PTAB (patent number) and CAFC (patent-number-in-opinion) search. My web-search method cannot guarantee completeness against sealed, settled, or non-indexed filings.

I'll verify the structured "no proceedings" signal against live sources before writing the analysis.

PTAB Proceedings on US 7,591,223 — Defendant's Analysis

Proceedings overview

Zero AIA trial proceedings. The canonical "PTAB proceedings on file" block (USPTO Open Data Portal ingest) returns no IPR, PGR, or CBM involving US 7,591,223, and that null result is corroborated by my independent web searches — which surfaced no petition, no institution decision, no Final Written Decision, no settlement, and no Federal Circuit appeal touching this patent. My searches on petitioner-side terms ("US 7591223 IPR," "7,591,223 PTAB petition," Ferag + IPR) returned only unrelated PTAB documents that happened to co-occur in the corpora; one third-party dataset (IP Verse competitive analysis for Ferag) separately lists Ferag with "PTAB Cases Faced: —", i.e., no PTAB hits from 2018 onward (https://ipverse.greyb.com/competitive-analysis/company/ferag).

Bottom-line defensive posture: "the patent was never tested at the PTAB — and it has now expired, so it never will be." This is the opposite of the "hardened patent" fact pattern. A defendant here does not face a claim that has been confirmed by the Board; it faces an untested, 19-claim patent whose adjusted expiration (2026-01-02, per Google Patents) has already passed as of today, 2026-09-30. The absence of any IPR on a patent this old, issued 2009-09-22, means every invalidity ground remains un-adjudicated and unencumbered by statutory estoppel — you can raise anything, anywhere, with no § 315(e)(2) bar and no Board record to overcome.

Per-proceeding entries

None. There are no proceedings to enumerate. I am deliberately not creating headings for proceedings that do not exist, because the instructive content here is the silence:

(Caveat on the modality: PGR/CBM were never legally available against this patent as of today's date — PGR's 9-month window long closed, and CBM's transitional program ended in 2020. The only AIA vehicle that was ever theoretically available is IPR, which is not time-barred by statute but is now practically moot given expiration. See "Recommended next steps.")

Strategic summary

Claim status: all 19 claims are UNTESTED. Independent claims 1, 6, 8, 11, 14, and 17 and all 13 dependents (2–5, 7, 9–10, 12–13, 15–16, 18–19) have never been construed, instituted, or adjudicated by the Board. Nothing is canceled; nothing is confirmed. That cuts both ways: the patent owner cannot point to a Board decision upholding its claims as a validity shield or as a § 315(e) estoppel sword, and you cannot point to a Board decision killing a claim. What you can point to is that the patent is expired as of 2026-09-30 (adjusted expiration 2026-01-02 per Google Patents), which converts the dispute from an injunction/ongoing-royalty fight into a past-damages fight capped by the 35 U.S.C. § 286 six-year lookback — and even that window is shrinking fast.

Estoppel landscape is maximally favorable to a defendant. Because no petitioner ever filed, 35 U.S.C. § 315(e)(2) estoppel attaches to no one. There is no "ground raised or reasonably could have been raised" bar, no Sotera/Sand Revolution stipulation, no § 325(e) PGR/CBM estoppel. Every prior-art combination — including art that Ferag cited during prosecution and art that was never before the examiner — is fully available to you in district court or in a DJ action. Contrast this with the typical scenario where a defendant inherits a well-mined prior-art record; here the field is completely open. Note, however, that this is a double-edged observation: there is likewise no third party's work product, no Board claim construction, and no invalidity expert report you can borrow.

Pattern signals are all negative (i.e., no adverse pattern).

• No serial petitioner — no party has filed multiple IPRs against this patent (trivially, because none has filed one).
• No defensive aggregator — there is no Unified Patents, RPX, or similar NPE-defense entity in the chain for this patent. Consistent with that, Ferag is an operating company (Swiss print-finishing/bindery equipment maker), not an NPE — the classic fact pattern that doesn't attract IPR filings.
• No aggressive PTAB appeal posture by the patent owner — the patent owner has never had to defend a Board decision, so there is no CAFC appeal to track. This also cross-confirms the earlier section's finding: no CAFC 2026 docket activity exists for US 7,591,223.
• Enforcement pattern is European, not American. Ferag's visible contentious activity in this space runs through EPO oppositions (the same dataset shows Ferag both filing and facing oppositions) rather than US PTAB trials. The family's parallel EP member, EP 1588971 B2, is the more likely theater if this technology is being enforced against your product abroad.
• Family caveat, flagged not resolved: Google Patents' "Family has litigation" flag (Darts-ip family ID 34935098) and the earlier summary's inability to identify a specific case remain unresolved. This does not indicate PTAB activity — the Darts-ip family ID is a litigation-database artifact and would capture foreign proceedings. Treat it as an open item, not as evidence of a US IPR.

Recommended next steps

1. Lead with expiration, not invalidity. The patent's adjusted expiration of 2026-01-02 (Google Patents; verify in USPTO PatentCenter/PAIR) means that as of 2026-09-30 there is no injunctive exposure and the damages case is a § 286 lookback question. If a demand letter is still citing US 7,591,223, the first thing to run is the accrual cutoff — not the merits.

2. Do not treat "no IPRs" as a signal that the family is a soft target. The same priority date (CH 707/04, 2004-04-22) produced US 8,342,317 B2 (continuation, filed 2011-09-23) and the abandoned US 2008/0149464 A1. The '317 covers the same disclosure and is a separate patent right. Any freedom-to-operate or invalidity workup should cover the '317's claims as well; "the '223 was never challenged" says nothing about the '317. Verify the '317's expiration independently — I do not have a reliable adjusted-expiration figure for it and will not guess one.

3. If a live assertion nonetheless exists, capture the record now. Because there is no PTAB file, there is no public Board record to cite. Pull the district court docket (via CourtListener, https://www.courtlistener.com, or PACER) and the Darts-ip family record behind ID 34935098 to determine whether the "family has litigation" flag reflects an active US case, a foreign case, or a stale entry. Report back if a case number surfaces — it changes the estoppel and stay analysis materially.

4. On the feasibility of filing your own IPR (if the matter is still live): treat it as unlikely to help and likely to be denied on discretionary grounds. Independent of the merits, a 2009-issued patent that expired in 2026-01-02 is a poor IPR candidate — the Board cannot be asked to cancel claims on an expired patent in a way that provides meaningful relief, and there are indications (from a 2026 Supreme Court amicus filing discussing a PTO "settled expectations" practice of denying petitions against older patents, https://www.supremecourt.gov/DocketPDF/25/25-1230/ — flagged as an unverified secondary source) that such petitions are being denied at the institution stage. If you need an invalidity adjudication, district court or an examiner-side vehicle is the better path for this patent.

5. Verify the status data directly before relying on it. Google Patents reports "Expired – Lifetime," adjusted expiration 2026-01-02, with maintenance fees paid through the 12th year on 2020-09-28. Confirm current status in USPTO PatentCenter and confirm the file wrapper contains no ex parte reexamination request (a reexamination would not appear in the ODP AIA-trial feed and would not have surfaced in my searches) before treating the record as clean.

Sources

• Authoritative full text of US 7,591,223: https://patents.google.com/patent/US7591223/en
• USPTO PTAB hearing/decision portal (confirm no proceeding exists on the actual docket): https://ptacts.uspto.gov
• Federal Circuit opinions and dockets: https://www.courtlistener.com
• Ferag PTAB/opposition footprint (third party): https://ipverse.greyb.com/competitive-analysis/company/ferag

I'll search the assignment and ownership records for this patent.

Let me dig further into the assignment record and Ferag's corporate status.

Let me try to pin down the assignment correspondent and check for any later transfers.

Assignment / Ownership Analysis — US 7,591,223 B2 ("Processing on a transporter")

Method note. My tooling is general web search, not an authenticated query into the USPTO Assignment Center API/PEDS. The reel/frame and assignment particulars below are grounded in the assignment record surfaced in the authoritative full text supplied to me (Google Patents legal events for US 11/111,948), which mirrors the Assignment Center data. I was not able to open the Assignment Center detail page for reel 016831/0368 directly, and I could not retrieve the correspondent of record — I say so where it matters rather than guessing.

Contradiction check with prior sections. The earlier Litigation summary flagged a date conflict (prompt said 2026‑04‑26; environment said 2026‑09‑30). The current header fixes today as 2026‑09‑30, which resolves the conflict in favor of the environment date. No other contradictions with the previously generated sections; the "current assignee: Ferag AG" call is confirmed below.

Inventors

Pattern notes:

• Single-inventor patent — no co-inventor fragmentation, so no "all inventors departed within 12 months" signal can even arise from the record. I found no evidence that Möckli left Ferag around the filing: a sibling Ferag application by the same inventor (US 2006/0179988 A1, filed 2005‑01‑21, later US 8,245,611) shows him still filing Ferag work contemporaneously. Beyond that I have no verified biography of his tenure or departure date, so I do not assert either.
• No inventor-held continuity (no inventor name reappears as an assignee downstream).

Original assignee

FERAG AG — a Swiss Aktiengesellschaft, "a corporation organized under the laws of Switzerland," Hinwil/Zurich. (The assignment instrument's own words, reel 016831/0368.)

• Primary line of business: Design and manufacture of conveying, processing and print‑finishing systems for the graphic arts industry, and — since roughly the mid‑2010s — intralogistics/material‑flow and e‑commerce fulfilment systems. Swiss commercial register purpose: manufacture and distribution of industrial products in processing conveyor technology, "in particular for the graphic arts sector."
• Did they ship a product embodying the claims? Yes, in substance. Ferag is a long‑established OEM of gripper conveyors ("Ferag Universal Conveyor") and inline finishing lines with rotary stapling/stitching stations — precisely the "transporter + connecting station that rotates around an axis" architecture the claims recite. Ferag's own literature describes its finishing systems as integrable, modular lines with connectable modules, consistent with the claim‑1/claim‑6 tangential‑vs‑spaced repositioning ("switch a workstation out without rebuilding the path") concept. The specification itself names Ferag-family products and workstations (rotary stapler/stitcher 38, benders 42, rotating support elements 34).
• Current status (as of 2026‑09‑30): operating. No insolvency, dissolution, or acquisition of Ferag AG is in evidence. The Swiss register (CHE‑102.207.498, Kt. Zürich) shows:
  • Capital increase to CHF 24,000,000 (SHAB 240112/2024, publ. 12 Jan 2024).
  • Corporate split into two legally independent companies, Ferag Solutions AG and Ferag AG, per the group's own "About us" page.
  • Name change FERAG AG → Ferag AG, statutes amended 2025‑12‑16, published SHAB 260112/2026 (publ. 2026‑01‑12, HR02‑1006533883); domicile changed to Industriestrasse 1, 8340 Hinwil; purpose restated to cover "integrated intralogistics, automation and supply chain solutions" and print‑finishing systems.
  • Still privately held ("Swiss family‑owned company"), ~600 employees, own sales/service companies in 18–50 countries; acquisitions incl. dereOida (Sept 2023) and integration of MacroSolutions (Leipzig).
  • Important for the ownership chain: this Swiss name change was not accompanied by any USPTO "Change of Name" recording on this patent — Google Patents' legal events show no post‑issuance assignment at all. Practically, title remains with the same legal person; formally, the U.S. record still names "FERAG AG."

Assignment timeline

Recorded assignments on this patent: one. It is the original inventor→employer assignment only.

1. 2005‑07‑14 (executed) / recorded 2005‑08‑02 — Reel 016831/0368
   • Conveyance: Assignment (assignment of assignors' interest)
   • Assignor: Möckli, Heinz
   • Assignee: FERAG AG, a corporation organized under the laws of Switzerland (Zürichstrasse 74, 8340 Hinwil)
   • Correspondent: Not retrievable with the tools available to me. I could not open the reel 016831/0368 record to read the recording attorney/firm, and no indexed copy surfaced a correspondent. Because this chain contains only one link, the "repeat correspondent" tell is untestable here regardless — see signal 3.
   • Context: Routine inventor‑to‑employer assignment perfecting title in the employer‑applicant, executed ~3 months after the 2005‑04‑22 U.S. filing and ~15 months after the 2004‑04‑22 Swiss priority date. This is a title‑clearing instrument, not a sale or transfer‑to‑asserter.

No further assignment records exist. Every other entry in the Google Patents legal events is a maintenance‑fee or administrative event, not a conveyance:

Continuation applications in the same family — no recorded re‑assignments either (so far as retrievable): US 12/043,618 (continuation filed 2008‑03‑06; published US 2008/0149464 A1; abandoned) and US 13/243,115 (continuation filed 2011‑09‑23; US 8,342,317 B2, expired‑fee related). These are same‑applicant continuations; nothing indicates title moved.

Plainly stated: the only ownership record for US 7,591,223 is the 2005 assignment from the inventor to Ferag AG. The original assignee still owns the patent, consistent with Google Patents listing Ferag AG as current assignee. The patent's term has ended (adjusted expiration 2026‑01‑02 per Google Patents), so no further transfer is commercially likely.

Timeline diagram

timeline
    title Ownership of US 7591223
    2004 : Swiss priority application filed
    2005 : US application 11 111 948 filed
         : Inventor Möckli assigns rights to Ferag AG
         : Assignment recorded at reel 016831 frame 0368
    2009 : Patent US 7591223 B2 issued to Ferag AG
    2013 : 4th year maintenance fee paid
    2017 : 8th year maintenance fee paid
    2020 : 12th year maintenance fee paid
    2025 : FERAG AG renamed Ferag AG in Switzerland
    2026 : Patent term ends with no further transfers

NPE / troll‑pattern signals

Cross‑reference note (litigation wording). Google Patents carries a generic "Family has litigation" banner tied to Darts‑ip family 34935098. That is a family‑level flag, unresolved in the prior Litigation section, and it does not manifest as any U.S. assignment, case, or party on this chain. The only concrete dispute signal found anywhere is the EPO opposition to the European counterpart EP 1 588 971 B2 (survived as B2, published 2019‑06‑19) — a European administrative proceeding, not a U.S. infringement campaign, and one that does not appear as a recorded U.S. assignment event.

Verdict

Insufficient data (per the definition "no records, or only the original assignment") — with the substantive caveat that the evidence that does exist is uniformly anti‑NPE.

Justification: the complete ownership record for US 7,591,223 consists of exactly one conveyance — inventor Heinz Möckli → FERAG AG, executed 2005‑07‑14, recorded 2005‑08‑02, reel 016831/0368 — and 21 subsequent years of maintenance‑fee events (2013‑03‑14, 2017‑03‑13, 2020‑09‑28) with no transfer, no LLC, and no asserter. All eight NPE signals score not present (signal 3 unclear solely because the reel's correspondent field could not be retrieved and there is only one link to test anyway), and the assignee is a going‑concern Swiss manufacturer that increased capital in 2024 and merely renamed itself in 2026. There is consequently no basis for an NPE verdict, and equally no assertion evidence for the "operating‑company assertion" box — so the honest label is insufficient data on a clean, non‑NPE chain.

Verify directly (search by patent number 7591223, or by assignor "MOCKLI" / assignee "FERAG"): https://assignment.uspto.gov/patent/index.html — and https://assignmentcenter.uspto.gov/

Residual verification gaps I cannot close with web search: (a) the correspondent of record on reel 016831/0368; (b) whether any unrecorded or non‑indexed transfer exists (an authenticated PACER / Docket Alarm / Lex Machina check for the '223 number and for Ferag AG as plaintiff would settle this, as would a PTAB and CAFC patent‑number search); (c) assignment activity on the two continuation siblings (US 12/043,618 and US 13/243,115) was not independently retrieved and lies outside this patent's chain.

I'll research the cited references to characterize each accurately, then map them to the claims of US 7,591,223.

Let me verify a few more of the key cited references.

Prior-Art Analysis for US 7,591,223 B2 ("Processing on a transporter")

Scope, method, and one important framing caveat

I worked from (a) the authoritative full text of US 7,591,223 supplied in this matter, whose face lists 28 patent citations and 2 non‑patent citations, and (b) corroborating records from Google Patents, Espacenet/EPO publication server, and the cited documents themselves (searches for US 4,408,755; US 4,729,554; US 5,324,014; US 5,951,744; EP 0 762 950; EP 1 254 857; EP 1 588 971, etc.).

Framing caveat you should keep front of mind. These 28 references are the backward citations of record. Almost none of them was cited as a §102 anticipation of an independent claim. Under 35 U.S.C. § 102, a reference anticipates only if it discloses every element of the claim in a single document. Patents 1, 6, 8, 11, 14 and 17 all require, in combination:

1. continuous conveyance of flexible flat products on a transport path;
2. a connecting station with a connecting element rotating about the station axis on a circle;
3. two selectable relative positions — (i) transport path tangential to the circle (processing) and (ii) transport path spaced from the circle (products pass unprocessed); and
4. a changing means / diverter / movable axis that switches between positions i and ii.

No single one of the 28 cited references discloses element (3). That feature is precisely the point of novelty. So the honest characterization is that the cited art is overwhelmingly § 103 (obviousness) material — individual claim elements and combination suggestions — not clean § 102 anticipations. Below I give the best-case § 102 mapping the citations support, and flag where the mapping fails on element (3).

Group A — The two references the specification itself identifies as the closest inflexible-path art

These are discussed in the "Related Art" section of US 7,591,223 and are the examiner's/patentee's own framing of the problem:

A1. EP 0 762 950 B1 (US counterpart US 5,950,511 A)

• Citation: Process and device for handling printed products, Ernst Lüthi, assignee Grapha‑Holding AG. EP 96908992.9, filed 1996‑04‑24; published EP 0 762 950 A1 1997‑03‑19; granted B1 1998‑12‑09. US counterpart US 5,950,511 A (priority 1995‑05‑02, published 1999‑09‑14) — this US counterpart is itself citation #18 on the '223 face.
• Description: Printed products are carried by grippers circulating on a first conveyor loop past an edge-cutting device; a second, co-rotating conveyor loop of clamps stabilizes the products during cutting, and the products are then removed from the clamps by the same grippers and conveyed away.
• Potential § 102 relevance: This is the strongest reference for the "same gripper transports a product to a workstation and away from it" limitation, which underlies dependent claims 5, 10, 13, 16, 19 (and is argued in the specification). It does not disclose a repositionable/tangential station, so it cannot anticipate independent claims 1, 6, 8, 11, 14, 17. The patentee expressly distinguishes it as a "fixed component of the transport device."

A2. GB 5861 ("Improvements in Printing Machines," Herbert Furnival et al., 9 Feb. 1910)

• Citation: GB 5861, published/issued 1910‑02‑09 (number and date reproduced literally from the specification; I did not normalize it to a modern GB‑year format).
• Description (per the '223 specification): A printed product to be cut is carried by the grippers of a gripper transporter to a cutting station, transferred to the station, and after cutting is transported away either by the same transporter's grippers or by a downstream transporter.
• Potential § 102 relevance: Again relevant only to the "same/next gripper services the station" idea (claims 5/10/13/16/19); no rotating connecting element and no tangential-vs-spaced repositioning → no anticipation of the independents.

Note: GB 5861 does not appear in the 28-entry Google Patents citation table; it is cited only in the body text. EP 0 762 950 B1 does appear in the table (as EP 0 762 950 A1), alongside its US counterpart US 5,950,511.

Group B — Ferag/Grapha rotating-stapler and processing-drum art (the "connecting element rotating on a circle" element)

Group C — Insertion / enclosures / collating / support-element art

Group D — Adhesive / ultrasonic binding art (spec's alternative "connecting" workstations)

Group E — Transport/bookbinding and non-analogous art

Non-patent citations (2)

1. English-language translation of the Office action from the Russian Patent Office, in Applicant's corresponding Russian application RU 2005111731, dated 26 Nov 2008. — This is prosecution material, not prior art; useful only to see how the RU examiner (family member RU 2,397,134 C2) assessed the same disclosure.
2. European Search Report issued in Applicant's corresponding EP 05 00 8056, issued 17 Jun 2005. — Also prosecution material; the EPO family member matured as EP 1 588 971 B1/B2, which, per the family-litigation record, was opposed at the EPO (the expanded EP record cites opponent art such as EP 1 310 444 A1, WO 96/34724 A1, EP 0 662 440 A1, EP 0 686 463 A1, EP 1 254 857 A2) and survived as B2. This is the only adjudicative/opposition activity I can tie to the family.

Synthesis: which cited art is actually "most relevant," and the § 102 bottom line

Element-by-element, the cited art maps as follows:

Bottom line under § 102. On the face of the citations, no cited reference anticipates any of claims 1, 6, 8, 11, 14, 17, because each lacks the defining two-position (tangential vs. spaced) limitation and the associated changing means. The two references the patentee itself frames as the closest art — EP 0 762 950 B1 (Lüthi/Grapha‑Holding; US 5,950,511) and GB 5861 (Furnival, 1910‑02‑09) — are § 103/distinguishing references: they show gripper transport to and from a station and a fixed cutting station, but expressly do not show a station that is repositionable relative to the path. The Ferag rotary-stapler family (EP 0 606 555; EP 0 691 215 / US 5,590,828; WO 02/36474; US 6,655,674) supplies only the "rotating connecting element on a circle" element, not the repositioning. The insertion/collection family (US 4,729,554; US 5,551,682; EP 1 254 857 / US 6,708,968; US 2003/0019722) and the support-element family (EP 1 418 146 / US 2004/0089991) supply process-step and hardware context only. US 5,863,571 (Sidel), US 6,948,292 (Tecnomeccanica) and UA 60258 (extruder) are not analogous art and have no § 102 bearing.

Most plausible § 103 combinations the citations could support (all still requiring the tangential/spaced teaching from elsewhere, e.g., the applicant's own admitted "diverters/points" and removable-workstation disclosure): a rotary gatherer‑stitcher (EP 0 606 555 or WO 02/36474) in view of a gripper conveyor that carries products to and away from a station (EP 0 762 950 / GB 5861), optionally combined with the removable/optional station art.

Caveats and flags

• "Potentially anticipates" is used above in the weak, best-case sense. Strictly, the citations are far better characterized as § 103 (and, for Groups D/E, merely background) art. If you need a formal § 102 chart, the answer is: no anticipation of the independents on this record.
• Two references lie outside the Google Patents citation table. GB 5861 appears only in the specification body; conversely, US 5,950,511 (US counterpart of EP 0 762 950) appears in the table. I have not treated the post‑dating forward citations (e.g., US 2006/0179988, US 2022/0219925, US 11,958,713) as prior art — they cannot be.
• Date/no-number integrity. I reproduced identifiers literally as they appear (e.g., "GB 5861 … 9 Feb. 1910"; "UA60258A"; "EP 0390734A2" vs. "EP 0390733B1" as separately listed in the family record) and did not normalize or "correct" any of them.
• Litigation/opposition. Consistent with the earlier sections, I found no U.S. litigation for US 7,591,223; the only adjudicative family event I can source is the EPO opposition to EP 1 588 971, which is not prior art but is relevant if you want to see how the closest art was actually argued against this family.
• Verification limits. I could not run authenticated PACER/PTAB/CAFC queries, and the search budget for this pass was exhausted before I could independently re-pull US 5,551,682, US 6,655,674 and GB 5861 full texts; the descriptions above for those rely on the patent's own citations plus indexed records rather than a fresh full-text read.

I'll ground this obviousness analysis in the prior-art references listed on the patent, checking a few key ones for their actual disclosures.

Obviousness Analysis — US 7,591,223 B2 under 35 U.S.C. § 103

Analyst work product. Scoped to the prior art listed in the patent's own Citations / Family Cites Families record and to the references the specification itself characterizes (Background of Invention). This builds on the earlier claim summaries and litigation sections rather than repeating them.

1. Legal framework and threshold caveats

Standard. Pre-AIA § 103 (the '223 was filed 2005-04-22), governed by Graham v. John Deere Co., 383 U.S. 1 (1966), and KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007). The four Graham factors: scope/content of the prior art; differences between the claims and the art; level of ordinary skill; and secondary considerations. Under KSR, a combination is obvious where the improvement is a "predictable variation," where there is a "design need or market pressure," and where a POSITA would have had "a reasonable expectation of success"; where there is a "finite number of identified, predictable solutions," the combination is obvious to try. MPEP §§ 2143, 2144.

Three caveats I want on the record before the analysis:

1. I do not have the full text of every cited reference. My mappings below are grounded in (a) the '223's own characterization of each reference, (b) the citation titles/assignees, and (c) live retrieval of EP 0762950 B1/US 5,950,511, US 2004/0089991 A1, US 20030019722 A1/US 6,672,447, and WO 02/36474 A1. Where I have not verified a reference's disclosure, I say so.
2. Common-ownership / § 103(c) risk. Nearly the entire Ferag-side citation list is commonly owned with the '223. Pre-AIA § 103(c) removes § 102(e)/(f)/(g) art from § 103 combinations where it was commonly owned when the invention was made. This matters for US 2004/0089991 A1 (filed 2003-11-05, published 2004-05-13 — i.e., a § 102(e) reference whose application predates the '223 priority date of 2004-04-22) and EP 1418146 A2 (published 2004-05-12, after the '223 priority date). Any § 103 ground leaning on those for the movable-station teaching is vulnerable. Grounds resting on § 102(b) references (published more than one year before 2004-04-22, or before the 2004-04-22 priority date) are not.
3. This is an analytical assessment, not an examiner's action. I did not have authenticated PAIR/Global Dossier file-wrapper access showing exactly which references the examiner actually applied; the earlier sections found no litigation that would have adjudicated validity.

Level of ordinary skill (POSITA). A bachelor's degree in mechanical engineering (or equivalent experience) plus roughly 3–5 years designing print-finishing / conveying machinery, familiar with gripper conveyors, rotary staplers/stitchers, saddle supports, and diverter/points ("Weiche") devices. I adopt this because the '223's disclosure is entirely mechanical, and no party has stipulated a level.

2. What is actually new in the independent claims

From the earlier claim breakdown, the independent claims 1, 6, 8, 11, 14, 17 share four building blocks, and only the fourth is an arguable point of novelty:

The dependent claims add only: multiple workstations serial/parallel (4, 9, 12, 15, 18); a gripper that carries one product to and another product away from a workstation (5, 10, 13, 16, 19); and electrical/hydraulic/pneumatic actuation of the changing means (7).

So the obviousness contest is essentially Block D — "make the rotary station switchable-in/switchable-out, either by displacing the station or by diverting the product stream." The rest of the claims ride along.

3. The prior-art arsenal available (as identified on this page)

Not available as prior art: US 8,245,611 B2 (Ferag; "…transporting flexible, two-dimensional products and simultaneously cutting these") has a 2005-01-21 priority — after the '223's 2004-04-22 priority date. Its teaching that cutting means move "from an idle position on a first movement path into a working position on a second movement path" is useful context for showing Ferag's own later approach, but it cannot be used against the '223.

4. Obviousness grounds

Ground 1 — Rotary processing station + movability of the station relative to the path → claims 1, 3, 6, 7, 14, 17 (and 4–5, 15–19)

Combination: WO 02/36474 A1 (or EP 0 606 555 / EP 0 691 215 for the rotary stapling head per se) in view of US 2004/0089991 A1's rail-mounted stapling apparatus and the admitted art of removing a workstation from a working path (specification Background: "when a workstation is removed from the working path, a transport path with a gripper transporter needs to span the space of the workstation that was removed").

Mapping:

Motivation to combine. (i) The references are all in the same field (rotary print-finishing stations feeding from and returning to a gripper conveyor) — KSR "familiar elements according to known methods." (ii) The art itself articulates the design need: US 2004/0089991's stated object is "even more flexible use of space," and WO 02/36474's modular stations exist to reconfigure a line. (iii) The '223's Background frames the market pressure expressly — "rapid resetting of the production line, e.g. a switch in production from large daily newspapers to weekly magazines… of varying size." (iv) Once one decides a station must be optional, moving it out of the path (the claimed solution) versus building a replacement transport path (the admitted prior practice) is one of a finite number of identified, predictable solutions — the KSR "obvious to try" scenario.

Reasonable expectation of success. High. Repositioning a machine unit on rails, a pivot, or a lift is routine mechanical engineering; no new principle of operation, no unexpected result. The patent's own figures (stapler lifted along arrow H; wheel-arrangement pivoted along arc S) confirm the solution is a simple relocation.

Ground 2 — Rotary processing station + switchable diverters into parallel lines → claims 2, 8, 11 (and 9–10, 12–13)

Combination: EP 0 606 555 / EP 0 691 215 / WO 02/36474 (rotary connecting element) in view of switchable-diverter/points art for gripper or imbricated streams (e.g., CA 2,260,219 A1 "Conveyor," US 2003/0019722 A1, and the diverter/points art that the '223's own specification treats as conventional: "parallel guided lines of the gripper transporter that are connected to one another via diverters, or points").

Mapping. Claims 8 and 11 do not require moving the station at all — they require (a) the same rotary connecting element geometry, and (b) a controllable diverter that routes products onto one of ≥2 parallel lines, one line being tangential to the circle (processed) and the other at a distance (unprocessed). The rotary-element half is taught by the same stapling references as Ground 1; the diverter half is old and conventional in print-finishing conveying (switchable points between parallel transport lines are standard).

Motivation. Routing a product stream around a station via a switch, rather than moving the station, is a well-known alternative for making a station optional; the two are functionally equivalent ways to achieve "process / do not process." KSR treats such known alternatives as an obvious design choice.

Caveat. A diverter/points document is described in the '223's specification but is not in the '223's own citation list. Any real § 103 ground for claims 8/11 would need an actual § 102(b) reference — e.g., a Ferag/Grapha points patent — identified by the examiner. I have not verified a specific one from the list on this page, so this ground is analytically sound but not yet anchored to a specific numbered reference.

Ground 3 — The "same-gripper / another-product" dependents are independently weak → claims 5, 10, 13, 16, 19

Claims 5/10/13/16/19 (a gripper carries one product to a workstation and another product away) are, in substance, the logical allocation concept already admitted in the art:

• EP 0762950 B1 / US 5,950,511 — grippers "remain allocated … to the printed products held by the tensioning clamps"; the same gripper conveyor supplies and leads away.
• GB 5861 — the product is carried to the cutting station by the grippers and taken away by the same transporter or a second downstream one.

The '223's only distinction in this respect — that the controller may instead hand off so that "a gripper which has conveyed a flexible, two-dimensional product to a workstation, conveying another … away" — is a control-logic choice expressly stated in the '223 as a way to "master speed differences." Selecting between (a) returning the same product with the same gripper and (b) picking up the next product with that gripper is an obvious engineering option once both are disclosed; the '223 itself presents them as alternatives. These dependents add essentially nothing patentable over the admitted art combined with Grund 1 or 2.

Ground 4 — Multi-workstation dependents → claims 4, 9, 12, 15, 18

Claims 4/9/12/15/18 (a plurality of workstations arranged serially or in parallel lines) are directly met by WO 02/36474 A1, which discloses a track with multiple stations (supply, processing, output) each having its own drive, and by the '223's own Background admission that multiple stations are arranged along a working path. Adding "serial or parallel" arrangement is a mere choice of line topology.

5. Counterarguments the patentee could raise — and their strength

6. Bottom line

• Independent claims 1, 6, 8, 11, 14, 17 are, on the available record, reasonably likely obvious under § 103 as a rotary-connecting-station reference (EP 0 606 555 / EP 0 691 215 / WO 02/36474) combined with known repositioning of a workstation out of a transport path (Ground 1) or known switchable diverters between parallel transport lines (Ground 2). The point of novelty — "make the rotary station switchable in/out by relative displacement or by diversion" — is a predictable variation serving an expressly recognized design need, with a finite number of solutions and a high expectation of success.
• Dependent claims 4, 5, 9, 10, 12, 13, 15, 16, 18, 19 add only line topology (serial/parallel) and gripper-allocation logic that EP 0762950 B1 / US 5,950,511 and GB 5861 already disclose; they are the easiest to invalidate. Claim 7 (electric/hydraulic/pneumatic actuation) is a routine actuator choice.
• The two soft spots are (a) the movable-station secondary reference and (b) the specific diverter reference. My best movable-station reference, US 2004/0089991 A1, is § 102(e) and commonly owned (Ferag), so § 103(c) may disqualify it from a § 103 combination; and my diverter support is drawn from the specification's narrative rather than a numbered citation. A rigorous § 103 rejection would need (i) a § 102(b) movable or bypassable rotary-station reference, and (ii) a numbered § 102(b) diverter/points reference, neither of which I can point to on this page with confidence.

Confidence: Moderate on the legal conclusion (the claimed advance is a predictable mechanical variation with articulated design motivation); Low-to-moderate on the evidentiary record, because I could not fully verify each reference's disclosure and the two most helpful secondary teachings carry § 103(c)/dating problems. I would not represent that a § 103 rejection is a certainty without pulling the full texts of EP 0 606 555, EP 0 691 215, WO 02/36474, and a § 102(b) diverter reference, and confirming the common-ownership status as of 2004-04-22.

Key sources: https://patents.google.com/patent/US7591223/en; https://patents.google.com/patent/EP0762950B1/en; https://patentimages.storage.googleapis.com/e9/88/f3/09265a9519be0f/WO2002036474A1.pdf; https://patents.google.com/patent/US20040089991 (grant US 7,581,724); https://patents.justia.com/patent/20030019722 (US 6,672,447); https://www.freepatentsonline.com/7591223.html.