US 11402419

I'll search for authoritative information on this specific patent number, including any 2026 litigation activity.

US 11,402,419 B2 — Analyst Summary

Important scoping note: You asked me to search USPTO and CAFC 2026 dockets for this specific number. I searched, but I can only report what my searches returned. I found no CAFC 2026 docket, appeal, or PTAB proceeding associated with 11,402,419. I did not find results for any other, similar patent number, so nothing here is a look-alike substitution. The absence of litigation hits is a negative search result, not proof of non-existence — PACER/CAFC docket full-text search and PTAB proceeding lists are not exhaustively reachable through the search tool I used, so treat "no litigation found" as low-to-moderate confidence.

---

Bibliographic data (from the Google Patents record at https://patents.google.com/patent/[US11402419B2](/patent/US11402419B2)/en)

Caution on assignee. The Google Patents page lists "National Instruments Corporation" as current assignee, but the underlying legal event is a release of a security interest, not an assignment of patent rights. Separately, NI acquired Optimal Plus via a May 27, 2020 share purchase agreement, closed July 2, 2020, for ~$365M (NI Form 8-K, SEC EDGAR, https://www.sec.gov/Archives/edgar/data/[935494](/patent/935494)/000093549420000029/a2020628-ksparta.htm and https://investor.ni.com/node/23126/pdf). Ownership of the patent today therefore most plausibly sits with NI (or an NI Israel subsidiary) via the equity acquisition, but the recorded chain-of-title on this specific patent does not itself show a patent assignment. Verify in USPTO Assignment Search before relying on it.

---

Abstract (verbatim)

"A method for identifying outlier devices during testing, includes: establishing binning limits for a device being tested based on one or more rules generated from external test results data of tests involving similar devices; receiving test results data in real time for the device being tested while the device is on a device tester; applying the one or more rules to the test results data for the device in real time; determining in real time, based on results of applying the one or more rules to the test results data, whether the device is an outlier with respect to the binning limits; and in response to determining that the device is an outlier, binning the outlier device separately from tested devices having test results data falling within the binning limits."

---

Independent claims — plain language

Claim 1 — Method (independent).

1. A local rule engine sets binning limits for a device under test, where those limits come from rule(s) generated from external test results data of tests on similar devices (i.e., data from elsewhere).
2. It then receives, in real time from a station controller, test results for the device while the device is still on the tester.
3. The local rule engine modifies the binning limits based on that test data, and transmits the modified limits back to the station controller.
4. It applies the rule(s) — now including the modified limits — to the test data in real time.
5. It decides in real time whether the device is an outlier against the modified limits.
6. If so, the outlier is binned separately from previously tested devices whose results fell within the (individually modified) limits for each of those devices.

The two features doing the heavy lifting are: (a) per-device in-flight modification of binning limits fed back to the station controller, and (b) the re-binning decision being made before the device leaves the tester.

Claim 5 — System (independent). The apparatus counterpart of claim 1: a central rule engine generates rule(s) from external test data of similar devices; a local rule engine receives them; a station controller receives real-time test results and forwards them to the local rule engine. The rules include binning limits; the local rule engine modifies the limits based on the test data, transmits the modified limits to the station controller in real time, decides outlier status against the modified limits, and bins outliers separately from previously tested devices that fell within their own per-device-modified limits.

Claim 9 — System for analyzing device test data (independent). A three-tier architecture, with the tiers differentiated by where and when they operate:

• First rule engine — modifies a first rule based on (i) test results of the device currently on the tester and (ii) a second rule, transmits the first rule to a station controller, and takes a first action while the device is on the tester.
• Second rule engine — generates the second rule from test results of other same-type devices tested at that same testing facility, plus a third rule, and takes a second action when the device is no longer on the tester.
• Third rule engine — generates the third rule from manufacturing data of same-type devices from multiple manufacturing facilities, applicable after the device leaves the tester.

Dependent claims add depth: excluded external-data types (claims 2, 6); central facility separate from test facility (3, 7); test add/omit instructions or SPL applied after the tester's own SPL (4, 8); online/offline operation of each engine and their physical placement (10–19); and the specific action menus — first action ∈ {rebinning, outlier detection, TTR, abort test}; second action ∈ {rebinning, outlier detection, shut down tester, maintenance, ML-based TTR} (claim 20).

---

Subject-matter overview (specification)

The patent addresses a real manufacturing problem: a company builds the same device type at multiple sites, but each site only sees its own test data, so identical parts get classified differently and yield inconsistent quality. The disclosed fix is a three-layer rule hierarchy:

• Core (central) analytics rule engine — sits at headquarters with access to pooled external data (other company sites, third-party test houses, higher-level assembly/product test data, failure-analysis data, MES data) and publishes static process limits (SPL) and "first rules" (weekly or on demand). Examples listed include adaptive test-time reduction, bin monitoring, freeze detection, Cpk/Gage R&R, parametric trend, site-to-site deviation, tester-settings validation, and ML-based system-level-test reduction.
• Edge analytics rule engine — site-local, generates narrower "second rules" from that facility's own data.
• Device analytics rule engine — real-time tier that computes e.g. Part Average Test (PAT) limits from the current lot as devices are tested, producing a "third rule." FIGS. 3–4 illustrate the progression: ATE SPEC limits (widest) → SPL from the first rule (narrower, catching outliers 355/357) → real-time dynamic limits from the third rule (narrowest, additionally catching outlier 455), allowing the station controller to re-bin a part before it is removed from the handler.

Notably, the patent frames the benefit around new product introduction (NPI), where little statistical data exists and the real-time device-level engine can establish limits on the fly instead of waiting for a large sample.

---

Points of uncertainty / things I'd verify before relying on this

1. Claim text caveat. The claim set I reproduced (claim 1's "modifying… the binning limits" and the "respective binning limits modified for each tested device" language in the final step; claim 9's "modify a first rule… transmit the first rule to a station controller") closely tracks the continuation family and reads as though it may reflect an amended/continuation claim set. The Google Patents page for US11402419B2 was the source, but I could not independently confirm through a separate full-text source that this is the exact issued claim 1 of the '419 patent as opposed to the '668/'305 continuation claims. Verify against the USPTO PatentCenter granted-claims document for US 11,402,419 B2. This is a material caveat, because the "modifying binning limits" limitation is what narrows claim 1 relative to the pre-grant publication's apparent scope.
2. Litigation/CAFC status. No CAFC 2026 docket or PTAB proceeding found for this number (see the scoping note above). Confidence: low-to-moderate.
3. Current ownership. As explained, the recorded document is a lien release, not a patent assignment; NI's ownership rests on the 2020 equity acquisition of Optimal Plus. Confidence: moderate.
4. Family. Three family applications are all shown as active (expiration 2040-10-07 for the '419; 2039-12-03 for US 12,339,305 B2). I did not independently audit their claim scopes.

Sources: https://patents.google.com/patent/US11402419B2/en; https://www.sec.gov/Archives/edgar/data/935494/000093549420000029/a2020628-ksparta.htm; https://investor.ni.com/node/23126/pdf

I'll search for litigation involving US Patent 11,402,419 across the sources you specified.

Let me run more targeted searches to confirm whether any litigation exists for this specific patent.

Litigation Search Report — U.S. Patent No. 11,402,419

Bottom line: Based on the searches performed, I found no known litigation (district court, ITC, or PTAB) involving U.S. Patent No. 11,402,419.

What I searched and what I found

I searched Unified Patents' litigation resources, PACER/RECAP-derived dockets (CourtListener), Docket Alarm materials, CAFC/PTAB materials, and general web sources for the specific number 11402419 / 11,402,419.

No case was identified in which U.S. 11,402,419 is a patent-in-suit, an asserted patent, or a challenged patent. Results returned were empty or unrelated.

Important — false positives that must not be confused with this patent

Several hits referenced a different "419," and I am flagging them so they are not mistakenly attributed to U.S. 11,402,419:

None of these involve U.S. 11,402,419, and none involve assignee Optimal Plus Ltd.

Available facts about U.S. 11,402,419 (for context, not litigation)

• Patent: U.S. 11,402,419 B2 — "Method and system for real time outlier detection and product re-binning"
• Inventors: Shaul Teplinsky; Arie Peltz; Dan Sebban
• Original assignee: Optimal Plus Ltd; a later assignment record (Oct. 13, 2023) shows National Instruments Corporation in a release of security interest against Wells Fargo Bank (i.e., the security interest was released; this is a lien/reassignment record, not a litigation event).
• Priority: Nov. 13, 2018 (Provisional 62/760,327); filed Nov. 13, 2019; granted Aug. 2, 2022.
• Family: continuations U.S. 11,852,668 (from 17/863,261) and U.S. 12,339,305 (from 18/511,705).
• The only "cited by" entry on the patent's face (US20220349930A1) is a family member publication, not litigation.

Caveats (per your strict rules)

• I did not auto-correct or substitute any patent number; 11,402,419 was searched literally as given.
• Litigation databases are not perfectly complete; a case could exist that is sealed, very recently filed, or not indexed in the sources reachable here. I did not exhaustively page through PACER for every district.
• I cannot confirm a negative with absolute certainty. Stated plainly: no litigation involving U.S. 11,402,419 is known from the available search results.

If you want, I can next attempt narrower docket searches (e.g., by assignee "Optimal Plus Ltd" or "National Instruments" as a party) to check whether the patent was ever asserted under a different caption.

I'll verify the structured "no proceedings" signal against public sources before reporting.

No proceeding surfaced. Let me run a few more targeted checks, including the family members and any litigation context, before concluding.

Let me check the family members and any PTAB docket entries directly.

Proceedings overview

Total AIA trial proceedings on file for US 11,402,419: 0 (0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution denied). The structured "PTAB proceedings on file" block sourced from the USPTO Open Data Portal returns an empty set, and my independent web searches surfaced no IPR, PGR, or CBM involving this patent or its family members (US 11,852,668, US 12,339,305). The bottom-line defensive posture: this patent is completely untested at the Board — there is no invalidated claim to lean on, and no petitioner has yet blazed a prior-art trail you could follow. A defendant facing assertion today has to build its invalidity case from scratch; there is no free option and no estoppel benefit inherited from anyone else's IPR.

---

No proceedings to report

There is nothing to list under the "most-impactful first" heading. Per the operating rule that the structured ODP block is canonical and the default is "no PTAB activity on file," I am not inventing a proceeding, a petitioner, a judge panel, or an FWD. Every field the task asks for (petition grounds, institution decision, claim-level FWD outcome, settlement, appeal) is inapplicable because no trial was ever instituted.

Caveats you should carry forward (stated honestly rather than papered over):

• The ODP ingest is a point-in-time snapshot. A recently filed petition that has not yet been indexed (or a proceeding filed in the last few weeks) would not appear in it and may not yet surface in general web search either. With today's date being 2026-09-29, an I would not expect a pre-2026 proceeding on this patent to be missing from both sources — but a filing within the last ~30–60 days is the realistic blind spot.
• My searches were general web searches, not a live query of PTAB E2E / PTAB Decisions. A general web search is a weaker instrument than the PTAB's own docket for confirming a negative. If this question drives a real invalidity budget, confirm by running the patent number directly in PTAB E2E and the USPTO PTAB Decisions search.

---

Strategic summary

Claim status: all 20 claims are UNTESTED. Claims 1–20 of US 11,402,419 stand exactly as issued on 2022-08-02. Not a single claim has been canceled, disclaimed via IPR, or adjudicated unpatentable by the Board. That cuts both ways: there is no dead claim to attack strawman infringement theories against, but there is also no adjudicated narrowing — no claim construction, no written-description or § 112 ruling, no prior-art-driven amendment — that you can exploit. The patent has never been stress-tested in a contested forum, which means its claim scope (notably the "modifying … the binning limits for the device being tested" limitation added into independent claims 1 and 5 during prosecution) is largely untested as a matter of validity. Note also that the file history shows a Non-Final Office Action mailed 2021-12-06 and a response entered 2022-03-10 before allowance — the prosecution record is thin, which historically makes a patent easier to challenge, not harder.

Estoppel landscape: empty. Because no IPR/PGR was ever instituted, there is no § 315(e)(2) estoppel running against any party. No petitioner, no privy, no real party in interest is barred from anything. Every prior-art ground — § 102 anticipation, § 103 obviousness, and any § 112 written-description or enablement attack — remains fully available to you, both at the Board and in district court. There is no "reasonably could have raised" shadow hanging over your invalidity contentions. Practically, that means the entire universe of art cited on the face of the patent (the patent's own citation list, which includes US 6,338,148; US 2003/0144810; US 2006/0085155; US 2011/0000829; US 2013/0275073) plus everything cited in the family prosecutions is open season — subject only to § 325(d) discretion at the Board, which is not estoppel and is applied case-by-case.

Pattern signals: none, with one structural caveat. No petitioner has filed even a single petition against this patent, so there is no serial-filer pattern, no joinder activity, and no defensive aggregator (Unified Patents, RPX, AST, et al.) visible anywhere in the chain. The patent owner (Optimal Plus Ltd., now associated with National Instruments Corporation following the 2023-10-13 assignment entry) has not needed to defend the patent at the Board at all. The more important signal is portfolio structure: the '419 is the root of a live continuation family — US 11,852,668 (filed 2022-07-12, Active) and US 12,339,305 (filed 2023-11-16, Active), with published applications US 2022/0349930 A1 and US 2024/0159813 A1. Because this family has multiple patents with claims sharing the 2018-11-13 priority date, invalidating the '419 alone would not clear the field. Any serious defense strategy has to assess the '668 and '305 claims in parallel; a successful IPR on the '419 that leaves continuation claims intact is a tactical win that may not reduce your exposure much.

---

Recommended next steps

If you are a defendant being asserted against today:

• The absence of PTAB activity is itself a signal — but read it correctly. Well-asserted, high-value patents typically attract IPRs. This patent has not been asserted widely enough (as far as public records show) to draw a petition. That does not mean it is strong; it usually means it has not been worth anyone's filing fee yet. You would be the first mover, which means you get to pick the art and the framing without having to distinguish a prior petition's record — a real advantage, since anything you file is a clean slate.
• Budget for a family-level challenge, not a single-patent challenge. Pull the full prosecution histories of US 11,852,668 and US 12,339,305, plus published applications US 2022/0349930 A1 and US 2024/0159813 A1, and map their claims against the '419's independent claims 1 and 5. The added limitation that drove allowance — the local rule engine modifying binning limits based on real-time test data and transmitting them back to the station controller — is the natural obviousness pressure point across all three patents.
• Mine the file history before you file. The 2021-12-06 Non-Final Action and the 2022-03-10 response are the key documents. Any argument the applicant made to overcome that rejection is a prosecution-history-estoppel trap for the owner and a roadmap for you. Start at USPTO PatentCenter for application 16/682,925.
• Note the statutory clock if you go to the Board. If you have been served with a district court complaint, § 315(b) gives you one year from service to file an IPR. Once instituted, the trial must be completed within 12 months of the institution decision (extendable up to 6 months for good cause). Those are your two hard dates.
• No claims are invalidated, so do not cite an FWD. There is no Final Written Decision to link to and no disposition to quote. If an adversary or a comparison chart tells you claims 1–20 have been "canceled" or "held invalid by the PTAB," that statement is false as of 2026-09-29 — every claim in this patent remains live and presumptively valid under 35 U.S.C. § 282.

Bottom line: the defensive toolkit here is your own invalidity work product, not someone else's PTAB win. Confirm the negative at PTAB E2E before you finalize the opinion, then build the challenge across the whole '419/'668/'305 family in parallel.

I'll research this patent's ownership chain. Let me start with several searches.

Let me dig into the assignment records and the security interest.

Let me look for the specific reel/frame records and the Wells Fargo security agreement.

I hit the search-step limit before I could pull the raw USPTO Assignment Center entries. Below is what I can state with confidence from the authoritative patent text plus the corroborating sources I did retrieve. I flag explicitly where a field (the correspondent of record) could not be verified and should not be treated as confirmed.

---

Inventors

The patent names three inventors, all obligated to the original assignee. The only recorded assignment in which they appear is the confirmatory inventor→company assignment signed 2020-05-25 to 2020-05-26 and recorded 2020-06-23 (reel 053020/0632).

Pattern note (present, but benign here): the inventor assignment was executed roughly five weeks before National Instruments closed its acquisition of Optimal Plus, and only two days before the Share Purchase Agreement was signed (SPA dated 2020-05-27; per NI's 8-K). That is the classic "clean up chain of title ahead of an M&A closing" step, not a fire-sale precursor. I found no evidence that any of the three inventors left Optimal Plus within 12 months of filing, so I will not assert that pattern. Note this patent's inventor assignment was recorded after the application was filed (application 2019-11-13), i.e., a post-filing confirmatory assignment rather than a pre-filing one.

---

Original assignee

Optimal Plus Ltd (Holon, Israel; formerly OptimalTest Ltd), the entity named on the face of the patent.

• Line of business: big-data / "Manufacturing Intelligence" analytics software for semiconductor, automotive and electronics manufacturing. Founded 2005 by Dan Glotter; ~240 employees; ~$51M revenue for 2019. HQ and R&D in Israel with offices in Asia, Europe and the US.
• Did they ship a product embodying the claims? Yes. Optimal Plus shipped the OptimalPlus analytics platform and the on-tester agent Proxy+, whose advertised capability list explicitly includes "Automated re-binning," "Adaptive test time reduction," and "Drift detection" (OptimalPlus/NI product deck, "Analytics Solutions For Semiconductor"). That maps directly onto claim 1's real-time re-binning of outlier devices.
• Current status: Acquired. NI acquired all of the share capital of Optimal Plus Ltd on 2020-07-02 through its wholly-owned subsidiary National Instruments Israel Ltd, for ~$365M in cash (NI 8-K dated 2020-07-06; SPA dated 2020-05-27). Optimal Plus Ltd therefore remained a going concern as an NI subsidiary. NI itself was then acquired by Emerson Electric Co., closing 2023-10-11 (equity value ~$8.2B; NI's stock suspended 2023-10-12). So the corporate parent today is Emerson, with the patents sitting in the NI/Test & Measurement group.

Important structural point: because the 2020 deal was a share purchase (not an asset purchase), Optimal Plus Ltd is still the record owner of the patent. There is no recorded assignment of this patent to National Instruments Corporation. The only NI-linked entry in the chain is a lien release, not a title transfer.

---

Assignment timeline

Only two post-filing assignment records are surfaced for this patent (via Google Patents legal events, which mirror USPTO Assignment Center). Both are set out below.

• 2020-05-25 / 2020-05-26 (executed) / recorded 2020-06-23 — Reel 053020/0632

  • Conveyance: Assignment (inventor → company confirmatory assignment)
  • Assignor: Shaul Teplinsky, Arie Peltz, Dan Sebban (individually)
  • Assignee: Optimal Plus Ltd. (Israel)
  • Correspondent: Not retrievable from the sources I could reach. Google Patents does not expose the recording correspondent, and my searches for the raw Assignment Center entry were cut off. Recommend a direct pull at https://assignmentcenter.uspto.gov/ — search US11402419 — to capture the attorney/firm of record for reel 053020/0632. I will not guess this field.
  • Context: M&A title-cleanup — confirmatory assignment executed two days before the SPA and ~five weeks before the July 2 2020 share-purchase closing.
  • Family note: the same reel/frame 053020/0632 is the common origin link for the whole family — continuations US 17/863,261 → US 11852668 B2 (filed 2022-07-12) and US 18/511,705 → US 12339305 B2 (filed 2023-11-16) all claim this 2018-11-13 priority and hang off the same inventorship/ownership.

• 2023-10-11 (effective) / recorded 2023-10-13 — Reel 065231/0561

  • Conveyance: Release of Security Interest in Patents (lien release, not an assignment of title)
  • Assignor: Wells Fargo Bank, National Association, as Administrative Agent
  • Assignee: National Instruments Corporation (Texas)
  • Correspondent: Not retrievable from the sources I could reach — same caveat as above; verify at Assignment Center.
  • Context: Securitization unwind. The released lien traces to NI's Amended and Restated Credit Agreement dated 2020-06-12 with Wells Fargo as administrative agent ($145M facility; NI drew a $70M term loan on 2020-06-30 specifically to fund part of the Optimal Plus purchase price — NI 8-K). The release is effective 2023-10-11, the exact date Emerson closed its acquisition of NI — i.e., Emerson paid off/terminated NI's debt at closing ("payoff of NI debt at closing" appears in Emerson's purchase-consideration table) and the patent security interest was released.

Lien-record gap to note: the release above implies an underlying Patent Security Agreement was recorded at some earlier reel/frame (likely in 2020, tied to the Wells Fargo credit facility). I could not locate or verify that reel/frame. Treat it as "a security agreement almost certainly exists on record, reel unverified."

No recorded transfer to any NPE, licensing entity, or aggregator exists for this patent. Chain is: inventors → Optimal Plus Ltd (2020) → [Wells Fargo lien, then release] → maintained in place; owner of record remains Optimal Plus Ltd / NI / Emerson group. Maintenance fee for year 4 was paid 2026-01-21, and adjusted expiry is listed as 2040-10-07 — the patent is alive and being maintained.

---

Timeline diagram

timeline
    title Ownership of US 11402419
    2018 : Priority filing by Optimal Plus Ltd
    2019 : US application filed by Optimal Plus
    2020 : Inventors assign rights to Optimal Plus
         : NI buys Optimal Plus in share deal
    2022 : US 11402419 granted
    2023 : Wells Fargo releases patent lien
         : Emerson acquires National Instruments

---

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. The only assignee of record is Optimal Plus Ltd, an operating software company with products (OptimalPlus platform, Proxy+), real revenue (~$51M in 2019) and ~240 employees. No "IP / Patents / Licensing / Holdings / Ventures" entity appears anywhere in the chain (compare reel 053020/0632 and reel 065231/0561). No registered-agent-service address appears as an assignee address.

2. Known asserter in the chain — Not present. None of the chain participants — Optimal Plus Ltd, National Instruments Corporation, Emerson Electric, or Wells Fargo — matches the Acacia / Marathon / IV / IPNav / Wi-LAN / Conversant / Vringo / Pendrell / Innovatio / MPHJ / Lumen View / Round Rock / DGC / Spangenberg lists, nor any Unified Patents or RPX high-frequency-plaintiff directory I can identify from the retrieved material. Note Wells Fargo appears only as a security-interest holder/administrative agent, which does not make it an owner or asserter.

3. Repeat correspondent across the chain — Unclear / unverified. This is the one signal I could not test: the correspondents of record for reel 053020/0632 and reel 065231/0561 were not surfaced in any source I reached. Do not score this as a finding either way. To close it out, run the two reel/frame entries at Assignment Center and compare the recording attorneys.

4. Cascading transfers — Not present. Only two post-filing records exist across ~3.4 years, and one of them is a lien release rather than a title transfer. There is no chain of LLCs, shared correspondent addresses, or common principals. The NPE-typical pattern (multiple transfers through chained LLCs in <24 months) is absent.

5. Pre-litigation transfer — Not present. I found no infringement litigation naming US 11402419, and consequently no assignment dated within six months before a first suit. There is no standing-record or venue-setting transfer in the chain.

6. Bankruptcy fire-sale — Not present. Optimal Plus was sold in an all-cash share purchase at a $365M valuation (NI 8-K, 2020-07-06), and NI itself was later acquired by Emerson at an $8.2B equity value (Emerson press release, 2023-10-11). No Chapter 7/11 proceeding, no distressed sale.

7. Privateering — Not present. NI is an operating company that ships test-and-measurement hardware/software; the patent was absorbed into an operating product line (NI/Emerson Test & Measurement), and there is no NPE asserting on NI's behalf.

8. Defensive aggregator (anti-NPE) — Not present. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. It terminates in ordinary corporate ownership (Emerson's Test & Measurement segment).

---

Verdict

Operating-company assertion — with the honest caveat that the "assertion" half is unproven; what the record actually shows is operating-company ownership, no NPE.

Justification: The only title-bearing assignment is reel 053020/0632 (recorded 2020-06-23), which moved the patent from its three inventors to Optimal Plus Ltd, an operating analytics vendor whose shipped product (Proxy+ / OptimalPlus platform) advertises the very "automated re-binning" and "adaptive test time reduction" functionality claimed. The only other entry, reel 065231/0561 (effective 2023-10-11), is a security-interest release by Wells Fargo, not a transfer — it merely confirms a prior credit-facility lien under NI's 2020-06-12 Wells Fargo facility. The chain runs inventors → Optimal Plus Ltd → (share purchase, no title change) → NI → Emerson, with zero shell entities, zero NPE assignees, zero cascading LLC transfers, no litigation, and the year-4 maintenance fee paid on 2026-01-21. One data gap remains: the recording correspondents for both reel/frame entries were not verifiable from my sources and should be pulled directly from Assignment Center before any final NPE screen is signed off.

Verify at: USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (search US11402419), with the Google Patents mirror at https://patents.google.com/patent/US11402419/en.

---

Confidence / honesty notes:

• The reel/frame identifiers 053020/0632 and 065231/0561, the dates, the entity names, and the "Release of Security Interest" characterization come from the authoritative patent text you supplied (Google Patents legal events). I did not independently re-pull them from Assignment Center.
• I could not verify the correspondent of record on either recording, nor locate the reel/frame of the underlying (pre-release) Wells Fargo patent security agreement. I have deliberately left those fields unfilled rather than guess.
• I did not find any litigation involving this patent; if one exists and I missed it, signal #5 would need re-testing.

I'll search the USPTO database for the specific patent and its cited prior art references.

Prior-Art Analysis — US 11,402,419 B2

Patent under review: US 11,402,419 B2 — "Method and system for real time outlier detection and product re-binning"

• Application: US 16/682,925; filed 2019-11-13; priority 2018-11-13 (Prov. 62/760,327)
• Granted 2022-08-02; Assignee: Optimal Plus Ltd (current assignment record: National Instruments Corporation, release of security interest 2023-10-13)
• 20 claims; independent claims 1 (method), 5 (system), 9 (three-rule-engine system)
• Source verified: https://patents.google.com/patent/US11402419/en

Sourcing note. I could not reach the live USPTO PatentCenter full-text page directly from this session; the citation table below is taken from the authoritative patent text supplied to me and corroborated against Google Patents / FreePatentsOnline / USPTO.report mirrors (links given). Per your strict rule, I have reproduced every alphanumeric identifier exactly as it appears and made no corrections.

---

A. The five references in the "Patent Citations" table of US 11,402,419

---

B. Which claims each reference could bear on — and under which statute

Important framing: these are examiner-cited references (marked * in the table, i.e., "cited by examiner"). In my assessment none of the five is a clean § 102 anticipation of independent claims 1, 5, or 9 as issued. Each is missing at least one element that the issued independent claims require, so their realistic role is § 103 (obviousness), typically in combination. I flag below where a single reference could arguably reach a dependent claim or a narrower reading.

1. US 6,338,148 B1 (Compaq, "Real-time test controller")

• Potentially relevant to: claim 4; claim 9 (action definitions); claim 20.
• It discloses real-time feedback of test results and re-sequencing / re-selecting which diagnostic tests to run based on accumulated failure history — this maps onto claim 4's "instructions to add or omit tests for the device," and onto the "test time reduction / aborting test" actions recited in claim 20.
• It does not disclose binning limits derived from external multi-site test data, outlier-based re-binning, or the three-rule-engine hierarchy of claim 9. Not an anticipation of claims 1, 5, or 9.

2. US 2003/0144810 A1 (Tabor, "Methods and apparatus for data analysis")

• Potentially relevant to: claims 1–2 (outlier determination from multi-dataset data); claim 9.
• Its composite analysis over more than one dataset is the closest of the five to claim 2's / claim 6's "external test results data … generated at a facility different from a facility at which the device is being tested," and to the notion of detecting outliers and changing a device's bin based on composite data.
• It is a batch/post-run analysis system (data "not usually analyzed at product run time" per the sibling disclosure) and does not disclose real-time receipt of test data while the DUT is on the tester, nor the local-rule-engine modification and re-transmission of binning limits. § 103 candidate at best; not § 102 for claims 1/5/9.

3. US 2006/0085155 A1 (Miguelanez, "local outlier detection")

• Potentially relevant to: the "determine … whether the device is an outlier" limitation of independent claims 1, 5, and 9; possibly claim 20's "detecting the device as an outlier."
• This is the most on-point reference for the outlier concept itself — it expressly identifies outliers, including local outliers within subsets of a larger population (analogous to detecting an outlier relative to the current lot).
• However, its own background states the data "is not usually analyzed at product run time" and is instead analyzed between runs/off-line. That undercuts any § 102 reading of claim 1's "in real time … while the device is on a device tester" and the re-transmission of modified limits to the station controller. Best characterized as § 103 art, potentially combinable with reference 4.

4. US 2011/0000829 A1 (Optimaltest, "binning at final test")

• Potentially relevant to: claims 1 and 5 (overriding/re-binning a DUT while still socketed); claim 20 ("rebinning").
• This is the closest reference on the binning mechanism: it discloses an evaluator that defines a binning assignment for a device while it is still in the test socket and that assignment may differ from the test program's provisional bin — exactly the "binning the outlier device separately" outcome, and it is from the same corporate family (Optimaltest / Optimal Plus). It also mentions D-PAT (real-time part-average-test limits), which is the mechanism the '419 patent's local rule engine uses.
• What it lacks relative to issued claim 1: (a) binning limits derived from external test data of similar devices across sites; (b) the local rule engine modifying the binning limits based on the current device's test results and transmitting the modified limits to the station controller; (c) the "respective binning limits modified for each tested device of the previously tested devices" accumulation. § 102 candidate for a broad "re-bin while socketed" reading only; not for claims 1/5/9 as issued. Strong § 103 building block.

5. US 2013/0275073 A1 (Arnold, "decision and execution tree of a test cell controller")

• Potentially relevant to: claim 9 (the rule/decision-engine architecture and initiation of actions); claims 10–12, 16–17 (rule-engine placement/on-line vs off-line).
• Its decision-and-execution tree executed by a test-cell controller corresponds to the "rule engine … define and initiate an action … on a tester" structure of claim 9.
• It does not disclose the specific three-tier split (first rule engine modifies a first rule from test results + second rule; second rule engine generates from same-type devices at the facility; third rule engine generates from manufacturing data of a plurality of facilities). § 103 candidate; not § 102.

---

C. Prosecution-history inference (material to § 102/§ 103 posture)

The issued claim 1 contains two limitations that do not appear in the published pre-grant disclosure (US 2020/0150173 A1) summary or abstract:

• "modifying, by the local rule engine and based on the test results data, the binning limits for the device being tested;" and
• "transmitting, by the local rule engine, the modified binning limits to the station controller,"

plus the "respective binning limits modified for each tested device of the previously tested devices" language. The record shows a Non-Final Action (2021-12-06), a response (2022-03-10), and allowance (2022-03-24). The natural reading is that these real-time modify-and-re-transmit limitations were added to distinguish the claims from the cited art above — which is consistent with my assessment that the five references are § 103 (obviousness) art whose combination was overcome by those added limitations, not § 102 anticipatory art for the independent claims.

---

D. Other documents worth noting (not in the citation table)

• References incorporated by reference into the patent (may be relevant art/priority context): US 9,885,751; US 8,872,538; US 8,421,494; US 7,969,174; US 7,567,947; US 7,528,622; and applications Ser. Nos. 12/497,789 and 12/497,798. Also provisionals referenced by US 2006/0085155 A1 (60/483,003; 60/546,088; 60/603,312).
• "Cited By" entries — US 2022/0349930 A1, US 11,852,668 B2, and US 12,339,305 B2 — are later family members/continuations (2022–2025) and are not prior art to the '419 patent.

---

E. Bottom line

No single cited reference anticipates issued independent claim 1, 5, or 9. The strongest § 102 exposure, if any, is on the narrow "re-bin a device while it is still in the socket/test program bin is overridden" concept via US 2011/0000829 A1 — but that reference does not teach the external multi-site rule generation or the local-rule-engine modification/re-transmission of binning limits that the issued claims require. The primary risk to the claims is a § 103 combination, most plausibly US 2011/0000829 A1 + US 2006/0085155 A1 (optionally + US 2003/0144810 A1 for the external/multi-dataset data, and US 2013/0275073 A1 for the rule-engine architecture).

Caveat: I was unable to pull the live USPTO PatentCenter prosecution documents or the examiner's stated § 102/§ 103 rejections in this session, so the statutory-characterization column is my analytical judgment based on the reference texts, not a quotation of an Office Action. If you need the examiner's exact grounds, the 2021-12-06 Non-Final Action and 2022-03-10 response in the file wrapper should be retrieved directly.

I'll ground the prior-art disclosures before building the § 103 analysis. Let me pull details on the five examiner-cited references from the patent's Prior Art section.

Obviousness Analysis — U.S. Patent No. 11,402,419 B2

"Method and system for real time outlier detection and product re-binning"

Statutory basis: 35 U.S.C. § 103 | Priority date: 2018‑11‑13 | Granted: 2022‑08‑02
(Analysis date for this section: April 26, 2026)

---

0. Source discipline and scope

Prior art set used here. Per your instruction, I used the Prior Art / "Citations" section of the patent page as the authoritative universe. That section lists exactly five references:

All five published/granted more than five years before the 2018‑11‑13 priority date, so each qualifies as prior art under AIA § 102(a)(1) (printed publication / patented) and independently under § 102(a)(2) where applicable. None is subject to the § 102(b)(2)(C) common-ownership exception relevant to R4 (see §9 note on shared provenance).

Explicitly not treated as prior art (and I do not auto-correct or re-characterize them):

• US20220349930A1, US11852668B2, US12339305B2 — same-family continuations/publications of the patent at issue. These are not prior art; they are the patentee's own later disclosures.
• The "Similar Documents" list (e.g., US11861739B2, US20170115658A1, EP3591592A1, Streitwieser 2016, Tahvili 2016). These appear on the page but are not in the examiner-cited Prior Art/Citations list; I flag them only as potential supplemental art and do not build the primary grounds on them.

---

1. Legal framework

Obviousness is assessed under the Graham v. John Deere factors (scope/content of prior art; differences; PHOSITA level; objective indicia) as refined by KSR Int'l v. Teleflex, which permits combinations where (a) the references are from the same field or analogous fields, (b) the combination is of known elements performing known functions yielding predictable results, or (c) there is a design incentive / market pressure / obvious-to-try rationale. A conclusory "motivation to combine" is insufficient; the rationale must be articulated with record support (KSR; In re Kahn).

Level of ordinary skill (PHOSITA). For this art I would posit a PHOSITA with a B.S. in EE/CS plus ~3–5 years in semiconductor ATE, test-program development, and statistical test-data analysis (SPL/PAT, SPC), or equivalent experience. The '419 specification assumes working knowledge of ATE, handlers/bins, SPEC/SPL/PAT limits, and STDF data — all of which appear in R1–R5. This is the level used below.

---

2. Key claim terms and how the prior art maps them

---

3. Reference-by-reference technical summary (grounded)

R1 — US6338148B1, "Real-time test controller" (Compaq)

• Test controller (1) controls testing across multiple test locations (TSA, TSB, TSC), connected via communication lines C1–C3 with real-time feedback lines FB1–FB3. It receives real-time feedback of test results and "will re-sequence the tests for a device … if the prior history of testing for such a device suggests the occurrence of test or tests which have a higher incidence of failure."
• Contains a failure database (40) serving "similar PWAs which have undergone testing previously," from which a "fault spectrum" (frequency of failures) is derived, and a diagnostic optimizer (20) that optimizes diagnostic sequence.
• Expressly supports remote/central control: "the present invention can control testing in one location while being positioned in another," and the controller may be "positioned in another" location from the tested device using modems/network.
• Relevance: (i) real-time test-result feedback loop controlling a tester; (ii) use of a historical/aggregate database of previously tested similar devices to set decision thresholds; (iii) adding/omitting/re-sequencing tests; (iv) central control of geographically separate test sites. These feed claims 3/7 (rule-generation facility distinct from test facility) and 4/8 (add/omit tests).

R2 — US20030144810A1, "Methods and apparatus for data analysis" (Tabor)

• "A method and apparatus for testing semiconductors … comprises a test system comprising composite data analysis element configured to analyze data from more than one dataset." Abstract and claim 1: "a composite analysis element configured to analyze the test data for common characteristics among the datasets."
• Operates with ATE (tester 102) and computer system 108; the storage system may be remote ("a hard drive array associated with a server"), or "a remote computer, such as an independent workstation or a third party's separate computer system."
• Performs supplementary analysis at run time ("within a matter of seconds or minutes") including outlier identification and classification, signature analysis, and statistical figures; it classifies parts into critical/marginal/good, defines outlier threshold limits (e.g., scaled sigma from mean/mode/median), and "may also identify user-defined good part patterns … for the purposes of dynamic test time reduction."
• Relevance: supplies the "external test results data of tests involving similar devices" element (multi-dataset/multi-lot/third-party sources) and the outlier-detection-at-run-time concept; supports claims 2/6 (external data types) and 20 (TTR action).

R3 — US20060085155A1, "Methods and apparatus for local outlier detection" (Miguelanez et al.)

• "A method and apparatus for data analysis … configured to identify statistical outliers in test data for components, including local outliers representing outliers within subsets of larger data populations." Claim 1: tester + "a local outlier identification system configured to select a data subset from the test data and automatically identify local outliers in the data subset."
• Test system 100 includes ATE tester 102, a device interface 104 "and/or a device handler"; the computer 108 classifies the component "into a particular bin classification, for example by comparison to the upper and lower test limits."
• Includes an outlier classification element 212 that "utilize[s] multiple candidate outlier identification algorithms and identif[ies] one or more algorithms suited for identifying outliers," automatically selecting algorithms based on the data population type; also a classification engine 3012 applying "predetermined and/or user-defined recipe-driven rules."
• Relevance: teaches re-computing local/dynamic outlier thresholds from the run's own data (the "modifying … based on the test results data" step) and bin classification; also teaches a rule-selection engine — a building block for a "rule engine."

R4 — US20110000829A1, "System and method for binning at final test" (Optimaltest Ltd.)

This is the most material reference and shares substantial disclosure with the '419 specification (same technical family lineage; see §9).

• Recites the same problem statement as the '419 background: ATE, a test station host computer, a handler placing devices into "bins," and binning signals passed ATE→handler. Uses identical terminology: final test, STDF, SPL-type limits.
• Core teaching: a system for sorting a device "into one of a plurality of bins comprising: an evaluator to define a permanent binning assignment for an electronic device inserted in a socket of a final test operation, wherein the permanent binning assignment may or may not concur with a provisional binning assignment … defined by a test program," and to output it while the device is still socketed so the handler bins accordingly.
• Discloses D-PAT (Dynamic Part Average Test) and feed-forward flows for defining the permanent binning, and lists algorithms: SPAT (static PAT), DPAT, moving-limit PAT, "statistical bin limits," "static part average testing," "manufacturing and test data feed-forward," and "feed-backward algorithms," computed from "distributions of parts of common lot, sub-lot, wafer, or sub-wafer origin," plus access to a Manufacturing Database Server for population data.
• Relevance: directly supplies (i) re-binning/overriding the ATE test-program bin, (ii) limits that may be recalculated from a running population (D-PAT/moving-limit PAT), (iii) population/external data sources, and (iv) the "binning while the device is still socketed" timing that maps to "while the device is on a device tester."

R5 — US20130275073A1, "Creation and scheduling of a decision and execution tree of a test cell controller" (Arnold)

• "A method for real-time statistical analysis of test results … After a quantity of test results have been collected, statistical analysis … may be performed and actions may be executed in response to any identified testing errors or defective devices."
• SPC rules "may be executed in a short loop on an exemplary test cell controller," "in synchronicity with normal test cell controller program activities, such as handler and prober equipment communications." Identifies SPC failures used "to trigger corrective or abortive actions," including re-test, needle clean, z-height adjustment, stop test (Table 1), and detects "historical results such as a bin that has had too many failures."
• Crucially: "SPC may also detect process issues on parts that are passing but are not within predicted process limits. For example, while parts may be passing within hard limits, violations of SPC rules, once detected in real-time, can be used to determine that process issues are present" — this is the '419 outlier concept in a nutshell (pass SPEC, fail tighter limit).
• Architecture includes test cell controller 102, testing apparatus 104, material handling equipment 106 (prober 108/handler 110), a database server 112, and a web server 114 "containing control rule recipes and reports." "SPC rules, result analysis, and reporting may be centrally managed," and rules are "created and edited with a web interface … and stored in a relational database"; decision trees are created/executed "during periods of test inactivity."
• Relevance: supplies (i) a rule engine co-resident with the tester/station controller executing rules in real time, (ii) rule action → handler binning/abort/re-test, (iii) centralized rule creation and distribution, and (iv) off-line vs. real-time rule generation — the exact architectural skeleton of '419 claim 9.

---

4. GROUND 1 — Claims 1–4 and 5–8

Proposed rejection: Claims 1–8 are obvious under § 103 over R4 (US20110000829A1) in view of R5 (US20130275073A1) and R3 (US20060085155A1), optionally further in view of R2 (US20030144810A1) and R1 (US6338148B1).

4.1 Element-by-element mapping (claim 1)

Conclusion for claim 1: Each element is disclosed, and the combination is a mere collocation of known ATE functions — dynamic limit computation (R4), real-time rule execution at the cell controller (R5), and local outlier classification (R3) — all operating on the same tester/handler data path. No new physical or functional interaction arises.

4.2 Dependent claims 2–4 (and 6–8)

• Claims 2/6 ("external test results data comprise … device test results at a different facility, higher-level assembly test results, product test results, FA test results") — R2 expressly analyzes multiple datasets across wafers/lots and can reside on "a third party's separate computer system"; R4 uses Manufacturing Database Server population data and feed-forward from wafer sort to final test (a previous test operation/higher-level data flow). R1 aggregates previously tested similar PWAs across multiple test locations. Motivation: an engineer wishing to catch reliability risk earlier would look to the broadest available population — the specification itself concedes this was known ("while typically available at a central 'headquarters' location").
• Claims 3/7 (rules generated at a facility different from the test facility) — R1: "control testing in one location while being positioned in another" via C1–C3; R5: "SPC rules, result analysis, and reporting may be centrally managed," rules stored in a relational DB served by a web server. Direct anticipation-by-combination.
• Claims 4/8 ("instructions to add or omit tests … or SPL applied subsequent to SPL applied by the device tester") — R1: diagnostic optimizer re-sequences/omits tests based on fault spectrum; R4: binning limits (SPAT/DPAT/moving-limit PAT) applied in addition to / subsequent to the test program limits; R5: rule actions include retest, stop, and TTR-type adjustments. R2 explicitly mentions "dynamic test time reduction."

---

5. GROUND 2 — Claims 9–20 (three-tier rule-engine architecture)

Proposed rejection: Claims 9–20 are obvious under § 103 over R5 (US20130275073A1) in view of R2 (US20030144810A1) and R4 (US20110000829A1), further in view of R1 (US6338148B1) and R3 (US20060085155A1).

Claim 9 recites three engines:

Dependent claims 10–20:

• 10 (first engine on-line) — R5 real-time SPC execution; the '419 spec's own "on-line" definition.
• 11–12 (second/third engines off-line) — R5 off-line decision-tree creation; R2 analysis "between test runs or in other batches."
• 13 (manufacturing data includes same-type test results) — R2 composite datasets; R4 feed-forward test data.
• 14 (plurality of facilities includes the testing facility) — R1 (TSA/TSB/TSC). Trivial.
• 15 (first engine at testing facility, generates during testing) — R5 cell-controller-resident execution.
• 16 (first = real-time edge; second = off-line edge) — R5/R3 vs. R2/R5 off-line modes; a routine architectural partition of one computer's functions (and the '419 spec itself concedes the two engines "may be performed by a single component").
• 17 (third engine off-line at central facility connected to multiple edge facilities) — R1/R2 hub-and-spoke.
• 18–19 (second/third engines communicate rules to the first engine) — R5 centralized rule distribution via web/DB server; R1 feedback/communication lines.
• 20 (first action = re-binning/outlier/TTR/abort; second action = re-binning/outlier/tester shutdown/maintenance/ML-TTR) — R5 Table 1 (retest, needle clean, z-height, stop test); R4 re-binning; R2 TTR and diagnostic corrective action. The ML-based variant is supported by R2's statement that the system "may employ … neural networks" and its classifier/feature-selection machinery.

---

6. Motivation to combine (articulated, non-conclusory)

The KSR/In re Kahn rationale is strong here for at least five independent reasons:

1. Same field / analogous art. All five references are in automated test of electronic/semiconductor devices, addressing the same components (ATE/tester, station/host controller, handler/prober, bins) and the same metrics (test limits, yield, bin counts, outliers). There is no field-of-invention barrier.

2. Same problem, expressly recognized in the art. The '419 Background and R4's Background are near-identical (ATE + test station host computer + handler + bins + SPL/PAT + STDF). R5 states its own driver: "to make process decisions beyond simple pass/fail results … before many testing hours have been expended." R3 targets parts that "pass " but stray beyond limits. A PHOSITA would combine references aimed at the same problem of test escapes.

3. Known elements performing known functions → predictable result. Each element does exactly what it did before: R4 defines/re-defines bins from population statistics; R5 runs rules in real time at the cell controller and actuates the handler; R3 classifies local outliers. Combining them yields nothing more than the expected sum of their functions (tight, self-updating limits applied in real time to the binning path). No functional interaction is created that did not previously exist.

4. "Obvious to try" — a finite number of predictable solutions. Given the known problem of tighter/adapting binning limits, the known technique of D-PAT/moving-limit PAT (R4) and the known technique of real-time SPC rule execution at the cell controller (R5), a PHOSITA would naturally place the limit-computation where the data is freshest — the local controller — which is exactly what R5 already does for other SPC parameters.

5. Architectural incentive is explicit in the art. R5 teaches centralized rule authoring/storage (web server + relational DB) with local execution; R1 teaches central control of geographically distributed test sites; R2 teaches multi-dataset/remote/third-party analysis with run-time or batch execution. The three-tier on-line/off-line split of claim 9 is the predictable top-down architecture these references collectively point to.

---

7. Secondary considerations ("objective indicia")

I found no evidence of secondary indicia in the material available to me, and I am explicitly noting the limits of that finding:

• No unexpected results are demonstrated in the specification for the claimed combination as a whole; the described benefits (tighter limits, earlier outlier detection, NPI benefit) are the natural consequences of the known dynamic-limit techniques.
• No evidence of copying, industry praise, licensing-nexus, or long-felt-but-unmet need was recovered (consistent with the earlier litigation-search section, which found no litigation involving this patent; a non-asserted patent typically provides no such record).
• Commercial success cannot be inferred from the assignment history alone (the Oct. 13, 2023 assignment record on the patent is a release of security interest — a financing event, not evidence of nexus to the claimed invention).

Caveat: objective indicia are fact-intensive and I could not exhaustively search for them; absence of evidence here is not proof of absence.

---

8. Where the non-obviousness arguments are strongest (candid counter-analysis)

Faithful § 103 analysis requires identifying the weakest links in the rejection, and I flag these because they are precisely why the examiner allowed claim 1 as amended:

1. "modified for each tested device of the previously tested devices" (claims 1, 5). The granted claim was amended to require per-device, respective modified limits — i.e., every previously tested unit was judged against its own updated limit value, not a single lot-level limit. R4's moving-limit PAT arguably recalculates as parts are tested, but whether it yields a distinct limit value per device (vs. a periodically-refreshed lot limit) is a genuine claim-construction dispute. A patentee would argue this is a specific temporal granularity not squarely disclosed. This is the strongest non-obviousness redoubt.

2. The combination's "external test results data of tests involving similar devices" as a source of the binning limits (as opposed to a diagnostic database). R1's failure database is used for test sequencing, not bin limits; R2's composite analysis is described for fault diagnosis; only R4 ties population data directly to binning limits, and R4 is the patentee's own antecedent work. A patentee could argue R4 does not teach the external/multi-facility dimension that claim 1 implies and claim 2/3 make explicit.

3. Claim 9's specific three-engine division of labor with the recitation that the second engine consumes third-rule input and the first consumes second-rule input. A PHOSITA might combine R2+R5, but the ordered, three-tier, bidirectional rule-dependency is a more specific architecture than any single reference shows. (Note, however, the specification's own concession that two of the engines "may be performed by a single component," which weakens a structural-difference argument.)

Net assessment: Grounds 1 and 2 are reasonably strong but not conclusive; the per-device-limit limitation in claim 1/5 and the recency of the "external-data-driven limit" linkage create a credible but not overwhelming non-obviousness position. I would place ~60–70% confidence that claims 1–8 would be held obvious on these references and ~55–65% for claims 9–20 (slightly lower due to the specificity of the three-engine dependency chain).

---

9. Notes, flags, and confidence

• Provenance flag (not a contradiction, but material): R4 (US20110000829A1) lists assignee "Optimaltest Ltd.," while the patent at issue lists "Optimal Plus Ltd." The earlier litigation section correctly treats the Optimal Plus ↔ National Instruments assignment record as a financing/release event rather than litigation. I make no legal conclusion about common ownership or the § 102(b)(2)(C) exception; I simply flag that R4 is patentee-adjacent art and that a real-world challenger or the patentee would litigate this characterization. This does not affect R4's status as prior art under § 102(a)(1) as a printed publication.
• I did not auto-correct any identifier. All patent/publication numbers, dates, assignees, and inventor names are reproduced literally from the page and from the cited search results.
• Where results contradict training data, I preferred the search results, per your rule — e.g., I relied on the fetched texts of R4 and R5 rather than recollection.
• What I could not verify: I did not open the full prosecution history (the Dec. 6, 2021 non-final action and Mar. 10, 2022 response) that produced the claim-1 amendment. The specific examiner rationale for allowance is therefore inferred from the claim language, not confirmed from the record. If you want, I can attempt to pull the prosecution documents to confirm which limitation the examiner relied on.
• The five references are the entire examiner-cited prior-art universe here; any invalidity theory would also want to consider the "Similar Documents" (e.g., US20170115658A1 "Use of Multivariate Models to Control Manufacturing Operations," EP3591592A1, Streitwieser 2016), which I have not analyzed because they are outside the page's Prior Art/Citations section as you scoped it.

---

Bottom line

Under KSR/Graham, claims 1–8 are prima facie obvious over US20110000829A1 (R4) in view of US20130275073A1 (R5) and US20060085155A1 (R3), with US20030144810A1 (R2) and US6338148B1 (R1) supplying the external-data and multi-facility/central-control teachings; claims 9–20 are prima facie obvious over R5 in view of R2 and R4, with R1/R3 as further support. The combinations rest on same-field references, a shared problem statement, known elements performing known functions with predictable results, and explicit architectural incentives — but the rejection is rebuttable primarily on the "binning limits modified for each tested device" limitation in claims 1 and 5.