US 11402434

I'll search for litigation and docket information on this specific patent number, then compile the summary.

Let me check specifically for any litigation or PTAB activity involving this patent.

Summary: US Patent No. 11,402,434 B2

Important note on scope of search: I was asked to search the USPTO database and CAFC 2026 dockets for patent number 11402434. I ran targeted searches and found no district-court litigation, ITC investigation, PTAB proceeding, or CAFC appeal naming US 11,402,434 (or the number "11402434") as of this review. That is a negative search result, not proof of absence — I do not have authoritative access to a complete litigation database, so I flag this as uncertain. The authoritative bibliographic and claim data below comes from the patent text itself (Google Patents: https://patents.google.com/patent/US11402434/en).

Bibliographic data

Minor discrepancy: Google Patents lists anticipated expiration as 2040-07-02 (20 years from the PCT filing date), while a third-party site lists an "adjusted expiration" of 2041-12-30. I cannot resolve this authoritatively and note it rather than assert one.

Abstract (as granted)

The application provides a battery state estimation method relating to the field of battery technology, including: obtaining a first cell state parameter of each of a plurality of cells in a battery; under a condition that there are multiple cell partitions, determining whether the first cell state parameter of each cell belonging to a cell partition is within a first cell state parameter range corresponding to the cell partition; when it is within the first cell state parameter range corresponding to the cell partition, estimating a cell state of each of the cell partitions based on a target cell state parameter corresponding to the cell partition; and estimating a battery state of the battery based on the cell state.

Plain-language overview of the independent claims

Claim 1 — Method. You first measure/derive a "first cell state parameter" (defined in claim 5 as an OCV curve and/or a cell capacity) for every cell in a multi-cell battery. If the battery's cells have already been organized into multiple "cell partitions," then for each partition you check that every cell assigned to it actually falls inside that partition's corresponding parameter range. Only if all cells pass that check do you estimate a cell state (SOC, SOH and/or SOP per claim 6) for each partition using that partition's own "target" parameter (e.g., a partition-representative OCV-SOC curve / capacity), and then combine the per-partition cell states into a battery-level state estimate. The express condition on partitions is that cells in the same partition must have a deviation of the first cell state parameter within a first deviation range. In short: don't apply one global OCV/capacity model to the whole pack — group like-behaving cells and estimate per group.

Claim 7 — Device. A battery state estimation device comprising a memory, a processor, and a program stored in the memory and executable on the processor, where the program causes the processor to carry out the same sequence of operations recited in claim 1 (obtain parameters → check each partition's cells against its range → estimate per-partition cell state → estimate battery state), subject to the same same-partition deviation condition.

Claim 13 — Non-transitory computer-readable storage medium. Instructions stored on a non-transitory CRM that, when executed by a processor, cause the processor to perform the same operations as claim 1, again subject to the same-partition deviation condition.

Dependent claim themes (mirrored across the method, device, and medium claim sets as claims 2–6, 8–12, and 14–18 respectively):

• Re-partitioning stragglers (claims 2/8/14): if a cell's first cell state parameter falls outside its current partition's range, move that cell to a different partition based on its parameter, then estimate the adjusted partition's cell state.
• Forming partitions on the fly (claims 3/9/15): if no partitions exist yet, partition the cells by grouping those whose parameter deviation falls within the first deviation range.
• How the parameter is obtained (claims 4/10/16): obtain a cell parameter — voltage, current, and/or temperature — and calculate the first cell state parameter from it.
• Definitions (claims 5/11/17): first cell state parameter = OCV curve and/or cell capacity. (claims 6/12/18): cell state = SOC, SOH, and/or SOP.

Practical observations

• Granted claims differ from the published application. The pre-grant publication's "aspects" describe an apparatus with an obtaining module, determination module, first/second estimation modules, etc. The granted independent claims instead cover a method (claim 1), a device with memory/processor/program (claim 7), and a storage medium (claim 13). Anyone analyzing the enforceable scope should use the granted claims, not the published summary.
• Drafting artifacts. Claim 7 uses mixed verb forms ("obtain"/"determine" alongside "estimating"/"estimate"), and the specification reuses "battery management system 400" interchangeably with "battery state estimation system 400," all pointing to a translated filing.
• Cited prior art of record includes US 2019/0190091 A1 (Samsung) and US 2013/0207462 A1 (Boehm), and the EPO search report cited US 2019/0190091 as an "X" reference against the EP sibling EP 3958005.
• Forward citation noted in the record: US 12,633,486 B2 (Milwaukee Electric Tool Corporation) is listed as citing this family, with a 2026-05-19 publication date. I take this identifier literally as recorded; I have not independently verified it.

Uncertainty statement: I could not locate any 2026 CAFC docket, district court case, ITC complaint, or PTAB filing specifically identifying US 11,402,434. If you need a definitive litigation-clearance check, that should be confirmed directly against USPTO Patent Center/Assignment records, PACER/CourtListener, and the ITC EDIS rather than relying on this search.

I'll search for litigation involving US patent 11402434 across patent litigation resources.

Let me search more targeted litigation resources.

Let me try more specific queries focused on US litigation databases.

Findings: Litigation involving US Patent No. 11,402,434

No litigation involving US Patent No. 11,402,434 was identified.

I searched for the specific number across litigation-oriented sources and general web results, and found no federal district court case, ITC investigation (Section 337), PTAB post-grant proceeding, or CAFC appeal naming this patent as an asserted patent.

What I checked and what came back

1. The patent's own Google Patents page (authoritative text provided at https://patents.google.com/patent/[US11402434](/patent/US11402434)/en) contains no "Litigation" section — only prosecution/assignment events and a single third-party citation. This is significant, because Google Patents auto-generates a "Litigation" panel when the patent number appears in US court/PACER records. Its absence is affirmative evidence that no US litigation has been indexed against this patent.
2. Case-list / litigation searches returned no matching case. Queries for the patent number returned only:
   • The unrelated Estonian business-registry code "11402434" (AS Teede REV-2 / AS TREV-2 Grupp), which is a company registration number, not the patent. Per your strict-ID rule, I am flagging this collision explicitly so it is not mistaken for litigation activity.
   • A single forward citation to US 11,402,434 from US 12,633,486 B2 (Milwaukee Electric Tool Corporation, "Device including a fuse"). A citation is not litigation.
3. No PTAB petition, IPR/PGR, or Federal Circuit docket referencing "11,402,434" appeared in results.

Case table

Important caveats and recommended verification

• I could not directly query PACER, Unified Patents' litigation portal, Docket Alarm, or Docket Navigator in this session; those are the sources most likely to reveal a recently filed complaint. My searches were limited.
• This patent issued 2022-08-02 and is owned by Contemporary Amperex Technology (Hong Kong) Limited (assignee; original assignee Contemporary Amperex Technology Co., Ltd., i.e., CATL). CATL has been an active enforcer of a large battery patent portfolio (e.g., its disputes with ATL / Amperex Technology Limited and with Zhuhai CosMX, announced in Chinese exchange filings). That broader portfolio litigation is real, but I found no evidence that this specific patent — US 11,402,434 — has been asserted.
• The patent has active family members (EP 3958005 B1, CN 115151832 B, WO 2022000415 A1). Foreign proceedings (e.g., in China or Europe) would not appear on the US Google Patents litigation panel and I did not identify any.

Bottom line

Based on the sources available to me, there is no known litigation — US or foreign — involving US Patent No. 11,402,434. If you need a litigation-grade clearance, I recommend running a direct PACER/docket search on the patent number and checking the Unified Patents litigation caselist, as those were not directly queryable here and could surface a very recent filing.

I'll verify the structured block's "no proceedings" result against live sources before writing anything down.

Nothing relevant surfaced. Let me run a few more targeted checks before concluding.

PTAB / AIA Trial Analysis — U.S. Patent No. 11,402,434 B2

Patent: Battery state estimation method, apparatus, device, battery system and storage medium
Assignee: Contemporary Amperex Technology (Hong Kong) Limited (originally Contemporary Amperex Technology Co., Ltd.; reassigned 2024-08-06)
Priority date: 2020-07-02 · Granted: 2022-08-02 · Application: US 17/565,460 (continuation of PCT/CN2020/099909)
Statutory disclaimer: I am not your counsel; this is a public-records survey, not a validity or freedom-to-operate opinion.

---

Proceedings overview

Zero AIA trial proceedings are on file for US 11,402,434. The structured PTAB block (USPTO Open Data Portal, canonical source for this task) returns no IPR, PGR, or CBM petition or proceeding, and independent web sweeps of the PTAB docket surface nothing to contradict that — so the breakdown is 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution-denied. The bottom line for a defendant: this patent is completely untested at the Board, un-narrowed by any FWD, and untouched by any § 315(e)(2) estoppel — every one of claims 1–18 is live and carries its original scope. That cuts both ways: the patent owner has no PTAB loss to apologize for, but you also face no adverse-judgment shortcut, and a well-built IPR is still fully available to you.

No proceedings to itemize

I cannot populate the per-proceeding template because there is nothing to populate it with. To be explicit about what I checked, so you can audit the negative:

• USPTO ODP structured block (canonical): no AIA trial proceedings returned.
• Open web search on the patent number, patent title, and assignee + "IPR": no petition, institution decision, FWD, or termination for this patent. Repeated hits resolved to unrelated matters — e.g., Fresenius USA v. Baxter's unrelated '434 hemodialysis patent, and IPRs against other battery patents (U.S. 10,401,350/10,401,351, U.S. 9,077,035).
• Searches for CATL/Contemporary Amperex-initiated or -facing IPRs in this art unit space: nothing matching this patent or this family.

One caveat I will not paper over: PTAB filings can lag public indexing by some weeks, and ODP ingest is not instantaneous. A petition filed in the last ~30–60 days could conceivably exist without appearing in either source. On the evidence available today, though, the correct answer is no PTAB activity.

---

Strategic summary

Claim-level status: everything is UNTESTED. No claim of US 11,402,434 has been canceled, confirmed, or even challenged. The claim set stands at claims 1–18: claims 1–6 (method, with independent claim 1), claims 7–12 (device, independent claim 7), and claims 13–18 (non-transitory CRM, independent claim 13). Each independent claim carries the same core architecture — obtain a first cell state parameter for each of a plurality of cells; where multiple cell partitions exist, verify each cell's parameter against the partition's corresponding range; where it verifies, estimate each partition's cell state from that partition's target cell state parameter; then estimate battery state from the cell state — with the wherein clause requiring the deviation of the first cell state parameter among cells in a same partition to fall within a first deviation range. Claims 2/8/14 (partition readjustment for an out-of-range cell) and 3/9/15 (partitioning from scratch when no partitions exist) are the meaningful narrowing fallbacks, and they are intact. Claims 4–6/10–12/16–18 add the voltage/current/temperature acquisition basis, OCV curve and/or cell capacity as the parameter, and SOC/SOH/SOP as the state.

Estoppel landscape: clean. Because no IPR or PGR was ever instituted — indeed, none was ever filed — no § 315(e)(2) estoppel attaches to anyone. There is no petitioner-side estoppel barring you or your privies, no Patent Owner-side estoppel on claim amendments, and no adverse judgment under § 42.73(b) narrowing the claim set. Practically: the full universe of § 102/§ 103 prior art, including system art and printed publications, remains available to you in a district court invalidity case, and you can file an IPR on any or all of claims 1–18 without arguing about what a prior petitioner "reasonably could have raised." Nothing is foreclosed.

Pattern signals: none of the usual ones. No serial petitioner, no follow-on petitions, no joinder activity, no defensive aggregator (I found no Unified Patents, RPX, or similar involvement) in the chain — the only post-grant-style activity associated with this family is a third-party Chinese prosecution citation (CN115343649B, cited by this family's CN counterpart, not against it) and one unrelated later patent citing the '434 as background. The patent owner, CATL (now Contemporary Amperex Technology (Hong Kong) Limited), has not had to litigate validity at the Board for this patent at all, and there is no Federal Circuit appeal to trace. The patent's expiration is calendared at 2040-07-02, so roughly 14 years of term remain — this is not a patent running out the clock.

Two grounded observations for a defendant. First, the patent's own face gives you a prior-art roadmap: 16–17 references of record (including US 2010/0085009, US 2012/0074898, US 2013/0207462, US 2013/0234719, US 2018/0267106, US 2018/0306865, US 2019/0190091, JP 2019-164969 A, and several CN filings on cell capacity and internal-resistance estimation), plus the extended EPO search report for EP 20923695 and the PCT/CN2020/099909 ISR — that is where a petitioner's Exhibit 1000-series should start. Second, the family is internationally granted and live (EP3958005B1 granted 2023-04-05; CN115151832B granted 2024-08-09), so if you are defending in Europe or China, the EPO opposition window and CN invalidation route are separate tracks from the PTAB analysis above and should be checked independently — I did not find European opposition data within the ODP scope.

---

Recommended next steps

1. If you are a defendant now: treat this as a clean-slate IPR opportunity. There is no FWD to link you to and no estoppel to navigate, so you may file on all of claims 1–18 with any art you can find. Budget for the standard statutory track: ~3 months from petition to institution decision (§ 314(b)), then a 12-month trial clock to FWD (§ 316(a)(11)), extendable up to 6 months for good cause. Serve a Sotera-style stipulation (or broader) to improve your § 314(a) / Fintiv-style discretionary-denial posture, since the Board continues to weigh parallel district court litigation.
2. Prioritize the claim-1 wherein clause. The "first deviation range" requirement is the claim's most constructional soft spot — it is a relative, result-oriented limitation ("within a first deviation range") with no numeric bound in the claims, and its scope depends on the specification's discussion of the deviation range and the "applicable range" of the first cell state parameter. Expect the Board and any district court to fight over indefiniteness or over the breadth of that term (BRI-style at the Board; Phillips-style in court). That asymmetry alone can justify a petition.
3. Run a family-level check. Because the WO application went "Ceased," the enforceable family members are the EP and CN grants. If your exposure is multi-jurisdictional, the EP3958005B1 opposition period and CN115151832B invalidation are distinct leverage points that a US-only PTAB strategy will not touch.
4. Do not assert or rely on the negative as a merits position. The absence of PTAB activity is not evidence the patent is strong; it more likely means no one has yet had commercial reason to spend $300K–$500K attacking it. Well-asserted patents eventually attract IPRs — if this patent enters an active campaign, expect a petition to follow. Monitor USPTO Patent Center and the PTAB docket periodically rather than treating today's clean record as permanent.

Sources / dockets to verify live: US11402434B2 on Google Patents · USPTO Patent Center · PTAB E2E / PTAB Decisions · USPTO Open Data Portal · CourtListener (no CAFC appeal on record for this patent).

I'll verify the assignment record and cross-reference the entities involved.

Assignment & Ownership Analysis — US Patent No. 11,402,434 B2

Patent: US 11,402,434 B2 — "Battery state estimation method, apparatus, device, battery system and storage medium"
App. No.: 17/565,460 (continuation of PCT/CN2020/099909, filed 2020-07-02)
Granted: 2022-08-02 | Priority: 2020-07-02 | Status: Active (4th-year maintenance fee paid 2026-01-21)

Sourcing note (read first): I could not load the USPTO Assignment Center UI (https://assignmentcenter.uspto.gov/) directly in this session. The assignment data below is taken from the patent's own record at https://patents.google.com/patent/US11402434/en#legal-events, which mirrors the USPTO assignment (AS) events for this patent, cross-checked against the EPO/UK-IPO ownership records for the same family. Only one assignment event appears in that record. The Assignment Center's correspondent-of-record field is not exposed in the sources I could reach, so I flag it as not determinable rather than guessing — see Signal 3.

---

Inventors

• Attribution of all five to CATL/Ningde is supported by (a) CATL's common inventor table for the sibling BMS patents these same engineers co-authored (e.g., US 11,366,170; US 11,536,772; US 11,598,811; US 11,609,276; US 11,819,071), and (b) third-party inventor records giving the Ningde location for Ruan/Du/Li (e.g. https://www.wikidata.org/wiki/Q122711257; https://www.patentleaderboard.com/contemporary-amperex-technology-co-limited/shichao-li/35792).
• Discrepancy flagged: one aggregator lists only four inventors for this patent (no Shichao Li). The patent front page and Google Patents list five. The authoritative patent text controls: five inventors.
• Pattern check — inventor departures: No evidence of departures. These five are a tight, recurring BMS co-inventor cluster at CATL who continue to appear on CATL patents issued well past 2022 (through US 12,181,533 in 2024). That is the opposite of the "all inventors leave within 12 months of filing" fire-sale tell. No inventor-to-company assignment was recorded at the USPTO for this case, which is normal for a PCT national-phase filing where the company was the applicant of record from the outset.

---

Original assignee

Contemporary Amperex Technology Co., Limited ("CATL") — Ningde, Fujian, China (addresses of record in family filings: No. 1 / No. 2 Xingang Road, Zhangwan Town, Jiaocheng District, Ningde City, Fujian 352100, CN).

• Business: the world's largest manufacturer of lithium-ion cells, battery packs and battery-management systems for EVs and stationary storage; publicly listed (Shenzhen Stock Exchange, 300750).
• Status: operating — not acquired, not dissolved, not in bankruptcy. It is a large-entity (undiscounted) filer per the 2021-12-30 FEPP event.
• Product practice: CATL ships BMS hardware/firmware across its EV and ESS battery products; the claimed subject matter (per-partition cell-state estimation using partition-specific target parameters) is directly implementable in its shipped BMS. I have not independently verified a specific product SKU practicing claim 1, so this is a "consistent with product practice" call, not a documented one.

The issued patent names CATL as assignee; the current assignee of record is its Hong Kong affiliate (below).

---

Assignment timeline

One post-issuance assignment is recorded.

• 2024-08-06 (executed/effective) / recorded 2024-08-06 — Reel 068338 / Frame 0723
  • Conveyance: Assignment (ASSIGNMENT OF ASSIGNORS INTEREST)
  • Assignor: CONTEMPORARY AMPEREX TECHNOLOGY CO., LIMITED (Ningde, CN)
  • Assignee: CONTEMPORARY AMPEREX TECHNOLOGY (HONG KONG) LIMITED (Level 19, China Building, 29 Queen's Road Central, Central and Western District, Hong Kong)
  • Correspondent: not determinable from the sources available to me. The Google Patents legal-events record for this patent does not expose the recording correspondent, and I could not reach the Assignment Center UI. I therefore cannot state who filed the recording, and I will not guess. (For context, the sibling CATL→CATL-HK recordings in the same batch carried the same reel series — e.g. US 12,113,234 B2 at reel 068338/0402 — which is consistent with a single agent handling the whole batch, but "consistent with" is not evidence of identity.)
  • Context: internal corporate reorganization / intra-group IP consolidation — part of a bulk transfer of CATL's portfolio to its wholly-owned Hong Kong affiliate, not a sale to a third party.

Corroborating family evidence (independent of the US record): the same wave of transfers was recorded at the EPO on the same dates for the sibling EP family members. UK IPO records for CATL EP patents show CATL-HK "registered as Applicant/Proprietor in place of CONTEMPORARY AMPEREX TECHNOLOGY CO., LIMITED … by virtue of deed of assignment dated 02.08.2024" with the "Application under Section 32 filed on 06.08.2024." Same assignee address in every case: Level 19, China Building, 29 Queen's Road Central, Hong Kong. (e.g. https://www.search-for-intellectual-property.service.gov.uk/EP3514879/history and …/EP3832773/history.)

No other assignment events. There is no inventor-to-company assignment, no security agreement, no license, no release, no change-of-name, and no second hop. The chain is a single link: CATL (operating) → CATL Hong Kong (group affiliate).

---

Timeline diagram

timeline
    title Ownership of US 11402434
    2020 : PCT filed by CATL of Ningde
    2021 : US national phase entry
    2022 : US 11402434 B2 granted
         : Published as US 20220120820 A1
    2024 : Assigned to CATL Hong Kong Ltd
         : Recorded 2024-08-06 at reel 068338 frame 0723
    2026 : Fourth year maintenance fee paid

---

NPE / troll-pattern signals

1. Shell-entity transfer — not present. The transferee is a real, long-established corporate affiliate of the assignor: CATL (Hong Kong) has been the group's Hong Kong entity since 2016, with a genuine commercial address at Level 19, China Building, 29 Queen's Road Central (not a registered-agent mail drop), and it is itself an active patent filer — roughly 1,000+ published US applications and 700+ US grants, with new filings as recently as 2025 (https://www.patents-review.com/assignee/[625975](/patent/625975)-contemporary-amperex-technology-hong-kong-limited-hong-kong-cn.html). A licensing-only shell does not generate its own R&D output. Its Chinese-registered business scope is equity/investment holding, which is consistent with a group IP-holding company — but "holding company" here is a wholly-owned subsidiary of an operating manufacturer, not a single-purpose assertion vehicle. Reel 068338/0723, 2024-08-06.

2. Known asserter in the chain — not present. Neither CATL nor CATL (Hong Kong) appears on the asserter lists maintained by RPX, Unified Patents, or any of the enumerated NPE families (Acacia, Marathon, IV, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Round Rock, Spangenberg entities, etc.). The chain contains exactly two entities, both CATL-group. Reel 068338/0723.

3. Repeat correspondent across the chain — unclear (data gap). With only one recorded assignment there is no cross-link recurrence to test in any event, but I also could not retrieve the correspondent of record for reel 068338/0723 from the sources available. The tight sequential frame numbering within reel 068338 (0402 … 0723 among CATL→CATL-HK records all recorded 2024-08-06) implies a single batch filer, but I have no name, so I decline to make this a finding. Recommend verifying directly in the Assignment Center's correspondent field.

4. Cascading transfers — not present. Exactly one assignment in the patent's life, executed and recorded the same day (2024-08-06). No chained LLCs, no rapid succession, no shared-principal intermediaries. The bulk nature of the transfer is breadth (many patents, one deed) not depth (patent passing through several hands).

5. Pre-litigation transfer — not present / not applicable. The prior litigation sweep for this analysis found no lawsuit naming US 11,402,434 in any US district court, the ITC, the PTAB, or the Federal Circuit, and the Google Patents record has no litigation panel. The 2024-08-06 transfer therefore cannot be characterized as pre-suit venue/standing engineering on this patent.

6. Bankruptcy fire-sale — not present. CATL is a solvent, publicly listed, market-leading operating company. No Chapter 7/11, no trustee sale, no distressed disposition. The assignor and assignee are affiliates.

7. Privateering — not present. Privateering requires transfer to an independent NPE asserting on the operating company's behalf. Here the "NPE-looking" recipient is CATL's own wholly-owned Hong Kong subsidiary; the group retains beneficial ownership end-to-end. Worth noting for context: CATL is an active first-party enforcer of its battery portfolio against competitors (publicized disputes with CALB/中创新航, Zhuhai CosMX, and ATL/Amperex Technology Limited). The 2024-08 consolidation of IP into the Hong Kong entity is consistent with a group-wide enforcement strategy — but that is direct operating-company assertion, which is the opposite of privateering. No suit on this patent is known.

8. Defensive aggregator (anti-NPE) — not present. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN; the patent remains with a CATL-group entity and stays Active.

---

Verdict

Operating-company assertion (with a label-fit caveat).

The chain is a single documented hop, CATL (Ningde) → Contemporary Amperex Technology (Hong Kong) Limited, per reel 068338/0723, executed and recorded 2024-08-06 — an intra-group IP consolidation by the world's largest battery manufacturer, corroborated at the EPO by a deed of assignment dated 02–03.08.2024 and a Section 32 filing on 06.08.2024 for sibling family members. Zero of the eight NPE signals are present, and the two strongest structural tells run the other way: the transferee is an active filer in its own right (not a shell) and the ultimate owner is a solvent operating manufacturer that asserts its own patents directly. Caveat on the label: the strict definition of "operating-company assertion" requires the current assignee to ship products and be suing competitors — here the current assignee is the group's Hong Kong holding affiliate and no suit on this patent is known, so a strict reader could also record "Insufficient data." Either way the substance is the same: no NPE pattern; normal corporate IP-holding reorg.

Verify at: https://assignmentcenter.uspto.gov/ — search patent number 11402434 (reel 068338 / frame 0723); Google Patents legal events: https://patents.google.com/patent/US11402434/en#legal-events

Residual data gaps: (1) the recording correspondent for reel 068338/0723, and (2) whether the underlying deed covers a defined patent schedule (the bulk pattern strongly suggests it does, but I have not seen the instrument).

I'll research the cited prior art references for US 11,402,434 to assess potential §102 anticipation. Let me search for key details on the most substantive references.

Let me research the remaining cited references.

Prior Art Analysis — US 11,402,434 B2

Analyst note on dates: The task header says "Current Date: April 26, 2026," while the system date is 2026-09-29 (the earlier sections also used mid-2026 dates). I flag the inconsistency and proceed; nothing in this prior-art analysis turns on which date is correct.

ID verification (strict rule). Searching the USPTO/patent databases for the literal number 11402434 returns exactly one patent: US 11,402,434 B2, "Battery state estimation method, apparatus, device, battery system and storage medium," App. No. US 17/565,460, granted 2022-08-02, assignee Contemporary Amperex Technology (Hong Kong) Limited. As flagged in the earlier Litigation section, the same numeric string also appears as an unrelated Estonian business-registry code (AS TREV-2 Grupp); that is not this patent and is excluded here per your no-auto-correction rule.

---

1. The claim elements that any §102 reference must disclose

Claim 1 (the method; claims 7 and 13 are the device and CRM analogues with the same elements) requires, in substance:

The single-reference anticipation test (35 U.S.C. § 102) requires every element, arranged as claimed, in one reference. As shown below, no cited reference discloses the (B)+(C)+(E) combination, so none is a clean § 102 anticipation of an independent claim; the strongest references are § 103 (obviousness) bases.

---

2. Complete citation-by-citation review

The 16–17 references of record (from the patent's own citation list) are analyzed below. Dates are as listed in the patent record / verified where searched. Where I could only rely on the record's (often machine-translated) titles, I say so.

2.1 The three closest references (deep dive)

R1 — US 2013/0207462 A1 — Boehm et al. (SB LiMotive / Bosch / Samsung SDI)

• Full citation: US 2013/0207462 A1, "Method for Determining at least one State of a Plurality of Battery Cells, Computer Program, Battery and Motor Vehicle." Priority 2010-06-24; published 2013-08-15. (Family: EP 2586090 B1.)
• Substance: Divides the totality of battery cells (e.g., 24 cells) into groups (e.g., 4 groups of 6) and observes each group with an observer/cell model to determine states (SOC and SOH) for the cells. It expressly analyzes the variation σ of a parameter vᵢ across the totality; if σ does not exceed a defined limiting value σ_max, an average change in the parameter Δυₖ is used to determine the state of all cells. Group observation is alternating to reduce computation.
• § 102 at-risk claims: Closest to claim 1's grouping, and to the "target = averaged parameter" concept of element (C); also touches claim 6 (SOC/SOH). However, it does not (i) partition cells so that the intra-partition parameter deviation is within a first deviation range as the defining condition (E), nor (ii) check each cell against its partition's range (B). It comes closest to a § 103 combination with a range-checking teaching rather than a standalone § 102 anticipation. Potentially relevant to claims 1, 3, 6 as a combination base.

R2 — US 2019/0190091 A1 — Kim et al. (Samsung Electronics)

• Full citation: US 2019/0190091 A1, "Method and apparatus estimating a state of battery." Priority 2017-12-18; published 2019-06-20; App. 16/182,938 (family US 11,073,563 B2 / US 11,549,988 B2; EP 3499257 B1).
• Substance: Groups subsets of plural batteries into groups and assigns respective battery models to the groups; determines state of a "target" battery per group from the model + sensing data and states of non-target batteries from a state variation; regrouping in response to a group update event (preset interval, travel distance, or the determined state). Grouping is based on previous state information and voltage values.
• § 102 at-risk claims: Closest to elements (A), (C) and the "adjust the partition" concept of dependent claim 2 (regrouping). It does not disclose partitioning such that intra-group parameter deviation is within a first deviation range (E), nor the per-cell range check (B), nor "target cell state parameter" in the sense of a partition-representative OCV curve/capacity. Note: The earlier sections already flagged that the EPO search report on the EP sibling cited US 2019/0190091 as an "X" reference — i.e., Europe treated it as the most material art — but an "X" citation is a novelty/inventive-step objection, not proof of US § 102 anticipation. Potentially relevant to claims 1, 2, 6.

R3 — US 2010/0085009 A1 — Kang Jung-Soo (LG Chem)

• Full citation: US 2010/0085009 A1, "Cell balancing apparatus and method." Priority 2008-08-08; published 2010-04-08.
• Substance: Measures voltage of each cell, estimates OCV of each cell from a voltage-variation pattern (correcting IR drop), maps OCV+temperature to SOC via a lookup table, and balances cells based on the estimated OCV/SOC.
• § 102 at-risk claims: Relevant to element (A) (obtaining OCV per cell) and to claim 4 (obtain voltage/current/temperature → compute parameter) and claim 6 (SOC). It does not partition cells at all, so it is not an anticipation of claim 1; it is background/cumulative art.

2.2 Remaining references of record

2.3 Family citation (not asserted art, listed for completeness)

• CN 107356874 B (prio. 2017-06-15; granted 2021-02-05; CATL) — "Estimation method, device and system for state of charge of battery pack." This appears in the record as a "Family Cites Families" entry (applicant's own related art), not as a citation against US 11,402,434. It does not form part of the § 102/§ 103 record against this patent.
• Forward citation (not prior art): US 12,633,486 B2 (Milwaukee Electric Tool Corp., "Device including a fuse") cites this patent; a later citation cannot be prior art.

---

3. Bottom-line §102 assessment

Conclusion: Based on the citations of record, no single reference anticipates any of claims 1, 7, or 13 under § 102. The references are overwhelmingly cumulative or § 103 (obviousness) art. The two references that come nearest to the core inventive concept — partitioning a pack into groups and estimating states per group using a group-representative parameter — are Boehm US 2013/0207462 A1 and Samsung US 2019/0190091 A1; a validity challenge to this patent would most plausibly be framed as an obviousness combination of one of these (for grouping + representative-parameter estimation) with a range-checking/partition-by-deviation teaching. The remaining US, CN, JP and the CATL utility-model references supply only isolated claim elements (OCV-curve modeling, per-cell capacity estimation, SOC/SOH estimation, voltage/current/temperature acquisition).

---

4. Caveats and limitations (per your no-fabrication rule)

1. I relied on the patent's own citation list for the full set of 16–17 references and, for several (R4–R16), on their titles/abstracts only. The Chinese and Japanese titles in the record appear machine-translated; I did not verify the full text of CN 103901354 A, CN 108549032 A, CN 109164398 A/B, CN 109596985/6 A, CN 110031770 A, CN 110568362 A, JP 2019-164969 A, or CN 206878453 U. My descriptions of those are therefore provisional.
2. I could not retrieve full-text detail for US 2012/0074898 A1 (R4) and US 2013/0234719 A1 (R5) within the search budget; descriptions rest on title/assignee only.
3. § 102 vs. § 103: The task asked specifically about § 102. For the reasons stated, the cited art is not a clean anticipation bar; treat the above as the maximum § 102 exposure, with § 103 as the realistic invalidity theory.
4. EPO status: The earlier sections noted an EP "X" citation of US 2019/0190091 against the EP sibling EP 3958005 B1. That is a European novelty/inventive-step objection and does not establish US § 102 anticipation; I have not independently re-verified which reference the EPO designated "X."
5. Timeliness: This review is current only as of the sources retrieved. A full prior-art/invalidity clearance should re-run the citation set against the complete USPTO Patent Center file wrapper (including any IDS references not surfaced on the public Google Patents page) and against non-DOCDB art.

I'll ground this analysis in the actual prior art of record. Let me pull details on the key references.

Let me pull detail on the remaining cited references that matter for grouping/partitioning.

Obviousness Analysis — U.S. Patent No. 11,402,434 B2 under 35 U.S.C. § 103

Patent: US 11,402,434 B2, "Battery state estimation method, apparatus, device, battery system and storage medium"
Priority date: 2020-07-02 (PCT/CN2020/099909) · Grant: 2022-08-02 · Assignee: Contemporary Amperex Technology (Hong Kong) Ltd.
Analyst's note: I am not counsel; this is a technical § 103 assessment, not a legal opinion. All § 103 quotations below are from the prior-art references of record on the face of the patent, as retrieved live, and are cited to the specific reference. Where I am inferring rather than quoting, I say so.

Ground rule applied: Per the strict-ID instruction, every alphanumeric identifier below is reproduced literally as it appears in the patent's Citations listing (e.g., US 2019/0190091 A1, US 2013/0207462 A1, US 2010/0085009 A1). I do not normalize or "correct" them. Where I build on the previously generated sections, I flag it rather than repeat it.

---

1. Legal framework and level of ordinary skill

Governing standard. Under Graham v. John Deere, obviousness turns on (1) scope and content of the prior art, (2) differences between the prior art and the claims, (3) the level of ordinary skill, and (4) secondary considerations. Under KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), a combination need not be taught by the references themselves where the improvement is "the product of ordinary creativity, not of invention," and a finite number of identified, predictable solutions creates a strong inference of obviousness.

PHOSITA definition (proposed). A person having ordinary skill in the art would be a battery-management engineer (or equivalent) with a bachelor's degree in electrical, electrochemical, or mechanical engineering and roughly 2–4 years of experience in battery management system (BMS) algorithm development, including hands-on work with state-of-charge (SOC) / state-of-health (SOH) estimation for multi-cell lithium-ion packs. The subject matter is a mature, well-populated art — evidenced by the fact that this very patent's background section concedes the well-known nature of the problem it addresses (cell-to-cell consistency differences and differential aging). The prior art is therefore dense, and a POSITA has strong incentives to combine teachings from within it.

Art analytics. Every reference of record predates the 2020-07-02 priority date. The patent is directed to estimation of battery state in a multi-cell pack, classified G01R 31/396, G01R 31/3842, G01R 31/392. All the substantive references below are in that same field.

---

2. What the claims require — element breakdown of claim 1

Claim 1 (independent, method) decomposes into five limitations:

Dependent claims add: partition readjustment (2/8/14), partition creation from scratch (3/9/15), V/I/T-based parameter derivation (4/10/16), first cell state parameter = OCV curve and/or cell capacity (5/11/17), cell state = SOC/SOH/SOP (6/12/18). Claim 7 is a device claim and claim 13 a non-transitory CRM claim, both reciting the same A–E operations.

Critical construction point (cross-referenced to the earlier PTAB section). As flagged there, limitation E — "within a first deviation range" — is a relative, result-oriented limitation with no numeric bound in the claims. Reclaiming it now: that breadth weakens limitation E as a § 103 gasket. Any prior-art grouping that is premised on parameter similarity will, on any reasonable reading, satisfy "a deviation … within a first deviation range," because the specification itself defines that range functionally ("the applicable range in which the difference values between all the first cell state parameter values in the corresponding cell partition falls"). A limitation that reads on any "similar-enough" grouping adds essentially nothing over the grouping concept itself, and is separately vulnerable under § 112(b) (see § 8 below).

---

3. The prior-art arsenal on the face of the patent

---

4. Ground 1 (primary): US 2019/0190091 A1 in view of US 2013/0207462 A1

This is the strongest ground and the one I would lead with. It renders claim 1 (and its mirrors in claims 7 and 13, and dependents 2, 3, 6/12/18) obvious.

4.1 Why these two references

Both references are in the identical field, address the identical problem, and disclose the identical architectural solution — grouping a multi-cell battery's cells and applying a group-specific model rather than a single global model.

• Samsung '091 discloses grouping cells into groups, assigning a respective battery model to each group, estimating state from the group's own model, regrouping on a group-update event, and using grouping criteria based on prior state information and voltage. Samsung's stated purpose is efficiency, but the architecture is claim 1's architecture.
• Boehm '462 supplies precisely the missing quantitative glue: it teaches grouping the totality of cells, assigning an observer structure (model) to each group, and applying a statistical range test on a cell parameter — "the variation σ … not exceeding a previously defined limiting value σmax" — before using a group-representative value. Boehm thus expressly discloses both limitation B (a range check on the cell parameter) and limitation E (a deviation-based grouping condition).

4.2 Element-by-element mapping for claim 1

Result: every limitation of claim 1 is disclosed by the Samsung '091 + Boehm '462 combination.

4.3 Motivation to combine

A POSITA would have been motivated to combine these references for at least five independent, evidence-based reasons — the KSR factors are all satisfied:

1. Same field of endeavor. Both are BMS algorithms for multi-cell lithium-ion packs in EV/ESS applications. Samsung's own text contemplates "a vehicle using the plural batteries as a power source"; Boehm's contemplates "a motor vehicle." Combining references within a single, narrow field is the paradigm of obviousness.
2. Same problem, articulated in both references. Both identify the identical problem: the computational cost of modeling every cell individually. Boehm states its object is that "the computational complexity and memory requirements of the battery management system can be significantly reduced" while "the observation quality is hardly decreased." Samsung's whole switching/grouping scheme exists to reduce per-period modeling load. When two references attack the same problem with the same solution class, the combination is obvious under KSR.
3. The references expressly invite the combination. Boehm hands the POSITA the exact validation criterion Samsung lacks: a parameter-variation threshold (σmax) defining when cells may be treated as a group. Samsung hands the POSITA a mechanism Boehm lacks: regrouping on a group-update event to keep groups current as cells age. Each reference supplies what the other is missing, with no change in principle of operation.
4. Predictable, finite, identified solutions. "Group cells by similarity and apply a group model" is one of a small number of predictable BMS design choices. KSR holds that where a finite number of predictable solutions exists and the reference points to them, the combination is obvious. Neither Samsung nor Boehm even needs to "point" — both are the solution.
5. The patent's own background concedes the motivation. The '434 specification states that "since each cell may have consistency differences in the manufacturing process, the basic parameters of respective cells may be different, and in actual use, the aging degrees of respective cells are different, so there may be a large estimation error." A reference's — or a patent's own — recognition of a problem is strong evidence of motivation to solve it. The '434 patent's stated motivation and the prior art's stated motivation are the same sentence.

4.4 Reasonable expectation of success

Both references successfully practice grouping-and-group-model estimation — Samsung for state estimation output, Boehm for SOC/SOH by observer. A POSITA combining them would expect at least Samsung's estimation performance with Boehm's efficiency improvement, and would have no technical reason to expect failure. There is no teaching away in either reference and no criticality that the combination would violate.

4.5 No teaching away

I found no statement in Samsung '091, Boehm '462, or any other reference of record that disparages grouping cells by parameter similarity, applying group-specific models, or validating group membership against a deviation range. The absence of a teaching away removes a common non-obviousness argument.

---

5. Dependent claims 2–6 (and mirrors 8–12, 14–18)

Because claims 7–12 and 13–18 mirror claims 1–6 with only a statutory-class change (device/processor vs. method; non-transitory CRM vs. method), the same grounds apply. The class-change claims are separately obvious because Samsung discloses "an apparatus with battery state estimation include[ing] a processor," and Boehm discloses "a computer-readable storage medium on which a program is stored." Reciting the method on a processor-readable medium is routine and carries no patentable weight.

---

6. Independent claims 7 (device) and 13 (CRM)

• Claim 7 recites "a memory, a processor, and a program … executable on the processor." Samsung '091 discloses a "processor-implemented battery state estimation method" and "an apparatus … includ[ing] a processor," with the identical operational content (elements A–E). The device claim therefore rises and falls with claim 1. Configuring a disclosed algorithm as a memory-stored program on a processor is the routine implementation of software, not a patentable advance.
• Claim 13 recites a "non-transitory computer-readable storage medium." Boehm '462 expressly provides "a computer-readable storage medium on which a program is stored which … permits said data processing device to carry out the method." Combined with the Samsung algorithm of claim 1, claim 13 is obvious; a non-transitory CRM is the standard statutory vehicle for such a program.

---

7. Alternative / backup grounds

Should the Samsung '091 + Boehm '462 pairing be resisted on the theory that Samsung's grouping is for computational scheduling rather than accuracy (a distinction the claim does not draw — the claim is silent on why partitions exist), the following fallbacks should be preserved:

Ground 2 — Boehm '462 as primary, Samsung '091 as secondary. Boehm alone discloses grouping the cell totality (limitation B/E), assigning observers/models groupwise (limitation C), and producing SOC/SOH (limitation D). Samsung supplies the processor/apparatus and the state-parameter acquisition. The motivation is unchanged.

Ground 3 — Adding the OCV/capacity art for claim 5 regardless of primary pairing. Kang '009 (OCV curve), GM '865 (OCV–SOC relationship), CN 110031770 A and CN 109164398 (cell capacity) independently supply the claim-5 parameter definitions. Since claim 5 is a dependent claim adding a parameter definition, and the claim-1 "first cell state parameter" is otherwise unbounded, claim 5's added matter is squarely disclosed by the references of record.

Ground 4 — CN 103901354 A / CN 108549032 A for claim 6's SOC/SOH. These supply the "cell state" definitions independently, though Samsung and Boehm already do.

What the secondary references do NOT cure. None of the fallback references independently supplies the full combination of (i) partition-specific target model, (ii) pre-estimation validation of each cell against the partition range, and (iii) the same-partition deviation condition. That combination is only fully assembled by the Samsung + Boehm pairing. The secondary references are therefore best used as dependent-claim grounds, not as standalone attack bases.

---

8. Secondary considerations and anticipated rebuttals

• Unexpected results / criticality: None apparent on the face of the patent. The described benefit (better estimation accuracy by not using "a unified standard") is the expected and inherent consequence of per-partition modeling, which both primary references already achieve or approximate. No data in the specification compares the claimed method against a global-model baseline, so no unexpected-results showing is preserved.
• Long-felt need / failure of others: Not supported by the record. The problem was recognized and solved in the prior art (Samsung 2017/2019; Boehm 2010/2013), so the need was not "long-felt" or unmet.
• Licensing / commercial success: No nexus evidence available from this record. (The earlier sections found no litigation, so there is no assertion history to mine for commercial-success indicia.)
• "Different purpose" argument: A defendant should expect the patentee to argue that Samsung groups cells for scheduling efficiency while the '434 groups for estimation accuracy, so the references teach away or are non-analogous. This should be rebutted on three grounds: (i) claim 1 is silent as to the reason for partitioning, so the distinction is not in the claim; (ii) both references state the same dual goals — Boehm explicitly claims "observation quality is hardly decreased" while reducing complexity; and (iii) grouping criteria based on parameter similarity is the same technique regardless of the label attached.
• Prosecution-history caveat (flagged honestly): The U.S. record shows a "RESPONSE TO NON-FINAL OFFICE ACTION ENTERED" on 2022-04-21 followed by a "NOTICE OF ALLOWANCE" on 2022-04-29 — some rejection was overcome. The provided record does not show which art drove that rejection, and the EPO Extended Search Report (2022-06-03) post-dates the U.S. allowance, so the "X" citation of US 2019/0190091 against the EP sibling may never have been before the U.S. examiner. That gap is a genuine opening for a § 103 challenge that the examiner may not have considered.

---

9. Rebuttal risk — the two genuinely contestable limitations

In candor, two limitations are the ones a patentee will defend, and each is the subject of a caveat I will not paper over:

1. Limitation B's "determining whether … within a first cell state parameter range." Samsung '091 describes grouping and regrouping but does not, in the abstract lines retrieved, expressly recite a distinct post-grouping validation query per cell against a stored range. The mapping relies on Samsung's similarity-based grouping plus Boehm's σ ≤ σmax test, read together. A patentee will argue these are different operations (group-forming vs. group-verifying). Counter: Boehm's range test is a validity determination on a parameter, and Samsung's "regrouping … based on the determined state information" is a re-check. The combination argument is strong but not airtight on this element alone.
2. Limitation E's "first deviation range." As the earlier PTAB section already flagged, and as I emphasized in § 2, this term has no numeric bound in the claims and is defined only functionally in the specification. It cuts both ways: it makes the claim broad and easy to read onto prior art (good for the challenger), but it also invites a § 112(b) indefiniteness counterattack that the patentee could use to argue the term must be read narrowly to the specification's "applicable range." Nautilus (2024) raises the bar for indefiniteness, but a BMS-relevant, purely functional "range" recitation remains a plausible § 112 target in parallel.

I have not verified the internal figures/tables of Samsung '091's grouping paragraphs line-by-line; the quotes above are from Google Patents and patnets-review.com abstracts of US 2019/0190091 A1, which are consistent with each other. If this analysis is used in an adversarial setting, the Samsung specification ¶¶ describing "grouping" and "group update" should be pulled in full and pin-cited.

---

10. Bottom line

US 11,402,434 is, in my assessment, vulnerable to a § 103 challenge on all 18 claims.

• Claims 1–3, 6 and their mirrors (7–9, 12–15, 18) would be obvious over US 2019/0190091 A1 (Samsung) in view of US 2013/0207462 A1 (Boehm). Samsung supplies A–D (group, group-specific model, re-group, output); Boehm supplies B and E (parameter-variation threshold σmax and deviation-based grouping) and adds SOC/SOH.
• Claims 4, 10, 16 would be obvious over the same primary combination in further view of US 2010/0085009 A1 (Kang), US 2018/0267106 A1 (Denso), and/or US 2013/0234719 A1 (Yazaki) for the V/I/T acquisition + calculation.
• Claims 5, 11, 17 would be obvious over the primary combination in further view of Kang '009 or GM US 2018/0306865 A1 (OCV curve) and/or CN 110031770 A / CN 109164398 A/B (cell capacity).
• The device (7–12) and CRM (13–18) claims are obvious over the same grounds plus routine software/medium implementation, since Samsung discloses a processor-implemented apparatus and Boehm discloses a computer-readable storage medium storing a program.

The single most important observation is that this is a "group-like-cells-and-use-a-group-model" invention. That concept, and the statistical criterion for deciding which cells are "like" enough, are both squarely in the prior art of record (Samsung's grouping/regrouping; Boehm's σmax test). The claimed advance is the application of a known grouping technique to a known estimation problem, achieving the expected improvement — the KSR definition of obviousness.

Caveats I will not soften: (i) the express per-cell range-validation step (limitation B) is the weakest link in the mapping and rests on a combination argument rather than a single-reference disclosure; (ii) limitation E is broad-but-attackable, so the § 112(b) and § 103 attacks should be developed together, not sequentially; and (iii) the U.S. examiner apparently overcame at least one rejection to allow this patent, so the prosecution history should be obtained and read before committing to this ground in a contested proceeding.