US 9625533

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

Let me get the exact claim text and check specifically for any litigation involving this patent number.

US Patent 9,625,533 B2 — Analyst Summary

Sources consulted: Google Patents (https://patents.google.com/patent/[US9625533](/patent/US9625533)/en), Justia Patents (https://patents.justia.com/patent/9625533), and FPO/Patents-Review/EPO (OEPM) records. I did not rely on web results where they conflicted with the authoritative full text.

1. Bibliographic Data (verified against the patent front page and Justia)

Foreign family members: WO 2014/172449 A1, EP 2986994 B1 (granted Mar. 24, 2021), CN 105122074 B.

2. Abstract (as issued)

A method including predicting terminal voltage of a battery module in a vehicle using a battery control module, where predicting includes determining a gassing current of the battery module using a gassing current model (the gassing current quantifies terminal current that is not used to charge the battery), and calculating the predicted terminal voltage based at least in part on a measurement model and the determined gassing current; measuring terminal voltage with a sensor; and determining a corrected state of the battery module by minimizing the difference between predicted and measured terminal voltage.

3. Technical Core (for context on the claims)

• Lead-acid batteries lose coulombic efficiency through a side gassing reaction (H₂O → H₂ + O₂ at high SOC / high charge voltage), so coulomb counting drifts — the patent's FIG. 8C example shows a coulomb-counted SOC erroneously exceeding 100%.
• The battery is modeled as an equivalent circuit (OCV source 48, resistor R₀ 50, RC pair C₁ 52 / R₁ 54) with a gassing branch 56 in parallel with the terminals, so V_g = V_t.
• Gassing current model, eq. (4): I_g(k) = V_t(k−1)·G_P0·exp[ V_t(k−1)/V_t0 + A_g(1 − θ(k−1)/θ_f) ], with V_t0 = gassing voltage, θ_f = electrolyte freezing temperature (characteristic parameters), and G_P0, A_g = tuning parameters.
• Measurement model, eq. (2)/(6): V_t(k) = OCV(k) + I_m(k)R₀(k) + V₁(k), rewritten as V_t(k) = H·x̂_k + I_m(k)R₀(k), where I_m = I_t − I_g ("main current" actually stored).
• A state-space model (eq. 5) plus the measurement model drive a Kalman filter recursion (predicted covariance eq. 7, gain eq. 8, corrected state eq. 9, corrected covariance eq. 10), producing a corrected SOC used for auto-stop control (engine disable/restart, charge voltage control, target 80–100% partial SOC).

4. Independent Claims — Plain-Language Overview

The granted patent has 24 claims. Based on the claim text I could directly verify, the independent claims are 1, 12 (medium), 14 (vehicle), 22 (vehicle), and 23 (method) — see the uncertainty note below.

Claim 1 — Method (core claim).
A method with three steps: (a) predict the battery module's terminal voltage using a battery control module, where predicting means (i) determining a gassing current with a gassing current model that quantifies terminal current not used to charge the battery, and (ii) calculating the predicted terminal voltage from a measurement model and that gassing current; (b) measure the terminal voltage with a sensor coupled to the battery control module; and (c) determine a corrected state of the battery module by minimizing the error between the predicted and measured terminal voltages. Plainly: build a physics-based voltage prediction that accounts for wasted gassing current, then use the mismatch against a real voltage measurement to correct the battery's estimated state (effectively a Kalman-type observer).

Claim 12 — Tangible, non-transitory computer-readable medium.
The same three-step logic (predict voltage using a gassing current model + measurement model; measure voltage via sensor; determine corrected state by minimizing error) expressed as instructions executable by a processor of a vehicle battery control module. Functionally the software counterpart of claim 1.

Claim 14 — Vehicle.
A vehicle having (i) a battery module, (ii) a battery control module that determines a corrected state of the battery module, and (iii) a vehicle control module that controls vehicle operation based at least in part on that corrected state, where the corrected state is determined by predicting terminal voltage using a measurement model and a gassing current model. Dependent claims 15–21 narrow this to specific controls: deciding whether to disable the engine at idle (15), when to restart an engine disabled in auto-stop (16), when to charge (17), at what voltage to charge (18), SOC as the corrected state (19), gassing reaction quantification (20), and lead-acid water dissociation (21).

Claim 22 — Vehicle (system-level).
A vehicle with (i) a battery module supplying electrical power to the vehicle's electrical systems, (ii) a battery control module that recursively determines a corrected state using a state space model and a measurement model predicting an operational parameter, the measurement model containing a gassing current module that lets the control module determine the current actually used to charge the battery, and (iii) an internal combustion engine that disables when the vehicle is stationary and restarts when propulsion is desired. This is the auto-stop system claim.

Claim 23 — Method (broader, with explicit recursion and calibration).
A method combining: predicting terminal voltage via the gassing current model + measurement model; measuring terminal voltage with a sensor; predicting battery state from a state space model, the prior time step's corrected state, and a prior-step operational parameter (with predicted voltage based on the predicted state); periodically calibrating the state space model by fully charging the battery; and determining the corrected state by minimizing predicted-vs-measured voltage error. Plainly: the recursive-observer method of claim 1, plus the periodic full-charge re-initialization.

Representative dependent claims: claim 2 (predict state via state space model, prior corrected state, prior operational parameter); claim 5 (corrected uncertainty fed forward to the next time step); claim 6/14-published (compute main current = terminal current − gassing current); claim 7 (control the vehicle based on corrected SOC); claim 10 (train the gassing current model offline by determining characteristic and tuning parameters).

⚠️ Uncertainty note on independent-claim identification: I directly verified the full text of claims 1 (method), 22 (vehicle), and 23 (method), and confirmed from the granted-claim listing that claim 14 is a "vehicle" claim (claims 15–21 all recite "The vehicle of claim 14"). The pre-grant publication lists claim 12 as the computer-readable-medium claim; the grant renumbered claims relative to the publication, so the medium claim in the issued patent may sit at claim 12 or 13. I recommend confirming the exact independent-claim set against the USPTO Patent Public Search / PatentCenter "Claims" tab or the granted PDF before relying on it for a validity or infringement opinion.

5. Litigation / CAFC 2026 Docket Check

I found no evidence of any litigation, CAFC appeal, ITC action, or PTAB proceeding involving U.S. Patent 9,625,533.

• Targeted searches for this exact number plus "lawsuit," "infringement," "complaint," "district court," and "CAFC 2026 docket" returned no docket, complaint, opinion, or case caption naming US 9,625,533. All hits were either the patent's own documents or unrelated cases that merely happened to contain similar strings (e.g., TTAB proceeding 92053383, a UK patent priority number 9625533.6 in JP 4032136, a table in WO 2011/022420 — none of these relate to this patent).
• The dominant hits for "9625533" in a litigation-style query were false positives from unrelated matters (Spencer Gifts/Urban Bratz TTAB, Lexmark v. Nuance, AuthenTec/Atrua, Cerence v. Microsoft/Nuance, Yondr counterclaims). None asserted this patent.

Caveats on this negative finding: (1) I could not directly query the CAFC's PACER/RSS docket system or USPTO PTAB/PatentCenter APIs in this session; my search relied on public web indexing, which is not a complete docket index. (2) Patents in this family are held by CPS Technology Holdings LLC (the former Johnson Controls Power Solutions business, later a Clarios-affiliated entity), which has been an active patent litigant in other matters; the absence of hits here should not be read as a definitive all-courts clearance. For a definitive check, query the CAFC docket, PACER, and the USPTO Patent Trial and Appeal Board End-to-End (PTAB E2E) system by patent number 9,625,533.

6. Bottom Line

US 9,625,533 is a CPS Technology Holdings (ex-Johnson Controls) patent, issued Apr. 18, 2017, to inventors Daniel B. Le and Brian C. Sisk, claiming a Kalman-filter-style battery observer for 12 V lead-acid auto-stop/start-stop systems whose distinguishing feature is a gassing-current model embedded inside the measurement model so that the current wasted on water dissociation is subtracted before state-of-charge correction. Its independent coverage spans a method (claims 1, 23), a computer-readable medium (claim 12), and a vehicle (claims 14, 22), with claim 23 adding periodic full-charge calibration and claim 22 tying the estimation to an engine that auto-stops when stationary. I found no litigation trace for this number, and I flag moderate uncertainty on the grant's exact independent-claim set (claims 12–14 renumbering) pending confirmation in USPTO PatentCenter.

I'll search for litigation records involving US patent 9625533.

Let me search more specifically for this patent in litigation databases and for the current assignee (CPS Technology Holdings/Clarios).

Let me try more targeted searches, including the assignee's enforcement activity and litigation databases.

Summary of Findings — U.S. Patent No. 9,625,533

No litigation involving U.S. Patent No. 9,625,533 was found.

Despite multiple targeted searches (including queries for the patent number, its asserted use, and the current assignee CPS Technology Holdings LLC / Clarios), I was unable to identify any district court case, Federal Circuit appeal, PTAB proceeding, ITC action, or foreign litigation in which U.S. Patent No. 9,625,533 has been asserted or challenged.

What the searches did surface (for transparency)

The search results that contained the numeral string "9625533" were not references to this patent, and I am flagging them so they are not mistaken for litigation:

These matches are coincidental digit-strings and should not be treated as litigation evidence for the '533 patent.

Patent posture (from the authoritative full text / Google Patents record)

• Patent: US 9,625,533 B2 — "Lead acid state of charge estimation for auto-stop applications"
• Inventors: Daniel B. Le; Brian C. Sisk
• Assignee chain: Johnson Controls Technology Company → CPS Technology Holdings LLC (assignment recorded 2019-06-20)
• Security interests recorded: Citibank N.A. as collateral agent (First Lien and ABL Patent Security Agreements, 2019-08-29) — these are financing instruments, not litigation
• Legal status: Active; adjusted expiration 2035-06-21
• Family: US 2014/0309911 A1; WO 2014/172449 A1; EP 2986994 B1; CN 105122074 B

Important caveat on completeness

I could not reach the dedicated litigation databases directly within this session (e.g., the Unified Patents litigation portal, PACER/CM-ECF dockets, Docket Alarm, Lex Machina, or Darts-ip), and I did not obtain a positive confirmation of "no cases" from those authoritative sources — only the absence of any results indicating otherwise. Accordingly, my conclusion is stated as "no litigation known/found," not as a certified negative.

To confirm before relying on this, I recommend verifying against:

1. Unified Patents Litigation Portal — https://portal.unifiedpatents.com/litigation (search by patent number)
2. PACER / CourtListener RECAP — full-text and party search for "9,625,533"
3. USPTO Patent Trial and Appeal Board — for any IPR/PGR/CBM involving the patent
4. USPTO Assignment Search — to confirm the CPS/Clarios ownership chain and any transfers that might precede enforcement
5. Commercial docket analytics (Lex Machina, Docket Alarm, Darts-ip) — for any sealed or recently filed complaints referencing this patent

If you would like, I can attempt a deeper pass specifically through the Unified Patents portal and CourtListener RECAP in a follow-up.

I'll verify the structured block's "no proceedings" signal against public sources before drafting.

Important: the top hits concern U.S. Patent 9,651,533 (Omni MedSci), a different patent from 9,625,533. Per my operating rules I won't conflate them. Let me search specifically for the lead-acid/SOC patent and its owner.

Proceedings overview

Zero AIA trial proceedings are on file for US 9,625,533 — no IPR, no PGR, no CBM — so the breakdown is 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution denials, and the defensive posture for a defendant is the unusual one: this patent has never been tested at the PTAB, so there is no IPR record, no claim-cancellation certificate, and no § 315(e)(2) estoppel in place — but equally, no petitioner has blazed a trail of prior art that you can reuse.

No proceedings to report

The canonical structured block from the USPTO Open Data Portal returns no AIA trial proceedings for US 9,625,533, and my independent web searches found no IPR, PGR, or CBM petition naming this patent. I found no institution decision, no Final Written Decision, no termination, and no Federal Circuit appeal arising from a PTAB trial on this patent. I am not going to manufacture a proceeding number to fill the template — there is nothing to fill.

Context worth knowing about the patent itself, from the authoritative specification text:

• Title: Lead acid state of charge estimation for auto-stop applications
• Inventors: Daniel B. Le; Brian C. Sisk
• Original assignee: Johnson Controls Technology Company → CPS Technology Holdings LLC (2019-06-20 assignment), the entity now associated with Clarios
• Priority date: 2013-04-16 (provisional 61/812,534); filed: 2014-04-15; granted: 2017-04-18
• Adjusted expiration: 2035-06-21
• Claim set: not reproduced in the fetched record I was given, so I decline to state which claims are independent or how many exist. The disclosure describes at least a method embodiment, a tangible non-transitory computer-readable medium embodiment, and a vehicle embodiment.
• Foreign family: WO2014172449A1, EP2986994B1 (granted 2021-03-24), CN105122074B. Notably, the Spanish register entry for the EP member records "Oposiciones no presentadas" (no EPO oppositions filed) as of 2022-01-04 — that is EPO post-grant data, not a PTAB proceeding, and I flag it only as corroboration that the family has not drawn concentrated validity attack.

Search-artifact caution: US 9,651,533 is a different patent

Nearly every search hit for a "533" patent at the PTAB concerns US 9,651,533 (Omni MedSci, Inc. — wearable physiological measurement), not US 9,625,533. The two numbers differ by one digit and are constantly confused. Per my operating rules I read the numbers literally and do not auto-correct; these proceedings are not proceedings on the patent you asked about. Do not import them into a validity analysis of 9,625,533.

For completeness, the '651' patent's PTAB history surfaced in search includes IPR2019-00913 and IPR2019-00916 (Apple Inc. v. Omni MedSci, Inc.), IPR2020-00029 and IPR2020-00175, and more recent Samsung filings IPR2025-01250 / IPR2025-01252 and PGR2025-00064. Reported in those exhibits: the Board in the '533 IPR FWD (Paper 39, 2020-10-14) held independent claims 5 and 13 unpatentable over Lisogurski and Carlson, and the appeal was docketed at the Federal Circuit as Omni MedSci, Inc. v. Apple Inc., No. 21-01229. I did not verify the disposition of that appeal, and none of it has any bearing on 9,625,533.

Strategic summary

Claim status on 9,625,533: entirely UNTESTED. Because no petition was ever instituted, there is no PTAB certificate under § 318(b) cancelling or confirming a single claim, and no claim of this patent has been held unpatentable or patentable by the Board. If you are weighing an IPR, you would be the first challenger — meaning you get no free roadmap of the examiner's and Board's treatment of the art, but you also face no adverse claim-construction record and no prior petitioner's § 315(e)(1) estoppel to inherit. Any assertion letter citing this patent today cites claims whose validity has never been adjudicated by the Office post-grant.

Estoppel landscape: empty, in both directions. § 315(e)(2) estoppel attaches only to a petitioner that obtains an FWD; with no petitioner there is no estoppel restricting anyone. Practically, the entire universe of prior art remains available to you — in an IPR, in an ex parte reexamination, or in district court. The absence of an estoppel bar cuts both ways: the patent owner has never been forced to defend the claims on a PTAB record, so there is no prosecution-history narrowing from an IPR to exploit, but there is also no Board decision you can point to as res judicata-adjacent leverage.

Pattern signals: none. There is no repeat-petitioner pattern, no defensive aggregator (no Unified Patents or similar) in the chain, and no PTAB appeal history for this patent. The ownership chain runs Johnson Controls Technology Company → CPS Technology Holdings LLC, with Citibank N.A. recorded as collateral agent under first-lien and ABL patent security agreements (2019-08-29) — a financing encumbrance, not a litigation signal. The most plausible reading of the zero-IPR record is simply that the patent has not been asserted aggressively enough in a venue that provokes IPR filings, or that its ~2035 expiry and battery-management subject matter keep it in licensing/defensive quiet.

Recommended next steps

• Treat the absence as the finding. There is no FWD to link and no disposition to quote, because none exists. Any AIA-trial summary you produce for this patent should state that plainly rather than padding the record.
• Don't let the near-miss number contaminate the file. If someone on your team hands you an IPR memo about "the '533 patent" citing IPR2019-00916 and canceled claims 5 and 13, check the number: that is US 9,651,533 (Omni MedSci). Asserting that claims of 9,625,533 were canceled would be a serious, sanctionable error.
• If you have been served with an infringement complaint on 9,625,533, note the § 315(b) clock: an IPR petition must be filed within one year of service of the complaint alleging infringement. That deadline is jurisdictionally enforced and is the single hardest constraint on your options. Confirm the service date before doing anything else.
• Pre-filing art search. Because no prior petitioner has litigated this patent's art, budget for a full prior-art and claim-construction study against the '533 specification (equivalent-circuit battery model with a gassing-current branch, Kalman-filter state estimation) rather than relying on an existing PTAB record. The gassing-current model of equations (3)–(4) and the measurement model of equation (2) are the natural focus points.
• Foreign exposure. If you sell into Europe or China, the EP2986994B1 and CN105122074B members of the same family are live and were granted on the same disclosure; EPO opposition (9-month window from grant, which closed in 2021) is no longer available, but national revocation actions in validated states may be. The Spanish register records the EP patent as not validated in Spain, which limits that exposure in ES.

Bottom line for a defendant today: this patent is a blank slate at the PTAB — nothing has been canceled, nothing has been sustained, and nothing has been estopped. That means no shortcut defense and no inherited roadmap; plan on a first-principles invalidity case.

I'll research the assignment record for US 9625533 before assembling the report.

Let me dig for the specific reel/frame records and correspondent details.

Let me try to pin down specific reel/frame numbers and the recorded correspondent.

Assignment & Ownership Analysis — US 9625533 B2

Patent: "Lead acid state of charge estimation for auto-stop applications"
Appl. No.: 14/253,720 · Filed: 2014-04-15 · Issued: 2017-04-18 · Priority: 2013-04-16 (US Prov. 61/812,534) · USPC/CPC: G01R 31/36, G01R 31/379

Sourcing caveat up front. The Google Patents legal-events record for US 9625533 (fetched 2026-09-26) supplies the existence, type, parties, and recording dates of four post-filing assignment events, but it does not expose reel/frame numbers or the correspondent-of-record. I could not retrieve the Assignment Center record itself in this session. I am not fabricating reel/frame or correspondent data — those fields are marked not retrieved below and must be pulled directly from USPTO Assignment Center (search by patent number 9625533) to complete the prosecution-grade record.

Inventors

Unusual-pattern check: Both inventors were JCI Power Solutions battery-controls engineers who assigned to the corporate IP holder in the ordinary course (recording 2014-05-02). There is no public record evidence of either inventor departing JCI within 12 months of the 2014-04-15 filing, and no inventor-initiated or inventor-owned follow-on assignment. That specific fire-sale precursor ("all inventors departed within 12 months") is therefore not determinable from available records — but nothing in the chain is inconsistent with a normal employee-inventor assignment.

Original assignee

Johnson Controls Technology Company (JCT) — the intellectual-property holding subsidiary of Johnson Controls, Inc. (now Johnson Controls International plc). JCT is a holding/record-owner entity, not itself a product-shipping company.

• Operating business embodying the claims: Johnson Controls Power Solutions, the world's largest automotive lead-acid battery manufacturer. The claimed subject matter (recursive SOC estimation via a gassing-current-augmented measurement model, used to gate auto-stop) is directed squarely at the product line JCI sold for micro-hybrid / start-stop vehicles — Absorbent Glass Mat (AGM) and Enhanced Flooded Battery (EFB) lead-acid batteries marketed under VARTA, OPTIMA, Heliar, LTH, MAC, Delkor, and OEM private label.
• Current status of the original assignee entity: Johnson Controls Technology Company remains extant as a JCI (Johnson Controls International plc) IP-holding affiliate. The Power Solutions business was divested to Brookfield Business Partners in a ~$13.2 B transaction that closed 2019-04-30; the carved-out business was rebranded Clarios, whose IP is held by CPS Technology Holdings LLC (the current assignee of record per Google Patents).

Assignment timeline

Chronological, per the Google Patents legal-events record for US 9625533. Dates shown below are recording dates; execution dates are noted where reasonably inferable and flagged otherwise.

• Executed ~2014-04-15 / recorded 2014-05-02 — Reel/Frame: not retrieved

  • Conveyance: Assignment of Assignors' Interest
  • Assignor: Daniel B. Le; Brian C. Sisk
  • Assignee: Johnson Controls Technology Company
  • Correspondent: not retrieved (verify at Assignment Center)
  • Context: Ordinary employee-inventor assignment to the corporate IP holder — internal, pre-issuance.

• Executed ~2019-04-30 (inferred from Brookfield/Clarios closing) / recorded 2019-06-20 — Reel/Frame: not retrieved

  • Conveyance: Assignment
  • Assignor: Johnson Controls Technology Company
  • Assignee: CPS Technology Holdings LLC
  • Correspondent: not retrieved
  • Context: Carve-out transfer of the Power Solutions patent portfolio to the Clarios IP holding entity pursuant to the Brookfield divestiture — corporate divestiture (internal reorg of the acquired business), not an NPE transfer.

• Executed ~2019-04-30 (inferred) / recorded 2019-08-29 — Reel/Frame: not retrieved

  • Conveyance: Grant of security interest — First Lien Patent Security Agreement
  • Assignor: CPS Technology Holdings LLC
  • Assignee: Citibank, N.A., as Collateral Agent
  • Correspondent: not retrieved
  • Context: Securitization — collateral grant to the first-lien lenders financing the Clarios acquisition.

• Executed ~2019-04-30 (inferred) / recorded 2019-08-29 — Reel/Frame: not retrieved

  • Conveyance: Grant of security interest — ABL Patent Security Agreement
  • Assignor: CPS Technology Holdings LLC
  • Assignee: Citibank, N.A., as Collateral Agent
  • Correspondent: not retrieved
  • Context: Securitization — collateral grant under the asset-based lending facility.

Correspondent recurrence: Cannot be assessed — no correspondent-of-record value was retrievable for any of the four events. This is the single most important field to backfill at Assignment Center; the "repeat lawyer across the chain" test is unanswered on current data.

Reel/frame gap: All four reel/frame identifiers are outstanding. This is a data-retrieval gap, not a substantive finding.

Timeline diagram

timeline
    title Ownership of US 9625533
    2013 : Provisional filed by Le and Sisk
    2014 : Utility application filed
         : Assigned to Johnson Controls Technology Co
    2017 : Patent issued
    2019 : Assigned to CPS Technology Holdings LLC
         : First lien security interest to Citibank
         : ABL security interest to Citibank

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. CPS Technology Holdings LLC carries the "Holdings" suffix and is a single-purpose IP-holding LLC, which is a watch item — but the concrete corroboration required for a finding is absent: it is the record IP owner for Clarios, an operating lead-acid battery manufacturer (a going concern with revenue, plants, and products), and there is no evidence its recorded address is a registered-agent mail drop. Post-divestiture IP holding companies standardly have exactly this name shape. Naming alone is not a finding.

2. Known asserter in the chain — Not present. No link in the chain (Johnson Controls Technology Company → CPS Technology Holdings LLC → Citibank, N.A. as Collateral Agent) matches Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Conversant/Mosaid, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, or any Erich Spangenberg entity. Citibank, N.A. is a secured lender / collateral agent, i.e., it holds a security interest rather than an assertion right — a common false positive when reading assignment records.

3. Repeat correspondent across the chain — Unclear / not determinable. Correspondent-of-record is not exposed in the retrieved record for any of the four events. Recurrence cannot be tested. This is the key open item.

4. Cascading transfers — Not present. The chain spans 2014 → 2019 (a five-year gap) and the only near-in-time cluster (the two 2019-08-29 recordings) is the same assignor (CPS Technology Holdings LLC) granting two parallel security interests to the same collateral agent — not a chain of successive LLC transfers. No shared-principal chained LLCs; no evidence of venue-shopping serial assignments.

5. Pre-litigation transfer — Not present / not evidenced. No infringement action naming US 9625533 was surfaced, so there is no assignment-to-suit temporal link to evaluate. The 2019-06-20 transfer predates nothing relevant; it is a divestiture milestone.

6. Bankruptcy fire-sale — Not present. The 2019 transfer arose from Johnson Controls' sale of its Power Solutions segment to Brookfield Business Partners — a going-concern, multi-billion-dollar carve-out, the opposite of a §7/§11 liquidating sale. No Chapter 7/11 docket is implicated for JCT or JCI in connection with this patent.

7. Privateering — Not present. No evidence (SEC filing, Patent Progress/EFF coverage, or litigation pattern) indicates the operating company placed the patent with a third-party enforcer to assert against competitors. The transferee is the acquirer's own IP-holding affiliate, i.e., a first-party successor in interest.

8. Defensive aggregator (anti-NPE) — Not present. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. It terminates inside Clarios' IP holding company, encumbered by blanket first-lien and ABL security interests in favor of Citibank, N.A.

Summary of signals: 0 present · 1 unclear (correspondent recurrence) · 7 not present.

Verdict

Operating-company assertion (nearest enumerated option — see caveat).

The chain shows zero NPE signals: US 9625533 was assigned in the ordinary course from its JCI employee-inventors to Johnson Controls Technology Company (recorded 2014-05-02), then passed as part of the Brookfield/Clarios carve-out to Clarios' IP holding company CPS Technology Holdings LLC (recorded 2019-06-20), and was simultaneously securitized to Citibank, N.A. as Collateral Agent under first-lien and ABL patent security agreements (recorded 2019-08-29). The current assignee is the IP holder for an operating lead-acid battery manufacturer whose products practice the claims (AGM/EFB start-stop batteries).

Caveat, stated plainly: the enumerated verdict "Operating-company assertion" has two prongs — ships products and sues competitors. Only the first is satisfied here; no assertion is on record. If the taxonomy is applied literally and the assertion prong is mandatory, the residual category is Insufficient data, driven by (a) missing reel/frame and correspondent fields and (b) no litigation evidence. Either way, the substantive finding is the same: this is a corporate divestiture plus financing chain, not an NPE/troll pattern. Backfill the four reel/frame entries and the correspondent names at Assignment Center — the correspondent-recurrence test is the only material unknown.

I'll research the prior art citations for US 9,625,533, focusing on the references cited during prosecution and the patent's own citation record.

Let me dig into the specific "References Cited" section on the patent's front page and its cited-by records.

I've located the "Referenced Cited" list. Let me pull the full citation details and confirm the front-page patent references.

Prior Art Analysis — U.S. Patent No. 9,625,533 B2

0. Identification & methodology (read first)

Patent confirmed: US 9,625,533 B2, "Lead acid state of charge estimation for auto-stop applications," inventors Daniel B. Le and Brian C. Sisk; applicant Johnson Controls Technology Company (now CPS Technology Holdings LLC); App. No. 14/253,720; filed 2014-04-15; priority 2013-04-16 (US 61/812,534); granted 2017-04-18; adjusted expiration 2035-06-21. Its independent claims are claim 1 (method), claim 12 (tangible, non-transitory computer-readable medium), and claim 14 (vehicle), with dependents 2–11, 13, and 15–21.

Sources used and limitations — stated plainly:

• I retrieved the "References Cited" list from the patent's face via the Justia record (https://patents.justia.com/patent/9625533) and the front-page continuation of the granted patent PDF (patentimages.storage.googleapis.com/.../US9625533.pdf).
• Caveat: I could not execute a live query against USPTO PatentCenter / the USPTO "Patent Full-Text" database within this session, and the authoritative full text supplied in the prompt omitted the front-page reference list. The list below therefore reflects the patent's printed face as mirrored by those sources. Verify against USPTO PatentCenter ("References Cited" / "Patent History") before relying on it for any filing. I flag every item where I am not fully certain of the exact title.

Important framing on § 102: Every reference below is an examiner-cited reference of record (cited in the "References Cited" section), not necessarily a reference that was used as an anticipation rejection. The '533 patent issued, which indicates the examiner did not find any of them to anticipate the claims as a whole. Below I distinguish "potentially anticipatory as to a claim" from "merely cumulative/general background," and I do not manufacture anticipation where the record does not support it.

1. U.S. Patent Documents Cited (56 References Cited)

2. Foreign Patent Document

Only CN 103323781 A (2013-09) appears on the cited list. See timing defect in row 8 — it is not available as § 102 prior art on the dates shown.

3. Non-Patent Literature Cited (the substantive art)

4. Bottom-line § 102 assessment

1. No cited reference anticipates independent claim 1 (or its CRM/vehicle counterparts, claims 12 and 14). The novelty-driving limitation of claim 1 is the gassing current model — "determining a gassing current of the battery module using a gassing current model, wherein the gassing current quantifies terminal current that is not used to charge the battery module" — and the corresponding measurement model. None of the cited references teaches treating a terminal current as split between a charge current and a gassing current that is subtracted before the state estimate. Absence of that element defeats anticipation under § 102 for claim 1.
2. Best § 102 candidates are only partial and target the dependent claims. The lead-acid + Kalman/EKF/observer references — Vasebi (2007), Bhangu (2005), Han (2008), Zhang, J. (2010), Choudhury (2009), Barbarisi (2005), Plett (US 6,534,954) — disclose recursive state/uncertainty estimation and would be the references mapped against claims 2, 4, 5 (state-space prediction, uncertainty/covariance, corrected uncertainty recursion). Even so, each still lacks the gassing-current teaching that claim 2 inherits from claim 1, so none anticipates the dependent claims as a whole. These are obviousness (§ 103) candidates, not § 102 art.
3. Two cited items are not available as prior art on their face:
   • CN 103323781 A — published 2013-09, after the 2013-04-16 priority date, and foreign publications cannot be § 102(a)(2) U.S. art.
   • Le & Sisk, SAE 2013-01-1532 — the inventors' own disclosure dated 2013-04-08, within the one-year grace period; presumptively excepted under § 102(b)(1)(A).
4. Berndt (2001) is the reference of record most apt to the lead-acid chemistry recited in claims 8–9 (water dissociation into oxygen and hydrogen), but it is a chemistry overview and is not an anticipating reference.
5. Welch & Bishop and the PCT ISR/WO are generic/enabling or procedural documents, not § 102 art.

5. Caveats and consistency flags

• Consistency with the previously generated litigation section: no contradiction. That section found "no litigation known/found" for the '533 patent; nothing in this prior-art review indicates otherwise. The family members identified earlier (US 2014/0309911 A1; WO 2014/172449 A1; EP 2986994 B1; CN 105122074 B) are consistent with the record here.
• Verification needed: (a) exact titles of US 5,321,627 (Reher), US 6,618,681 (Hoenig), US 6,927,554 (Tate), US 7,471,090 (Atehortua), US 2012/0041698 (Zhang), and CN 103323781; (b) whether the granted patent's front-page NPL list is exactly the union of the Justia list and the PDF list I found — the PDF front page yielded Choudhury, Zhang (J.), Barbarisi, and Han, while the Justia list yielded Wang, Bhangu, Le, Schiffer, Berndt, Vasebi, Welch, Rynkiewicz, and the ISR/WO; they overlap incompletely and should be reconciled against the printed patent.
• What I did not do: I did not run a live USPTO PatentCenter / PatFT "References Cited" retrieval or an examiner's "Notice of References Cited" (PTO-892) pull, and I did not reconstruct the actual rejection history. If a rigorous § 102/§ 103 opinion is required, the prosecution file wrapper (NPL and PTO-892 citations, and any prior-art statements) should be pulled from USPTO PatentCenter for App. 14/253,720.

Practical takeaway: The '533 patent's cited art is dominated by lead-acid/automotive Kalman-filter SOC literature. That art supports the recursive estimator scaffolding but not the patent's point of novelty — the gassing current model. On the cited record, the references are § 103 (obviousness) material against the estimation-dependent claims and are not standalone § 102 anticipatory references against independent claims 1, 12, or 14.

I'll start by locating the actual prior-art references cited on this patent's face and in its family, plus the key non-patent literature in this field.

I found the examiner-cited non-patent literature. Now let me get the complete "References Cited" list (US patent documents) from the front page, plus the key gassing-model prior art surfaced in the family's search reports.

I have the examiner-cited NPL and key gassing-model art. Now let me pin down the granted claim set (the numbering appears to differ from the pre-grant publication) and the U.S. patent documents cited.

Obviousness Analysis — U.S. Patent No. 9,625,533 B2

("Lead acid state of charge estimation for auto-stop applications") under 35 U.S.C. § 103

0. Evidence base and two required flags

What I used as "the Prior Art section." The patent record supplied to me does not contain a rendered "Citations / References Cited" table, so I reconstructed the prior-art section from (a) the examiner-cited references that appear on the face of the patent (the "cited by examiner" NPL list, verified in the granted PDF at patentimages.storage.googleapis.com/.../US9625533.pdf), and (b) art surfaced through the international search report of a related lead-acid SOC family (WO 2011/004249 A3) and through cross-citations in the Plett observer family. I mark verified-from-the-record items [VERIFIED] and secondary-sourced items [UNVERIFIED — flagged].

⚠️ Flag 1 — claim numbering contradicts the earlier sections. The previously generated summary states that claim 12 is the computer-readable-medium claim and claim 14 is a "vehicle" claim. The granted claim text I retrieved contradicts this:

"10. A tangible, non-transitory, computer-readable medium storing a plurality of instructions executable by a processor of a battery control module in a vehicle…" [VERIFIED — granted PDF]
"11. The tangible, non-transitory, computer-readable medium of claim 10…"
"12. The tangible, non-transitory, computer-readable medium of claim 10…"

So in the granted patent the medium claim is claim 10, and claims 11–12 are dependents of it. The earlier section's mapping (claim 12 = medium; claim 14 = vehicle) reflects the pre-grant publication US 2014/0309911 A1, not the grant. All claim analysis below uses the granted numbering, and I could not retrieve granted claims 13–24 within this session — those are analyzed provisionally.

⚠️ Flag 2 — OCR artifact in the granted claim text. Claims 5–8 of the granted PDF render the word "gassing" as "passing" (e.g., "determining amount of current actually used to charge the battery module by subtracting the passing current from the terminal current"). Per the operating rules I do not auto-correct identifiers. I read "passing current" literally as the OCR rendering of "gassing current," but I note the discrepancy rather than silently normalizing it.

1. Legal framework and level of ordinary skill

The governing standard is Graham v. John Deere Co., 383 U.S. 1 (1966) as refined by KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007). Under KSR, a claimed combination is obvious where a POSITA would have had a reason to combine known elements, and the reason may come from the references themselves, the nature of the problem to be solved, or the ordinary creativity/common knowledge of the artisan. The enumerated KSR rationales I apply below are:

Level of ordinary skill (POSITA). A person with an M.S. in electrical, chemical, or mechanical engineering (or equivalent experience) plus roughly 2–5 years in automotive battery-management/controls, or a B.S. with substantially more experience. Such a person would be conversant with: (i) equivalent-circuit battery models (OCV source + series resistance + RC polarization pair); (ii) Kalman/EKF/UKF state observers for SOC; (iii) lead-acid electrochemistry, including the H₂O → H₂ + O₂ side reaction; and (iv) automotive start-stop/micro-hybrid control strategy. The filed specification is written at exactly this level — it assumes Kalman filtering and equivalent-circuit modeling are known and spends its text on how to fold gassing into the observer.

2. The prior-art references in play

2a. Examiner-cited on the face of US 9,625,533 [VERIFIED]

These were listed under "* cited by examiner" on the granted front page:

The examiner thus had on the record: (a) Kalman SOC observers, (b) specifically for lead-acid, (c) specifically for automotive use, and (d) in nonlinear/adaptive form. That is the entire observer half of the asserted invention.

2b. Observer architecture — the Plett family [VERIFIED as to existence and text]

• US 6,534,954 B1 — Plett, "Method and apparatus for a battery state of charge estimator," Compact Power Inc., issued Mar. 18, 2003. Uses a Kalman/Extended Kalman filter with SOC as an internal system state; expressly claims "an uncertainty associated with its SOC estimation." (Notably, the primary examiner of US 6,534,954 is Edward H. Tso — the same examiner of record on US 9,625,533, per the earlier section. The examiner was therefore personally familiar with the Plett observer art.)
• US 2006/0111870 A1 — Plett, "Method and system for joint battery state and parameter estimation." Sets out, in table form, exactly the recursion of the '533 patent: state-space model χ_k+1 = F(χ_k, u_k), output y_k = g(χ_k, u_k) + v_k, time update Σ⁻ = A Σ⁺ Aᵀ + Σ_w, gain L_k = Σ⁻ C_kᵀ [C_k Σ⁻ C_kᵀ + Σ_v]⁻¹, corrected state χ̂⁺ = χ̂⁻ + L_k [y_k − g(χ̂⁻, u_k)], corrected covariance Σ⁺ = (I − L_k C_k) Σ⁻. Compare the '533 patent's equations (7)–(10). These are the same equations.
• US 2006/0111854 A1 — Plett, "Method and system for battery parameter estimation." Its parameter vector expressly includes η_i,k — "an efficiency factor(s) such as Coulombic efficiency" alongside capacity, resistance and hysteresis terms.

The last item is the crux: the Plett parameter-estimation disclosure already names Coulombic efficiency as a modeled parameter. The '533 patent's "gassing current" is the physical implementation of that Coulombic-efficiency loss in a lead-acid cell.

2c. Gassing-current modeling — the electrochemical half

2d. Context reference (not prior art)

SAE Technical Paper 2013-01-1532, "Lead-Acid State of Charge Estimation for Start-Stop Applications" (Apr. 8, 2013) is authored by Daniel Le — the named inventor. It is a § 102(b)(1)(A) excepted inventor disclosure, but it usefully establishes that accurate lead-acid SOC under start-stop was a recognized industry problem immediately before the Apr. 16, 2013 priority date. That supports a motivation to combine, while itself not being citable art.

3. Element-by-element mapping of granted claim 1

Granted claim 1 requires: (1a) predicting terminal voltage of a vehicle battery module using a battery control module; (1b) determining a gassing current via a gassing current model that quantifies terminal current not used to charge; (1c) calculating the predicted terminal voltage based at least in part on a measurement model and the gassing current; (1d) measuring terminal voltage with a sensor coupled to the BCM; (1e) determining a corrected state by minimizing error between predicted and measured terminal voltage; (1f) controlling vehicle operation via a vehicle control module based on the corrected state, the corrected state comprising SOC.

Every element of granted claim 1 is disclosed. The only question is motivation to combine, which I address next.

4. The combinations that render the claims obvious

Combination A — Plett observer + Schöner-type gassing current + lead-acid start-stop control (primary; kills claim 1 and claim 10)

References: Plett US 6,534,954 / US 2006/0111870 / US 2006/0111854 in view of Schöner-1988 (as embodied in the Jülich/MIT lead-acid model and Schiffer/Sauer 2007; equivalently EP 1 688 722) in further view of EP 1 562 049.

Why a POSITA would combine them (KSR rationales A, C, D, F):

1. The problem supplies the motivation. Plett's observer is only as accurate as its input current. Plett himself models a Coulombic-efficiency factor η_i in his parameter vector (US 2006/0111854 ¶ describing θ_k = [… η_i,k …]). A POSITA seeking to apply Plett to a lead-acid cell (which Han 2008, Zhang 2010 and Barbarisi 2005 show was the then-current automotive trend) must confront the fact that a lead-acid cell gasses at high SOC / high charge voltage, so not all terminal current is stored. Physically parameterizing η_i as 1 − I_g/I_t is the most natural, and arguably the only principled, way to close that loop. → Rationale A (known elements combined for predictable result) and Rationale B (substitution of a physically-derived gassing term for an abstract scalar efficiency).
2. The Schöner equation was the standard, off-the-shelf gassing-current formula in lead-acid modeling (it appears verbatim in MIT's/PVSyst's published battery model and in the Jülich quasi-static lead-acid model, both pre-2013, both accompanied by the statement I_q = I_bat − I_gas). A POSITA would not have to invent it; he would look it up. → Rationale E (finite number of identified, predictable solutions).
3. EP 1 688 722 already made the conceptual move for an automotive battery: it states the terminal current is only partly used in the main reaction and is partly consumed by a side reaction (water decomposition), and it quantifies the water-decomposition side reaction. Combining that recognition with a Kalman observer is a use of a known technique (side-reaction accounting) to improve a similar device (an automotive battery monitor). → Rationale C.
4. Rennison EP 1 097 388 provides yet another express hook: it models decreasing charging efficiency as SOC approaches full and applies a "gassing voltage correction." Feeding such a charge-acceptance/gassing correction into a voltage-model-based SOC estimator is exactly claim 1's architecture, merely with a Kalman filter substituted for Rennison's recursive tank model. → Rationale B/D.
5. Market force — fuel-economy/emissions regulation driving auto-stop (see the inventor's own SAE 2013-01-1532 framing of the industry need, and EP 1 562 049's explicit engine-standstill SOC monitoring). → Rationale F.

Predictability. The '533 patent's own FIG. 8C proves the result is predicted, not surprising: coulomb counting drifts above 100% SOC, and subtracting the gassing current keeps SOC at ≤100%. Removing a known, quantified error source from an observer and observing that the error disappears is the definition of a predictable result. There is no teaching-away anywhere in the art (nothing in Schöner, EP 1 688 722, or Plett says "do not model gassing in an SOC observer").

Combination B — Lead-acid Kalman SOC art (examiner-cited) + gassing-current model

References: Han 2008 (adaptive EKF, lead-acid) or Zhang 2010 (multi-state UKF, VRLA) or Barbarisi 2005 (automotive) in view of Schöner/EP 1 688 722, in view of EP 1 562 049.

This is the structurally simplest attack because it stays entirely within the field the examiner already cited. Han's adaptive EKF is designed to track changing parameters; the gassing-induced loss of coulombic efficiency at high SOC is precisely the kind of slowly-varying parameter adaptive observers are built for. A POSITA would add the known Schöner term to Han's measurement/state model to fix the known high-SOC inaccuracy. → Rationales A, C, D, E.

Combination C — Charge-efficiency/gassing modeling + Kalman filtering (narrow form)

Reference: EP 1 097 388 (Rennison) in view of Plett US 6,534,954.

Rennison teaches a lead-acid SOC model with a gassing voltage correction and a charge-acceptance correction, but uses a non-Kalman recursive estimator. Plett teaches that Kalman/EKF estimation of a voltage-model SOC gives an associated uncertainty. Substituting a Kalman filter for Rennison's estimator yields claim 1. → Rationales A, C, D.

Combination D — for the "subtract" limitation (granted claim 5 / claim 12)

References: the Jülich/MIT lead-acid model and PVSyst documentation in view of Han 2008/Zhang 2010.

The identity I_q = I_bat − I_gas (main reaction current = terminal current − gassing current) is literally printed in the published lead-acid models. Combined with any of the cited Kalman observers, claim 5's "determining amount of current actually used to charge the battery module by subtracting the [gassing] current from the terminal current" is squarely met. → Rationale A.

5. Dependent claims

6. Secondary considerations (objective indicia)

On the record reviewed, the objective indicia cut toward obviousness or are neutral, which is unusual and worth stating plainly:

• No unexpected results. The '533 specification's FIG. 8B/8C comparison shows the claimed method converging to the expected physical answer (SOC ≤ 100%, predicted V_t ≈ measured V_t). The patent argues accuracy improvement, but that improvement is the predictable consequence of removing a known error source. Under KSR, "a predictable variation" of prior art is "likely to be obvious."
• No teaching-away in any reference.
• No litigation or PTAB record (per the earlier sections) means there is no adjudicated secondary-considerations record — no nexus evidence, no commercial-success findings, no copying, no unexpected-results testimony. Absent a litigation record, the obviousness case rests on the art alone, which on this record is strong against claim 1/10.
• Industry-long-felt need is satisfied but points the other way: the SAE 2013-01-1532 abstract and EP 1 562 049 both show the auto-stop SOC problem was recognized and being actively worked, and the Kalman-for-lead-acid literature (Han 2008, Zhang 2010, Barbarisi 2005) shows the field was converging on observer-based solutions years before the priority date. A recognized, actively-pursued problem with finite predictable solutions is a KSR "design incentive" scenario.

7. Where the patent would most plausibly survive

To be balanced, the strongest (though on this record still modest) nonobviousness arguments are:

1. No express teaching to embed a gassing-current model inside the measurement model. The prior art discusses Coulombic efficiency (Plett), side-reaction current (EP 1 688 722), and gassing voltage correction (Rennison) — but I did not find a single reference that expressly says "compute I_g and subtract it within the observer's output/measurement equation." A patentee would argue this is a specific architectural insight, not a mere substitution. Counter: Plett's output equation y_k = g(χ_k, u_k) is agnostic as to what u_k is, and feeding a corrected input current into it is a textbook substitution → Rationale B.
2. The specific functional form I_g(k) = V_t(k−1)·G_P0·exp[ V_t(k−1)/V_t0 + A_g(1 − θ(k−1)/θ_f) ] — in particular the use of the previous time-step terminal voltage and temperature and the freezing-temperature-normalized term (1 − θ/θ_f) — appears distinctive; I could not verify a reference teaching that exact exponential-with-freezing-temperature parameterization. However, no granted claim recites this equation. Claims 1, 10 and (provisionally) the vehicle/method claims recite only "a gassing current model" generically. The narrow formula is therefore an unclaimed embodiment, and it will not save the claims. (If a dependent claim elsewhere recites the equation, that claim — not claim 1 — is the only realistic shelter.)
3. Claim 23 of the earlier summary (periodic calibration by fully charging) is likelier to survive, because periodic full-charge re-initialization of a Kalman SOC observer is a specific calibration step; but the '533 specification itself describes it as routine ("the battery control module 32 may periodically re-initialize (e.g., calibrate) the state of charge, for example, by fully charging the battery module and initializing the state of charge as 100%") — an admission that it is conventional. That admission weakens even this defense.

8. Bottom line

• Granted claims 1, 10 (and their dependents 2–9, 11–12) are, on the prior-art record identified, vulnerable to § 103. The observer half (claims 2–4, 9, 11, and the prediction/measurement/minimization steps of claims 1 and 10) is squarely met by the art the examiner himself cited plus the Plett family (US 6,534,954; US 2006/0111870; US 2006/0111854, which even names Coulombic efficiency as a modeled parameter). The gassing half (claims 5–7, 12) is met by the Schöner-1988 gassing-current equation and its published implementations (I_q = I_bat − I_gas), by EP 1 688 722 (main vs. side-reaction current; water-decomposition quiescent voltage), and by EP 1 097 388 (gassing voltage / charge-acceptance correction).
• The most damaging single combination is: Plett US 6,534,954 or US 2006/0111870 (EKF SOC observer with voltage output model and covariance propagation) + Schöner-1988 gassing-current model as applied to lead-acid + EP 1 562 049 (engine-standstill SOC monitoring for restart) as the vehicle-control element. The motivation is supplied by the references themselves, by the known high-SOC coulombic-efficiency loss in lead-acid cells, and by the market pull toward auto-stop fuel economy.
• Claim 8 (offline training of characteristic/tuning parameters) and any claim reciting periodic full-charge calibration are the comparatively strongest, but even those are described as routine in the specification and would face a strong "routine parameter fitting" counter-argument.

9. Confidence and verification gaps

If you want, I can next (a) build a formal element-by-element invalidity chart against the Combination A references, or (b) attempt to retrieve granted claims 13–24 so the vehicle claims can be analyzed on the same footing.