US 5544721

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

US Patent 5,544,721 — Analyst Summary

Source of record: Full patent text fetched from Google Patents (https://patents.google.com/patent/[US5544721](/patent/US5544721)/en), corroborated by FreePatentsOnline (https://www.freepatentsonline.com/5544721.html) and the USPTO-style front page reproduced at patentimages.storage.googleapis.com (US5544721.pdf). Verified as patent number 5,544,721 exactly; results for adjacent numbers (e.g., 5544469, 5544521) were excluded.

---

Bibliographic Data

Foreign family / also-published-as: JPH04338083A, KR920017932A, CA2062644C, AU653783B2, AU1217592A, CA2062644A1, JP2756208B2.

Note on dates: The July 25, 1994 date is the filing date of this application (the third in the chain); the substantive priority goes back to March 13, 1991. The "priority date 1991-03-13" on Google is expressly labeled an assumption, so treat it as highly likely but not a legal conclusion.

---

Abstract (as issued)

"Elevator horizontal vibrations are actively controlled by retrieving data stored in memory indicative of the out-of-straightness of the elevator car's guide rails. The stored data is compiled during a learning run by summing the car's relative displacement with respect to the rail with a doubly integrated acceleration signal indicative of the car's deviation from true vertical."

---

Plain-Language Overview of the Independent Claims

There are three independent claims — 1, 5, and 6 (claims 2–4 depend from claim 1). All three share the same core limitation, expressed in apparatus, system, and method form respectively.

Claim 1 — Apparatus (open-loop active suspension).
An apparatus for reducing horizontal movement of an elevator car on opposed guide rails. It has two parts:

• an actuator that, in response to an actuator drive signal, horizontally repositions the car relative to the rails; and
• a signal processor that receives a signal telling it where the car vertically is in the hoistway and, based on that, produces the actuator drive signal.
  The processor includes memory that holds information about both the sensed horizontal displacement of the car relative to the rails and the sensed horizontal acceleration of the car, stored at a memory address that is selected by the car's vertical position. The actuator drive signal is thus a function of (i) vertical position, (ii) sensed horizontal displacement relative to the rails, and (iii) sensed horizontal acceleration.
  Key point: the stored "learned rail" record itself contains the displacement term and the acceleration term — the departure over the Otala-style approach of storing acceleration alone.

Claim 5 — Elevator control system (structural/system framing).
A four-element system:

• (a) vertical position sensing means producing elevator car vertical position signals;
• (b) "learned-rail" memory means that, on a read command, outputs learned-rail information signals containing learned information about the car's sensed horizontal displacement relative to the opposed rails and its sensed horizontal acceleration, indexed by car vertical position;
• (c) car control means that, in response to the vertical position signals, issues the memory read signals, and in response to the learned-rail information signals issues horizontal adjustment control signals; and
• (d) actuator means that horizontally adjusts the car relative to the rails in response to those control signals.
  Again, the control signals are stated to depend on vertical position, sensed horizontal displacement, and sensed horizontal acceleration.

Claim 6 — Method.
A method mirroring claim 5, comprising the steps of: (1) generating elevator car vertical position signals with vertical position indicating means; (2) generating learned-rail memory read signals from car control means in response to those position signals; (3) retrieving, from a learned-rail memory, learned-rail information signals containing learned information on the car's sensed horizontal displacement relative to the opposed rails and its sensed horizontal acceleration, the information being indexed by the car's vertical position along the rails; (4) generating horizontal adjustment control signals in response to the retrieved information; and (5) horizontally adjusting the car relative to the rails via actuators. The dependence on vertical position + displacement + acceleration is repeated in the control-signal limitation.

Dependent claims 2–4 (all from claim 1) add: (2) sensing means supplying one combined "sensed information signal" that carries both the displacement and acceleration components for storage during a rail-profile run; (3) a double integrator on the acceleration signal plus a summing means that adds the sensed displacement to the doubly-integrated acceleration to create the stored value; and (4) a detailed signal-conditioning pipeline — twice integrate and detrend the acceleration, high-pass filter with time reversal, detrend, twice high-pass filter with time reversal; separately detrend and twice low-pass filter the displacement signal; then sum the two filtered signals for storage.

---

Context That Matters to Claim Scope

• The patent is expressly a continuation/companion to the rail-profile estimation/storage invention of co-pending application U.S. Ser. No. 07/668,544 — which corresponds to US 5,524,730 ("Method and apparatus for storing sensed elevator horizontal displacement and acceleration signals"). The '721 patent is directed to the retrieval-and-actuate half (open-loop control), while '730 covers the storage half.
• The specification distinguishes U.S. Pat. No. 4,750,590 to Matti Otala (assigned to Kone Elevator GmbH), which allegedly discloses storing only an acceleration-derived "deviation table." The '721 specification argues that acceleration alone is non-repeatable across loads/distributions, and that only the sum of relative displacement + doubly integrated acceleration yields a repeatable rail profile — this is the stated novelty hook.
• Corresponding foreign/EP family member: EP 0 503 972 B1 ("Elevator rail profile estimation and elevator control method").
• Applicant's own prior art cited as related: US 5,308,938; US 5,321,217; US 5,322,144; US 5,329,077 (all Otis, Roberts et al.).

---

USPTO / CAFC Litigation Check — Explicit Uncertainty

I found no evidence of any CAFC 2026 docket, appeal, IPR, or district-court action involving US 5,544,721. My searches of open web sources returned only patent-database, patent-office-journal, and unrelated hits (including coincidental uses of the string "5544721" as a New York attorney registration number, a Ricoh service-contract number, and an unrelated ID in a Belgian patent gazette).

Caveats I want to be explicit about:

• The tools available to me in this session are general web search only. I cannot directly query PACER, the CAFC docket system, or USPTO PatentCenter/Assignment records. A negative web-search result is therefore not authoritative proof that no litigation exists — only that none surfaced publicly.
• The patent's legal-status field on Google Patents reads "Expired – Lifetime" with anticipated expiration 2013-08-13, and the last recorded legal event is a 12-year maintenance-fee payment dated 2008-01-07. Given expiration in 2013, any 2026 CAFC activity would be unusual absent a damages-tail or declaratory-judgment scenario, but I cannot rule it out.
• I have not independently verified the current ownership chain (e.g., any assignment to Otis Worldwide Corporation or a successor) beyond Google's "Current Assignee: Otis Elevator Co." listing.

If you need a definitive litigation answer, that requires a PACER/CAFC docket query or a commercial litigation database (LexisNexis, Docket Navigator, Unified Patents portal), which I can't run here.

I'll search for litigation involving US patent 5544721 across patent litigation sources.

Initial results are mostly irrelevant — the Blue Spike cases reference a different patent (7,346,472). Let me search more precisely.

Let me search more targeted litigation databases and check for any assertions of this specific patent.

Litigation Search Report: U.S. Patent No. 5,544,721

Patent Identification (Verified)

Before reporting on litigation, I confirm the patent in question from the authoritative full text provided to me:

• Patent No.: US 5,544,721 (do not confuse with the "…472" patents discussed below)
• Title: "Method and apparatus for adjusting an elevator car based on stored horizontal displacement and acceleration information"
• Inventor: Randall K. Roberts
• Assignee: Otis Elevator Company
• Priority date: 1991-03-13 (continuation of Ser. No. 08/067,414, filed 1993-05-25, which was a continuation of Ser. No. 07/668,546, filed 1991-03-13)
• Filing date: 1994-07-25
• Grant date: 1996-08-13
• Anticipated expiration: 2013-08-13
• Status: Expired – Lifetime
• Foreign family: JP 2756208B2, KR 920017932A, CA 2062644C, AU 653783B2, JPH04338083A

Litigation Findings

No litigation involving U.S. Patent No. 5,544,721 was found in the sources searched.

I searched for litigation records tied to this specific patent number (Google Patents litigation links, general patent‑litigation aggregator content, and references to "5544721"/"5,544,721"). All references that surfaced to a "…472" patent concerned a different patent entirely — U.S. Patent No. 7,346,472, which was asserted by Blue Spike, LLC against 3M Cogent and Lumidigm in the Eastern District of Texas (e.g., Blue Spike, LLC v. 3M Cogent and Blue Spike, LLC v. Lumidigm, E.D. Tex., docket No. 6:12‑cv‑00499‑type filings in the gov.uscourts.txed.138874 docket). Those cases are not related to US 5,544,721 and should not be attributed to it.

I was unable to retrieve a dedicated Unified Patents litigation caselist entry, a PACER docket, or a CAFC appeal for US 5,544,721 within my search. I cannot rule out the theoretical existence of an unindexed or pre‑1990s‑era dispute, but I found none, and I will not fabricate case details.

Context That Supports a "No Litigation" Conclusion

• The patent is an Otis Elevator Company internal engineering patent (inventor Randall K. Roberts) relating to elevator ride quality / active suspension. The subject matter was implemented by the patent owner itself.
• The disclosed "rail profile estimation" approach was the subject of a copending, commonly owned Otis application (U.S. Ser. No. 07/668,544) — i.e., the related activity was intra‑company prosecution, not litigation.
• The patent expired on 2013-08-13, and the family members (JP, KR, CA, AU) similarly show no litigation indicators in the materials reviewed.

Related Post‑Grant Proceedings (Not Litigation Against the Patent)

For completeness — and to be explicit that these are citations/related documents, not litigation involving the '721 patent:

• The patent is cited by later third‑party patents, including Otis's own US 5,955,709 (1999) and US 6,305,502 (2001), Inventio AG's US 5,896,949 (1999), and various Mitsubishi/Thyssenkrupp/Kone/Fujitec documents. Being cited by a later patent is not evidence of litigation.
• No IPR, PGR, CBM, ex parte reexamination, or other post‑issuance challenge to US 5,544,721 appeared in the searched records.

Status Summary Table

Caveats

1. My search was limited to publicly indexed patent‑litigation content reachable via the queries run; I was unable to directly query the Unified Patents caselist, PACER, or Docket Navigator for this specific number within the available steps.
2. I found no evidence of any assertion of US 5,544,721, but the absence of indexed records is not a legal‑certainty representation. A definitive negative clearance would require a paid database query (e.g., Docket Navigator / Lex Machina) across all U.S. district courts and the ITC.
3. Per the strict rule, I have not auto‑corrected any identifiers. The "'472" references located in the Blue Spike documents refer to US 7,346,472, a distinct patent, and are excluded from this report.

Proceedings overview

Zero AIA trial proceedings are on file for US 5,544,721. The USPTO Open Data Portal returns no IPR, PGR, or CBM proceeding for this patent as of the most recent ingest, and my independent web searches of PTAB/PTAB E2E-indexed material, CAFC opinions, and CourtListener surfaced no petition, institution decision, final written decision, or appeal referencing US 5,544,721 — breakdown: 0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denials. The defensive posture this gives a defendant is therefore not "hardened patent" or "dead claims" but something more basic: the patent expired on 2013-08-13, and the reason it never attracted an IPR is that it was already past its commercial life before AIA trials became available. That is a much stronger defense than any FWD.

Per-proceeding detail

There is no proceeding to itemize. I will not manufacture a {PROCEEDING_NUMBER} for a filing that does not exist. What I can give you is the negative-search record, which is the substantive answer:

Search performed (2026-09-30):

• USPTO ODP structured "PTAB proceedings on file" block for US 5,544,721 — empty.
• Web search for PTAB/IPR/PGR filings citing US 5,544,721 or 5,544,721 — no hits.
• Web search for litigation asserting 5,544,721 — no hits; the only "5544721" string match returned was an unrelated New York attorney registration number.
• Google Patents "Legal Events" for the patent as reproduced in the authoritative text — only grant (1996-08-02), a payor-number event (1999-12-01), and three § 41 maintenance fee payments (2000-01-21, 2004-02-09, 2008-01-07). No reexamination certificate, no reissue, no disclaimer, no adverse post-grant event of any kind. The record shows "Anticipated expiration 2013-08-13 / Expired - Lifetime."

Caveat on the negative finding: a clean Google Patents transaction history is strong but not conclusive evidence of the absence of an ex parte reexamination, because reexam certificates are sometimes imperfectly surfaced there. If the patent matters, pull the full file wrapper at USPTO PatentCenter for application 08/279,826 and confirm. The AIA-trial absence itself, however, is corroborated by two independent sources (ODP + web) and I regard it as reliable.

Why the absence is expected here (practitioner read): IPR petitions became available 2012-09-16. This patent's term ran out 2013-08-13 — roughly an eleven-month window, on a 1991-priority mechanical/systems patent that had no known assertion campaign. The absence of PTAB activity on this patent is only weakly informative. Contrast a modern software patent with the same clean record: that would be a signal that nobody in the industry cares enough to invalidate. Here, the patent simply aged out first. Notably, the same analysis presumably holds for the sibling family members — US 5,524,730, US 5,309,938, US 5,321,217, US 5,322,144, US 5,329,077 — though I did not exhaustively verify each, and you should run the same ODP check on any of them if they are in a demand letter.

---

Strategic summary

Claim status. All six claims (1–6) are UNTESTED before the PTAB. None is canceled; none has been confirmed either. Claim 1 (apparatus with memory storing sensed horizontal displacement and sensed horizontal acceleration indexed by vertical position), claim 5 (elevator control system with learned-rail memory means), and claim 6 (method counterpart) all stand exactly as they issued on 1996-08-13, unamended, with no prosecution-history narrowing from any post-grant proceeding. Claims 2, 3, and 4 depend from claim 1 and recite the sensing/double-integration/filtering refinements drawn from FIG. 9 and Table I of the specification.

The real status is expiry, not validity. The patent's term ended 2013-08-13 (consistent with a pre-1995-06-08 filing chain — the application is a continuation of Ser. Nos. 08/067,414 and 07/668,546 — and therefore a 17-years-from-issuance term rather than 20-years-from-filing). That has two consequences that dwarf everything else in this memo:

1. 35 U.S.C. § 286 bars recovery of damages for infringement more than six years before the filing of a complaint. For any complaint filed on or after 2026-09-30, the recoverable window would have to start no earlier than 2020-09-30 — nearly seven years after the patent ceased to exist. On a straightforward reading, there is no actionable damages window left at all. This is the answer you give to a demand letter, ahead of any invalidity argument.
2. There is no injunctive relief available. The statutory right to exclude ended 2013-08-13. A demand letter citing claims 1–6 of this patent is a demand for past damages in a period that is already time-barred.

Estoppel landscape. Because no IPR or PGR was ever filed, no § 315(e)(2) estoppel attaches to anyone with respect to this patent. That cuts both ways: no petitioner is boxed out of prior-art grounds, but there is also no petitioner whose invalidity work-product you can borrow. Every prior-art ground remains theoretically available to a defendant — § 102/§ 103 over the cited art of record (US 4,750,590 to Otala; GB 2 181 275; US 4,757,849; US 5,020,639; US 5,027,925; the Hitachi JPH03-series filings; the Rudd/Skalski/Brandenburg inertial-profilometer literature), plus the § 112(a) enablement attack the specification itself invites against Otala; plus § 101/§ 112(b) challenges to the purely functional "means for" language in claims 1–5, which is aggressive functional claiming even by 1991 standards. Realistically, you will never need any of it.

Pattern signals. No serial petitioner exists. No defensive aggregator (Unified Patents, RPX, or similar) appears anywhere in the chain — there is no IPR, no ex parte reexam, and no litigation history I could locate tying 5,544,721 to an industry-wide validity campaign. Otis Elevator Company is the assignee of record throughout, and it did litigate actively in the elevator ride-quality space in this era (for example, Otis was the accused infringer/appellant in Inventio AG v. Otis Elevator Co., the RFID "elevator destination-entry" dispute — Fed. Cir. appeal decided on the '094 patent; the Harvard/CourtListener PDF is available at storage.courtlistener.com/harvard_pdf/812554.pdf). That case concerns a different patent owned by Inventio, not 5,544,721, and I flag it only to make clear that I found no assertion history for the patent at issue. That family's contemporaneous Otis patents on active suspension (US 5,309,938; US 5,321,217; US 5,322,144) were filed in 1990 — earlier than this application's 1991-03-13 priority date — and they, not this patent, are the lineage you would expect to see asserted first.

---

Recommended next steps

1. If you are a defendant and have received a demand citing US 5,544,721, lead with expiration, not invalidity. Quote the record: "Anticipated expiration 2013-08-13 — Expired - Lifetime" (Google Patents US5544721A). Then run the § 286 six-year bar against the actual filing date of any complaint. If the assertion is in a complaint filed in 2026 or later, the damages window is empty on its face; move to dismiss for failure to state a claim, or seek summary judgment on § 286. Note that I cannot link you to an FWD for this patent because none exists — the strongest citation here is the patent's own expiration date, not a Board disposition.
2. Pull the file wrapper before you respond in writing. Confirm at PatentCenter whether any ex parte reexamination or reissue was ever filed and whether a terminal disclaimer or § 41 lapse affected the term. The Google Patents transaction history reproduced above shows fee payments through the 12-year payment in 2008-01-07 and no adverse entry, but the file wrapper is the authoritative source.
3. If you need an invalidity theory anyway (e.g., to defeat a claim for past damages, or because a licensee is seeking indemnity), an IPR is the wrong vehicle and an ex parte reexamination is the right one. Ex parte reexam under 35 U.S.C. § 302 is available to any third party at any time, has no § 315(b) one-year bar, and — unlike IPR — is not subject to § 315(e)(2) estoppel. For a patent that expired in 2013, this is the only cost-justified path, and only if real money turns on the past-damages theory. Verify the patent's reexam eligibility and status through PTAB E2E and the USPTO PTAB Decisions library before filing anything.
4. Monitor rather than prepare. There is nothing to calendar — no institution deadline, no oral hearing, no statutory one-year FWD clock, because there is no trial. If you want ongoing confirmation, set a periodic ODP query on US 5,544,721 and a CourtListener docket alert at courtlistener.com for the patent number. New PTAB activity on a patent that expired in 2013 would be unusual, but not impossible; a newly indexed proceeding would change this memo's bottom line and nothing else would.

Bottom line for a defendant: this is not a "the patent survived IPRs and is hardened" situation, and it is not a "claims 1–5 are canceled" situation either. It is an expired patent with an untested but commercially dead claim set — which, for assertion-defence purposes, is the easiest position you will ever be in. The absence of PTAB activity on file is real, and it is a consequence of expiry, not of validity strength.

I'll investigate the assignment record for US 5,544,721, starting with USPTO Assignment Center and corroborating sources.

Let me search for the assignment/ownership chain specifics and Otis corporate history.

Let me look for the specific recorded assignment and any Otis→Otis Worldwide change-of-name recordings.

Scope and evidence limitation (read first)

This task depends on USPTO Assignment Center reel/frame data. My available tooling in this session is general web search only — I cannot query assignmentcenter.uspto.gov / assignment.uspto.gov directly, nor PEDS/ODP, PACER, or paid databases. I ran multiple targeted queries for 5544721, 5,544,721, and Otis-associated assignment phrasing; no recorded assignment (reel/frame, execution date, recording date, or recording correspondent) for this patent was retrievable. I therefore will not invent reel/frame entries. Everything below is grounded in the patent front matter, Google Patents legal events, and corroborated corporate-history sources, with the gaps flagged explicitly.

---

Inventors

• Sole named inventor on US 5,544,721 (Google Patents lists Roberts alone; Radaris's aggregated Otis inventor page reproduces his Amston, CT address and this patent's abstract/8/279826 data).
• Do not confuse with Clement A. Skalski (Avon, CT), who appears adjacent to Roberts in aggregated Otis-inventor listings (e.g., the rope-sway / magnetic-suspension work) but is not a named inventor on '721. The Blue Spike "…472" confusion already flagged in the litigation section is similarly excluded.
• Roberts is a prolific Otis inventor in the same active-suspension/ride-quality program (also named on the '730 companion and on US 5,308,938; 5,321,217; 5,322,144; 5,329,077 — all Otis).
• Unusual departure pattern: not determinable / no evidence. I found no record of Roberts (or any co-inventor, since there is none) departing Otis within 12 months of filing, and the repeat appearances across Otis filings argue against a fire-sale precursor.

---

Original assignee

• Entity on the issued patent: Otis Elevator Company, One Carrier Place, Farmington, Connecticut (a New Jersey corporation per USPTO assignment documents in the Otis family).
• Primary line of business: design, manufacture, sale, installation and service of elevators, escalators and moving walkways — a large operating manufacturer (~US$13.0 bn revenue in 2014 per Wikipedia's Otis page), not a holding/licensing vehicle.
• Product embodying the claims: Otis is an operating manufacturer and the specification reports validation on Otis's own Bristol, Connecticut Test Tower (Hoistway No. 4) and states the active suspension "may be used on new elevators or may be retrofitted to existing installations." I cannot confirm that a commercial Otis product specifically practices claims 1/5/6 — treat "shipped product" as likely but unverified.
• Ownership corporate status:
  • Otis Elevator Company was acquired by United Technologies Corporation (UTC) in 1976.
  • UTC separated its businesses; Otis was spun off as an independent public company, Otis Worldwide Corporation (NYSE: OTIS), effective April 3, 2020. The separation was effected under an Intellectual Property Agreement dated April 3, 2020 among United Technologies Corporation, Otis Worldwide Corporation and Carrier (filed as Exhibit 10.4 to Otis Worldwide's Form 10 / 8-K; UTC press release March 11, 2020).
  • Status today: operating (Otis Worldwide Corporation). Not dissolved, not in bankruptcy. (UTC later merged into Raytheon Technologies / RTX; Carrier was also spun off — neither is the elevator business.)
• Bankruptcy check: No Otis/UTC bankruptcy. (The 1997 Evans Lifts bankruptcy noted in Otis corporate history is an Otis acquisition target, unrelated to this patent.)

---

Assignment timeline

Plain finding: I could not retrieve any recorded assignment from USPTO Assignment Center for US 5,544,721. What is evidenced is (a) an original inventor→Otis assignment implied by the patent's front-page "Assignee: Otis Elevator Company" field, and (b) a corporate reorganization (UTC → Otis Worldwide, 2020) that is not documented for this patent in any source I could reach.

Google Patents' Legal Events for this patent list only: 1996-08-02 STCF (patent grant), 1999-12-01 FEPP, and FPAY maintenance-fee payments for years 4, 8 and 12 (2000-01-21, 2004-02-09, 2008-01-07). No "AS" (assignment) event appears post-grant, which for a Google-indexed US patent is meaningful corroboration that no transfer was recorded after issuance.

• ~1991-03-13 (executed) / recorded not retrieved — Reel not retrieved/Frame not retrieved

  • Conveyance: Assignment (inventor → company; implied by the front-page assignee field; not independently verified)
  • Assignor: Randall K. Roberts
  • Assignee: Otis Elevator Company
  • Correspondent: not retrieved. The only attorney identified anywhere in the record is the prosecution attorney of record on the patent face — Francis J. Maguire, Jr. (FreePatentsOnline "Attorney, Agent or Firm: FRANCIS J MAGUIRE JR"). Maguire is an Otis-side prosecution attorney of record, not established as the assignment-recording correspondent; do not treat him as an NPE recording agent.
  • Context: original prosecution-time assignment (routine employee→employer), not an acquisition or fire-sale.

• 2020-04-03 (corporate reorganization) / recorded not retrieved — Reel not retrieved/Frame not retrieved

  • Conveyance: Change of name / internal reorganization (UTC group separation into Otis Worldwide Corporation) — assumed only
  • Assignor: Otis Elevator Company / United Technologies Corporation
  • Assignee: Otis Worldwide Corporation
  • Correspondent: not retrieved
  • Context: internal reorg (UTC three-way separation). Caution: this patent expired 2013-08-13, so it very likely was not carried into the 2020 spin-off IP schedules (or transferred only by operation of law). I have no document tying this specific patent to the 2020 IP Agreement.

If Assignment Center in fact shows only the original Roberts→Otis record (or none at all), that is the expected finding for a 1996 Otis engineering patent that expired in 2013: the original assignee retained it end-to-end.

---

Timeline diagram

timeline
    title Ownership of US 5544721
    1991 : Inventor Roberts assigned to Otis
         : Priority application filed
    1993 : First continuation filed
    1994 : This application filed
    1996 : Patent granted Aug 13
    2008 : Final maintenance fee paid
    2013 : Patent expired Aug 13
    2020 : Otis separated from United Technologies

Note: the 1991 "assigned to Otis" event is inferred from the issued patent's assignee field; the reel/frame could not be retrieved (see above).

---

NPE / troll-pattern signals

1. Shell-entity transfer — not present. No evidence of any transfer to an "IP/Patents/Licensing/Holdings/Ventures" LLC; Google Patents legal events show no post-grant assignment event and the current assignee remains Otis Elevator Co.
2. Known asserter in the chain — not present. No assignee matches any public NPE roster (Acacia, Marathon, IV, WiLAN/Conversant, Vringo, Pendrell, Round Rock, Spangenberg entities, etc.). The only named assignee is Otis, an operating OEM.
3. Repeat correspondent across the chain — not present / not assessable. Only one chain link is even hypothesized, and the sole attorney in the record (Maguire) is prosecution counsel, not an evidenced recording correspondent. No recurrence pattern exists to evaluate.
4. Cascading transfers — not present. No chained LLC-to-LLC transfers; no timeline of transfers at all in the retrievable record.
5. Pre-litigation transfer — not present. No infringement suit naming this patent exists (established in the prior litigation section), and no transfer within 6 months of any suit.
6. Bankruptcy fire-sale — not present. Assignee (Otis/UTC) never filed; ownership passed through a solvent corporate separation, not insolvency.
7. Privateering — not present. No transfer to a third-party assertion vehicle; the same corporate family (Otis) retained the rights.
8. Defensive aggregator (anti-NPE) — not present. Chain does not terminate at RPX / AST / LOT / Unified / OIN.

---

Verdict

Insufficient data — (only the original assignment is evidenced).

Justification: The retrievable record shows a single, ordinary inventor→employer assignment to Otis Elevator Company (inventor Randall K. Roberts; prosecution correspondent Francis J. Maguire, Jr. on the patent face) with no post-issuance assignment event in Google Patents Legal Events and no assertion history; the patent expired 2013-08-13 after the 12-year fee payment of 2008-01-07. No shell LLC, known asserter, cascading transfer, or pre-litigation transfer is evidenced. I could not reach USPTO Assignment Center to confirm reel/frame entries — the one missing input that would convert this from "insufficient data" to a definitive "original operating-company assignee retained ownership, no NPE pattern."

Verify at: https://assignmentcenter.uspto.gov/ (search patent number 5544721; also indexed at https://assignment.uspto.gov/patent/index.html). Patent of record: https://patents.google.com/patent/US5544721/en.

Bottom line: On present evidence this is a pure operating-company (Otis) engineering patent, never asserted and never transferred to an NPE — but that conclusion is inferred from absence of records and should be confirmed against Assignment Center reel/frame before it is relied on.

I'll verify the patent identification and pull authoritative details on the key cited references before doing the § 102 mapping.

Prior-Art Analysis — US 5,544,721

Task scope: search/confirm the record for US 5,544,721 and evaluate each cited reference for potential 35 U.S.C. § 102 anticipation of its claims. This builds on (and does not repeat) the bibliographic, litigation, PTAB, and assignment sections already generated.

---

1. Patent identification (confirmed — no re-search needed)

The authoritative full text supplied reproduces the Patent Citations (40) and Non-Patent Citations (10) lists that were printed on the face of US 5,544,721 (the citation lists on the face of the patent are the examiner‑cited "References Cited" set). I did not substitute any similar number:

Critical‑date framework applied below. Because this application is a continuation claiming benefit back to 1991‑03‑13, the § 102(a)/(b) critical date is 1991‑03‑13 (with the § 102(b) one‑year line at 1990‑03‑13). A reference therefore can only be § 102(a)/(b) art if it published before 1991‑03‑13; a U.S. patent/application that published later can still be § 102(e) art only if it was filed in the U.S. before 1991‑03‑13. Foreign (JP/GB) published applications are § 102(a)/(b) art only as of their publication date — § 102(e) does not reach foreign applications.

Verification sources used: the supplied Google Patents full text (USPTO‑derived), Justia Patents mirror (https://patents.justia.com/patent/5544721), and the EPO family member EP 0 503 972 search report (data.epo.org/.../EP0503972A3.pdf), which is a primary record of how the examiner family categorized the same art. I could not directly query USPTO PatentCenter within this session — see caveats.

---

2. Master citation table — all 40 patent citations, with date‑qualified prior‑art status

Legend: PA = qualifies as § 102(a)/(b) art by publication date; (e) = only reachable as § 102(e) art via pre‑critical U.S. filing; NOT PA = postdates 1991‑03‑13 in every respect.

Non‑patent citations (10 entries; 6 unique works):

---

3. The strict § 102 question: does any cited reference anticipate claims 1–6?

Short answer: none of the cited references discloses every limitation of any of claims 1–6, so there is no clean § 102 anticipation based on the cited record. The structural limitation that defeats anticipation is that each independent claim (1, 5, 6) requires the memory to contain information about both (i) the sensed horizontal displacement of the car in relation to the opposed guide rails and (ii) the sensed horizontal acceleration of the car, stored at an address according to vertical position — with the drive/control signal depending on all three. The cited elevator references store a position-indexed rail deviation/acceleration table, but none discloses storing the car-to-rail relative-displacement term together with the acceleration term. That is precisely the point the specification makes against Otala, and it is what makes the cited art § 103 rather than § 102 material.

Below, however, is the reference‑by‑reference § 102 assessment the task requests — including the several references that are the closest candidates and the claims they are most relevant to.

---

4. Detailed § 102 analysis — most relevant references

4.1 US 4,750,590 A — Otala (Kone Elevator GmbH) — the primary reference

• Full citation: US 4,750,590 A, "Counteracting elevator car oscillation," inventor Matti Otala, assignee Kone Elevator GmbH; filed 1985‑09‑27; published 1988‑06‑14.
• Disclosure (verified): open‑loop elevator lateral compensation; a deviation value table relating local guide‑rail deviations to car position along the path of travel is set up in the memory of a computer controlling the guide shoes; the table is derived from data from acceleration meters temporarily attached to the car during an uncorrected trial run; the microprocessor "constantly address[es] (READ command) … an EPROM‑type memory … via an address bus" and controls the guide shoes to compensate local deviation at any car position.
• § 102 mapping:
  • Claim 1: reads on actuator (guide‑shoe operating means) + processor responsive to vertical position + position‑indexed memory + control of the car. BUT the stored content is an acceleration‑derived deviation value, not a sensed car‑to‑rail horizontal displacement and a sensed horizontal acceleration. → Not anticipatory as written. This same defect carries through claims 5 and 6 (the "learned‑rail information … containing … a sensed horizontal displacement … and a sensed horizontal acceleration"). Dependency of claims 2–4 on claim 1 likewise defeats them.
  • Strongest use: § 103 primary reference (combined with a displacement‑sensing guide, e.g., US 4,754,849, US 5,086,882, or the inertial‑profilometer literature) to render the "store rail profile indexed by position and recall it to actuate" concept obvious. Note the EP 0 503 972 search report (the foreign family counterpart) flagged GB 2 181 275 A as an "X" (alone‑relevant) reference — independent corroboration that the Otala/Kone disclosure is the closest art.

4.2 GB 2 181 275 A — Elevator GmbH (Otala counterpart)

• Full citation: GB 2 181 275 A, "Compensating for lateral oscillation of lift car," Elevator GmbH; filed 1985‑09‑27; published 1987‑04‑15.
• Disclosure (verified): the British counterpart to § 4.1 — maps out‑of‑straightness of the guide rails over the lift path, sets up a "deviation value table in terms of the lift car position … in the memory of a computer controlling the guide shoes," table set up from acceleration meters on the car during an uncompensated trial run; guide shoes then controlled from the table.
• § 102 mapping: Identical analysis to § 4.1. Potentially relevant to claims 1/5/6 under a very aggressive reading that an acceleration‑derived "deviation value table" is "information indicative of … a sensed horizontal acceleration," but it lacks the claimed sensed horizontal displacement of the car in relation to the opposed guide rails. → Not a clean § 102 reference; primary § 103 reference. (Examiner's "X" citation in EP 0 503 972.)

4.3 US 5,086,882 A — Hitachi, Ltd.

• Full citation: US 5,086,882 A, "Elevator apparatus provided with guiding device used for preventing passenger cage vibration," Hitachi, Ltd.; filed 1989‑08‑30; published 1992‑02‑11 (§ 102(e) reachable via 1989 filing).
• Disclosure (verified): roller‑type guide devices each with an actuator, actuator controller and sensor, keeping contact pressure constant; output of a preceding guide device is used to control a subsequent one. Its background expressly describes a prior proposal (Japanese Laid‑Open 74,897/1987) in which "an actuator is directly attached to the rollers … operated on the basis of the data on the bend on the guide rails which are measured and stored in advance by an acceleration detector."
• § 102 mapping: The reference's own system is closed‑loop pressure control — no position‑indexed stored rail profile, and no stored car‑to‑rail displacement + acceleration record. → Anticipates no claim. Its discussion of the 1987 Japanese proposal is highly relevant § 103 background (and the actual anticipatory document would be JP Laid‑Open 74,897/1987 itself, which is not in the '721 citation list). Relevant to claim 1's actuator + rail‑relative sensing environment.

4.4 US 5,020,925 A — Kone Elevator GmbH

• Full citation: US 5,020,925 A, "Procedure and apparatus for damping the vibrations of an elevator car," Kone Elevator GmbH; filed 1988‑09‑23; published 1991‑07‑02 (§ 102(e) reachable via 1988 filing).
• Disclosure: elevator‑car vibration damping (mechanical/active). Institution‑level detail not confirmed from the available text.
• § 102 mapping: Same field as the claimed invention; no indication of a stored rail profile indexed by position with combined displacement + acceleration content. → Anticipates no claim; § 103 context on the vibration‑damping aspect.

4.5 US 5,020,639 A — Inventio AG

• Full citation: US 5,020,639 A, "Method of, and apparatus for, absorbing vibrations in cars of high‑speed elevators," Inventio AG; filed 1988‑11‑02; published 1991‑06‑04 (§ 102(e) reachable via 1988 filing).
• Disclosure (verified): floatingly supports the car body on a support frame via hydraulic suspension units, isolating the body from horizontal shocks; includes actuating cylinders for horizontal positioning and an electrical control unit responsive to position‑sensor means between car body and support frame, and it varies adjusting force as an inverse function of travel speed.
• § 102 mapping: Discloses an actuator + horizontal position sensing + controller (relevant to claim 1/5's actuator and sensing environment), but there is no rail‑profile memory indexed by vertical position and no stored combined displacement + acceleration record. → Anticipates no claim; § 103 context.

4.6 US 4,754,849 A — Mitsubishi Denki K.K.

• Full citation: US 4,754,849 A, "Control system for elevator cage guide magnets," Mitsubishi Denki K.K.; filed 1986‑09‑29; published 1988‑07‑05.
• Disclosure: magnet‑type (non‑contact) elevator cage guidance with displacement detection relative to a reference line and control of the magnets.
• § 102 mapping: Relevant to the "sensed horizontal displacement of the car in relation to the guide rails" element and to an electromagnetic actuator; no position‑indexed stored rail profile. → Anticipates no claim; § 103 candidate for the sensing/actuation elements. (Also cited as "D,A" in the EP 0 467 673 search report.)

4.7 US 4,899,852 A — Otis Elevator Company

• Full citation: US 4,899,852 A, "Elevator car mounting assembly," Otis Elevator Company; filed 1988‑11‑03; published 1990‑02‑13.
• § 102 mapping: Structural car‑mounting art. Anticipates no claim; background/§ 103 context only. (Cited "D,A" in EP 0 467 673.)

4.8 Non‑patent literature — inertial profilometry (N1 Rudd; N2 Brandenburg)

• Full citations: Rudd et al., "Inertial Profilometer as a Rail Surface Measuring Instrument," 1973; Brandenburg et al., "Development of an Inertial Profilometer," ENSCO/NTIS, 1974.
• Significance: These are the references directed to the core measurement principle of the specification — combining an inertially‑referenced (doubly integrated accelerometer) measurement with a relative displacement transducer to yield a rail profile. That principle is what claims 3 (double integration + summing) and 4 (the FIG. 9/Table I pipeline) are built around, and it is the subject matter of the co‑pending storage application (US 5,524,730).
• § 102 mapping: Potentially anticipatory only of the profile‑making concept, not of claims 1–6, because these are rail‑measurement instruments, not elevator control systems — they disclose no elevator actuator and no vertical‑position‑indexed retrieval that drives a car actuator. → No anticipation of claims 1–6; high‑value § 103 art against claims 2–4 (and central art for the '730 sibling).

4.9 N6 — Derwent WPIL AN 90‑224718 / FI A 8804380 (Kone Elevator GmbH)

• Citation: Database WPIL, Week 9030, Derwent, AN 90‑224718 & FI A 8804380 (Kone Elevator GmbH), 1990.
• Significance: This Kone document was cited as an "X" (alone‑relevant) reference against claims 1, 7 and 8 of EP 0 467 673 (the Otis active‑suspension family) — i.e., an external examiner treated it as anticipating an active elevator suspension claim set. It is thus the most important non‑Otis reference on the active‑suspension actuation side.
• § 102 mapping: Relevant to the actuator/active‑horizontal‑control concept (claim 1/5 environment). No evidence it discloses the position‑indexed displacement + acceleration store. → No anticipation of claims 1–6; strong § 103 reference.

4.10 US 5,308,938 / US 5,321,217 / US 5,322,144 — Otis Elevator Company (active suspension)

• Full citations: all filed 1990‑07‑18; published 1994‑05‑03, 1994‑06‑14, 1994‑06‑21 respectively.
• § 102(e) analysis: Each was filed before 1991‑03‑13, so each is date‑eligible for § 102(e). However, these are the applicant's own active‑suspension family (the same Otis ride‑quality program, with overlapping inventive entity), raising the "by another" requirement of § 102(e)/§ 102(a) and, for any § 103 use, the § 103(c) common‑ownership exception. Even setting that aside, they disclose horizontal actuators and accelerometer/position feedback for an elevator car, not a rail‑profile store indexed by vertical position containing combined displacement + acceleration data. → Anticipates no claim; § 103 art with entity/ownership caveats.

Note on US 5,329,077 ("Elevator ride quality," Otis): filed 1991‑10‑24 and published 1994‑07‑12 — both after the 1991‑03‑13 critical date. It is therefore not prior art to US 5,544,721 under any subsection of § 102, notwithstanding its appearance in the face citation list. This is a same‑era Otis sibling, not a reference.

---

5. § 102 disposition table — remaining cited references

---

6. Bottom line

1. No cited reference is a clean § 102 anticipation of any of claims 1–6. The claim‑critical limitation — a memory storing both a sensed car‑to‑guide‑rail horizontal displacement and a sensed horizontal acceleration, indexed by vertical position, whose retrieval drives the actuator — is not disclosed by any single item in the citation list.
2. The closest § 102 candidates are the Otala/Kone pair (US 4,750,590 and GB 2 181 275). They disclose the entire "learn the rail, store a position‑indexed table, recall and actuate" architecture — but they derive that table from acceleration meters only and lack the claimed car‑to‑rail relative‑displacement term. Under strict claim construction they are § 103 references, not § 102 references — consistent with the specification's own accent on Otala as the distinguishing art and with the examiner's "X" categorization in the EP 0 503 972 family search report.
3. The most technically on‑point art for the claimed storage/processing content (claims 3–4) is the non‑patent inertial‑profilometer literature (Rudd 1973; Brandenburg 1974) — combining a doubly integrated accelerometer signal with a relative displacement transducer to obtain a rail profile. It cannot anticipate claims 1–6 (no elevator actuator, no position‑indexed control retrieval) but is the key § 103 art, and it is the substantive art of the co‑pending US 5,524,730 sibling.
4. Date defects worth flagging: US 5,329,077, JP H03‑115076 and JP H03‑124683 postdate the 1991‑03‑13 critical date and are not prior art to '721. US 5,020,925, US 5,020,639 and US 5,086,882 are reachable only under § 102(e) (pre‑critical U.S. filings, post‑critical publication). The Otis active‑suspension patents (US 5,308,938 / 5,321,217 / 5,322,144) are date‑eligible only under § 102(e) and carry same‑entity / § 103(c) common‑ownership caveats.

---

7. Caveats and limits (explicit)

• Source of record. The 40 patent citations and 10 non‑patent citations are taken from the authoritative full text of US 5,544,721 (Google Patents, USPTO‑derived) and corroborated against Justia. I could not directly query USPTO PatentCenter/ODP in this session; the face‑citation list is the examiner‑cited "References Cited" set and is reliable as printed.
• Full‑text vs. title. For the majority of the 40 references I have only the bibliographic record and title (plus, for the elevator items, the abstract/material returned by search). § 102 anticipation requires element‑by‑element comparison against the full disclosure; my "no anticipation" conclusions for the references marked "on available description" reflect the absence of any disclosure of the claim‑critical stored displacement + acceleration limitation, but a definitive § 102 call on any single reference requires pulling its complete specification — which I flag rather than overstate.
• Two apparent citation anomalies preserved verbatim (no auto‑correction): JP S58‑39753 (zinc recovery, Sumitomo) and JP S61‑22675 (thermopile manufacture, Tatsumo) bear no discernible relation to elevator ride quality and appear to be citation entries of record as printed; I have not altered them.
• Identifier discipline. I did not substitute any similar patent number. Where searches surfaced a "…472" patent (e.g., US 7,346,472) it was excluded, as were adjacent‑number results (US 5,524,730 is the sibling, not the patent at issue).

I have the full patent text, including the complete Prior Art citations list. Let me ground the key references before building the § 103 case.

Obviousness Analysis — U.S. Patent No. 5,544,721 under 35 U.S.C. § 103

Analyst note on framing. This section builds on the previously generated bibliographic, claim-construction, litigation, PTAB, and assignment sections and does not repeat them. It uses the Prior Art of record — the 39 patent citations and 10 non-patent citations listed on the Google Patents page for US5544721A (https://patents.google.com/patent/[US5544721](/patent/US5544721)/en), plus the specification's own admitted prior art — as the reference universe, as instructed.

Minor contradiction flag: the task header states "Current Date: April 26, 2026," while the session context states "Today's date is 2026-10-01." This has no bearing on the § 103 analysis (both dates post-date the 2013-08-13 expiration), but I flag it per the operating rules.

Evidentiary caveat (read first). I have full text for the '721 patent and I verified the actual disclosures of the two most load-bearing references — Otala (US 4,750,590) and the inertial-profilometer literature (Rudd/Brandenburg 1973; Brandenburg/Rudd 1974) — via the search results reproduced below. For the remaining cited references (the Japanese Hitachi/Mitsubishi filings, Inventio '639, Kone '925, and the Derwent abstract), I am working from title, date, assignee, and the specification's characterization, not from full-text review. Where a reference's disclosure depth is assumed rather than verified, I say so.

---

I. The claims and the dispositive limitation

There are three independent claims (1, 5, 6) and three dependents (2–4, all from claim 1). As summarized in the prior claim-construction section, all three independents recite the same core: an actuator that repositions the car horizontally, driven by a signal processor that reads a memory indexed by car vertical position, where the stored record contains both a sensed horizontal displacement of the car relative to the rails and a sensed horizontal acceleration of the car.

The dispositive departure over the closest prior art is therefore narrow and structural:

The stored "learned-rail" record is a function of two distinct sensed quantities — relative car-to-rail displacement and horizontal acceleration — and the actuator command depends on both, indexed by vertical position.

Everything else in claims 1, 5, and 6 — actuator, signal processor, vertical-position input, memory addressed by position, open-loop retrieval — is old in the elevator art. Claims 2–4 add standard signal-processing refinements (a shared "sensed information signal"; a double integrator plus summer; a detrend/filter/reverse-time pipeline).

---

II. The prior art of record, grouped by the claim element it supplies

---

III. The governing legal standard

• 35 U.S.C. § 103(a) asks whether the subject matter as a whole would have been obvious at the time of the invention to a person having ordinary skill in the art (PHOSITA). The person is presumed to know all the relevant art. In re Winslow, 365 F.2d 1017 (CCPA 1966).
• Graham v. John Deere, 383 U.S. 1 (1966): scope/content of the prior art; differences; level of ordinary skill; secondary considerations.
• KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007): a combination is obvious where the elements were known, the combination does no more than yield predictable results, and a PHOSITA had a reason to combine — including design incentives, market demand, the "finite number of identified, predictable solutions," and "obvious to try."
• Analogous art (In re Clay, 966 F.2d 656 (Fed. Cir. 1992)): a reference is analogous if it is from the same field of endeavor or reasonably pertinent to the problem the inventor faced. The profilometer references satisfy both prongs, as shown below — and the applicant's own listing of them as prior art is itself evidence of pertinence.
• MPEP § 2144 et seq.: combination of known elements; routine optimization (2144.04(III)); design choice; "obvious to try."

The level of ordinary skill here is a mechanical/controls engineer with several years of elevator suspension or vehicle-dynamics experience, familiar with accelerometers, relative-displacement transducers (LVDTs/potentiometers), digital filtering, and active suspension — the very profile of the named inventor's own Otis colleagues (Roberts et al., US 5,308,938; 5,321,217; 5,322,144).

---

IV. Ground 1 (primary): Otala in view of the inertial-profilometer art of Group B — renders claims 1, 2, 3, 5, and 6 obvious

IV.A Otala discloses every element except the two-sensor combination

I verified Otala's actual text (US4750590A, https://patents.google.com/patent/US4750590A/en). Otala's own claim 1 recites:

"ascertaining the magnitudes of local deviations of said guide rails from a straight condition thereof and setting up in the memory of a computer controlling the guide shoes a deviation table relating said local deviations to the position of said elevator car along a path of travel… and subsequently controlling said shoes by said computer in dependence on said table to displace said elevator car laterally to compensate for a local deviation at any position…"

And Otala's specification describes the hardware chain: "a microprocessor 11, mounted in the elevator car… constantly addressing (READ command)… an EPROM-type memory 12 via an address bus 13. The memory circuit sets the contents of the memory position addressed at each moment, e.g. on an eight-bit data bus 14, the value of this byte being converted into an analog voltage by a D/A converter 15 and said voltage being amplified by amplifiers 16. The amplified voltage serves as loading value for motors 17 of the hydraulic system, these motors controlling hydraulic pumps 18 which in turn operate the hydraulic cylinders 9 controlling the guide shoe 7…"

Otala's claim 2 adds: "temporarily attaching acceleration meters to the elevator car and performing an uncorrected trial run… the deviation value table being derived from information obtained from said acceleration meters." Otala's claim 3 adds that the guide-shoe processor "obtain[s] information on the position of said elevator from a control computer of said elevator."

Mapping to claim 1:

So Otala anticipates the architecture and the learn/recall indexing, but on the specification's reading it lacks the two-sensor content of the stored record.

IV.B The profilometer art supplies exactly the missing element

The Rudd/Brandenburg literature — cited by the applicant as prior art — is not merely analogous; it is the identical measurement problem in a rail-vehicle setting. The FRA/ENSCO treatment of the same technology states the principle verbatim:

"This situation is resolved by adding the displacement of the mass relative to an inertial reference position to the relative displacement between the wheel and mass. The displacement of the mass relative to the inertial plane is determined by double-integrating the output of the accelerometer." (Use of Automatically Acquired Track Geometry Data, FRA, 1980 — describing the Rudd/Brandenburg profilometer.)

And the '721 specification's own governing equation, x_r = x_a + x_b + r_w (rail displacement = car displacement from plumb line + wheel-to-car relative displacement), is the same sum the profilometer art teaches, with x_a being the double integral of the sensed acceleration and x_b the relative displacement. The profilometer art further teaches the high-pass filter to strip integration drift — the exact function of the '721 claim-4 filtering steps:

"Profile data is high-pass filtered to eliminate integration errors…" (FRA, 1980); and, from the UMTRI profilometer study, "Thermal drift, resolution limits, and hysteresis result in offsets, which when double integrated cause erroneous indications of elevation. This problem is controlled in profilometers by passing the accelerometer output through a high pass filter to strip off low frequency drift." (https://deepblue.lib.umich.edu/bitstream/handle/2027.42/227/74520.0001.001.pdf)

Thus the very combination the '721 specification calls its novelty — displacement (x_b) + doubly-integrated acceleration (x_a), stored as the profile — is the standard inertial-profilometer algorithm, published 18 years before the '721 priority date.

IV.C The combination

A PHOSITA tasked with implementing Otala's deviation-table scheme would confront Otala's own stated method ("the deviation value table being derived from information obtained from the acceleration meters") and would immediately know its deficiency: acceleration-only double integration drifts and is corrupted by unrepeatable, non-rail-induced accelerations (the specification's own FIGS. 10(a)–12(a) show precisely this non-repeatability across loads). The known, textbook cure in the rail-profile-measurement art is the two-sensor inertial profilometer: sum the relative displacement with the doubly-integrated acceleration. Applying that known cure to Otala's known system yields the '721 stored record with predictable results.

Claim-by-claim result of Ground 1:

• Claim 1 — obvious: Otala (actuator + position-indexed memory + recall + drive signal) + Rudd/Brandenburg (store a profile that is the sum of relative displacement and doubly-integrated acceleration).
• Claim 2 — obvious: the "combined sensed information signal" captured "during a rail profile run" is Otala's "uncorrected trial run," in which both an accelerometer and a relative-displacement transducer are sampled; using a single data channel carrying both is a trivial engineering choice.
• Claim 3 — obvious: the "double integration means" and "summing means" are the express, literal teaching of Rudd/Brandenburg (double-integrate the accelerometer output; add to the relative displacement).
• Claim 5 — obvious for the same reasons, in system form; the "learned-rail memory means" is Otala's EPROM deviation table, and the "actuator means" are Otala's guide-shoe hydraulic cylinders.
• Claim 6 — obvious as the method counterpart (Otala claims 1–4 are themselves method claims performing exactly these steps).

---

V. Ground 2: Otala in view of the elevator rail-measurement art of Group C

If the examiner or a defendant prefers elevator-specific art (to defeat any "non-analogous art" argument), the Mitsubishi JP H01-321283 / H01-321285 ("Instrument for measuring installation accuracy of elevator guide rail," 1988-06-22) and Hitachi JP H03-124683 ("Mounting accuracy measuring device of guide rail," 1989-10-11) references supply the elevator-context teaching of measuring guide-rail deviation directly, rather than inferring it from acceleration alone.

• Motivation: Otala's stated object is to ascertain the magnitudes of local deviations of the guide rails from a straight condition and to relate them to car position. The Mitsubishi/Hitachi instruments are dedicated to that precise measurement in the same field of endeavor (elevators). A PHOSITA implementing Otala with an accuracy requirement (Otala's own goal is to double the interval between rail-straightening operations, which demands an accurate deviation measurement) would consult the known rail-accuracy measurement instruments, then combine their relative-displacement output with an accelerometer-based inertial reference to produce a full-length profile.
• Result: the stored table becomes a function of both a rail-referenced displacement measurement and a car-referenced acceleration measurement, indexed by position — claims 1/5/6.

This Ground is weaker than Ground 1 only because the Japanese abstracts as cited do not, on their face, recite the summing step; I have not verified their full texts, so I present Ground 2 as a secondary/backup combination, not the lead theory.

---

VI. Ground 3: Otala in view of the Otis active-suspension patents (Group D)

If the § 103 challenge is directed specifically at the actuator and control-architecture limitations, the Otis patents cited of record — US 5,308,938 ("Elevator active suspension system"), US 5,321,217 ("Apparatus and method for controlling an elevator horizontal suspension"), and US 5,322,144 ("Active control of elevator platform"), all filed 1990-07-18, i.e., before the 1991-03-13 priority date — disclose actuator-based horizontal suspension with position/acceleration sensors and a controller. Combining Otala (open-loop learned-table recall) with the Otis actuator architecture yields claim 1's "actuator means responsive to an actuator drive signal" and claim 5's system framing. These references are commonly owned with the '721 patent, so a § 103(c) (pre-AIA) common-ownership disqualification could be raised against using them as prior art in a patentability challenge — but they remain available as evidence of the ordinary skill level and the state of the art, and they are not disqualified for the analogous-art/obviousness-to-try analysis if used only for what a PHOSITA knew.

(Note: the specification's own FIGS. 13–15 show the actuator arrangement — actuators 180/182/236/238, tie rod 184/244, springs 190/192/240/242 — which the Otis patents disclose in like form, reinforcing that this element required no invention.)

---

VII. Claim 4 in particular: the detrend/filter/reverse-time pipeline

Claim 4 is the most detail-heavy claim but the least defensible under § 103, for two independent reasons:

1. The applicant admitted the technique is known. The specification states: "The multi-pass filtering technique described is one method of smoothing data and is discussed in some detail in Applied Optimal Estimation, Chapter 5, 'Optimal Linear Smoothing', Gelb, A., editor MIT Press 1974. Other methods of data smoothing also exist, as discussed for example in Optimal Filtering, Chapter 7, 'Smoothing of Discrete-Time Signals', Anderson, B. D. O., and J. B. Moore, Prentice-Hall, 1979, which could also be utilized." That is a classic admission that the claimed smoothing steps were within the ordinary skill level.
2. Every sub-step is standard and the parameters are routine optimizations:
   • "integrate and detrend (subtract mean and first moment)" = standard removal of integration drift;
   • "high pass filter (third order Butterworth with 0.5 Hz breakpoint)… reversing time after the filtering operation" = zero-phase (forward-backward) filtering, the standard technique for avoiding the phase distortion the UMTRI profilometer study warns about ("phase shift effects in filtering cause a different profile to be obtained when measured from opposite directions");
   • the 0.5 Hz / 20 Hz breakpoints are expressly justified by the specification in terms of sensor noise and ISO/DIS 2631 human-sensitivity — i.e., routine optimization (MPEP § 2144.04) against a published standard.

The profilometer literature supplies the filtering rationale, and the two admitted smoothing texts supply the method. Claim 4 is obvious over Otala + Rudd/Brandenburg + Gelb/Anderson & Moore, with the breakpoint selection an obvious design choice driven by ISO 2631.

---

VIII. Motivation to combine (the KSR factors)

A PHOSITA would have been motivated to combine Otala with the profilometer art for several independently sufficient reasons:

1. Same problem, same field of endeavor (or at minimum "reasonably pertinent"). Otala is about ascertaining guide-rail out-of-straightness and using it to counteract lateral car oscillation. Rudd/Brandenburg is about ascertaining rail-surface deviation from an inertial reference — the identical measurement problem for a guided vehicle. The applicant's own citation of the profilometer papers as prior art confirms pertinence.
2. Otala explicitly invites the improvement. Otala derives its table "from information obtained from the acceleration meters" and concedes the table must be rebuilt when the rails move. A designer improving Otala's accuracy is drawn directly to better rail-profile measurement, and the established method for that is the inertial profilometer.
3. The deficiency of the single-sensor approach was known and documented. Acceleration-only double integration drifts (thermal, gravity-misalignment) and is corrupted by non-rail accelerations. The patent's own FIG. 10(a)/11(a)/12(a) set — non-repeatable acceleration vs. position — demonstrates the motivation. This is the classic case where the specification's "teaching of why the prior art fails" supplies the very motivation to combine.
4. Finite, predictable solutions. By 1991 there was essentially one known way to recover a true rail profile from on-board sensors: combine a relative-displacement measurement with a doubly-integrated inertial measurement. That is the "finite number of identified, predictable solutions" of KSR.
5. Predictable result, no unexpected behavior. The combination simply yields the rail profile the art defines; the ~90% vibration reduction claimed is a magnitude benefit from accurate open-loop compensation, not a kind of result that differs from what one would expect from accurate profiling.
6. Design incentive / commercial pressure. Both references frame their work around accelerating vehicles (Otala on high-speed elevator ride quality; the profilometer papers on high-speed rail). The need to raise elevator speed while keeping lateral ride quality is a classic design incentive.

---

IX. Reasonable expectation of success

The combination is a predictable, mechanical/controls combination with a high expectation of success:

• The profilometer algorithm is a closed-form identity (γ = ∬z dt² + W), not a speculative technique.
• The Otis patents establish that elevator actuator/suspension hardware capable of executing the retrieved commands already existed.
• Otala establishes that position-indexed memory recall driving actuators was already reduced to practice.

There is nothing in the combination that the art taught would not work.

---

X. Counterarguments and how a patent owner would respond (and why they likely fail)

---

XI. Conclusion

Primary § 103 ground. US 4,750,590 (Otala) in view of Rudd & Brandenburg, "Inertial Profilometer as a Rail Surface Measuring Instrument" (1973) and Brandenburg & Rudd, "Development of an Inertial Profilometer" (1974) — optionally with the Skalski/Groom magnetic-suspension papers — renders claims 1, 2, 3, 5, and 6 obvious. Otala supplies the actuator, the position-indexed memory, the learn/recall architecture, and the acceleration-meter-derived table; the profilometer literature supplies the missing limitation — that the stored profile is the sum of a relative displacement measurement and a doubly-integrated acceleration measurement — with a documented motivation (integration drift and non-repeatability) and a predictable result.

Secondary grounds.

• Claim 4 is obvious over Otala + Rudd/Brandenburg in further view of the applicant's admitted smoothing art (Gelb 1974; Anderson & Moore 1979) and ISO/DIS 2631, with the filter breakpoints a routine design choice.
• Claims 1/5/6 are additionally obvious over Otala in view of the elevator-specific rail-measurement instruments (JP H01-321283 / H01-321285; JP H03-124683) and/or the Otis active-suspension patents (US 5,308,938; 5,321,217; 5,322,144), subject to the § 103(c) common-ownership caveat on the Otis patents.

Confidence. Moderate-to-high on Ground 1 as a legal theory: the two load-bearing disclosures (Otala and the inertial profilometer) are verified verbatim and align with the '721 claims almost element-for-element. Moderate on the breadth of the Japanese-language references, which I did not read in full. High on the § 103 framework and the motivation analysis. Note that this is a theoretical invalidity analysis prepared for completeness only — as the earlier sections establish, the patent expired 2013-08-13, so any § 103 ground would now be relevant only to a past-damages or indemnity dispute, and the § 286 six-year bar likely forecloses even that.

---

XII. What I could not verify (explicit uncertainty)

1. Full text of the 12 Japanese-language cited references (Hitachi/Mitsubishi guide, damping, and rail-accuracy filings). Their element-level disclosures are drawn from titles/dates and the specification's characterizations.
2. Full text of US 5,020,639 (Inventio), US 5,027,925 (Kone), US 5,086,882 (Hitachi), US 4,757,849 (Mitsubishi). Treated as cumulative hardware/control art; not read in full.
3. The Derwent abstract AN 90-224718 / FI A 8804380 — only the citation-level identity is known.
4. Whether any of the cited references was applied in a rejection during prosecution of application 08/279,826. I have not retrieved the file wrapper; if a § 103 rejection on Otala was made and overcome, the prosecution history would materially affect the analysis above. Recommend pulling the application file at USPTO PatentCenter (application 08/279,826).

Key URLs: US5544721A — https://patents.google.com/patent/US5544721/en · Otala US4750590A — https://patents.google.com/patent/US4750590A/en · Rudd/Brandenburg 1973 — https://trid.trb.org/view/16107 · Brandenburg/Rudd 1974 — https://trid.trb.org/View/22515 · UMTRI profilometer filtering discussion — https://deepblue.lib.umich.edu/bitstream/handle/2027.42/227/74520.0001.001.pdf · FRA 1980 profilometer principle — https://railroads.fra.dot.gov/sites/fra.dot.gov/files/fra_net/16059/1980_USE%20OF%20AUTOMATICALLY%20ACQUIRED%20TRACK%20GEOMETRY%20DATA.PDF