US 6718277

I'll verify the patent details against live sources, including the USPTO record and any 2026 CAFC activity.

Let me check for any Federal Circuit 2026 activity and the PTAB proceeding referenced in the record.

Let me do one more targeted check for Federal Circuit activity before finalizing.

US Patent 6,718,277 — Verified Summary

Identification note: The patent is properly cited as US 6,718,277 B2 (often written "6718277" or "6,718,277"). Everything below reflects that exact number; I excluded results for neighboring numbers (e.g., 6,854,287, 6,868,682, 7,031,870), which belong to the same HPE/Valtrus family but are different patents.

Bibliographic data

Abstract (verbatim)

"A method and system for controlling atmospheric conditions within a building. A conditioned fluid is supplied inside of the building and one or more atmospheric parameters in various locations inside of the building are sensed. An empirical atmospheric map is then generated and compared to a template atmospheric map. Pattern differentials are identified therebetween and corrective action to reduce the pattern differentials is determined. One or more of the quantity, quality, and distribution of the conditioned fluid is varied in accord with the corrective action determination."

Plain-language overview of the independent claims

Claim 1 — Method of controlling atmospheric conditions in a building (the broadest claim). Five steps: (a) supply a conditioned fluid inside the building; (b) sense at least one atmospheric parameter at a plurality of indoor locations; (c) generate an empirical atmospheric map from the sensing results; (d) compare that empirical map to a template atmospheric map; and (e) identify pattern differentials between the two maps. Note: Claim 1 stops at identifying the differentials — it does not require taking corrective action. Corrective action and fluid adjustment are added only in dependent claim 2, and the specific control variables (vents, blowers, conditioned-air source) appear in claims 3–5.

Claim 12 — Method of cooling a data center containing equipment. The data-center-specific sibling of claim 1: (a) supply a cooling fluid within the data center to cool the equipment; (b) sense temperature at a plurality of locations; (c) generate an empirical thermal map from those readings; (d) compare it to a template thermal map; and (e) identify pattern differentials between them. Like claim 1, it does not itself require remediation; that is added by dependent claim 13. Claim 17 adds thermal-mapping software that triangulates hot-spot locations; claim 21 adds extrapolation/triangulation to locate hot spots.

Claim 22 — System for controlling atmospheric conditions in a building. A means-plus-function system claim: means for supplying conditioned fluid; means for sensing at least one atmospheric parameter at a plurality of locations; means for generating an empirical atmospheric map from the sensing means; means for comparing the empirical map to a template map; and means for identifying characteristics of pattern differentials — explicitly including location, size, and intensity. Claim 23 adds means for determining corrective action and means for varying quantity/quality/distribution; claim 24 specifies an air-conditioning system with vents/blowers/conditioned-air source and adds that the generating means triangulates hot spots.

Inventive core (as described in the specification): discrete, location-specific sensor points (temperature, and optionally humidity, pressure, velocity, particle, smoke) are converted into continuous map information; hot spots are triangulated between sensors (including at various elevations); the resulting empirical map is compared to a CFD-derived or real-time template map of an optimally operating facility; and cooling output is then steered preferentially to the hot spots — i.e., directed, location-specific cooling instead of undifferentiated room-level cooling. The specification expressly contemplates alternatives to discrete sensors: infrared thermography and inferring temperature from rack power draw.

Litigation and PTAB activity (as of the latest retrievable data)

• PTAB — IPR2025-00668, Vertiv Corporation v. Valtrus Innovations Limited / Key Patent Innovations Ltd., filed Feb. 26, 2025, on the '277 patent. Institution decision July 31, 2025 → Director discretionary denial: Deny (not instituted – procedural). Notably, the Google Patents "family has litigation" entry shows the petitioner field blank for this IPR; the petition caption and PTAB-tracker records identify Vertiv Corporation as petitioner. Refund of post-institution fees requested Sept. 2025.
• E.D. Tex. campaign (Valtrus Innovations Ltd. / Key Patent Innovations Ltd.): the '277 patent appears in the Valtrus assertion set alongside 6,854,287; 6,868,682; 6,862,179; 7,031,870; 7,339,490; 9,310,855; and 6,868,683. Specific dockets referenced include 2:24-cv-00139 (Digital Realty — dismissed with prejudice per Jan. 23, 2025 order), 2:24-cv-00142 (Dawn Acquisitions d/b/a Evoque), 2:24-cv-00361 (NTT Data — lead case, with Vertiv's declaratory-judgment action 2:24-cv-00907), 2:24-cv-00776 (TierPoint — dismissed without prejudice), 2:24-cv-00777 (DataBank), plus 2:24-cv-00534/00535, 3:24-cv-01152.
• Claim construction (E.D. Tex., 2:24-cv-00361-JRG): the term "empirical atmospheric map" / "empirical thermal map" is disputed. Vertiv proposed "visual representation of an area showing measured [atmospheric/temperature] conditions corresponding to locations"; Valtrus argued plain and ordinary meaning (a "correlation or compilation of data with location"). Vertiv additionally asserted indefiniteness against several claims. Joint P.R. 4-3 statement dated Aug. 11, 2025.
• 2026 filings: the record lists newer suits, including N.D. Ill. 1:26-cv-03926 / 1:26-cv-03928 / 1:26-cv-03945 (e.g., NTT Global Data Centers Americas), D.N.J. 2:26-cv-03884 / 2:26-cv-03886 / 2:26-cv-03890, E.D. Tex. 2:26-cv-00287, and W.D. La. 3:26-cv-01346 — the latter reportedly against Lumen Technologies Inc. over the '277, '287, and '682 patents.

Explicit uncertainties

1. Federal Circuit: I found no 2026 CAFC docket or appeal naming US 6,718,277 in any source retrieved. That is a negative result, not proof of absence — a CAFC appeal could exist but be indexed only in PACER/CM-ECF or not yet appear in the sources I could reach. Treat "no CAFC activity" as unconfirmed.
2. Petitioner identity for IPR2025-00668: Google Patents' litigation block lists the IPR under Unified Patents' PTAB data with an empty petitioner field; the petition caption names Vertiv Corporation. These do not conflict on the patent number, but the petitioner attribution comes from secondary trackers rather than the Google record itself.
3. Assignee string: Google Patents lists both "Valtrus Innovations Ltd" and "Hewlett Packard Enterprise Development LP" under "Current Assignee," while the legal-events history shows the 2021 chain ending at Valtrus Innovations Limited (Ireland). The assignment record (not Google's summary) is the more reliable indicator that Valtrus is the current owner.
4. Dates: some retrieved material is dated June 2026, which is later than the April 26, 2026 date stated in your prompt. I have reported the litigation data as retrieved rather than reconciling those dates.

Sources: https://patents.google.com/patent/US6718277/en · https://uspto.report/patent/grant/6718277 · https://patents.justia.com/patent/6718277 · https://ipverse.greyb.com/ptab-web/cases/case-details/IPR2025-00668 · https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1557497](/patent/1557497) · https://www.courtlistener.com/docket/68526786/1/1/valtrus-innovations-ltd-v-ntt-data-services-llc/ · https://storage.courtlistener.com/recap/gov.uscourts.txed.[230090](/patent/230090)/gov.uscourts.txed.230090.117.0.pdf · https://www.archive.org/download/gov.uscourts.txed.230090/gov.uscourts.txed.230090.148.0.pdf · https://www.patsnap.com/de/resources/blog/litigation/valtrus-v-digital-realty-trust-data-center-cooling-ip-dismissed-patsnap/ · https://www.legalnewsline.com/louisiana-record/suit-la-data-center-company-infringing-on-cooling-patents/article_09c9c16a-be3b-48e6-b7e3-b1cdfca365a3.html

I'll search live sources for litigation specifically involving US 6,718,277.

Let me search for the newer 2025–2026 filings and any additional defendants.

Let me pull the MDL 3190 case list and remaining 2026 dockets.

Litigation involving US Patent 6,718,277 (as retrieved 2026-09-11)

Scope note / rule compliance: Every case below was retrieved by searching the literal string 6718277 / 6,718,277. I excluded look-alike numbers that appear in the same results (e.g., 6,718,415 in Acqis v. IBM, and 8,277,377 in Philips v. Fitbit) — those are different patents and are not on this list.

Date-consistency flag (read first): Your prompt states today is April 26, 2026, but the live records I retrieved are dated June 2026, August 2026 and September 2026 (e.g., the MDL 3190 conditional transfer order of Sept. 9, 2026, and a Patent Lawyer Magazine piece of Aug. 21, 2026). I am reporting what the sources actually say rather than reconciling the two dates. This is consistent with the contradiction already flagged in the previously generated summary.

A. The enforcement campaign

All district-court cases below are brought by Valtrus Innovations Limited (Ireland) and, in the 2026 wave, co-plaintiff Key Patent Innovations Limited ("KPI," Ireland), successors-in-interest to the HPE portfolio. The '277 patent was asserted as part of a multi-patent set. Important limitation: the '277 patent reached its anticipated expiration on April 17, 2022, so any recovery is confined to pre-expiration conduct.

E.D. Tex. wave (2024) — all before Judge Rodney Gilstrap, Marshall Division

2026 wave — all consolidated/transferring into MDL No. 3190

MDL No. 3190, In re: Valtrus Innovations Ltd. Patent Litigation — motion for transfer filed April 28, 2026; MDL number assigned April 30, 2026; venue designated E.D. Tex. Valtrus and KPI moved to centralize 12 actions on overlapping patents/technology/facts. A joint opposition was filed June 5, 2026 by defendants including Iron Mountain Data Centers, Lumen, EvoDC, H5 Data Centers, Netrality Properties and Prime Data Centers.

Additional defendants named in MDL 3190 briefing but for which I did not retrieve a case number: Iron Mountain Data Centers LLC, H5 Data Centers LLC, Cologix, and CoreSite. Patent Lawyer Magazine (Aug. 21, 2026) lists all of Lumen, EvoDC, Prime Data Centers, Cologix, H5, Iron Mountain, CoreSite, Netrality and NTT Global Data Centers Americas as defendants in the campaign. Treat the case-number mapping for Cologix, H5, Iron Mountain and CoreSite as unconfirmed.

B. Related administrative proceeding (not district-court litigation)

IPR2025-00668 — Vertiv Corporation v. Valtrus Innovations Ltd., PTAB, filed Feb. 26, 2025, on U.S. Patent No. 6,718,277 (application 10/123,403). Institution decision July 31, 2025 → Director discretionary denial: Deny (not instituted – procedural), on Fintiv/§314(a) and §325(d) grounds. Post-institution fee refund requested Sept. 17, 2025; refund approved Sept. 23, 2025. This is an agency proceeding, listed here only because it is patent-specific and directly relevant to enforcement of the '277 patent.

C. Claim-construction exposure specific to the '277 patent

In E.D. Tex. 2:24-cv-00361-JRG, the term "empirical atmospheric map" / "empirical thermal map" is disputed. Vertiv proposed a "visual representation of an area showing measured [atmospheric/temperature] conditions corresponding to locations"; Valtrus urged plain and ordinary meaning. Vertiv also raised indefiniteness against several claims. Vertiv's DJ complaint pleads non-infringement element-by-element against claim 1 (no "generating an empirical atmospheric map," no "comparing… to a template atmospheric map," no "identifying pattern differentials").

D. Explicit uncertainties

1. Defendants for 1:26-cv-03926, 03928, 03958, 2:26-cv-03884, 03890, 2:26-cv-00287, 2:24-cv-00534, 3:24-cv-01152 — I could not retrieve the named defendants from the sources reached. Do not treat these as "no defendant"; they are simply unretrieved.
2. Whether every 2026 case asserts the '277 patent — the MDL covers the portfolio generally. I confirmed '277 assertions in the NTT ('361), Digital Realty ('139), TierPoint ('776), Lumen ('1346) and NTT Global Data Centers ('3945) complaints. For the remaining 2026 dockets, '277 assertion is inferred from MDL membership, not individually verified.
3. Vertiv's DJ posture — it is simultaneously a defendant-side intervenor in customer suits and a plaintiff in 2:24-cv-00907. Valtrus' motion to dismiss that DJ complaint was pending as of the last retrieved document (Aug. 11, 2025).
4. No Federal Circuit activity naming US 6,718,277 was found in any retrieved source. As previously flagged, that is a negative result, not proof of absence.
5. The Google Patents record for the '277 patent is itself consistent with this list — its "family has litigation" block carries the E.D. Tex. dockets (2:24-cv-00534, 00535, 00907, 01152, 00776, 00777, 00361, 00142, 00139, 00259) plus IPR2025-00668. Note that 2:24-cv-00259 appears there as well and is not in my table above — it is an additional Valtrus E.D. Tex. member case I could not otherwise detail. Flag as unconfirmed.

Primary sources: https://patents.google.com/patent/US6718277/en · https://portal.unifiedpatents.com/litigation/caselist · https://ipverse.greyb.com/ptab-web/cases/case-details/IPR2025-00668 · https://www.courtlistener.com/docket/68526786/1/valtrus-innovations-ltd-v-ntt-data-services-llc/ · https://storage.courtlistener.com/recap/gov.uscourts.txed.[230090](/patent/230090)/gov.uscourts.txed.230090.117.0.pdf · https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1557497](/patent/1557497) · https://dockets.justia.com/docket/multi-district/jpml/MDL%20No.%[203190](/patent/203190)/[1752936](/patent/1752936) · https://dockets.justia.com/docket/multi-district/jpml/TXE/2:26-cv-00286/[1752946](/patent/1752946) · https://dockets.justia.com/docket/multi-district/jpml/ILN/1:26-cv-03945/[1752953](/patent/1752953) · https://cdn.pacermonitor.com/public/case/66137018/Valtrus_Innovations_Ltd_et_al_v_Lumen_Technologies_Inc · https://cdn.pacermonitor.com/public/case/66698503/Valtrus_Innovations_Ltd_et_al_v_Cogent_Communications_Holdings_Inc_et_al · https://www.legalnewsline.com/louisiana-record/suit-la-data-center-company-infringing-on-cooling-patents/article_09c9c16a-be3b-48e6-b7e3-b1cdfca365a3.html · https://www.harrismartin.com/publications/27/MDL/articles/89664/jpml-asked-to-centralize-data-center-cooling-patent-infringement-lawsuits/ · https://patentlawyermagazine.com/data-center-boom-creates-new-patent-battleground-around-ai-infrastructure/ · https://www.patsnap.com/de/resources/blog/litigation/valtrus-innovations-v-tierpoint-data-center-cooling-patent-dispute-patsnap/ · https://www.patsnap.com/de/resources/blog/litigation/valtrus-v-digital-realty-trust-data-center-cooling-ip-dismissed-patsnap/

Bottom line: Litigation is known for US 6,718,277. In short: a 2024 E.D. Tex. campaign by Valtrus against data-center operators (NTT, Digital Realty, TierPoint, DataBank, Evoque, plus a Vertiv DJ action), one dismissal with prejudice (Digital Realty) and one without prejudice (TierPoint); a 2026 expansion joined by Key Patent Innovations against Lumen, NTT Global Data Centers Americas, EvoDC, Netrality, Cogent, Prime Data Centers and others, now being consolidated as MDL No. 3190 in the Eastern District of Texas; and a related PTAB IPR (IPR2025-00668) that was denied institution in July 2025.

I'll reconcile the structured ODP block against what the earlier summary found, and verify the one proceeding it identified.

Now let me check for any additional PTAB proceedings on the '277 patent and any appeal of the denial.

Let me pin down the challenged claims and grounds in the petition, and confirm no second petition exists.

Proceedings overview

One AIA trial proceeding has ever been filed against US 6,718,277 — IPR2025-00668, and no trial was ever instituted: the Director discretionarily denied institution under 35 U.S.C. § 314(a) on 2025-07-31, so the tally is 1 total / 0 active / 0 claims canceled / 0 claims sustained / 1 institution denied (no merits FWD, no settlement of an instituted trial). The defensive posture this gives a defendant is "untested but unhardened": Valtrus cannot point to a PTAB FWD upholding any claim, and a defendant cannot point to a PTAB FWD canceling any claim. All 24 claims remain live, still bearing the 2004 HPE priority date, and the Valtrus/KPI assertion campaign (now including Lumen in W.D. La.) is proceeding entirely in district court.

Contradiction to flag. The structured "PTAB proceedings on file" block states that the USPTO ODP API returns no AIA trial proceedings for this patent. That block is wrong or not yet ingested. Live retrieval confirms IPR2025-00668 exists, was filed 2025-02-26, and was terminated by a Director decision on 2025-07-31. I am treating the web-sourced record as controlling and flagging the structured block as stale. Separately, the earlier-generated summary correctly identified IPR2025-00668 but left the petitioner attribution hedged; the Director's own decision caption confirms Vertiv Corporation as petitioner, so that hedge can be retired.

Note on proceeding numbers. Only IPR2025-00668 is directed to the '277 patent. IPR2025-00667 ('287) and IPR2025-00669 ('179) are sibling petitions filed the same day against different Valtrus patents; they were decided in the same joint Director decision. Do not attribute their outcomes to the '277 patent as if they were separate '277 proceedings.

IPR2025-00668 — Vertiv Corporation v. Valtrus Innovations Ltd. (and Key Patent Innovations Ltd.)

• Type: Inter Partes Review (35 U.S.C. §§ 311–319). Decided jointly with IPR2025-00667 and IPR2025-00669.
• Filed: 2025-02-26 (per PTAB docket data; the § 315(b) clock had been running from the 2024 Valtrus filings against Vertiv's customers).
• Status: Discretionary Denial (verbatim from PTAB docket data); docket entry styled "Director Discretionary Decision: Deny," 2025-07-31. Plain English: the petition cleared no merits threshold and no trial was instituted; the proceeding is over.
• Judge panel: No three-judge APJ panel decided this. The decision issues from the Office of the Under Secretary/Director — captioned "Before COKE MORGAN STEWART, Acting Under Secretary of Commerce for Intellectual Property and Acting Director of the United States Patent and Trademark Office." This is a product of the 2025-03-26 bifurcated institution framework, under which the Director decides discretionary denial before any merits panel touches the petition. No APJ panel composition is therefore public for the discretionary phase.
  • Petitioner counsel: Timothy Maloney, Daniel Schwartz, Matthew Werber, Allison Strong (Nixon Peabody LLP).
  • Patent owner counsel: Patrick Colsher, Matt Berkowitz, Connor Houghton, Khue Hoang, Aaron Morris (Reichman Jorgensen Lehman & Feldberg LLP).
• Petition grounds: Three grounds. Challenged claims appear to be claims 1–21 (independent claims 1 and 12 plus dependents 2–11 and 13–21, as reflected in the petition's claim-by-claim analysis), with independent system claims 22–24 not appearing in that analysis — flag as unverified, as I could not retrieve the petition cover page listing the challenged claims.
  • Ex. 1005 — U.S. Patent No. 6,283,380 ("Nakanishi"), cited on an IDS during prosecution but never used as a basis for rejection.
  • Ex. 1006 — U.S. Patent Pub. No. 2002/0149911 A1 ("Bishop"), never cited or considered during prosecution. This is the petition's primary reference.
  • Ground 2: Bishop alone, as an anticipatory § 102 reference.
  • Ground 3: Bishop in combination with Nakanishi, § 103 obviousness.
  • (A third ground is referenced; the petition's § 103/§ 102 structure and its Advanced Bionics arguments confirm Bishop is the load-bearing reference in at least two of the three grounds.)
  • Prosecution history the petition exploited: a 2003-06-16 non-final rejection of claims 1–5, 8–14, 17–20 over Kiser (US 5,290,200); claims added as 21–24 on 2003-09-16 without amendment to the originals; allowance 2003-12-24 with an Examiner's Amendment to claims 1, 10, 15 and 18; and a Certificate of Correction on 2004-11-16 amending claims 1, 12 and 17. The petition argued Office error under Advanced Bionics Part 2.
  • Claim construction positions (as presented to the PTAB, contrasted with the later district-court fight):

    Term	Petitioner	Patent Owner (implied)
    "empirical [atmospheric/thermal] map"	"data displayable as a visual representation of an area showing measured … conditions corresponding to locations"	"a set of measured … conditions corresponding to locations"
    "template [atmospheric/thermal] map"	"data displayable as a visual representation of an area showing model … conditions corresponding to locations"	"a set of model … conditions corresponding to locations"

• Institution decision: DENIED — 2025-07-31, under 35 U.S.C. § 314(a), on discretionary grounds. The Acting Director granted Patent Owner's request for discretionary denial and denied all three petitions; "no trial is instituted." The reasoning, verbatim in substance:
  1. Trial-timing / duplication — "the projected final written decision due date in the Board proceedings is in October 2026" while two parallel proceedings (one against Petitioner, one against Petitioner's customers) were set for trial 2026-04-06 and 2026-06-22, so "it is unlikely that a final written decision in this proceeding will issue before district court trial occurs, resulting in significant duplication of effort, additional expense for the parties, and a risk of inconsistent decisions."
  2. No likely stay — "insufficient evidence that the district court is likely to stay its proceedings even if the Board were to institute trial."
  3. Settled expectations — "each of the challenged patents has been in force for approximately twenty years, creating strong settled expectations."
  4. The denial was expressly "based on a holistic assessment of all of the evidence and arguments presented," i.e., the merits were never reached. Vertiv's Fintiv and Advanced Bionics arguments (including its contention that the Fintiv factors "as a whole favor institution" and that the merits were "particularly strong") were rejected without a merits ruling.
• Final Written Decision: None. No FWD exists and none can exist, because no trial was instituted. No claim of the '277 patent has ever been canceled or sustained by the PTAB. Do not represent otherwise in any brief or demand-letter response.
• Settlement / termination: The petitions were denied, not settled; termination was by Director decision, not by the parties. On 2025-09-17 Vertiv filed a Request for Refund of Post-Institution Fees, and the Board issued a Notice: refund approved on 2025-09-23 — a mechanical consequence of non-institution, not a settlement term.
• Appeal: No Federal Circuit appeal found. This is a negative search result, not proof of absence. Substantively, a § 314(a) discretionary denial is the paradigm case for the § 314(d) bar on judicial review (Cuozzo Speed Techs. v. Lee; Thryv, Inc. v. Click-To-Call), so appeal was not a viable path.
• Defensive value: A defendant gets essentially no leverage here. Nothing was invalidated, and the petition's Bishop/Nakanishi art was never adjudicated — but the denial is also not estoppel-bearing, and it costs a defendant nothing to re-urge that art in district court. The one real takeaway is a cautionary template: the Acting Director accepted "settled expectations" for a ~20-year-old patent and treated an October-2026 FWD date as too late against spring-2026 trial settings, so any follow-on IPR by a different defendant on this expired patent faces a steep discretionary-denial headwind.

Strategic summary

Claim status: all 24 claims are UNTESTED at the PTAB. There are no canceled claims, and there are no claims sustained on the merits. Claims 1–21 were challenged but never decided; claims 22–24 (means-plus-function system claims, with claim 24 expressly reciting that the generating means "triangulates hot spots") do not appear to have been challenged at all. If you are a defendant, you cannot tell a court or an adversary that "the PTAB killed claim 1" — it did not. Equally, Valtrus cannot wave a PTAB FWD as a validity shield. The entire validity fight is still in district court, which is why the E.D. Tex. claim construction of "empirical atmospheric map" / "empirical thermal map" (claims 1, 12, 22) is now the single most consequential event for this patent: the PTAB never construed it, so the Federal Circuit standard of review on that term is de novo/Phillips, not PTAB deference (Williams v. Levi framing works against the patent owner here).

Estoppel landscape: there is none, from anyone. Section 315(e)(2) estoppel attaches only after a final written decision; IPR2025-00668 never instituted, so Vertiv is not estopped — and neither is any other party. Practically, this means every ground is still available to a current defendant: Bishop (US 2002/0149911 A1) alone or in combination, Nakanishi (US 6,283,380), Kiser (US 5,290,200) and the other prosecution-era references, plus § 112 indefiniteness (which IPRs cannot reach at all — note Vertiv asserted indefiniteness against 29 claims of the '870 patent and pressed indefiniteness positions across the '277 family in district court). The prior denial creates no estoppel and no res judicata against a differently situated defendant, but it does hand the patent owner a ready-made discretionary-denial narrative under the same "settled expectations" and Fintiv-factor reasoning if a new petition is filed while E.D. Tex. trial dates are pending.

Pattern signals. This is a coordinated three-petition salvo by a single petitioner (Vertiv), filed 2025-02-26 against the '287, '277 and '179 patents in parallel with a declaratory-judgment action against Valtrus and an intervention campaign on behalf of Valtrus's customers — and it lost all three at the discretionary gate on the same day. There is no defensive aggregator in this chain: IPR2025-00668 was filed by a commercially interested defendant, not by Unified Patents or RPX (note that Google Patents' litigation block lists the IPR beneath Unified's PTAB data feed with an empty petitioner field — that is a data-attribution artifact, not a Unified filing). Valtrus/KPI, meanwhile, has shown no appetite for PTAB appeals, consistent with a patent owner that has now won three discretionary denials for free. The '277 patent expired 2022-04-17, so the exposure is past damages only (with the § 286 six-year lookback), not an injunction — which meaningfully changes your settlement math even though all claims are theoretically alive.

Recommended next steps

1. Do not plead or argue that any claim of US 6,718,277 has been canceled. It has not. The correct citation is the opposite: cite the 2025-07-31 Director decision — Vertiv Corp. v. Valtrus Innovations Ltd., IPR2025-00668 (P.T.A.B. July 31, 2025) (Director Discretionary Decision) — for the proposition that the Office declined to institute and made no validity determination. Sources: PTACTS petition viewer at https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1557496](/patent/1557496) (Director decision) and https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1557497](/patent/1557497) (Petitioner's Corrected Opposition to Patent Owner's Request for Discretionary Denial); docket summary at https://ipverse.greyb.com/ptab-web/cases/case-details/IPR2025-00668; official docket at https://e2e.uspto.gov/.
2. Preserve the Bishop/Nakanishi art as your lead § 102/§ 103 theory, but build it for Article III, not the PTAB. Bishop was never before the Examiner and was never adjudicated; it remains wholly available. Given the Director's "settled expectations" reasoning (≈20 years in force), assume a follow-on IPR will be discretionarily denied if parallel trial dates remain outstanding. Where timing permits, consider whether an ex parte reexamination under a different standard, or straight district-court invalidity, is the better vehicle — the 2025-2026 Office guidance has been openly skeptical of routine IPR institution on old patents.
3. Watch the § 315(b) one-year bar. Any defendant served with a Valtrus/KPI complaint on this patent has one year to file; the 2025 environment rewards early filing, and the March 26, 2025 interim-process memo plus the February 28, 2025 withdrawal of the Vidal Fintiv guidance mean the Office will not rescue a late petitioner.
4. Docket the E.D. Tex. claim construction. The "empirical atmospheric map" / "empirical thermal map" dispute (Vertiv's narrower "visual representation of an area" versus Valtrus's plain-meaning "correlation or compilation of data with location") is determinative now that no PTAB FWD exists; Valtrus's own briefing argues Vertiv's changed position was untimely "trial by ambush" gamesmanship. Verify the outcome and the indefiniteness rulings against the operative docket before relying on either party's framing. See the Joint P.R. 4-3 statement at https://www.courtlistener.com/docket/68526786/117/valtrus-innovations-ltd-v-ntt-data-services-llc/ and the claim construction briefs at https://www.courtlistener.com/docket/68526786/148/valtrus-innovations-ltd-v-ntt-data-services-llc/.
5. Confirm two open items before relying on this memo: (a) whether any second petitioner has filed against the '277 patent after the 2025-07-31 denial (my search surfaced none, but follow-on filings lag indexing); and (b) the exact challenged-claim set from the petition cover page — I report claims 1–21 with claims 22–24 apparently unchallenged, but flag that as reconstructed from the petition's claim-by-claim sections rather than verified from the cover page.

Explicit uncertainties: I could not retrieve the full petition cover page, so the challenged-claim list is reconstructed; I could not verify the proceeding's recorded termination date (the docket field returned blank); and the absence of a Federal Circuit appeal is a negative search result only. Institution denials under § 314(a) are procedurally unreviewable, so the absence of an appeal is expected rather than surprising.

I'll reconstruct the assignment chain from the USPTO Assignment Center records and cross-reference with the legal-events data I already have. Let me pull the recorded reel/frame data, correspondents, and any NPE-directory listings.

Let me pull the specific recording details, including reel/frame and correspondent, from the assignment records and PlainSite.

Strong signals emerging. Let me pin down the exact reel/frame for the '277 patent's Valtrus link, the correspondent of record, and the Ocean Tomo / OT Patent Escrow structure.

Assignment & Ownership Reconstruction — US 6,718,277 B2

Note on tooling: I hit the retained-search limit on my last four queries, so a few fields (notably the correspondents of record on the HPE-side recordings) could not be pulled directly from the Assignment Center and are marked not retrieved rather than guessed. Everything I do state is tied to a reel/frame or a named secondary source.

Cross-check vs. the previously generated summary: no contradiction found. The prior summary's chain (HPE Development LP → OT Patent Escrow, LLC → Valtrus Innovations Ltd) matches the recorded legal-events history and the RPX assignment record. One addition: the Valtrus entity is now documented as f/k/a Dolya Holdco 9 Limited, controlled by Key Patent Innovations Limited, which the earlier section did not capture.

Inventors

Pattern notes:

• The assignment was executed 16 days before the 2002-04-17 filing date, i.e., a standard pre-filing employment obligation, not a distress transfer.
• No inventor-departure signal. Sharma is the only named inventor, and there is no recorded reassignment, release, or inventor-side conveyance in the chain. The "all inventors gone within 12 months" fire-sale tell is not present on this record.
• No joint inventors, so no ownership-fragmentation issue (relevant because fragmentation is a common precursor to a portfolio sale).

Original assignee

• Named on the issued patent (2004-04-06): Hewlett-Packard Development Company, L.P. (via the 2003-06-18 reassignment, Reel 013776/0928).
• Original applicant/assignee at filing (2002): Hewlett-Packard Company (Colorado), per Reel 013127/0865.
• Primary line of business: HP was a full-stack operating company — computers, printers, servers, storage, and data-center infrastructure. This patent sits in its data-center thermal-management program; the specification cross-references HP's sibling filings "Smart Cooling of Data Centers" (Ser. No. 09/970,707, later US 6,574,104) and "Designing Layout for Internet Datacenter Cooling" (Ser. No. 10/076,635).
• Did it ship a product embodying the claims? HP/HPE is unquestionably an operating company that sold servers, storage and data-center cooling infrastructure, and HP commercialized related "smart cooling" technology out of the sibling filings. Whether a commercial HP/HPE product practiced the specific claims of the '277 is not documented in the retrieved record — I am flagging this as unclear rather than asserting it.
• Current status of the original assignee: Hewlett-Packard Company no longer exists as an independent entity; it split in 2015 into HP Inc. and Hewlett Packard Enterprise (HPE). The data-center side (HPE) is operating and solvent. No bankruptcy at any point — this matters for the fire-sale analysis below.

Assignment timeline

Five recorded assignments. Reel/frame for links 1–4 come from the Google Patents legal-events history of the '277 (which mirrors the Assignment Center); the Valtrus-side reel comes from RPX's assignment database.

1. 2002-04-01 (executed) / recorded 2002-07-30 — Reel 013127/0865

• Conveyance: Assignment of assignor's interest (employment/employment-obligation assignment)
• Assignor: Ratnesh Sharma (individual inventor)
• Assignee: Hewlett-Packard Company (Colorado)
• Correspondent: not retrieved (Assignment Center holds this field; typically HP's in-house patent department on an in-house-recorded assignment — do not treat that as confirmed)
• Context: Initial inventor→employer assignment, executed pre-filing as a condition of employment.

2. 2003-01-31 (effective) / recorded 2003-06-18 — Reel 013776/0928

• Conveyance: Assignment (intra-corporate)
• Assignor: Hewlett-Packard Company
• Assignee: Hewlett-Packard Development Company, L.P. (Colorado)
• Correspondent: not retrieved
• Context: Internal corporate reorganization — HP's standard 2003 transfer of IP into the HP Development Company vehicle. Not a sale; no third party.

3. 2015-10-27 (effective) / recorded 2015-11-09 — Reel 037079/0001

• Conveyance: Assignment of assignor's interest (intra-corporate, spun out with the HPE separation)
• Assignor: Hewlett-Packard Development Company, L.P.
• Assignee: Hewlett Packard Enterprise Development LP (Texas)
• Correspondent: not retrieved
• Context: Internal reorganization tied to the HP Inc./HPE split; the patent stays inside the operating company.

4. 2021-01-15 (effective) / recorded 2021-01-26 — Reel 055269/0001

• Conveyance: "PATENT ASSIGNMENT, SECURITY INTEREST, AND LIEN AGREEMENT" (a bundled assignment + security-interest/lien recording, not a plain assignment)
• Assignors: Hewlett Packard Enterprise Development LP and Hewlett Packard Enterprise Company
• Assignee: OT Patent Escrow, LLC (Illinois)
• Correspondent: not retrieved
• Context: Portfolio monetization transfer via an escrow intermediary — HPE moved a large patent tranche into an Ocean Tomo–affiliated escrow vehicle. This is the first break from an operating-company owner. Note the "escrow" here is a transaction structure, not an insolvency vehicle (see signal 6).

5. 2021-05-06 (recorded) — Reel/Frame not visible in the retrieved record; RPX associates Valtrus's acquisition tranches with Reel 055360/0424 (RPX-normalized "55360/424")

• Conveyance: Assignment of assignor's interest (portfolio sale out of escrow)
• Assignor: OT Patent Escrow, LLC
• Assignee: Valtrus Innovations Limited (Ireland) — f/k/a Dolya Holdco 9 Limited; registered address The Glass Houses, 92 Georges Street Lower, Dún Laoghaire, Dublin, A96 VR66, Ireland
• Correspondent: BOTOS CHURCHILL IP LAW LLP, 430 Mountain Avenue, Suite 401, New Providence, NJ 07974 (per RPX's assignment record for the Valtrus tranche — I could not confirm that Reel 055360/0424 is the specific recording covering the '277, so treat the correspondent as the firm of record for the Valtrus acquisition, pending Assignment Center confirmation)
• Context: Transfer-to-asserter / NPE acquisition. RPX: Valtrus "received at least 800 former HP Enterprise (HPE) patents through four assignments [in 2021], one direct and the others via Ocean Tomo, LLC (as OT Patent Escrow LLC)."

If, on verification at the Assignment Center, the '277's Valtrus recording carries a different reel than 055360/0424, only the reel number changes — not the conveyance, parties, or pattern.

Timeline diagram

timeline
    title Ownership of US 6718277
    2002 : Inventor assigns rights to Hewlett-Packard
         : Application filed on 17 April
    2003 : Reassigned to HP Development Company LP
    2004 : Patent issues on 6 April
    2015 : Moved to HP Enterprise Development LP
    2021 : HPE transfers to OT Patent Escrow LLC
         : Escrow transfers to Valtrus Innovations
    2022 : Valtrus files first portfolio suit
    2024 : Valtrus asserts 277 against NTT Data
    2025 : Vertiv files IPR2025-00668

NPE / troll-pattern signals

Corroborating assertion profile (context, not a separate signal): the '277 is Exhibit 1 in the NTT Data complaint (2:24-cv-00361); it sits in the Valtrus set alongside 6,854,287; 6,868,682; 6,862,179; 7,031,870; and others; it drew IPR2025-00668 (petition captioned to Vertiv Corporation), disposed of by Director discretionary denial; and KPI/Valtrus's Managing Director has stated in a PTAB declaration that Valtrus acquired the former-HP portfolio "in May 2021" — consistent with the 2021-05-06 recording.

Verdict

NPE — high confidence.

The chain leaves the operating company at Reel 055269/0001 (HPE → OT Patent Escrow, effective 2021-01-15) and lands at a licensing-only Irish vehicle, Valtrus Innovations Limited f/k/a Dolya Holdco 9 Limited, via the 2021-05-06 recording; Valtrus and its parent Key Patent Innovations are independently identified by both RPX and Unified Patents as high-frequency asserting entities, satisfying the "known asserter" signal. With the shell-entity transfer, cascading transfers through an escrow intermediary in under four months, and known-asserter signals all present — and with no bankruptcy, no defensive-aggregator terminus, and no products at the current owner — this is the textbook monetization posture. The two signals I could not cleanly establish (repeat correspondent and the strict 6-month pre-litigation test) are marked unclear rather than asserted in either direction.

Verify at: USPTO Assignment Center (search patent 6718277) — also indexed at assignment.uspto.gov/patent. Sources for this reconstruction: Google Patents legal events (https://patents.google.com/patent/US6718277/en); RPX Insight, "Irish NPE Files First Suit over Large Portfolio Pickup" (2022-01-16) and RPX patent-assignment record 055360/0424 (https://litigation.rpxcorp.com/patent_assignment/56848817); Unified Patents, "Key Patent Innovations entity, Valtrus Innovations, litigation analysis available" (2026-01-22); CourtListener docket 68526786 (Valtrus v. NTT Data, E.D. Tex. 2:24-cv-00361).

Two flags for the record: (1) the correspondents of record on Reels 013127/0865, 013776/0928, 037079/0001 and 055269/0001 were not retrieved and should be pulled from the Assignment Center before the "repeat correspondent" signal is finalized; and (2) the reel number for the Valtrus link (055360/0424) is sourced from RPX, not confirmed as the exact recording covering the '277.

I'll verify the patent and dig into its cited prior art. Let me run targeted searches.

Let me gather details on the remaining key references.

Prior-art analysis for US 6,718,277 B2

Scope note. The authoritative source for this task is the patent's own front-page citation record (the "Patent Citations (21)" table in the Google Patents text), which I reproduce and analyze below. The 21 references cited against US 6,718,277 are the universe the examiner (and, for four of them, third parties) placed of record. I verified several of the higher-relevance ones against live sources (FreePatentsOnline, uspto.report, Google Patents).

Two disclaimers before the analysis:

1. § 102 anticipation requires a single reference that discloses every limitation of the claim. My assignment of a claim to a reference below means "this claim's subject matter is most relevant to / potentially disclosed by this reference," not that I have found standalone anticipation. As explained in the summary, no cited reference appears to disclose all limitations of independent claim 1 or claim 12 in a single document, which is consistent with the patent having issued.
2. All dates are copyright/issue-record dates from the patent text; the "first date" column is the reference's own filing/priority date, the second is its publication date. Every reference predates the '277 filing date of 2002-04-17, so each is available as prior art under § 102(a)/(b) — except US 6,574,104, which qualifies only under § 102(e) (see item 21).

Master table — the 21 cited references

(Four references — US 5,074,137; US 5,290,200; US 5,769,315; US 6,574,104 — are designated "cited by third party"; the remainder are examiner citations.)

Detailed analysis — the references that matter most

1) US 4,737,917 A — Emhart Industries (Tier 1). Method and apparatus for generating isotherms in a forehearth temperature control system. Filed 1986-07-15; issued 1988-04-12.
This is the closest single reference to the map-generation concept. It discloses planar (and discussed 3-D) arrays of thermocouples, interpolation of discrete point readings into a continuous temperature field (Lagrangian polynomial fit), and generation of "isotherms" — bands of equal temperature — displayed as temperature profiles. It even addresses extrapolating beyond the sensor grid and interpolating between sensor planes (i.e., 3-D).

• Potentially anticipates / most relevant to: claim 1(c) ("generating an empirical atmospheric map"), claim 10 / 20 ("map composed of temperature contours that define various isothermal regions"), and claims 11 / 21 ("extrapolation … to determine locations" of hot regions). Emhart's interpolation-between-sensors is conceptually the same technique the '277 specification attributes to thermal-mapping software.
• Does not show: comparison to a template map, identification of pattern differentials, or HVAC control responsive thereto. It is a glass-forehearth PID control, not building atmospheric control.

2) US 5,506,768 A — Johnson Service Company (Tier 1). Pattern recognition adaptive controller and method used in HVAC control. Filed 1992-10-29; issued 1996-04-09. (See also its sibling US 5,355,305 A, Pattern recognition adaptive controller, and EP 0 628 181 B1.)
This is the closest single reference to the pattern-recognition element. It explicitly applies pattern recognition to HVAC control: it samples a feedback signal (temperature), derives a smoothed signal, characterizes the disturbance by a damping factor and closed-loop response time, and adapts the PI controller's gain/integral time. It contrasts its "PRAC" approach with model-reference adaptive control and self-tuning control.

• Potentially anticipates / most relevant to: claim 1(e) ("identifying pattern differentials") and claim 7 / 18 ("using pattern recognition software"). It is the strongest § 102(b) reference on the pattern-recognition-in-HVAC limitation.
• Critical gap: Johnson's "pattern" is the shape of a single time-domain feedback response curve, not a spatial map compared to a template map. It does not generate or compare areal/thermal maps, and it has no notion of "pattern differentials" between an empirical and a template field. So it defeats claim 1 only if one reads "pattern differentials" broadly enough to cover response-curve characterization — which is exactly the kind of dispute the E.D. Tex. claim construction (noted in the earlier section) is litigating.

3) US 6,574,104 B2 — Hewlett-Packard Development Co. (Tier 1, § 102(e)). Smart cooling of data centers (Patel et al.). Filed 2001-10-05; issued 2003-06-03.
This is the same-assignee cross-referenced application (Ser. No. 09/970,707) that the '277 specification expressly incorporates and identifies as the "smart cooling system" of FIG. 1. Because it was filed before the '277 filing date (2001-10-05 < 2002-04-17) but published after, it is § 102(e) prior art only.

• Potentially anticipates / most relevant to: claims 12–16 and 22–24 (the data-center cooling and system claims), since Patel describes location-aware, distributed delivery of cooling to racks in a data center — i.e., the "supply a cooling fluid within said data center" and "vary the distribution of conditioned fluid" concepts.
• Does not show: the empirical-map / template-map comparison that is the '277 point of novelty. It cools based on location-specific needs but not via atmospheric mapping/pattern recognition. (Note the potential § 102(e)/§ 103 interplay with the '277's own priority — worth flagging as a family-overlap issue, not full anticipation.)

4) US 5,326,028 A — Sanyo Electric (Tier 1). System for detecting indoor conditions and air conditioner incorporating same. Filed 1992-08-24; issued 1994-07-05.
Discloses a light-beam/PSD triangulation system that computes three-dimensional coordinates of reflecting positions across a room (to determine room area and occupant number/position) and feeds the result to an air conditioner's control circuit for "optimal" conditioning.

• Potentially anticipates / most relevant to: claims 11 / 21 ("performing … triangulation to determine locations") and claim 1(b) (spatial sensing). It is cited precisely for the triangulation/3-D-location idea.
• Does not show: atmospheric or thermal maps, isotherms, template comparison, or pattern differentials. Its "triangulation" locates objects, not hot spots.

5) US 4,823,290 A — Honeywell Bull (Tier 1). Method and apparatus for monitoring the operating environment of a computer system. Filed 1987-07-21; issued 1989-04-18.
Sensor array (temperature, humidity, liquid level, etc.) monitoring a computer room with polling, alarm thresholds, rate-of-change checks, time/date-stamped history, and histogram generation of environmental trends.

• Potentially anticipates / most relevant to: claim 1(b) and claim 9 (temperature/humidity sensors, etc.) and the "means for sensing … in a plurality of locations" of claim 22. It's a strong teaching of distributed environmental sensing in a computer facility.
• Does not show: map generation, template comparison, or automated corrective fluid adjustment (it raises alarms for human "remedial action").

Detailed analysis — the control-variable / distribution references

These are cited mainly against the dependent control-variable claims (2–5, 13–16, 23–24) and the "supply a conditioned fluid" limitation of claims 1(a)/12(a):

• US 5,177,972 A — Liebert (filed 1983-12-27; issued 1993-01-12): energy-efficient AC using a variable-speed compressor and integrally related expansion valves → relevant to "at least one source of conditioned air" and varying its output (claims 3–5, 14–16, 24).
• US 5,331,825 A — Samsung (filed 1992-03-07; issued 1994-07-26): air-conditioning system with controllable airflow → claims 3–5.
• US 5,478,276 A — Samsung (filed 1993-06-14; issued 1995-12-26): air-conditioner operation control apparatus/method → claims 2–5.
• US 5,249,741 A — IBM (filed 1992-05-04; issued 1993-10-05): automatic fan-speed control → claims 5/16 (blower speed).
• US 5,687,079 A — Sun Microsystems (filed 1994-04-08; issued 1997-11-11): improved computer-cooling fan-speed control → claims 5/16; shows temperature-driven modulation of a cooling blower in an electronics context.
• US 6,009,939 A — Sanyo (filed 1996-02-29; issued 2000-01-04): distributed air conditioning system → claims 2–5, 13–16 (distributed/zonal cooling).
• US 5,769,315 A — Johnson Service Co. (filed 1997-07-08; issued 1998-06-23): pressure-dependent variable-air-volume control strategy → claims 4/15/16 (adjusting vent/air distribution).
• US 6,296,193 B1 — Johnson Controls (filed 1999-09-30; issued 2001-10-02): controller for a dual-duct VAV terminal unit → claims 4/5 (controller responding to sensed condition to modulate air delivery).
• US 6,283,380 B1 — IBM (filed 1999-03-25; issued 2001-09-04): air conditioning system and method → generally claims 1–5 (the most "modern" examiner-cited AC control reference besides the '104).

Detailed analysis — remaining references

• US 2,991,405 A — General Motors (1960-02-19 → 1961-07-04): transistorized motor control responsive to temperature → only claims 5/24 (temperature-responsive actuation); very remote.
• US 5,074,137 A — Harris (1989-01-24 → 1991-12-24): Programmable atmospheric stabilizer → claims 1/9/22 (programmable control of an atmospheric parameter); third-party cited.
• US 5,290,200 A — Professional Supply, Inc. (1991-03-06 → 1994-03-01): detection and evacuation of atmospheric pollutants from a confined workplace → claims 1(b)/9 (particle/smoke sensing and conditioned-fluid response); third-party cited.
• US 5,372,426 A — Boeing (1992-03-11 → 1994-12-13): Thermal condition sensor system for monitoring equipment operation → claims 1(b)/8/9 (multi-point thermal sensing, incl. elevation).
• US 5,828,572 A — Canon (1995-07-07 → 1998-10-27): processing system with air-conditioning control in operational zones → claims 1/12 (location-specific/zonal atmospheric control); notable for the "operational zones" concept.
• US 6,080,060 A — ABB Fläkt (1996-04-01 → 2000-06-27): equipment for air supply to a room → claims 1(a)/3 (supplying conditioned fluid via room air-distribution hardware).
• US 5,709,100 A — Liebert (1996-08-29 → 1998-01-20): air conditioning for communications stations → claim 12 (cooling a data-center-type equipment room).

Bottom line on § 102

• Strongest § 102(b) candidates for individual limitations:
• US 4,737,917 (Emhart) → map/isotherm generation (claim 1(c), 10, 11).
• US 5,506,768 (Johnson Service) → pattern recognition in HVAC (claim 1(e), 7).
• US 4,823,290 (Honeywell Bull) → multi-point environmental sensing (claims 1(b), 9, 22).
• US 5,326,028 (Sanyo) → triangulation to fix 3-D locations (claims 11, 21).
• Strongest § 102(e) candidate: US 6,574,104 (Patel/HPE) → data-center cooling system (claims 12–16, 22–24), but it is the same family and lacks the map-vs-template core.
• No single cited reference discloses the full combination of claim 1 or claim 12. The distinguishing feature of the '277 — generating an empirical spatial map, comparing it to a template map, and identifying pattern differentials between the two fields — is not found in any one of the 21 references. Each reference supplies at most one or two claim elements; the examiner cited them for those pieces, not as anticipatory art. The real prior-art fight over this patent (as the earlier section shows) is therefore a § 103 / claim-construction contest around "empirical atmospheric map," not a § 102 knockout.

Caveats / explicit uncertainties:

• The descriptions of items where I could not pull the specification text (e.g., US 5,828,572; US 6,009,939; US 6,283,380; US 5,074,137) are based on the reliable titles/assignees and their evident claim relevance; I did not independently read those specifications, so my limitation-level mapping for them is inferential, not verified. The higher-tier references (Emhart, Johnson, Honeywell Bull, Sanyo, HPE '104) I did verify against live full-text sources.
• I searched the USPTO/Google record for patent 6,718,277 specifically; the results returned that exact patent (e.g., FreePatentsOnline 6718277, uspto.report grant/6718277). I did not substitute neighbor numbers such as 6,754,873 or 7,031,870.
• The final two batch searches (US 6,283,380 and US 5,074,137 details) did not return before the tool-step limit was reached; those entries rest on the authoritative citation table rather than independently retrieved full text.

Sources verified this session: https://patents.google.com/patent/US6718277/en · https://FreePatentsOnline.com/6718277.html · https://uspto.report/patent/grant/6718277 · https://patents.google.com/patent/[US4737917A](/patent/US4737917A)/en · https://www.freepatentsonline.com/[4737917](/patent/4737917).html · https://uspto.report/patent/grant/4737917 · https://uspto.report/patent/grant/[5506768](/patent/5506768) · https://patentimages.storage.googleapis.com/.../EP0628181B1.pdf · https://www.freepatentsonline.com/[4823290](/patent/4823290).html · https://uspto.report/patent/grant/4823290 · https://patents.google.com/patent/[US5326028A](/patent/US5326028A) · https://patentimages.storage.googleapis.com/7f/99/17/cd1e34e1ccc42f/US5326028.pdf

I'll ground the obviousness analysis in the actual prior-art references listed on the patent page, verifying their disclosures where possible.

Obviousness Analysis — US 6,718,277 B2 under 35 U.S.C. § 103

Building on the prior sections: This analysis uses the verified claim set, specification, and the 21 examiner/third-party citations plus the 4 foreign family citations listed on the Google Patents page. I do not repeat the bibliographic summary. Two points flagged earlier are directly material here and are developed below: (i) the disputed construction of "empirical atmospheric map," and (ii) the common-ownership / incorporation-by-reference problem with US 6,574,104.

Date discrepancy flag (carried forward): the environment header gives today as 2026‑09‑11, the task prompt gives April 26, 2026, and the prior section noted retrieved material dated June 2026. I report no new 2026 activity below and flag that all litigation statements are as of the prior section.

1. Legal framework and the correct prior-art regime

Pre-AIA § 103(a) governs. The '277 application was filed April 17, 2002, before the March 16, 2013 AIA change, so § 102(a)/(b)/(e)/(f)/(g) and pre-AIA § 103(a) apply. Every reference in the "Patent Citations (21)" and "Family Cites Families (4)" lists qualifies as prior art: all issued before April 17, 2002, except US 6,574,104 B2, which issued June 3, 2003 but is § 102(e) art via its October 5, 2001 filing date (App. 09/970,707) and its April 10, 2003 pre-grant publication — but see § 5 below, where I explain why 6,574,104 is a poor primary reference.

PHOSITA (as of April 2002): a mechanical or electrical engineer with a B.S. plus roughly 2–5 years in HVAC controls, computer-room/facility cooling, or building automation; familiar with (a) multi-point temperature/humidity sensing and sensor networks, (b) PID/PI control and adaptive tuning, (c) interpolation/contouring of scalar fields from sparse measurements, and (d) the economics of CRAC-unit compressor and fan power in data centers. The '277 specification itself supplies the last point as an admission (compressor power ≈ 30% of required cooling capacity; fans/condensers another ≈ 20%).

KSR standard applies: the question is whether the claimed subject matter as a whole would have been obvious, permitting rationales such as (A) known elements combined by known methods yielding predictable results, (C) use of a known technique to improve a similar device in the same way, (D) applying a known technique to a known device ready for improvement, and (F) design incentives/market forces. KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007).

2. What the claims actually require (the narrow frontage)

Two structural observations drive the whole analysis:

1. Claim 1 is a monitoring/diagnostic claim, not a control claim. It ends at "identifying pattern differentials." That removes the strongest non-obviousness argument a patent owner usually has (closed-loop control strategy), and it means the art need only show map generation + template comparison + differential identification.
2. Dependent claims 2–5 and 13–16 carry the closed-loop content, and that content maps almost element-for-element onto US 6,574,104 B2 (which is itself cited on the face of the '277 and is the '707 application the '277 expressly incorporates).

3. The prior-art reference set, classified by function

Verification note: I retrieved and read the full text of US 4,737,917, 5,506,768, 5,326,028, 6,009,939, and 6,574,104. For the remainder I rely on titles, assignees, and the fact of examiner citation; specific mappings to those references should be treated as provisional until their full texts are pulled.

4. Primary § 103 combinations

Combination A (strongest for claims 1, 6–11, 17, 21): Emhart '917 + Johnson '768, optionally + Sanyo '939

Mapping:

Why a POSITA would combine (motivation):

• Emhart itself supplies the motivation. Its stated problem is that "it is difficult for the operator to effectively visualize the temperature profile… based upon the limited number of discrete readings," and that where sensor readings are used in a control loop "the sophistication of the control technique may be limited by the use of only a discrete number of inputs." The '277 patent asserts the identical advantage as its invention: "a more precise method of tracking and using actual temperature measurement as an input to cooling system control," enabling "use of the spaces between the sensor locations… resulting in more accurate hot spot reduction than the prior art allows for." Emhart therefore teaches the very improvement the '277 claims, and teaches it for use in a heating/cooling control loop.
• Emhart also supplies the cost incentive that the '277 relies on: "thermocouples or other temperature sensor devices are very costly, it is accepted practice to use a limited number of temperature probes." Combined with the '277's admitted energy-cost problem in data centers, this is a textbook design incentive under KSR rationale (F).
• Johnson '768 is in the same field (HVAC control) and supplies the automated "decoding process in which discriminations in patterns are made without human intervention" that the '277 specification describes as its pattern-recognition step. Substituting an automated pattern-recognition comparator for Emhart's operator-observed display is a substitution of a known element for another known element to obtain a predictable result (KSR rationale (B)).
• Sanyo '939 adds the "plurality of locations" plus the comparative-differential logic (a detector whose reading deviates from the others beyond a threshold is flagged) and the A/C actuation, supplying claims 2–5 and 9 (temperature and humidity).
• Predictability: fitting a parametric/analytic field through discrete sensor points and comparing that field to a prescribed reference field is exactly the kind of predictable engineering step Emhart shows (Lagrangian polynomial of x, y and optionally z), and the '277's own alternative techniques (steepest-gradient, triangulation, MATLAB) are conventional numerical methods.

Result: Combination A renders claim 1 obvious (and claims 6, 7, 8, 10, 11, 17, 18, 21, which merely recite the software/technique labels and the 2-D/3-D/escalation/normalized-band details Emhart discloses). Claim 12 follows with Sanyo '939 or with 6,574,104 as the data-center context.

Combination B (strongest for claims 2–5 and 13–16, the closed-loop claims): A + US 6,574,104, or A + Liebert '972 / VAV '315 / '193

Mapping: all of claims 3–5 and 14–16 land on 6,574,104: "a plurality of vents" with independent vent controllers; "at least one blower" as a variable output fan; "at least one source of conditioned air" as a variable capacity compressor/chiller; and correlation of a localized condition (rack temperature above/below a set point) to the nearest vent and to compressor/fan output. The '277's own claim 4 recites exactly this correlation, and the '277 specification's worked example ("if one corner of the data center is demanding the most significant portion of the cooling needs… the most proximate chiller(s), blower(s), and vent(s) can be selected") is a paraphrase of the '104 control scheme.

Motivation: the '104 and '277 applications share an assignee and the '277 expressly identifies '104's parent ('707) as the smart cooling system to be used with the claimed method. Where the '277 specification itself says the atmospheric control system "can be a smart cooling system, exemplified by copending… Ser. No. 09/970,707," the patent owner has conceded the actuation half of claims 2–5 as known. That is an admission-based obviousness argument that does not even require § 103 combination logic.

Caveat — read this before deploying Combination B: see § 5.

Combination C (for claims 9, 10, 20 and the "quality" limitations): A + US 5,074,137 + US 5,290,200 + US 5,177,972

• Claim 9's laundry list (temperature, humidity, pressure, particle, smoke, velocity sensors) is met by Emhart '917 + Sanyo '939 (temperature, humidity) + Harris '137 (programmable atmospheric stabilizer — atmospheric-parameter sensing/regulation outside pure temperature) + Professional Supply '200 (particle/pollutant and smoke-type detection in a confined workplace atmosphere). Choosing among well-known sensor types to monitor a given environmental parameter is routine and predictable.
• Claim 10/20 (isothermal contour map) is directly disclosed by Emhart '917.
• The "quality" of the conditioned fluid (chiller cycle, output temperature) as distinct from quantity/distribution is met by Liebert '972's variable-speed compressor, which changes the refrigeration capacity/delivered air temperature.

Combination D (alternative, for the data-center claims 12–16): US 4,823,290 + US 5,372,426 + US 5,682,079 + US 5,506,768

Honeywell Bull '290 ("monitoring the operating environment of a computer system"), Boeing '426 ("thermal condition sensor system for monitoring equipment operation"), and Sun '709 ("improved control of computer cooling fan speed") together establish that (i) the "building" of claim 1 / the "data center" of claim 12 was a known target for multi-point environmental monitoring and (ii) sensed thermal condition was known to be fed back to the cooling actuator. Johnson '768 supplies the automated pattern comparison. The field is not merely analogous — it is the same problem (keeping heat-dissipating electronics within a thermal envelope), which defeats any non-analogous-art argument. See In re Bigio, 381 F.3d 1320 (Fed. Cir. 2004) (two-part test: same field of endeavor, or reasonably pertinent to the problem).

5. Two serious caveats that a challenger must confront

(a) US 6,574,104 / App. 09/970,707 is a risky primary reference. It is (i) expressly incorporated by reference into the '277 specification, which makes its disclosure part of the '277's own disclosure, and (ii) commonly owned with the '277 (both HP at the time the invention was made) and cited only under § 102(e) — so pre-AIA § 103(c) disqualifies it as § 103 prior art. Practically, 6,574,104 is best used as (1) background/admitted prior art establishing the known actuation hardware, and (2) support for the motivation to combine other references, not as a § 103 reference in its own right. The same logic applies to the '635 and '010 applications cross-referenced in column 1 of the '277. A petitioner relying on 6,574,104 for the map/template/pattern elements invites a § 103(c)/incorporation-by-reference knock-out.

(b) The entire 21-reference set is already of record, and the claims issued over it. This is a real headwind:

• Of record and cited by the examiner (per Google's key): US 5,074,137, US 5,290,200, US 5,769,315, US 6,574,104.
• The remaining references were cited in the patent (including by third parties during prosecution).
• A district court or PTAB challenge built solely on the references of record must explain the examiner's apparent error. The most defensible explanation is that the examiner never articulated the Emhart-to-Johnson-to-Sanyo combination, and in particular never relied on Emhart's express statement that its interpolated isotherms "may be included in an adaptive control loop to regulate devices for heating and cooling" — a teaching that converts Emhart from a glass-forehearth display patent into a § 103 linchpin for claims 1/6/8/10/11/21.
• Countervailing procedural point: IPR2025-00668 was denied institution on discretionary/procedural grounds (per the prior section, petitioner identified as Vertiv Corporation). A renewed IPR on the same art would face § 325(d) (art previously presented to the Office) and discretionary-denial exposure. A challenge that adds new art (e.g., non-U.S. building-automation/CRAC literature, or the Canon '572 zone-control patent worked up in detail) is materially stronger.

6. The claim-construction dispute is outcome-determinative

The E.D. Tex. dispute over "empirical atmospheric map" / "empirical thermal map" (Vertiv: "visual representation of an area showing measured conditions"; Valtrus: plain meaning — "a correlation or compilation of data with location"; joint P.R. 4-3 statement Aug. 11, 2025) maps directly onto obviousness:

• If Valtrus's position prevails, the term covers any location-tagged sensor data set. Emhart's interpolated grid, Sanyo '939's centrally collected detector readings, and 6,574,104's rack-location temperature table all read on it — and Combinations A–D become very difficult to distinguish.
• If Vertiv's "visual representation" position prevails, the art must produce a display, which Emhart '917 does (isotherm-band displays, FIGS. 3–6) but Johnson '768 and Sanyo '939 largely do not. Under that construction, Emhart becomes indispensable, and a patent owner can argue Emhart's forehearth display is not a building atmospheric map and that no reference teaches comparing two such displayed maps.
• Vertiv's separate indefiniteness challenge, if successful on "template atmospheric map" (never defined with numerical specificity) or on claim 22's means-plus-function terms, would moot the § 103 fight. Note that a § 112 ¶ 6 analysis of claims 22–24 requires identifying the corresponding structure in the '277 specification — CPU 14 + thermal mapping software 18 + pattern recognition software 20 + strategic software 22 — which are the same generic components Emhart's CPU 53 and Johnson's microprocessor 42/memory 48 disclose.

7. Claim-by-claim bottom line

Overall assessment: The '277 patent is, on this record, a substantially vulnerable § 103 patent if challenged with the combination Emhart '917 (field reconstruction/contouring + express control-loop use) + Johnson '768 (pattern recognition in HVAC) as the core, plus Sanyo '939 / Sanyo '028 for distributed multi-location sensing and triangulation, and 6,574,104 or the VAV/compressor art for the actuation-dependent claims. The weak link is not the field-map concept — that is squarely in Emhart — but the "template" comparison and whether a POSITA would have automated it, which is where Johnson '768 and Sanyo '939's cross-detector deviation threshold do the most work. Conversely, the strongest defenses are (i) the narrow "visual representation" construction of the map terms, (ii) the absence of any reference that compares two complete field maps rather than a map to a setpoint/band, and (iii) the fact that all of the identified art is already of record and the Office allowed the claims over it.

Caveats on my own confidence:

1. I verified full text for US 4,737,917, 5,506,768, 5,326,028, 6,009,939, 6,574,104 only. Mappings involving US 4,823,290, 5,074,137, 5,290,200, 5,177,972, 5,769,315, 6,296,193, 5,828,572, 6,283,380, 6,080,060, 5,372,426, and 5,682,079 are inferred from titles/assignees and should be re-verified against full texts before being asserted in any filing.
2. I did not retrieve the office actions or the examiner's reasons for allowance; the statement that the examiner never articulated the Emhart+Johnson combination is an inference from the absence of any "examiner's statement of reasons" or combination rationale in the retrieved record — treat it as unverified.
3. Litigation status is carried forward from the prior section and is not re-verified here; the 2026 dockets (E.D. Tex. 2:26-cv-00287, N.D. Ill. 1:26-cv-03926/03928/03945, D.N.J. 2:26-cv-03884/03886/03890, W.D. La. 3:26-cv-01346) are reported as previously retrieved.
4. Nothing here is a legal opinion or a freedom-to-operate or validity opinion; it is technical prior-art analysis.

Sources: https://patents.google.com/patent/US6718277/en · https://patents.google.com/patent/[US4737917A](/patent/US4737917A)/en · https://uspto.report/patent/grant/[4737917](/patent/4737917) · https://patents.google.com/patent/US5506768 · https://uspto.report/patent/grant/5506768 · http://www.everypatent.com/comp/pat5506768.html · https://patents.google.com/patent/[US5326028A](/patent/US5326028A) · https://patents.google.com/patent/[US6009939A](/patent/US6009939A) · http://www.everypatent.com/comp/pat6009939.html · https://patents.google.com/patent/[US6574104B2](/patent/US6574104B2)/en · https://webapp1.dlib.indiana.edu/virtual_disk_library/index.cgi/[5705037](/patent/5705037)/FID2772/OG/html/1271-1/us06574104-20030603.html · https://www.freepatentsonline.com/[8554515](/patent/8554515).html (US 6,574,104 citation context) · https://patents.justia.com/patent/[6886353](/patent/6886353) (co-pending HP family cross-references)