US 7555357

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

Let me check for CAFC 2026 appellate activity and verify claim language.

Let me verify whether a CAFC appeal specifically naming the '357 patent exists.

US Patent 7,555,357 — Analyst Summary (as of April 26, 2026)

Bibliographic data (from the patent record and USPTO/WIPO family data)

Abstract (verbatim from the patent)

"A method of forming a three-dimensional object using an extrusion-based layered deposition system, the method comprising generating a build path for building a layer of the three-dimensional object, where the build path defines a void region. The method further comprising generating at least one intermediate path in the void region, and generating a remnant path based at least in part on the at least one intermediate path."

Technical thrust

The patent addresses porosity caused by small unfilled "void regions" in FDM-type build data. Conventional build paths are generated on a coarse "first road width resolution" (e.g., ~250–1,020 µm / 10–40 mils), so regions narrower than that resolution are ignored and become cavities. The disclosed method identifies those void regions (step 28 of method 22), generates a higher-resolution intermediate path such as a raster path within the void (FIG. 3, element 50; method 52), computes void widths at raster-leg center points, labels vertices when width change exceeds a threshold (Δ), and then generates a "remnant path" whose deposition rate — and thus road width — varies with the measured void dimensions. A second embodiment (method 214, FIGS. 7–11) switches between two orthogonal raster paths at branch/curvature locations using a length threshold, enabling filling of branched or curved void regions (FIG. 12).

Independent claims — plain language

Claim 1 — A method of forming a 3D object with an extrusion-based layered deposition system, comprising:

1. generating a build path for a layer, where the build path defines a void region;
2. generating at least one intermediate path in the void region;
3. determining dimensions of the void region based at least in part on the at least one generated intermediate path; and
4. generating a remnant path in the void region based at least in part on the intermediate path, wherein the remnant path comprises deposition rates that are configured to vary based on the determined dimensions of the void region.

Claim 8 — A method comprising:

1. generating a build path for a layer, the build path being based on a first road width resolution;
2. identifying a void region in the build path;
3. determining dimensions of the void region; and
4. generating a remnant path in the void region, wherein the remnant path comprises deposition rates configured to vary based on the determined void dimensions.

Claim 15 — A method comprising:

1. generating a build path for a layer with a plurality of first deposition roads based on a first road width resolution, wherein the build path defines a void region having dimensions smaller than the first road width resolution along at least one axis; and
2. generating a remnant path in the void region for filling at least part of a cavity corresponding to the void region with a second deposition road, based on deposition rates that vary based on the void region's dimensions.

(Note: one public claim transcription renders claim 15 as "at least pan of a cavity"; the patent PDF reads "at least part of a cavity." I flag this as a transcription artifact, not a claim-text discrepancy.)

Interpretive caveat: the abstract/summary in the specification describes a broader two-step concept (build path → intermediate path → remnant path). The issued independent claims are narrower, adding "determining dimensions of the void region" and "deposition rates that are configured to vary based on the determined dimensions." That narrowing language appears to have been added during prosecution, consistent with the examiner's citation of Jamalabad et al., U.S. Pat. No. 6,823,230 in the PCT search report (WO 2007/089576 A3, category "X" against claims 1–20).

Litigation posture — and the 2026 question you asked about

District court (this is where the '357 patent has actually been adjudicated):

• Stratasys, Inc. v. Shenzhen Tuozhu Technology Co. Ltd. et al., No. 2:24-cv-00644 (E.D. Tex., Marshall Division, Chief Judge Rodney Gilstrap). Google Patents also lists W.D. Tex. 1:24-cv-01511 and E.D. Tex. 2:25-cv-00465.
• The '357 patent is one of ten patents-in-suit asserted against Bambu Lab. A claim construction order (Dkt. 169) construed "determining dimensions of the void region"/"determining the dimensions of the void region" and "void region" to have their plain meaning, rejecting Bambu's indefiniteness arguments.
• Bambu filed a § 101 summary-judgment motion against the '357 patent (Dkt. 174, April 2026); the docket shows a sealed memorandum opinion and order dated July 1–2, 2026 addressing it and other sealed motions. I could not retrieve the substance of that ruling.
• On September 17, 2026, an E.D. Tex. jury found all four patents presented at trial — US 9,421,713, US 7,555,357, US 9,168,698, and US 10,556,381 — valid and infringed, awarding approximately $27.6 million in past damages, and found willfulness. US 9,592,660 was not presented (PTAB had invalidated its challenged claims; Stratasys did not appeal). Reports indicate the court entered judgment September 18, 2026, denied enhanced damages and a permanent injunction, and that Bambu stated it will seek post-trial review and appeal. Chinese-language coverage (IPRdaily) reports the '357 patent accounted for nearly half of the damages figure — I could not independently verify that allocation.
• The '357 patent expires (adjusted) on March 3, 2027, which is roughly six months after the verdict.

CAFC — what I can and cannot confirm:

• I found no Federal Circuit docket for 2026 that specifically names US 7,555,357. The only 2026 CAFC appeal I could identify in this dispute concerns US 9,421,713 (Bambu's appeal of the PTAB decision upholding the purge-tower patent's validity), which is a different patent from a different family.
• Bambu's public statement — "Bambu intends to avail itself of the legal processes," i.e., post-trial review and appeal — is a stated intention. As of today no notice of appeal or CAFC case number naming the '357 patent has been located. Treat any claim that a CAFC 2026 appeal over 7,555,357 is pending as unverified. An appeal, if filed, would likely follow disposition of post-trial motions and would be docketed under a 26- or 27-series case number.

Parallel proceedings (not involving the '357 patent, but relevant context for the same portfolio): the UPC Local Division The Hague dismissed Stratasys's preliminary-injunction application against Bambu's H2C printer on April 24, 2026, holding EP 2964450 (the European family member of US 9,421,713) not infringed — the court adopted Bambu's construction of "in a layer by layer manner." Stratasys did not appeal. PTAB invalidated challenged claims of US 9,592,660 and US 8,562,324; Stratasys's Director Review request on the latter was rejected.

Uncertainty flags

1. No CAFC docket for the '357 patent located — the 2026 Federal Circuit activity in this dispute concerns the '713 patent's PTAB appeal only.
2. The July 2026 sealed summary-judgment ruling on the '357 patent's § 101 validity is on the docket but I have not seen its contents.
3. Per-patent damages allocation of the $27.6M is not in the sources I reviewed; the "nearly half from the '357 patent" figure comes from secondary reporting.
4. Damages/validity figures above are from a jury verdict, not a final, non-appealable judgment.

I'll research litigation involving US 7555357 using the sources you specified. Let me run several searches.

Let me dig into the specific case numbers and the litigation portal.

Summary

US 7,555,357 ("Method for building three-dimensional objects with extrusion-based layered deposition systems," inventor Donald J. Holzwarth, assignee Stratasys, Inc.) is being asserted in the ongoing Stratasys v. Bambu Lab patent campaign. All known litigation traces to that single dispute, which spans three district court dockets (two in the Eastern District of Texas and one transferred out of the Western District of Texas) that have since been consolidated. I found no other litigation — past or present — involving this patent.

I did not auto-correct any identifiers below. Case numbers and patent numbers are reproduced exactly as they appear in the sources.

Litigation involving US 7,555,357

1. Stratasys, Inc. v. Shenzhen Tuozhu Technology Co., Ltd. et al.

Sources: Complaint, Dkt. 1; AO‑120 notice.

2. Stratasys, Inc. v. Shenzhen Tuozhu Technology Co., Ltd. et al. (second suit)

Source: PTAB mandatory notice in IPR2025-00257 referencing 2:24-cv-00645.

3. BambuLab USA, Inc. et al. v. Stratasys, Inc. (declaratory judgment action)

Sources: Docket (Justia); Case Readiness Status Report (Dkt. 34); Ex Parte complaint analysis.

4. BambuLab USA, Inc. v. Stratasys, Inc. (transferred action)

Sources: Case Readiness Status Report, 2:25-cv-00465; Order (Dkt. 54), Feb. 24, 2026.

Consolidated action: outcome / current status

The three dockets now proceed as a single consolidated case (lead: 2:24-cv-00644-JRG; member: 2:25-cv-00465-JRG). Key developments touching the '357 patent:

• Claim construction: On February 18, 2026, Judge Gilstrap issued a Claim Construction Order (Dkt. 169) in 2:24-cv-00644-JRG, listing US 7,555,357 among the ten patents-in-suit. (CourtListener)
• Jury verdict: Following a week-long trial in the E.D. Tex. (Marshall Division), a jury returned a verdict for Stratasys, finding four patents valid and infringed — US 9,421,713; US 7,555,357; US 9,168,698; and US 10,556,381 — and awarding approximately $27.6 million in past damages, with a finding of willful infringement. (3DPrintingIndustry; TCT Magazine; 3DPrint.com; McDermott press release)
• '357 § 101 challenge: During summary judgment (Dkt. 401, signed by Judge Gilstrap on July 1, 2026), the court addressed "Defendants' Motion for Summary Judgment of Invalidity of U.S. Patent No. 7,555,357 Under 35 U.S.C. § 101." The court reportedly held the claims are directed to an abstract idea (Alice step 1) but found an open question of fact as to whether the claims add an inventive concept (Alice step 2). (Justia Dkt. 401; Google Patents record)
• Post-verdict posture: Bambu Lab has disputed the verdict (the "first of two" cases). Per reporting, a separate trial on the remaining asserted patent has not yet been scheduled, and Stratasys did not appeal as to that fifth patent.

Related PTAB proceedings (not district court litigation, but relevant)

Shenzhen Tuozhu Technology Co., Ltd. filed a series of IPR petitions against the Stratasys patents-in-suit in late 2024 / early 2025, including IPR2025-00257, IPR2025-00311, IPR2025-00321, IPR2025-00354, IPR2025-00438, IPR2025-00531, IPR2025-00532, IPR2025-00585, and IPR2025-00611. I could not confirm from the retrieved material which (if any) of these specifically challenges US 7,555,357 — I am flagging this rather than guessing. (Sources: 2:25-cv-00465 status report; PTAB correspondence.)

Notes, caveats, and non-matches

• Do not confuse with earlier Stratasys suits. The 2013 Stratasys, Inc. v. Microboards Technology, LLC d/b/a Afinia action (D. Minn., Case No. 13-cv-03228-DWF-JJG) asserted different patents ('925, '058, '124, '239) — not the '357 patent. It is not relevant here.
• Excluded litigation matched by number, not patent. Several search hits involved other patents (e.g., Mobile Data Technologies v. Samsung, Realtime Adaptive Streaming) and are not related to US 7,555,357.
• Date discrepancy: The searched web material contains events dated up to late September 2026, while the current date given in the prompt is April 26, 2026. I have reported the items as they appear in the sources rather than reconciling the dates.
• Google Patents litigation links (from the patent page) — W.D. Tex. case 1:24-cv-01511, E.D. Tex. case 2:25-cv-00465, and E.D. Tex. case 2:24-cv-00644 — are all accounted for above and are consistent with the district court records.
• Confidence: High on the three/four dockets, parties, case numbers, and the '357's role. Moderate on the exact procedural details of the § 101 ruling and on whether an IPR specifically targets the '357; those should be verified against PACER and the PTAB docket before being relied upon.

I'll research this systematically. Let me start with searches on PTAB proceedings for this patent.

Let me dig deeper into any IPR petitions specifically challenging the '357 patent.

Let me verify whether any petition was ever filed against the '357 patent specifically.

Proceedings overview

Total AIA trial proceedings on US 7,555,357 (the '357 patent): 0. The structured USPTO ODP block provided in this prompt returns no IPR, PGR, or CBM proceeding for this patent, and my web search surfaced no petition, institution decision, or FWD naming the '357 patent as the challenged patent. The breakdown is therefore 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution denials — nothing to break down.

Bottom line for a defendant: the '357 patent has never been through an AIA trial, so there is no PTAB invalidation to lean on and no § 315(e)(2) estoppel constraining anyone — but that cuts both ways. Stratasys took the '357 patent to a jury in September 2026 and won a validity-and-infringement verdict on it. You are not facing a hardened, IPR-tested claim set; you are facing a 2006-priority software/method patent whose only validity test to date has been a district-court § 101 fight and a jury verdict, with roughly six months of term left (adjusted expiration 2027-03-03).

No proceeding entries to report

There are no ### {PROCEEDING_NUMBER} — {Petitioner} v. {Patent Owner} entries because there are no proceedings on this patent. To be transparent about my method: I could not independently query PTAB E2E or the ODP API in this session; I am relying (a) on the canonical structured block in your prompt, which reports zero AIA trials, and (b) on targeted web searches of the PTAB public petition/decision corpus. Zero indexed proceedings is consistent with the fact that a denied-institution petition would still appear as a terminated proceeding in ODP — none does.

Related PTAB activity on sibling patents (NOT the '357 patent)

This is context, not a proceeding on the '357 patent. In the 2024–2026 Stratasys v. Bambu Lab campaign, petitioner Shenzhen Tuozhu Technology Co., Ltd. filed IPR petitions against numerous Stratasys patents — its own briefing describes "IPR petitions challenging nine different patents spanning six patent families." The '357 patent is not among those identified as challenged. Confirmed sibling proceedings include:

I am deliberately not assigning proceeding numbers to the '660 and '324 matters — I did not confirm those numbers from a primary source, and you instructed me not to invent them.

Strategic summary

Claim status: the '357 patent's claims are neither canceled nor confirmed by any PTAB decision — they are entirely UNTESTED at the PTAB, and TESTED-and-SURVIVING in district court. Per the PTAB docket there is no FWD on this patent, so I will not claim any claim number was canceled or sustained. What has happened: (1) Judge Gilstrap denied Bambu's summary-judgment motion that the '357 claims are invalid under 35 U.S.C. § 101, finding at Alice step one that the claims are not directed to an abstract idea (ruling in the 2026-07-01 pretrial order, Dkt. No. 401 in E.D. Tex. 2:24-cv-00644-JRG); and (2) on 2026-09-17 an E.D. Tex. Marshall jury found the '357 patent valid and infringed, in a verdict awarding Stratasys ≈$27.6M in past damages across four patents (the '357, '713, '698, and '381 patents). Press reporting (ifeng, 2026-09-20) states the '357 patent accounted for close to half the damages figure — treat that apportionment as a press characterization, not a court finding. The verdict is not final; post-trial motions and a Bambu appeal are expected.

Estoppel landscape: there is no IPR estoppel on the '357 patent at all. Because no IPR was instituted on this patent, no petitioner — Bamu included — is barred under § 315(e)(2) from raising § 102/§ 103 grounds against the '357 claims. Note that Bambu's pre-institution Tesla-type stipulation in the sibling IPRs (e.g., IPR2025-00321) expressly covers only the patent at issue in that proceeding; it does not sweep in the '357 patent. Practical consequence: a new defendant accused of infringing the '357 patent has a completely open prior-art runway — every reference remains available, and you are not the party bound by Bambu's contentions. Two caveats: (i) you must still meet your district court's invalidity-contentions deadline and Local Patent Rule requirements; (ii) a jury's validity finding in Stratasys v. Bambu is not binding on you as a non-party, but Stratasys will cite it, and the Bambu record (including the § 101 expert fight and any IPR-art overflow) is now discoverable ammunition.

Pattern signals. The same petitioner (Shenzhen Tuozhu / Bambu Lab) filed many IPRs against this portfolio but skipped the '357 patent — a notable omission that likely reflects the patent's 2006 priority date, its expiration on 2027-03-03 (per the structured data), and the fact that its method claims are hard to kill with patent/printed-publication art (the validity attack that actually reached trial was § 101, not § 103). There is no defensive aggregator (e.g., Unified Patents) in the chain for this patent — the litigation-docket links in your structured block are district-court cases (Texas W.D. 1:24-cv-01511; Texas E.D. 2:25-cv-00465 and 2:24-cv-00644), not aggregator-filed IPRs. Historical contrast worth knowing: Stratasys's earlier Stratasys v. Afinia campaign (IPR2015-00284, -00287, -00288, on the '058, '124, and '239 patents) produced three institution denials — Stratasys has a track record of defeating IPR petitions, but again on other patents and other claim sets.

Recommended next steps

1. Do not assume invalidation exists. There is no FWD to cite and no canceled claim to point to. If you are drafting a demand-letter response premised on "this patent has been invalidated," stop — that premise is false. The correct framing is: untested at the PTAB, but a Texas jury just upheld it and Judge Gilstrap rejected the § 101 attack.

2. Exploit the clean-slate IPR position if you want to file. Because no IPR has been instituted on the '357 patent, a fresh petition is not foreclosed by any estoppel and is not a "follow-on" petition, so the § 325(d) / General Plastic discretionary-denial risk is lower than for a serial petitioner — but Fintiv-style § 314(a) discretion is the live threat, since Stratasys will argue the E.D. Tex. case is close to resolution. If you file, consider the Tesla-type stipulation (as Acting Director-endorsed in IPR2025-00217) to neutralize Fintiv. Statutory mechanics: institution decision due within 6 months of the petition's filing date (§ 314(b)), and the FWD within 1 year of institution (§ 316(a)(11)) — about 18 months total.

3. Calendar the patent's death. The structured ODP block lists an adjusted expiration of 2027-03-03. With that little term left, the economic exposure is backward-looking (past sales) rather than forward-looking, and it sharply reduces the value of an injunction. That also means an IPR filed now may not mature before expiration — weigh whether a district-court § 101/§ 112 defense (cheaper, faster) is the better vehicle than an IPR.

4. Mine the Bambu record. Request or obtain via PACER: Dkt. No. 401 (the 2026-07-01/02 pretrial order ruling on the § 101 motions and MILs — govinfo copy and the CourtListener docket), the final judgment, and the claim-construction order (Dkt. No. 169). The construction order's treatment of "determining dimensions of the void region" (cls. 1, 8) / "determining the dimensions of the void region" (cls. 4, 12) is where a non-infringement or § 112 indefiniteness theory likely lives.

5. Track the parallel appeals. Bambu has appealed the PTAB's upholding of the '713 patent to the CAFC, and the '660 invalidation is final. None of those dispositions touches the '357 patent, but if the CAFC narrows or reverses on the '713 (same patent family as EP2964450), it will not help you on the '357 — they are separate families. Do not conflate them. Monitor the CAFC docket and CourtListener for the Stratasys v. Bambu appeal once the E.D. Tex. judgment is entered.

Confidence note: I am highly confident that no AIA trial proceeding exists on the '357 patent — the ODP block is unambiguous and my searches corroborate it. I am less certain about the precise proceeding numbers mapping to the '660 and '324 IPRs, and I have flagged those gaps rather than guess.

I'll research the assignment record and ownership chain for US 7,555,357.

Let me dig into the assignment record specifics and Stratasys corporate history.

I'll research the prior-art citations for US 7,555,357. Let me pull the patent's front-page references and related family documents.

Let me pull the front-page "References Cited" list and the PCT search report that examined this family.

Let me pull the specific IDs, dates, and content of the key cited references, plus the PCT search report categories.

Prior-Art Analysis — US 7,555,357 B2 (Holzwarth / Stratasys)

Method: Everything below is grounded in the reference list retrieved for US‑7555357‑B2 (PubChem patent record), the WO 2007/089576 A3 international search report (PDF), and the reference's own specification text (supplied in the user message, authoritative). I did not auto‑correct any patent number.

Date note (resolves a flag in the earlier sections): the previously generated Litigation Summary flagged a discrepancy between "April 26, 2026" and search material dated to late September 2026. Today's date is September 29, 2026, which reconciles the record — the Sept. 17, 2026 E.D. Tex. verdict is contemporaneous, not anomalous.

1. What the citation set actually is

US 7,555,357's citation list (as published on the patent record) contains ~49 U.S. patent documents plus four Wohlers non‑patent items. The PubChem export tags each entry either (APP) or (SEA). Only three entries carry the (SEA) tag:

The remaining 46 U.S. documents carry (APP). I read this as an applicant‑vs‑examiner source distinction (i.e., Jamalabad was the examiner's/hired‑searcher's reference), which is corroborated independently: the PCT search report for this exact family cites US 6,823,230 B1 (Jamalabad et al.), 23 November 2004, "whole document," category X, against claims 1–20.

This is the single most important fact in the whole analysis: the only reference the international examiner used to reject every pending claim was Jamalabad '230. It is also the reference the applicant discusses by name in the '357 specification (FIG. 14 "bat ear" protrusions):

"…the reduction of void regions in this manner is generally discussed in Jamalabad et al., U.S. Pat. No. 6,823,230 ('the '230 patent'). … This is undesirable because the large volumes of build material deposited at the protrusions tend to force the previously deposited perimeter roads … to bulge out…"

2. Independent‑claim element map (the § 102 yardstick)

Any § 102 anticipation must meet every element. The issued independent claims require:

Claim 1: (a) build path defining a void region → (b) at least one intermediate path generated in the void region → (c) determining dimensions of the void region based at least in part on that intermediate path → (d) remnant path whose deposition rates are configured to vary based on the determined dimensions.

Claim 8: build path based on a first road width resolution → identify void region → determine dimensions → remnant path with deposition rates varying on determined void dimensions.

Claim 15: build path of first deposition roads at first road width resolution, defining a void region smaller than that resolution along at least one axis → remnant path filling at least part of the cavity with a second deposition road based on deposition rates that vary based on the void region's dimensions.

Elements (b)+(c)+(d) are the differentiators. Nearly every pre‑2006 FDM reference below teaches (a) and sometimes (d), but essentially none teaches using a separate higher‑resolution intermediate path to measure the void and then deriving a variable deposition rate from that measurement.

3. Tier 1 — references capable of supporting a § 102 or lead § 103 challenge

3.1 US 6,823,230 B1 — Jamalabad, V. R., et al. (Stratasys)

• Full citation: US 6,823,230 B1, "Tool path planning process for component by layered manufacture," issued Nov. 23, 2004 (examiner‑cited; SEA).
• Disclosure (per the reference text): improves tool‑path generation for layered manufacture; positions raster tool‑path vertices to eliminate voids; automatically relocates outer‑boundary vertices outward toward the design perimeter; extends inner boundary vertices toward enclosing vertices "thereby eliminating some internal voids"; and expressly addresses "narrow regions between combinations of outer and/or inner perimeters," and "perimeter gaps or sub‑perimeter voids where raster tool paths meet perimeters or contour beads."
• § 102 analysis:
  • Against original/pending claims 1–20 (as examined in PCT): the international search report says X, whole document. On that record, Jamalabad anticipates the as‑filed claim set.
  • Against issued claims 1, 8, 15: likely insufficient for pure anticipation. Jamalabad modifies/extends vertices and boundaries but the retrieved disclosure does not show (i) a second, higher‑resolution intermediate path generated inside the void and used to (ii) determine the void's dimensions, with (d) a deposition rate that varies as a function of those determined dimensions. It computes a boundary offset from a bead width, but that is the coarse first‑resolution geometry, not a measured void profile.
  • § 103: strongest primary reference; the natural combination is Jamalabad + a variable‑extrusion‑rate teaching (see 3.5, 3.6).
• Why this matters practically: the narrowing that produced the issued claims (adding "determining dimensions of the void region" / "deposition rates … configured to vary") is precisely the language that distinguishes over Jamalabad. This is consistent with the interpretive caveat recorded in the earlier Patent Summary section.

3.2 US 5,121,329 — Crump, S. S.

• Full citation: US 5,121,329, "Apparatus and method for creating three‑dimensional objects," issued June 9, 1992 (filed Oct. 30, 1989); Stratasys.
• Disclosure: the foundational fused‑deposition "road‑by‑road, layer‑by‑layer" patent — extrusion head, x‑y raster, z‑increment, slicer→tool‑path build data.
• § 102: No. Discloses element (a) at most, using a constant road width. No intermediate path, no void measurement, no variable deposition rate tied to void dimensions. Anticipates nothing in claims 1/8/15; background for § 103.

3.3 US 5,340,433 — Crump, S. S. (continuation family)

• Full citation: US 5,340,433, "Modeling apparatus for three‑dimensional objects," issued Aug. 23, 1994; Stratasys.
• § 102: No — apparatus/control disclosure; fixed road geometry. Same conclusion.

3.4 US 5,503,785 — Crump, S. S., et al.

• Full citation: US 5,503,785, "Modeling apparatus for three‑dimensional objects," issued Apr. 2, 1996; Stratasys.
• § 102: No — same reasoning.

3.5 US 5,402,351 / US 5,764,521 — Batchelder, J. S. (Stratasys control patents)

(Titles/dates: US 5,402,351, "Modeling method…", issued Mar. 28, 1995; US 5,764,521, "Method and apparatus for solid prototyping," issued June 9, 1998 — treat exact titles as not independently verified in this session.)

• Disclosure: Stratasys control architecture for computing and executing deposition tool paths and material feed. Batchelder‑era patents are the natural place to look for extrusion rate/velocity control.
• § 102: No as standalone. § 103: relevant secondary — these supply the "deposition rate is a controllable parameter" teaching that, combined with Jamalabad, frames the obviousness case.

3.6 US 6,067,480 — Stuffle / Stratasys (variable extrusion control)

• Full citation: US 6,067,480 — Stratasys‑era patent cited by the applicant (APP). Exact title and issue date not verified in this session; do not quote a title without checking the front page.
• § 102: No as standalone; § 103 secondary on the variable deposition‑rate element (d).

3.7 US 6,261,077 B1 — Ensz, M. T., et al.

• Full citation: US 6,261,077 B1 (APP). Title/date not verified in this session. (Do not confuse with US 6,261,077 vs. the similarly numbered 6,241,945/6,211,136‑family patents I did not retrieve.)
• § 102: No on the retrieved record; listed for completeness.

3.8 US 6,129,872 and US 6,165,406 — Jang, et al.

*(Jang‑family extrusion/apparatus patents; exact titles/dates not verified this session.)*

• § 102: No as issued‑claim anticipation. Note: the Afinia IPR expert declaration against Stratasys' US 8,349,239 relied on "Jang, et al." as a § 103 primary reference in combination with Kulkarni et al. and/or Kao (see § 5) — i.e., Jang is treated in the art as geometry/path‑pattern art, not void‑measurement art.

3.9 US 5,587,913 — Cline / sequential 2D geometry for shells

• Full citation: US 5,587,913, "Method employing sequential two‑dimensional geometry for producing shells for fabrication by a three‑dimensional solid freeforming apparatus," issued Dec. 24, 1996 (APP).
• Disclosure: generating 2‑D shell geometry before freeforming.
• § 102: No — teaches shell/contour generation, not void profiling or variable deposition rate.

3.10 US 5,594,652 — Penn / computer‑controlled manufacture from computer data

• Full citation: US 5,594,652, issued Jan. 14, 1997 (APP).
• § 102: No — build‑data/apparatus background.

3.11 US 5,738,817 and US 5,900,207 — Danforth, S. C., et al.

• Full citation: Danforth et al., "Solid freeform fabrication methods," issued Apr. 14, 1998 and May 4, 1999 respectively (APP).
• § 102: No — materials/process background.

3.12 US 6,823,230's siblings in the same family (Stratasys path‑generation line)

For completeness, the applicant also cited US 6,004,124, US 6,054,077, US 5,968,561, US 6,022,207, US 5,866,058, US 6,085,957, US 6,132,355, US 6,197,393, and US 6,494,096/6,722,872/6,813,594/6,814,907/6,938,212/6,998,087 (all APP). Titles/dates for this block were not verified in this session. None was tagged (SEA), i.e., the examiner did not rely on them.

4. Tier 2 — the full applicant‑cited U.S. list (completeness requirement)

These are the remaining U.S. documents appearing in the '357 citation list. I list them verbatim as retrieved; I am not vouching for issue dates or titles for those I could not verify this session, and none is tagged (SEA):

US 4,665,492 · US 4,749,347 · US 4,961,154 · US 5,059,266 · US 5,121,329 · US 5,134,569 · US 5,136,515 · US 5,140,937 · US 5,149,548 · US 5,204,055 · US 5,216,616 · US 5,303,141 · US 5,340,433 · US 5,402,351 · US 5,474,719 · US 5,503,785 · US 5,587,913 · US 5,594,652 · US 5,695,707 · US 5,738,817 · US 5,764,521 · US 5,807,437 · US 5,866,058 · US 5,900,207 · US 5,939,008 · US 5,943,235 · US 5,968,561 · US 6,004,124 · US 6,022,207 · US 6,054,077 · US 6,067,480 · US 6,085,957 · US 6,129,872 · US 6,133,355 · US 6,162,378 · US 6,165,406 · US 6,175,422 B1 · US 6,193,923 B1 · US 6,214,279 B1 · US 6,261,077 B1 · US 6,490,496 B1 · US 6,606,528 B1 (SEA) · US 6,722,872 B1 · US 6,813,594 B2 · US 6,814,907 B1 · US 6,823,230 B1 (SEA) · US 6,859,681 B1 (SEA) · US 6,898,477 B2 · US 6,936,212 B1 · US 6,998,087 B1

Group‑level § 102 conclusion for Tier 2: these are stereolithography (4,961,154 Pomerantz; 5,059,266; 5,695,707), 3‑D‑printing/binder‑jet (5,204,055; 5,807,437), ballistic/particle deposition (4,665,492; 5,134,569; 5,216,616), fused deposition and its Stratasys control line (5,121,329 et al.), and liquefier/filament‑handling apparatus patents. None discloses element (b)–(d) in combination. They are § 103 background only.

5. Non‑patent literature cited on the '357

Note on § 102(b) status: the 1991 Wohlers items are printed publications more than one year before the Jan. 31, 2006 filing date, so they qualify as § 102(b) art — but they are industry surveys and do not disclose the void‑measurement/variable‑deposition‑rate method. They are useful only for establishing the state of the art / level of ordinary skill.

6. Art not cited in the '357 that is arguably more pertinent than most of Tier 2

These surfaced in the parallel Stratasys v. Afinia IPR record (declaration of T. A. Campbell, IPR against US 8,349,239, PTAB petition):

• Kao, J., "Process Planning for Additive/Subtractive Solid Freeform Fabrication Using Medial Axis Transform," Stanford Univ. dissertation, June 1999 — used as a § 102(b) anticipatory reference against Stratasys path‑planning claims.
• *Kulkarni, P., et al., "Deposition Strategies and Resulting Part Stiffness in Fused Deposition Modeling," J. Mfg. Sci. & Eng., Feb. 1999, vol. 121, pp. 93–103* — deposition‑strategy work; the closest known non‑patent art to variable road/void‑filling strategy.
• Ruan, J., et al., "2‑D Deposition Pattern and Strategy Study on Rapid Manufacturing," ASME DETC2006‑99326 (2006) — dated 2006; post‑dates the '357 priority date (Jan. 31, 2006) and would not qualify as § 102 prior art. Do not cite it as '357 prior art.
• Alexander (Exh. 1008) in that IPR; possibly corresponding to one of the Stratasys‑cited US 6,8xx,xxx documents — not verified.

If you are building an invalidity position, Kao (1999) and Kulkarni (1999) are the references that most plausibly combine with Jamalabad '230 to reach the "determine void dimensions → vary deposition rate" limitation. I flag them as candidate art, not as § 102 anticipators, because I have not read their full texts.

7. Bottom line

No single cited reference anticipates issued claims 1, 8, or 15. The closest is Jamalabad '230, and its failure is precisely on the two elements the applicant appears to have added during prosecution — using a generated intermediate path to determine the void's dimensions, and configuring deposition rates to vary as a function of those dimensions.

8. Uncertainty flags (please verify before relying)

1. I could not open the USPTO PatentCenter/PAIR front page for 7,555,357 in this session, so I could not confirm which references the examiner applied versus merely listed, nor the exact (56) sub‑headers. The (APP)/(SEA) split above is inferred from the PubChem export and corroborated only for Jamalabad by the PCT search report.
2. Issue dates and titles for the Tier‑2 block (esp. US 5,402,351, 5,764,521, 6,067,480, 6,261,077, 6,129,872, 6,165,406, 6,606,528, 6,859,681, 6,898,477) are NOT verified here. I deliberately did not invent titles/dates for them. Check each on its own front page before quoting.
3. § 102 vs. § 103 framing depends on which claim version you are attacking. The PCT rejection was against the original claims; the analysis above addresses the issued claims.
4. The Kao/Kulkarni/Ruan/Alexander references come from a different Stratasys patent's IPR, not from the '357 or its family. Treat them as research leads.
5. No contradiction detected with the earlier Patent Summary / Litigation Summary / Assignment sections, other than the now‑resolved date discrepancy (item 4 of the earlier Litigation Summary caveats).

I'll research the prior art landscape relevant to this patent before building the § 103 analysis. Note: the fetched patent text provided does not itself contain a "Prior Art" listing (Google Patents' citations section wasn't captured), so I'll ground the analysis in the references that are identified in the patent text (the Jamalabad '230 patent) plus the admitted background art, and supplement with search.

Good — I now have the key reference (Jamalabad '230) and the prosecution context. Let me pull a few more specifics on the cited art and any validity challenges to the '357.

Let me pull the full claim set and verify the primary reference's disclosure, plus check the '357 prosecution history.

Obviousness Analysis — US 7,555,357 B2 under 35 U.S.C. § 103

Analyst note on the "Prior Art section of this page": the Google Patents page as fetched for this analysis (https://patents.google.com/patent/US7555357B2/en) did not render its Citations / Similar Documents panel — only the front-page data, the "Prior art keywords" tag list (path; raster; void; void region; leg), the specification, and the abstract. I therefore used the parallel PubChem record for the same patent (https://pubchem.ncbi.nlm.nih.gov/patent/US-7555357-B2), which reproduces the front-page "References Cited" list (≈51 U.S. patent references plus non-patent literature), and I supplemented with the references that the '357 specification itself discusses. Where I could not verify a reference's actual disclosure, I say so rather than asserting it.

1. Legal frame

• Effective filing date: January 31, 2006 (no earlier domestic priority; the whole family claims priority to US 11/343,355). Pre-AIA § 103 applies; § 102(a)/(b)/(e) govern which art qualifies. Nothing in the file suggests a pre-AIA § 103(c) common-ownership proviso is outcome-determinative, though several cited references are Stratasys's own.
• The claims are method claims directed to computer-implemented generation of build data — a "how do I compute the tool path" problem. That matters: § 103 here turns on what the tool-path-planning art already taught, not on machine hardware.
• POSITA: a person with a B.S. in mechanical, computer, or electrical engineering (or equivalent) and ~2 years' experience in CAD/CAM or additive-manufacturing tool-path generation, or an M.S. with 0–1 years; conversant with FDM build-data generation (slicing, contour/perimeter paths, raster/infill paths, G-code), and with the extrusion-control relationship between material feed rate, head speed, and deposited road width/height. The '357 specification treats that relationship (road width ≈ 1.2 × road height; width scales with feed rate) as background knowledge, which supports this level of skill.

2. The claimed subject matter, distilled

Per the previously generated claim analysis (taken as authoritative here):

• Claim 1: build path defining a void region → generate ≥1 intermediate path in the void region → determine dimensions of the void region based on that intermediate path → generate a remnant path in the void region whose deposition rates are configured to vary based on the determined dimensions.
• Claim 8: same core, but the build path is expressly based on a first road width resolution.
• Claim 15: build path of first deposition roads at a first road width resolution, the void region having dimensions smaller than the first road width resolution along at least one axis; remnant path fills "at least part of a cavity" via a second deposition road at deposition rates that vary based on the void dimensions.
• Dependents add: higher-resolution intermediate path (cl. 2); raster-type intermediate path (cl. 3); raster-leg center points, per-leg void widths, vertex labeling and a Δ threshold for deposition-rate changes (cl. 4); plus further dependents I could not retrieve in full text (claims 5–7, 9–14, 16–19).

The inventive kernel is narrow and identifiable: (i) use of a computational intermediate path (finer than the deposition resolution) to measure a sub-resolution void, and (ii) depositing a variable-rate / variable-width remnant road keyed to those measurements.

3. The prior art

3.1 Primary reference — Jamalabad et al., US 6,823,230 B1 ("the '230 patent")

• Honeywell International; filed Sept. 7, 2000; granted Nov. 23, 2004 → § 102(b) art.
• It is cited on the face of the '357 (marked "SEA" in the PubChem citation list) and discussed in the '357 specification at FIG. 14 — i.e., applicant-admitted prior art. (The previously generated summary also states the '230 was cited as category "X" against claims 1–20 in the PCT search report for WO 2007/089576 A3; I could not independently verify that categorization, but I treat the earlier section as authoritative.)
• Disclosure (from the abstract and body, https://patents.google.com/patent/US6823230):
  • "the tool path should not come closer than about half a bead width to an inner perimeter defining an interior feature… can cause interior solid features of less than one bead width to be under-defined. In one example, two holes separated at their perimeters by less than a bead width will be formed as a single oblong hole…"
  • "the inside surfaces of the contour tool path beads may come closer together than one bead width. A void will result at this location, as no tool path following along the contour tool path can enter this narrow region…"
  • "Boundary vertices within parts can be improved by extending more internal boundary vertices outward toward enclosing vertices, thereby eliminating some internal voids."
  • "Narrow regions between combinations of outer and/or inner perimeters can be filled through improved tool paths. Layer regions near inner voids can receive consistent filling through an improved raster tool path method."
  • "Sub-perimeter voids between raster bead pair turn-arounds and contour beads can be eliminated by automatically adjusting the raster tool path vertices toward the adjacent raster bead pair and/or contour bead."

What '230 squarely supplies: identification of void regions narrower than the bead width; determination of their local dimensions in bead-width units; and generation of modified tool-path geometry (raster vertices extended/adjusted) to eliminate those voids.

What '230 does not squarely supply: an intermediate, finer-resolution path used purely as a measuring construct, and a variable deposition rate / variable road width on the fill road. The '357 specification itself characterizes the '230 as "depositing build material along the protrusions at the first road width resolution" — a characterization that, taken at face value, means '230's void cure is geometry-based at a single resolution, not rate-based. That is where the combination is needed.

Caveat: the '230 is a § 102(b) reference that also anticipates-or-renders-obvious much of claims 1/8/15's elements (a)–(c). Its deficit is limited to element (d)'s "deposition rates… configured to vary."

3.2 Secondary reference — Crump, US 5,121,329

• Filed 1989, issued June 9, 1992; the foundational Stratasys FDM patent, cited on the '357 face and referenced in the IPR art on sibling Stratasys patents.
• Disclosure: methods of moving a dispensing nozzle to produce curved trajectories and of "dispensing materials of variable thickness by changing material feed rate." (So characterized in the sibling-art record at https://patents.google.com/patent/US20240359401A1/en, which quotes Crump FIGS. 10 and 12.)

What Crump supplies: the mechanism and the known relationship — feed rate → deposited road width/thickness. This is the missing "(d)" element: once a void's dimensions are known, an ordinarily skilled artisan knows how to make a road that varies in width to match, by modulating feed rate and/or head speed. The '357 specification confirms exactly this: "the deposition rate may be decreased by decreasing the extrusion rate of the build material, increasing the rate of movement of the extrusion head, or a combination thereof."

3.3 Secondary references — Kao (1999) and Kulkarni & Dutta (1999)

Both are § 102(b) printed publications and both were later used as primary art in IPRs against Stratasys's other patents (see the petition documents at https://ptacts.uspto.gov/ptacts/public-informations/petitions/1461132/download-documents). Their content there is quoted at length:

• Kao, J.-H., "Process Planning for Additive/Subtractive Solid Freeform Fabrication Using Medial Axis Transform," Stanford Univ. dissertation, June 1999:
  • "recursive offsetting in spiral paths may result in incomplete filling… voids may exist and materials can be deposited non-uniformly."
  • "moving the deposition head at different feed-rates allows different widths of deposits."
  • The adaptive-offsetting method "produce[s] connected spiral deposition paths with varying step-over distances to accommodate non-uniform 'thickness' of the region," expressly "to overcome" void/gap formation.
• *Kulkarni & Dutta, "Deposition Strategies and Resulting Part Stiffness in Fused Deposition Modeling," J. Manuf. Sci. & Eng., Feb. 1999, 121:93–103:* expressly addresses FDM and notes that raster paths "can suffer from inaccuracies in the deposition, e.g., voids from the deposition process"; teaches modifying path geometry/step-over.

What Kao/Kulkarni supply: the "intermediate path" concept in all but name — a computed medial-axis/center-line representation of a region whose local width ("feature radius") is derived from the path, and the express teaching to vary deposited width/step-over based on that local width to eliminate voids. Kao's medial-axis representation is literally "determining dimensions of the void region based at least in part on the at least one generated intermediate path."

3.4 Secondary reference — Honeywell, US 2005/0131570 A1 ("'570 publication")

• Published June 16, 2005; priority Sept. 7, 2000 (same Honeywell layered-manufacturing program as the '230; the '570 publication states it is related to co-pending application "TOOL PATH PLANNING PROCESS FOR COMPONENT BY LAYERED MANUFACTURE"). Qualifies as § 102(a)/(e) art.
• Disclosure: flow charts show "Create Rasters from Fill Set," "Store Raster Paths in Toolpath Set," "Create Rasters in Curve Boundary," and "Copy Curve to Fill Set" — i.e., computing raster fill within bounded sub-regions of a layer. (https://patents.google.com/patent/US20050131570)

What '570 supplies: the routine practice of generating a raster fill path inside a bounded region identified by the planner — supporting the "intermediate path" and "raster path" elements, and reinforcing that the Honeywell art treated narrow/small regions as something to be filled by computed rasters.

3.5 Applicant's own admitted background (the '357 specification)

The Background section is itself prior-art-by-admission and is unusually helpful to an obviousness attack:

• FDM build data uses perimeter paths and bulk raster paths at a constant "road width resolution."
• "Such void regions are typically smaller than the constant road width resolution, and therefore, are ignored during data generation. This may result in small cavities being formed between the deposited roads… thereby reducing the structural integrities and sealing properties."
• "As such, there is a need for a method of generating build data that is effective for depositing roads of build material in small void regions."

That is the problem statement, with the motivation supplied by the patentee.

4. Combinations and the motivation to combine

Ground 1 (strongest): '230 in view of Crump

Rejection: Claims 1, 8, 15 obvious over US 6,823,230 in view of US 5,121,329.

Motivation / KSR rationales:

1. Same field, same problem, same solution family — both are layered-manufacturing tool-path/ extrusion-control art; both address insufficient material in narrow regions. KSR ("familiar elements… according to known methods").
2. Express problem recognition in the primary reference — the '230 identifies the void and states the goal of "eliminating some internal voids" and "fill[ing]… narrow regions." That is an articulated reason to do more/better in those regions.
3. Known, predictable relationship — Crump teaches that road width follows feed rate; using that knob to make a fill road narrower/wider to fit the measured void is the application of a known technique to a known problem with a reasonable expectation of success.
4. Design need / market pressure — the '357's own Background states the industry need to avoid cavities (porosity, sealing). That is a recognized design incentive.

Ground 2: '230 in view of Kao (optionally with Crump)

Rejection: Claims 1–4, 8, 15 obvious over '230 in view of Kao (1999), further in view of Crump.

Kao closes the remaining gaps more precisely than Crump alone: it teaches computing a width-descriptive path (medial axis, local "feature radius") inside a region, and then "mov[ing] the deposition head at different feed-rates [to] allow[] different widths of deposits," producing "varying step-over distances to accommodate non-uniform 'thickness' of the region" expressly to avoid voids. That is a two-step compute-then-deposit-at-variable-width teaching — structurally the same insight the '357 claims. Motivation: Kao states the same problem ('"portion of regions may not be completely filled with materials; voids or gaps are often present") and the same goal ("paths with varying step-over distances allow the layer deposited… seamle[ssly]").

Ground 3: '570 in view of Kao/Kulkarni and Crump

Rejection: Claims 1, 3, 8, 15 obvious over US 2005/0131570 in view of Kao and Crump. The '570 supplies the bounded-region raster generation; Kao supplies the variable-width-against-local-thickness algorithm; Crump supplies the hardware mechanism. This ground is useful mainly as a hedge if the '230 is disqualified on some technicality (e.g., if the '357's own FIG. 14 characterization of the '230 is used against the reference).

Dependent claims

• Claim 2 ("intermediate path… higher than the road width resolution used to generate the build path"): the closest call. Kao's medial-axis/adaptive-offset computation and the general CAD/CAM practice of computing at finer resolution than deposition supports it, but a direct, express teaching of a numerical resolution limit is not something I can point to in the references I verified. This is best argued as an obvious design choice (KSR: "a finite number of identified, predictable solutions"), with the '357's own rationale undercutting any claim of unexpectedness — the '357 says depositing at the fine resolution would be too slow and nozzle-limited, which is a reason to compute finely and deposit coarsely, exactly as claimed.
• Claim 3 (intermediate path = raster path): squarely met — '230 ("improved raster tool path method"), '570 ("Create Rasters from Fill Set").
• Claim 4 (raster legs, center points, per-leg void widths, vertex labeling, Δ threshold): the center-line/center-point measurement and vertex-based variable-width construction are taught or suggested by Kao's medial-axis framework. The Δ threshold (a data-compression deadband) I found no direct teaching of; it is obvious only as routine optimization of data volume, à la KSR, absent evidence of unexpected results.
• Claims 5–7, 9–14, 16–19: I could not retrieve their text, so I will not opine on them. Any of these that recite purely conventional tool-path parameters (materials, road dimensions, path ordering) would likely fall with the independent claims; those reciting specific numerical ranges or threshold values would need a KSR optimization argument.

5. Counterarguments the patentee will press

1. § 103 "teaching away." The '357 argues that depositing directly along the fine raster would be time-consuming, would reduce layer thickness (road width ∝ road height), and is limited by nozzle size. But this teaches away only from direct deposition at the fine resolution — not from the claimed approach of using the fine path as a measuring construct and depositing a variable-rate remnant. The specification's own reasoning supplies the motivation for the claimed architecture.
2. "The '230 is at the first road width resolution." The applicant's FIG. 14 characterization, if accepted, limits the '230's disclosure — hence the need for Crump/Kao on the variable-rate element. Expect the patentee to argue this is exactly why the combination isn't "reasonably expected." The counter is that varying feed rate to vary bead width was a known, one-variable, predictable control (Crump), and Kao had already applied it to region "thickness."
3. No explicit motivation to compute rather than deposit the intermediate path. This is the strongest non-§ 101 defense. It is a "why would you compute a path you don't print?" argument, answered by Kao's express compute-then-adapt approach and by the '357's own admission that the fine raster is unsuitable for direct deposition.

6. Secondary considerations

I found no evidence of unexpected results, long-felt unmet need, or industry praise in the material retrieved that is specific to the '357's claimed subject matter. The specification frames every benefit (reduced porosity, uniform layer thickness, preserved sealing) as a predictable consequence of filling a cavity — and "'230's goal of eliminating voids and improving strength" is precisely the same articulated benefit. Commercial success, if any, would be difficult to nexus to this patent given it is one of ten Stratasys patents asserted in the same campaign.

7. Uncertainty flags and contradictions with earlier sections

1. Flag — contradiction on the § 101 ruling (does not change the § 103 analysis, but matters for litigation posture). The earlier "Patent summary" states that in the July 2026 SJ order the court "held the claims are directed to an abstract idea (Alice step 1) but found an open question of fact as to whether the claims add an inventive concept (Alice step 2)." The CourtListener summary of Dkt. 401 actually contains two paragraphs both bearing the '357 label: one says the court found the claims "not directed to an abstract idea under Alice Step 1," citing DDR Holdings/Enfish; a second (referring to "collecting, organizing, and displaying data" and testimony about "querying a target printer for configuration information") appears, from its subject matter, to concern a different patent despite the '357 caption. Either the earlier summary is based on the mismatched paragraph, or the docket index is mislabeled. Treat the '357 § 101 outcome as unverified. Source: https://www.courtlistener.com/docket/69028070/401/stratasys-inc-v-shenzhen-tuozhu-technology-co-ltd/.
2. Flag — date inconsistency. The task states the current date is April 26, 2026, but the sources returned carry July–September 2026 events (the Dkt. 401 order signed July 1, 2026; the Sept. 17, 2026 verdict). I have reported the sources as found, as the earlier sections did.
3. Unverified: the PCT search report's "X" categorization of the '230 against claims 1–20 (asserted in the earlier summary; I could not retrieve WO 2007/089576 A3's search report in this session).
4. Not verified by me, cited on the '357 face, possible further secondary art: US 5,503,785 (Crump et al.), US 5,738,817 (Danforth et al.), US 5,939,008 and US 6,054,077 (Comb et al.), US 6,814,907, US 6,839,594 (Guertin et al.), US 6,998,087, and US 2003/0236588 A1 (Jang et al., cited by an examiner in the sibling '239 patent). I did not retrieve their disclosures and therefore do not rely on them. I would not assert them as § 103 grounds without reading them.
5. No IPR specifically against the '357 was confirmed. The Bambu IPR filings identified earlier (IPR2025-00257, -00311, -00321, -00354, -00438, -00531, -00532, -00585, -00611) are not mapped to the '357 in anything I retrieved. A PTAB challenge would be the more natural forum for the '230 + Crump/Kao combination above, since the '230 is § 102(b) art and Crump and Kao are both § 102(b) and were already used successfully against other Stratasys patents.

Bottom line: Claims 1, 8, and 15 are, in my assessment, vulnerable to § 103 over Jamalabad '230 in view of Crump, and more strongly over '230 in view of Kao (1999) — with the Honeywell '570 publication as a hedge for the raster-generation elements. The single element doing the most work for the patentee is the "deposition rates that vary based on the determined dimensions" limitation, and that element is squarely met by Crump's feed-rate/road-width teaching as applied by Kao to non-uniform region "thickness." The dependent claims are the real defensive depth, and I cannot assess most of them because I could not retrieve their text.