US 10234294

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

Let me verify the claim text and check for any litigation or PTAB/Court dockets.

US Patent 10,234,294 B2 — Analyst Summary

Search coverage note: I searched for this exact number (10,234,294 / 10234294 / US10234294B2) only. I found no CAFC docket, IPR/PGR, or district court litigation involving this patent. Results returned for "2026 Federal Circuit" searches were unrelated cases (e.g., Yechezkal Evan Spero v. Mercedes-Benz USA, No. 2025‑1470, re U.S. 10,894,503; Exafer v. Microsoft), which do not involve 10,234,294. Treat "no litigation found" as a search-coverage statement, not a certified litigation clearance.

---

1. Bibliographic Data (from the patent front page / USPTO record)

Family / related filings (from Google Patents legal events):

• U.S. continuation app. 16/259,636 → US 10,921,134 B2
• EP 3436781 B1 (EP17714710.5)
• WO 2017/167831 A1 (PCT/EP2017/057462)

⚠️ Discrepancies between sources (flagged, not auto-corrected)

• Unified Patents portal lists priority date 2016‑03‑31, grant date 2019‑03‑18, and expiration 2036‑03‑31. This conflicts with the patent front page (filed Apr. 1, 2016; issued Mar. 19, 2019). The one-day offsets are consistent with a time-zone/EPO-record artifact carried over from the EP family member. The front page is authoritative; I have not confirmed a separate earlier priority document.
• RPX (insight.rpxcorp.com) lists priority date and status consistent with the front page, and reports "Filed: 04/01/2016; Issued: 03/19/2019; Status: Active Grant."

---

2. Abstract (verbatim)

"In one embodiment, a method for road geometry matching with componentized junction models is provided. The method includes receiving a plurality of predefined road component models, receiving map data representing a physical road junction, selecting a subset of the plurality of road component models to characterize the physical road junction represented by the received map data, and defining a road junction configuration for the physical road junction with the selected road component models."

---

3. Plain-Language Overview of the Independent Claims

Claim 1 — Method

Best-available granted text (via RPX; see confidence caveat below):

accessing a plurality of predefined road component models of short road segments with lanes varying by a shape, by a lane count, or by the shape and the lane count; receiving map data comprising line segments representing lanes of a physical road junction; selecting a subset of the plurality of road component models to characterize the physical road junction by matching the subset of road component models to the line segments represented by the received map data; accessing a table for an order for the subset of the plurality of the road component models; assembling the subset of the plurality of the plurality of road components in response to the order for the subset of the plurality of the road component models; modeling the physical road junction using the selected subset of the plurality of road component models; and rendering the model of the physical road junction to generate an image of the junction view from the selected subset of the plurality of road component models.

Plain language: Keep a library of pre-built "snap-together" 3D road pieces (short segments that differ by shape and/or lane count). Take digital map data expressed as line segments tracing the lanes of a real-world junction. Pick the subset of pre-built pieces whose geometry best matches those line segments. Consult a lookup table that specifies the order in which to connect the chosen pieces. Snap that subset together in the table-specified order to build a model of the junction, and render it to produce the junction-view picture shown to a driver. The key inventive hooks are the library of shape/lane-variable modular segments, matching them to raw map line data, and using a table to dictate assembly order.

Claim 8 — Apparatus

at least one processor; and at least one memory including computer program code for one or more programs; the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to at least perform [receiving a plurality of predefined road component models; receiving map data comprising line segments representing a physical road junction; selecting a subset of the plurality of road component models to characterize the physical road junction; and modeling the physical road junction using the selected subset of the plurality of road component models].

Plain language: The same concept implemented as a device/system — a server or in-vehicle navigation unit with a processor and stored program code that (a) holds/receives the modular road-segment models, (b) receives map data with line segments for a physical junction, (c) selects the subset of models that characterize that junction, and (d) builds the junction model from the selected subset. This claim tracks the specification's FIG. 27 acts more closely than the abstract does, and it does not expressly require the assembly-order table or the rendering step recited in claim 1.

Claim 11 — Non-Transitory Computer-Readable Medium (algorithm-specific)

receive a plurality of road component models; receive map data comprising line segments representing a physical road junction; select, using the received map data, a fork component model for characterizing a divergence or a convergence in the physical road junction; select a plurality of straight component models for characterizing one or more sequences of connected straight line segments; select a straight or curved component model for each line segment leading from the divergence or convergence that is not part of the straight sequences; and select a straight or curved component model for each line segment not leading from the divergence/convergence and not part of the straight sequences;
wherein selecting the straight component models for the straight sequences comprises: establishing a reference angle from a first line segment's direction; determining an angle for each subsequent segment relative to that reference angle; determining whether the sum of the angles is less than a first threshold; determining whether the sum of the angles for each pair of consecutive segments is less than a second threshold; and, if both thresholds are satisfied, selecting the straight component models for that sequence.

Plain language: A software-product claim directed to the specific matching algorithm. It requires handling three categories of map geometry differently:

1. Forks — pick a fork model at each split/merge point (a divergence or convergence).
2. Straight runs — find the longest stretches of connected segments that are "straight enough," using a two-tier angle test: total accumulated bend across the run must stay under a first threshold (spec gives an example of ~10°), and local bend between each pair of consecutive segments must stay under a second, tighter threshold (spec example ~5°). If both tests pass, model the whole run with straight segments.
3. Everything else — for each leftover segment branching off the fork, and for each leftover segment elsewhere, pick whichever straight or curved model best fits (the specification describes angle-based selection for fork branches and distance-based selection for other segments).

Claim 17 — Non-Transitory Computer-Readable Medium (endpoint/angle selection detail)

Reported as: "…wherein selecting the straight component model or the curved component model for each line segment leading from the divergence or convergence that is not part of the one or more sequences of connected straight line segments comprises: identifying an endpoint of each line segment leading from the divergence or convergence; determining an angle between the end point of each line segment … relative to a direction of a corresponding output path of the fork model; and selecting the straight component model or the curved component model … based on the determined angle, wherein determining the endpoint … comprises adding the length of a road component model to the selected fork component model."

Plain language: This recites the "direction algorithm" in detail: extend from the fork model by a component-model length to fix the candidate endpoint, measure the angle between that endpoint and the fork's output-path direction, and choose the straight/curved model based on that angle.

---

4. Confidence and Uncertainty Notes (important)

1. Claim-set reading. The front page states 17 claims. Based on the retrieved text, I read the independent claims as claims 1, 8, and 11. Claim 17 opens with a "wherein…" clause that strongly suggests it depends from claim 11 (or from claim 16) rather than being independent — i.e., only three independent claims. I could not retrieve the complete, verbatim claim set (claims 2–7, 9–10, 12–16 were not returned), so I cannot authoritatively describe those dependents or confirm the dependency of claim 17. Do not treat the claim listing above as a certified full claim set; obtain the granted claims from USPTO Patent Center / the printed patent for that.

2. Claim 1 text confidence: moderate. The claim 1 wording I quote comes from a third-party aggregator (RPX). It differs materially from the patent's own abstract and from the FIG. 27/FIG. 23 summary in the specification, which use "receiving… selecting… defining a road junction configuration." The two are not inconsistent as a legal matter (granted claims frequently narrow relative to the as-filed summary, and the added "accessing a table for an order," "assembling," and "rendering… junction view" limitations look like claim-narrowing amendments), but I have not verified the RPX text against the USPTO certified copy. Verify against the official record.

3. No litigation / CAFC 2026 activity found. I located no 2026 CAFC docket, PTAB proceeding, or district court case naming U.S. 10,234,294. The only related litigation hit near HERE Global B.V. involved a different patent (a HERE indoor-positioning complaint referencing U.S. 8,xxx,487 in Metataste General Trading LLC v. HERE Global BV, D. Del. 1:19-cv-01379) and does not implicate 10,234,294.

4. Status. Google Patents records the patent as Active, with anticipated expiration 2036-04-01. Unified Patents lists lifetime renewal fees of $11,000 (patent) and $57,430 (family). I have not independently confirmed current fee/maintenance status at the USPTO.

5. What the invention covers (technical context), for completeness: The specification describes assembling a complete 3D junction view from a finite library of "snap-together" straight, curve, and fork component models, with a naming syntax (e.g., SXYZ for straight, CUV.W.X for curve, FA.B.C.DE for fork, and junction templates like R321h2h1) plus a look-aside table (LAT) mapping each route through a junction archetype to the ordered component-model parts, including fillers and overpass/underpass (cloverleaf u/d) handling. The patent claims space-efficiency: ~2,500 component models supporting >10²⁴ junction configurations, or ~750 models supporting ~10²³.

Sources: USPTO front page PDF (patentimages.storage.googleapis.com/…/US10234294.pdf); Google Patents, https://patents.google.com/patent/US10234294B2/en; Unified Patents portal, https://portal.unifiedpatents.com/patents/patent/US-10234294-B2; RPX Insight, https://insight.rpxcorp.com/patent/US10234294B2; Justia Patents (family member US 10,921,134).

I'll search for litigation involving US patent 10234294.

Let me dig deeper into litigation-specific databases.

Let me check the Unified Patents litigation database and other sources specifically for this patent.

Summary of Findings

Based on my searches of patent-litigation sources (including the Unified Patents patent and litigation portals, RPX/Insight, PTAB/PTACTS filings, Docket Navigator-derived data, and general web results), I found no litigation involving U.S. Patent No. 10,234,294 ("Road geometry matching with componentized junction models"). I could not identify any district court complaint, ITC action, PTAB proceeding, or CAFC appeal in which this specific patent number appears.

What I checked and what came back

1. Unified Patents patent page — US-10234294-B2
   https://portal.unifiedpatents.com/patents/patent/US-10234294-B2
   The page returns bibliographic/valuation data only (assignee HERE Global BV, application 15/088,317, grant 2019-03-18, expiration 2036-03-31, family of 6). No litigation entries were returned for this patent.

2. Unified Patents litigation case list (https://portal.unifiedpatents.com/litigation/caselist) — no case listing for 10,234,294 was surfaced.

3. Broader searches for the patent number plus "litigation," "complaint," "infringement," "asserted," "defendant," and for HERE Global B.V. patent suits returned no complaint asserting U.S. 10,234,294.

Important caveat — related HERE cases that do NOT involve this patent

My searches did surface other litigation in which HERE entities appear, but none of these assert U.S. 10,234,294, and I am flagging them only so you do not mistake them for matches:

• Facet Technology Corp. v. HERE Global B.V., Case No. 2:24-cv-00269 (E.D. Tex.). The complaint appears to assert different patents (e.g., U.S. 9,335,255 and U.S. 9,671,3xx) directed to LiDAR/road-marker and HD-map technology — not the '294 patent.
• Omnitek actions against HERE Global asserting U.S. Patents '057 and '569 (navigation/routing) — not the '294 patent.
• Basarsoft v. HERE (Rechtbank Oost-Brabant, Netherlands) — a copyright/map-watermark dispute, not a U.S. patent case, and not involving this patent.

Conclusion

To the best of my search-based knowledge, there is no known litigation involving U.S. Patent No. 10,234,294. If you need a definitive negative confirmation, the authoritative places to verify would be:

• PACER (all U.S. district court dockets) — full-text/party search for the patent number
• Unified Patents Portal — https://portal.unifiedpatents.com/litigation/caselist (filter by patent number 10234294)
• RPX Insight / Docket Navigator — patent-level litigation lookup
• USPTO PTAB (PTACTS) — for any IPR/PGR on this patent
• CAFC docket search — for any appeal

Please note the databases I could reach via search may be incomplete or delayed, and I could not run a direct PACER/Unified query in this session; the absence of results here should be treated as "no litigation found," not as a certified negative.

PTAB proceedings on file

Verification note first. I ran fresh web searches to supplement the ODP block: for the patent number, for the invention title ("road geometry matching with componentized junction models"), for HERE Global as patent owner, and for litigation/PTAB docket references. I found no PTAB petition, institution decision, Final Written Decision, or settlement involving US 10,234,294. Two near-misses that should not be mistaken for this patent:

• Several PTAB papers discuss a "'342 patent" in IPRs filed by RPX against a content‑management / wireless content access patent (claim construction and APA arguments about "plain and ordinary meaning"). That is a different '342 patent, not US 10,234,294.
• PTAB Ex parte Appeal 2019‑006421 (Application 14/836,392, HERE Global B.V. as real party in interest, decided by APJs Bahr, Osinski, O'Hanlon) involves HERE road-database/intersection data, but it is an ex parte prosecution appeal, not an AIA trial, and a different application. It is not a proceeding "on" the '294 patent.
• Unified Patents' patent portal has a profile page for US‑10234294‑B2, but it shows only analytics (RUBY/CITX/BRIX/PVIX ratings) — no IPR/PGR/CBM record. The existence of a profile page is not evidence of a challenge.

---

Proceedings overview

Total AIA trial proceedings on US 10,234,294: 0 — 0 active, 0 with claims invalidated, 0 with claims sustained, 0 settled, 0 institution denials. The patent is completely untested at the PTAB, so the defensive posture is the inverse of the usual IPR story: there is no FWD to cite, no canceled claim to point to, and no § 315(e)(2) estoppel to weaponize — a defendant must build its invalidity case from scratch.

---

No proceedings to report

There is no {PROCEEDING_NUMBER} — {Petitioner} v. {Patent Owner} entry to populate for this patent. Per the structured ODP block (which is the canonical source) and confirmed by web search, no petition has ever been filed, instituted, terminated, or decided against US 10,234,294.

Relevant public-record coordinates for the patent itself:

Patent page: https://patents.google.com/patent/US10234294/en · PTAB E2E: https://ptab.uspto.gov/ · PTAB Decisions: https://www.uspto.gov/patents/ptab/decisions

---

Strategic summary

Claim status: all claims UNTESTED. The entire claim set stands as issued and unamended. Nothing has been canceled by the Board, and — importantly — nothing has been affirmed by the Board either. Note that the granted claim set is not the same as the published pre-grant claims: the published application (US 2017/0284812 A1) carried 20 claims, while the granted patent presents a narrower independent claim 1 that adds "accessing a table for an order," "assembling the subset … in response to the order," and "rendering the model … to generate an image of the junction view." Treat the granted text as the operative scope; per the RPX Insights patent profile the granted patent shows 17 claims. I flag that count as sourced from a secondary database rather than the USPTO face record — verify against the printed patent before relying on it.

Estoppel landscape: empty. Because no IPR/PGR has been instituted, § 315(e)(2) estoppel constrains nobody. A defendant today may raise any § 102/§ 103 ground, including art that the examiner already considered — subject only to the ordinary § 325(d) discretion of the Board and the district court's own validity rules. There is also no General Plastic serial-petition problem and no "same petitioner filed twice" dynamic. The converse risk: without an FWD, the patent's asserted claims have never been subjected to the Board's claim-construction and obviousness analysis, so prior-art strength is genuinely unquantified. The field is open but unproven.

Pattern signals: none yet. No petitioner has filed even once. HERE Global B.V. has not been an aggressive PTAB appellant — the only appellate-type activity I found in the vicinity is an ex parte prosecution appeal on a different application. No defensive aggregator (Unified Patents, RPX, etc.) appears in the chain for this patent; Unified merely maintains an analytics profile. HERE (and its successor in the mapping business) is a product company, not an NPE — so the usual "well-asserted patents eventually attract IPRs" heuristic is weaker here, which makes the total absence of PTAB activity less diagnostic of invalidity and more diagnostic of the patent simply never having been litigated.

---

Recommended next steps

• If you are a defendant: there is no FWD to link or quote — do not build a memo around a "canceled claims" theory, because no claims have been canceled. Instead, verify currency yourself: run the patent through PTAB E2E and USPTO PatentCenter (application 15/088,317) for any proceeding filed since the ODP ingest, and docket-watch the Federal Circuit for appeals. Absence of PTAB activity is a snapshot, not a permanent state.
• Mind the calendar. Because there is no prior petition, the § 315(b) one-year clock starts when you are served with a complaint alleging infringement of this patent. Preserve your IPR window and start prior-art collection now; the earlier you file, the less exposure to § 325(d) "same art previously presented" discretion.
• Sweep the family. A single-patent IPR on US 10,234,294 may be strategically incomplete — US 10,921,134 B2 is a continuation with substantially the same disclosure and a later expiration, and EP 3436781 B1 covers the same subject matter. Any freedom-to-operate or invalidity analysis should cover the whole family before you commit to a petition.
• Target the grant-time narrowing. The granted claim 1's added table/order/assembling/rendering limitations are the natural focus for both infringement avoidance and invalidity. Prior art addressing map-data segmentation, look-up-table-driven map rendering, and spline/component-based junction rendering in navigation systems is the highest-value search space.
• Bottom line for a demand letter: a plaintiff asserting this patent has no PTAB loss to recover from and no PTAB win to rely on — the patent's validity is an open question in every forum. That is a genuine defensive opening, but it is not a "no case" scenario, and nothing in the public record lets you tell a troll its claims are already dead.

I'll research this patent's ownership chain across the specified sources before reconstructing the record.

Ownership & Assignment Analysis — US 10,234,294 B2

Analyst note on sources (read first): I was able to retrieve the patent's full text, its Google Patents "Legal Events" table, and third-party portfolio records (Unified Patents portal, RPX Insight). I was not able to retrieve the record abstracts from USPTO Assignment Search / Assignment Center in this session. Consequently, reel/frame numbers and the correspondent-of-record are unavailable and are not reported below. I have not fabricated them. Everything below is either (a) stated in the authoritative patent text supplied, or (b) sourced to a citable third-party record. Where a data point is missing, it is marked "not retrievable."

---

Inventors

• Both inventors are named on the face of US 10,234,294 B2 and in the corresponding pre-grant publication US 2017/0284812 A1 (published 2017-10-05).
• Employer attribution is supported by the assignment record in the patent's own legal events table: both BUDAJ, Maros and KIM, Robin appear as assignors on the 2016-07-19 record running to HERE GLOBAL B.V. — i.e., they personally executed an inventor-to-company assignment ~3.5 months after the 2016-04-01 filing date. This is the ordinary employment-obligation pattern, not an inventor-initiated transfer.
• Unusual patterns: none detected. There is no evidence in any retrieved record that either inventor departed HERE within 12 months of filing, and no inventor-to-third-party conveyance appears in the chain. Note that the Google Patents rendering garbles the diacritic ("Maro{hacek over (S)} Budaj"); the correct legal name is Maroš Budaj — the encoding artifact is a rendering issue, not a different person.
• Corporate context at filing: the application was filed 2016-04-01, roughly four months after Nokia's sale of the HERE business (announced 2015-08-03) to the AUDI AG / BMW Group / Daimler AG consortium closed. HERE Global B.V. was therefore already consortium-owned, not Nokia-owned, on the filing date. See Nokia's listing prospectus describing the EUR 2.8bn enterprise-value transaction: https://live.euronext.com/sites/default/files/final_-_nokia_-_listing_prospectus_october_23_2015.pdf

---

Original assignee

HERE Global B.V. (trading as HERE Technologies), Kennedyplein 222 / Kennedylein 222, 5611 ZT Eindhoven–Veldhoven, the Netherlands; Dutch Commercial Register no. 17070436.

• Line of business: digital mapping, location intelligence and navigation. HERE builds HD maps and location platforms for automotive OEMs, mobile devices and enterprise customers; it is the successor to Nokia's NAVTEQ/Gate5/Earthmine mapping assets.
• Does it ship a product embodying the claims? Yes. Claim 1 recites selecting component road models, assembling them per a table-defined order, and rendering the model to generate a junction-view image. HERE's commercial products — embedded/cloud map data, the HERE SDK, HERE Navigation-on-demand and the HERE WeGo app — include turn-by-turn lane guidance and junction views; the specification itself frames the invention as an improvement to those offerings ("There remains a need for navigation systems to be able to display junction view images ... prior to performing driving maneuvers"). The patent is squarely a product patent, not a licensing vehicle.
• Ownership: HERE Global B.V. remains privately held and operating. Following the 2015 car-consortium acquisition, additional minority stakes were later taken by Intel (2017) and by Mitsubishi Corporation / NTT (2020). The company has not been dissolved, and I found no Chapter 7/11 proceeding involving it.
• Status of the portfolio's enforceability posture: HERE has historically been a licensing/partnering mapping supplier, not a high-frequency patent plaintiff. I did not locate any infringement suit naming US 10,234,294, and the Unified Patents portal entry shows no litigation or PTAB event and lists "Parent Company: Here Global B.V." with 6 family members — a benign, non-asserted profile. https://portal.unifiedpatents.com/patents/patent/US-[10234294](/patent/10234294)-B2

---

Assignment timeline

Direct USPTO Assignment Center records: not retrievable in this session. The chain below is reconstructed from the Google Patents Legal Events table embedded in the authoritative patent text, which is a mirror of USPTO assignment data. Reel/frame numbers are not shown by that source and I will not invent them. Correspondent of record is likewise unavailable.

• 2016-07-19 (recorded; execution date not exposed) — Reel not retrievable

  • Conveyance: Assignment of assignors' interest ("ASSIGNMENT OF ASSIGNORS' INTEREST (SEE DOCUMENT FOR DETAILS)")
  • Assignor: BUDAJ, Maros; KIM, Robin (the two inventors)
  • Assignee: HERE GLOBAL B.V.
  • Correspondent: not retrievable from the sources available to me
  • Context: standard inventor-to-employer assignment — the ordinary perfection step for a company-owned application; not an acquisition or reorg.

• 2016-07-21 (recorded; execution date not exposed) — Reel not retrievable

  • Conveyance: Assignment of assignors' interest
  • Assignor: NESS GLOBAL SERVICES LIMITED
  • Assignee: HERE GLOBAL B.V.
  • Correspondent: not retrievable
  • Context: internal/corporate assignment into the HERE entity. ⚠️ Flag: the assignor name "NESS Global Services Limited" is anomalous and I could not independently verify this entity or its relationship to HERE/Nokia. It does not match Nokia's known "NSN — Nokia Solutions and Networks," nor any HERE-affiliated entity I could confirm. Treat the assignor identity as unverified; the assignee side (HERE Global B.V.) is consistent with an internal chain-of-title clean-up following the Nokia→consortium carve-out of the HERE business. This is a data-integrity flag, not an NPE signal.

• No post-issuance assignment of US 10,234,294 appears in any record retrieved. The two remaining 2017/2019 entries in the legal-events table are not ownership transfers and should not be read as such:

  • 2017-03-29 — "Priority to EP17714710.5A" and "Priority to PCT/EP2017/057462" (foreign filing/priority linkage; the resulting EP3436781B1 / WO2017167831A1 are HERE-owned family members).
  • 2019-01-28 — "Priority to US16/259,636," which issued as US 10,921,134 B2 — a continuation in the same HERE-owned family, filed ~6 weeks before grant of the '294 patent. A continuation filing is a prosecution decision, not a conveyance.

---

Timeline diagram

timeline
    title Ownership of US 10234294
    2015 : Nokia sells HERE to car consortium
    2016 : Application filed by HERE Global BV
         : Inventors assign rights to HERE
         : Second assignment recorded to HERE
    2017 : PCT and EP filings claim priority
         : US application published
    2019 : US 10234294 granted
         : Continuation filed leads to US 10921134
    2026 : No transfer to an NPE recorded

---

NPE / troll-pattern signals

Summary of the affirmative finding: the chain never left the original operating company. There is no evidence of any NPE involvement, shell-entity layering, or assertion activity on US 10,234,294.

---

Verdict

Insufficient data — but in the clean sub-sense defined by the rubric: there are no post-issuance assignments at all, only the two original 2016 conveyances into the original assignee.

The only recorded assignments are 2016-07-19 (inventors BUDAJ and KIM → HERE Global B.V.) and 2016-07-21 (NESS Global Services Limited → HERE Global B.V.); there is no subsequent reel/frame moving the patent away from HERE, no transfer to any entity on the NPE lists, and no suit or PTAB proceeding naming the patent (Unified Patents portal shows no assertion events). HERE Global B.V. still owns the patent and ships mapping/navigation products embodying the rendered-junction-view claim, so every NPE-specific signal above scores not present — the "Insufficient data" verdict reflects the thinness of the record (2 pre-grant assignments, no reel/frame or correspondent retrievable), not any suspicion of assertion. The one loose end worth a manual pull is the anomalous 2016-07-21 assignor NESS Global Services Limited, whose identity I could not verify.

Verify directly: https://assignmentcenter.uspto.gov/ (search patent number 10234294) — confirm reel/frame, correspondent of record, and the true identity of the 2016-07-21 assignor.
Corroborating records used: https://patents.google.com/patent/[US10234294B2](/patent/US10234294B2)/en (legal events) · https://portal.unifiedpatents.com/patents/patent/US-10234294-B2 · https://insight.rpxcorp.com/patent/US10234294B2

---

Analyst caveats

1. Do not treat the missing reel/frame and correspondent as "no assignment recorded." The Assignment Center itself was not queryable for me; the Google Patents legal-events mirror does record two 2016 assignments. The correct statement is: records exist; their reel/frame and correspondent fields were not retrievable.
2. The 2017/2019 "Priority to …" entries are not assignments and must not be counted as chain links. US 16/259,636 is a same-owner continuation that issued as US 10,921,134 B2.
3. Do not infer NPE status from the "B.V." structure. HERE Global B.V. is a Dutch operating company with employees, commercial products and automotive customers — the opposite of a single-purpose licensing vehicle. Naming alone is not a finding, and here the concrete evidence (products, no transfers, no suits, consortium ownership) runs against an NPE characterization.
4. Nokia is not in this chain. Nokia's patent licensing program (Nokia Technologies / Advanced Technologies) is a real and active assertion operation, but the retrieved records show no Nokia entity assigning this patent; Nokia divested the HERE business in 2015 and retained its own separate portfolio.

I'll research the prior art and citations for US 10,234,294. Let me run several searches.

Let me get the full "References Cited" list from the patent's front page and the PDF.

Let me locate the complete reference list from the patent's front page.

Prior Art Analysis — U.S. Patent No. 10,234,294

Important scope note before the findings

Two things I must state up front, per the "do not fabricate" rule:

1. I could not retrieve the complete "(56) References Cited" list for US 10,234,294 in this session. The examiner-cited list is long (the front page carries a multi-column U.S. patent documents list that continues onto a second page, plus foreign documents and "Other Publications"). My searches returned only a partial front-page extract. I am therefore reporting (a) the references I could concretely verify, and (b) an analytical framework for the § 102 question — but I am not claiming the list below is exhaustive, and I have not invented descriptions for references whose content I could not verify.
2. Search noise was heavy. Many hits for "10234294" were unrelated patents (e.g., intraocular-pressure tonometers, bibliometrics papers). I filtered those out; the authoritative targets were the Google Patents record, the USPTO-hosted patentimages PDF, and the family/continuation records.

---

1. Patent identification (verified against the record)

Sources: https://patents.google.com/patent/[US10234294B2](/patent/US10234294B2)/en ; https://patentimages.storage.googleapis.com/95/79/7e/1e7f376f78ef90/US10234294.pdf ; https://patents.justia.com/patent/[10921134](/patent/10921134) ; https://portal.unifiedpatents.com/patents/patent/US-10234294-B2

---

2. Cited references retrieved (partial)

2A. U.S. Patent Documents (verified from the patent's own front page PDF)

The extract ends mid-column at "6,405,131 B1 6/200…", so additional U.S. references beyond this point were not recovered.

2B. Foreign Patent Documents (verified partial)

Again, the foreign-document column was truncated; more foreign references almost certainly follow.

2C. Other Publications (verified)

• Dal-Poz et al., "Automatic Extraction of Road Seeds from High-Resolution Aerial Images," Annals of the Brazilian Academy of Sciences, Jan. 5, 2005 — cited on the front page. This is a non-patent-literature reference directed to automated road extraction from imagery.

2D. References I could NOT verify (so I am not describing them)

The remainder of the U.S. list (post-6,405,131), the remainder of the foreign list, and any additional NPL items are unretrieved. Note that HERE's navigation-junction patent family is known to cite dense Japanese/European map-display art (Pioneer, Denso, Aisin, Nissan, Hitachi, Alpine, etc.), but I will not attribute specific titles or inventors to those entries without the source text — doing so would risk fabrication. The Google Patents "Cited By" snippets I saw (e.g., WO1999054848A1's page listing US10234294B2 alongside FR2674652A1, EP0738876A2, JPH09171348A, JPH09318380A) reflect those documents' citation tables and are not confirmed members of the '294 front-page list.

---

3. § 102 analysis — what the cited art can and cannot anticipate

3.1 The independent claims define a specific architecture

From the record, the independent claims are:

• Claim 1 (method): accessing a plurality of predefined road component models of short road segments with lanes varying by shape and/or lane count → receiving map data comprising line segments representing lanes of a physical road junction → selecting a subset by matching the component models to the line segments → accessing a table for the order of the subset → assembling the subset in that order → modeling the junction → rendering the model to generate a junction-view image.
• Claim 8 (apparatus): processor + memory counterpart (receive models, receive map data with line segments, select subset, model the junction).
• Claim 11 (CRM): fork-component selection for a divergence/convergence, straight-component selection for sequences of connected straight line segments, straight/curve for each line segment leading from the divergence, and straight/curve for each line segment not leading from the divergence — plus the specific straight-sequence test (reference angle; sum-of-angles vs. first threshold; pairwise consecutive-angle sum vs. second threshold).
• Claim 17 (CRM): the angle-based fork/model selection with endpoint computation ("adding the length of a road component model to the selected fork component model").

3.2 Assessment

Best § 102 candidates among the verified references. The verified cited references are, at face value, directed to navigation display and intersection-guidance systems — e.g., US 6,134,501 ("Vehicle travel-route guidance apparatus with internal intersection discount feature") concerns route guidance through/around intersections, not the assembly of a junction view from a library of shape-and-lane-count-variable component models. None of the verified references that I can characterize appears to disclose the component-model + look-aside-table + rendering pipeline that claim 1 recites, nor the fixed-length line-segment matching / threshold-based straight-sequence detection of claim 11.

Conclusion on anticipation: On the evidence I recovered, no single cited reference, taken alone, appears to disclose every element of claim 1, 8, 11, or 17 — which is consistent with the patent having issued without a § 102 rejection over these references (the record shows it granted on the first action, with the "0 days" term adjustment indicating no significant prosecution delay). If any of the unretrieved references does anticipate, the most likely candidates on subject matter would be those in the Japanese intersection-display families, but I could not verify their content.

Caveat. Because I could not retrieve the full (56) list, this § 102 conclusion is provisional, not certified.

---

4. What I recommend for a definitive answer

Within USPTO (and free) channels, the authoritative "References Cited" and any examiner's § 102/§ 103 rejections are at:

1. USPTO Patent Center — https://patentcenter.uspto.gov — application 15/088,317, "References Cited" and the full Image File Wrapper (office actions, IDS forms PTO-1449/SB-08).
2. USPTO Patent Public Search — https://ppubs.uspto.gov — retrieve the front-page "(56)" and the PDF full text directly.
3. Global Dossier (via Patent Center) — for the EP 3436781 and WO 2017/167831 family, whose EPO/ISA search reports list the closest prior art with X/Y significance codes. That search report is likely the single richest source of § 102-category art for this family.
4. EPO Register / Espacenet — EP 3436781 B1 and WO 2017/167831 A1 (PCT/EP2017/057462), including the ISR and written opinion.

---

5. Explicit limitations of this report

• The complete "(56) References Cited" list was not recovered; only the first ~14 U.S. entries, 2 foreign entries, and 1 NPL item are verified.
• I did not fabricate titles, inventors, or dates for the unretrieved references.
• No § 102 rejection over a specific reference can be asserted from the data available here; the analysis in § 3 is a subject-matter comparison against the references I could verify.
• Nothing here should be read as a legal opinion on validity or infringement.

I'll verify the subject matter of the cited references from the Prior Art section and check for closely analogous art in the same field before building the § 103 combinations.

I have good verification on several cited references. Let me check a few more from the list and look for closely analogous art on junction-view generation and road-geometry matching.

Obviousness Analysis — U.S. Patent No. 10,234,294 B2

"Road geometry matching with componentized junction models" (HERE Global B.V.)

Analysis date: 2026 · Statutory basis: 35 U.S.C. § 103; Graham v. John Deere Co., 383 U.S. 1 (1966); KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007); MPEP §§ 2141–2144.04

---

0. Bottom line up front

1. On the record I was able to verify, no combination of the cited references establishes a complete prima facie case for independent claim 1. The reason is structural, not doctrinal: the verified prior art (mostly 1990–2002 vehicle-navigation and map-database patents) discloses lane-level map data, junction pattern-matching, geometric abstraction of junctions, and threshold-based geometric classification — but not a library of "short road segment" component models that vary by shape and lane count, and not the assembly-order look-up table.

2. The strongest verified combination — Denso US 6,134,501 + Visteon US 5,902,350 + Visteon US 5,910,177 — does render obvious a substantial portion of the specification's technical content and the narrower algorithm-specific limitations of claims 11 and 17 (threshold-based geometric classification, angle comparisons, lane-count-driven modeling), and gets close on claim 8.

3. The single most powerful § 103 lever is not in the prior-art list at all — it is in the patent's own specification. The '294 disclosure contains multiple express admissions that (a) unique junction-view images/models per junction were known, (b) generic 2D junction-view images and generic 3D junction models that "approximate actual junction configurations" were known, (c) storing junction images is "space-intensive," and (d) the look-aside table is "only one exemplary implementation" and "not always necessary." Those admissions are prior art under MPEP § 2129 and they convert the claimed invention into a storage-vs.-fidelity optimization over admitted known techniques — precisely the KSR "design incentive" and "known technique applied to a known device ready for improvement" fact pattern.

4. What is missing to complete the case is a primary reference teaching template/component-model-based junction depiction. That reference is very likely in the unretrieved portion of the (56) list (the front page carries a long multi-column reference list that I could only partially recover) or in the EPO/ISA search report for family member EP 3436781 / WO 2017/167831. See § 9.

---

1. Threshold constraints on this analysis (read before relying on any conclusion)

---

2. POSITA and prior-art status of the verified references

Level of ordinary skill. A POSITA here is a navigation/map-data engineer or applied computer-graphics engineer with (i) a bachelor's degree in computer science, electrical engineering or geomatics (or equivalent experience), (ii) 2–4 years of experience with digital map databases (node/link or segment topology, lane attributes) and turn-by-turn guidance display, and (iii) familiarity with 2D/3D graphics rendering and map-data-to-image pipelines. The field of the invention is G01C 21/30 (map/contour matching), G01C 21/32 (structuring/formatting of map data) and G01C 21/36 (I/O arrangements for on-board computers) — i.e., navigation map data structures and their graphic presentation. All of the verified references sit in this same field or in the immediately adjacent vehicle-navigation display field, so the "analogous art" threshold is comfortably met for all of them.

§ 102(b) status. Application filed April 1, 2016. Every verified reference granted or published between 1990 and 2005 — more than a decade before the critical date. All qualify as prior art under § 102(b) and are therefore available for § 103. (Verified grant dates: '350 → 1999‑05‑11; '177 → 1999‑06‑08; '501 → 2000‑10‑16/17; '128 → 2002‑06‑11; NPL Dal-Poz → Jan. 2005.)

Examiner search profile. The patent face lists prior-art keywords road / straight / junction / component / model — confirming the examiner searched exactly the conceptual space of the claims. That cuts both ways: it supports that the field was active, and it explains why the record shows an apparently clean allowance (0 days PTA).

---

3. The claim architecture to be tested

Key analytic point: claim 1 is the broadest conceptual claim but recites the most specific pipeline; claims 11 and 17 are the most vulnerable because they recite quantitative geometric decision rules (thresholds, angles, distances) of exactly the kind heavily represented in 1990s navigation art.

---

4. Element-by-element mapping of the verified art

4.1 Claims 1 and 8 (the pipeline)

4.2 Claim 11 (algorithm) and claim 17 (endpoint/angle)

4.3 Unverified cited art

The remaining recited entries — US 4,907,160 (Duncan), 5,928,308 (Nanba), 6,018,697 (Morimoto), 6,249,740 (Ito), 6,269,305 (Yamashita), 6,278,943 (Yamauchi), 6,292,752 (Takake), 6,385,536 (Kimura), 6,405,131, EP 0 965 970 A1, EP 1 096 229 A1 — were not retrieved in full. I do not attribute subject matter to them. Given that the examiner allowed without a § 102/§ 103 rejection over them (0 days PTA), the presumption is that none of them discloses the missing 1a/1d/1e limitations — but this is an inference, not a verified fact, and it is exactly what a challenger should re-test.

---

5. The patent's own admissions — the real § 103 lever

The authoritative specification contains four statements that are prior-art admissions under MPEP § 2129 / In re Nomiya:

(A) "Some current navigation systems use unique two-dimensional (2D) junction view images or 2D renderings of three-dimensional (3D) junction models for every junction in a map database… Storing unique junction view images or unique models for each junction is very space-intensive."

(B) "Other navigation systems use a limited set of generic 2D junction view images or 3D junction models to approximate actual junction configurations. Using generic junction view images or generic models is space-efficient, but yields limited fidelity with respect to the accuracy of junction topology, junction geometry, lane counts, and lane connectivity…"

(C) "Using a LAT is only one exemplary implementation, thus listing the component models in LAT columns is not always necessary. For example, component models may be derived from the junction template name."

(D) "The definition of a 'good' match is subjective, and the algorithm employed may balance aesthetic considerations with more easily quantifiable measures of accuracy…"

Why this matters enormously. Admission (B) concedes that the two-step approach the claim recites — (i) represent a real junction by a stored, generic model, and (ii) accept that the model "approximate[s] actual junction configurations" — was already known in navigation systems. Admission (A) concedes that per-junction image/model generation was known, and identifies the precise problem the invention addresses: storage. Admission (C) concedes that the table-for-order limitation of claim 1 is optional and interchangeable with deriving components from a template name — i.e., a mere design choice among equivalent data organizations. Admission (D) concedes that the matching step is a subjective/empirical tuning problem, which is the classic § 103 fact pattern ("the improvement is a matter of design choice/hand-tuning, not invention").

Stated in Graham framework terms: the scope and content of the prior art includes generic junction models and per-junction junction views; the differences between that art and claim 1 are the componentization (short-segment models parameterized by shape and lane count) and the table-ordered assembly; and the level of ordinary skill is high (map-data + graphics engineering in a mature field). The invention is a space/fidelity optimization on a known architecture — the paradigm case for § 103.

---

6. Combinations and motivations to combine

Below, each combination is stated with: the references, the claim limitations each supplies, the KSR / MPEP § 2143 rationale, and a candid strength rating.

6.1 Combination 1 (core navigation/junction combination)

US 6,134,501 (Oumi, Denso) + US 5,902,350 (Tamai, Visteon) + US 5,910,177 (Visteon) — alternatively, '501 as primary with '350 and '177 as secondary.

What it supplies:

• '501 supplies the lane-level, junction-aware, node/link map database (lane counts, link angles, road width, three-dimensional/multi-lane intersection node attributes), the route/geometry processor, and the on-board display of the road map — i.e., the data substrate and display platform for claims 1 and 8.
• '350 supplies (i) matching a real intersection representation against a stored pattern ("does intersection fit a turn-lane pattern?"), (ii) abstracting/collapsing real junction geometry into a canonical modeled form for presentation, and (iii) computing the angle between entry and exit road segments and classifying against threshold ranges.
• '177 supplies threshold-based quantitative decision-making using junction/leg parameters including number of lanes, road width, and speed limit, and confirms the field's practice of encoding junction geometry with lane-count metadata.

Motivations to combine (each independently sufficient):

• Same field of endeavor / same problem (MPEP 2141.01(a)). All three are vehicle-navigation systems whose stated purpose is to convert a map database into guidance and display output for an intersection. A POSITA seeking to improve junction depiction would look to exactly these documents.
• KSR rationale (A) — known elements combined by known methods, predictable result. Combining a lane-level map DB ('501), a junction pattern-matching/abstraction step ('350), and quantitative geometric thresholds ('177) yields nothing more than the expected sum: a canonical, parameterized representation of a real junction derived from map data.
• KSR rationale (C)/(D) — known technique applied to a known device ready for improvement. '350 already teaches that a "double maneuver" for a turn-lane connector is "potentially confusing to the driver" and requires abstraction; the '294 patent asserts the identical problem (junction views that "may not accurately match real world intersections"). The improvement sought is the same, in the same way, in the same devices.
• Common-ownership/prior-collaboration signal. '350 and '177 share the same assignee (Visteon Technologies) and are cross-incorporated by reference ('177 expressly incorporates '350-assignee-family patents such as Desai 5,515,283 for highway-ramp identification). A POSITA in the same industrial community would have had these in view — a recognized KSR motivation (common ownership / same product line).
• Examiner already considered them. All three are on the '294 face, so they are the admitted closest art; the Federal Circuit routinely treats the cited art as the presumptive starting point.

Strength: Moderate for claim 8; moderate-to-good for the threshold/angle elements of claims 11 and 17; insufficient, standing alone, for claim 1 (no component-model library; no table-ordered assembly; no junction-view rendering from the assembled components).

6.2 Combination 2 (adds automated road-geometry derivation)

Combination 1 + Dal-Poz et al., "Automatic Extraction of Road Seeds from High-Resolution Aerial Images," Ann. Braz. Acad. Sci. (Jan. 5, 2005).

Rationale: Dal-Poz supplies the concept of deriving vector road primitives ("seeds") from image data by automated geometric analysis — which supports the claim 1/claim 11 notion of converting raw spatial data into ordered geometric line segments and matching primitives to them, and supports the specification's "divide the map data into line segments of a fixed length" preprocessing as a routine vectorization step. Motivation: KSR rationale (C) — the same known technique (automated extraction/vectorization of road geometry) applied to the same kind of data (road imagery/map structure) in the same field.

Strength: Adds support to 1b/1c and to the preprocessing rationale; does not fill 1a/1d/1e.

6.3 Combination 3 (the combination that would actually reach claim 1)

Combination 2 + a template/component-model-based rendering or map-symbol-assembly reference.

This is the combination that must exist for a complete claim-1 case, and it is the one I could not verify from the record. The required teaching is: a finite library of predefined, parameterized graphic/map primitives that is retrieved and assembled in a defined sequence (optionally via a table) to render a map feature. That teaching is a common pattern in (i) the map-symbol/landmark-rendering art, (ii) the "tile/parcel" map-organization art (note: EP 2 175 385 A1, "Spatial index for locating geographic data parcels stored on physical storage media," cites the Bechtolsheim reference US 6,405,128 B1, showing that table/index-driven geographic-data organization was itself a known field), and (iii) the Japanese intersection-display families that the record indicates populate the unretrieved portion of the (56) list.

Framing if such a reference is found: the combination would be KSR rationale (A)+(D): arranging a known template-driven map-feature renderer with a known lane-level junction database to obtain a space-efficient, fidelity-adjustable junction view — a predictable result with a clear design incentive (the patent itself frames the problem as the storage cost of unique junction images). This is the theory a challenger should build; I flag it as an open item rather than a completed showing.

6.4 Combination 4 (claim 1's table limitation as a naked design choice)

Even without a template reference, the "accessing a table for an order" limitation is independently vulnerable on three grounds:

• The specification's admission (C) that the LAT is "only one exemplary implementation" and that "component models may be derived from the junction template name" is a concession that the ordering information has an equivalent, known alternative form — i.e., the choice of table versus derivation is a design choice with no unexpected result (MPEP § 2144.04(IV)).
• Storing an assembly sequence in a look-up table indexed by junction type is routine data structuring — the archetypal "arranging old elements with each performing the same function" fact pattern (KSR, 550 U.S. at 417).
• The patent's own naming syntax (R321h2h1, A BCD EFGH J K LMN PQRS Z) is an explicit, deterministic function from template name to component list, which undercuts any argument that the table adds inventive weight.

6.5 Summary combination table

---

7. Rebutting the likely patentee responses

"No reference discloses the component-model library." Correct as to the verified record, and this is the plaintiff's strongest validity argument. The counter is twofold: (i) KSR does not require a single reference to disclose the claimed arrangement where the arrangement is a predictable optimization of admitted known structures; and (ii) the record is incomplete — the (56) list and the EP ISR remain unretrieved, so the absence of a teaching is not evidence of absence of the teaching in the art of record. Do not overstate this; a court will require something in the prior art corresponding to the componentized library.

"The references teach away / are non-analogous." Weak. All verified references are in the same field (G01C 21/xx) and directed to the same problem (representing and displaying a junction to a driver). '350 in particular is directly on point: it abstracts real junction geometry into a canonical form for driver-facing output.

"Unexpected results / secondary considerations." There is no evidence of record of unexpected results, commercial success with nexus, long-felt but unmet need, failure of others, or industry praise. Two points of caution: (i) the specification's space-efficiency numbers (~2,500 models → >10²⁴ configurations; ~750 → ~10²³) are asserted, not demonstrated, and a combinatorial count is a mathematical consequence of the design, not an unexpected result; (ii) any commercial-success argument from HERE's junction-view products would face a nexus problem because junction views and 3D junction models are admitted prior art (admissions A and B), so the success must be tied specifically to the componentization + assembly-table feature.

"The claims are narrow because of grant-time amendments." True and worth conceding: claim 1's added "accessing a table for an order," "assembling," and "rendering … junction view" limitations were added relative to the published application (US 2017/0284812 A1, 20 claims). But note the irony for the patentee — the amendment imports the very limitations the specification calls optional (admission C), which makes them easier to attack as design choices even as they narrow infringement scope.

---

8. Validity posture by claim

---

9. What would complete the § 103 case (recommended next steps)

1. Re-pull the complete (56) list from the certified front page. Target the U.S. entries after 6,405,131 and the full foreign column. Use the examiner's keywords (road / straight / junction / component / model).
2. Pull the EPO/ISA search report for EP 3436781 B1 and WO 2017/167831 A1 (PCT/EP2017/057462). This is the single richest unretrieved source: it will contain X/Y-coded closest art selected by an examiner for the same disclosure, including Japanese intersection-display (Pioneer/Denso/Aisin/Alpine/Nissan) and European map-display art.
3. Verify US 6,405,128 B1 (Bechtolsheim et al.) subject matter and resolve the US 6,188,697 B1 attribution discrepancy against the printed patent. If 6,405,128 is map-data/tile-organization art, it becomes a candidate secondary reference for the table-driven assembly element.
4. Search the specific missing element — a finite library of predefined, shape- and lane-count-parameterized road-segment graphic models assembled in a defined sequence to render a map feature — in: map-symbol/landmark rendering; map tile/parcel indexing (cf. EP 2 175 385 A1's citation of Bechtolsheim); CAD/parametric-primitive assembly in navigation graphics; and the pre-2000 Japanese junction-display literature.
5. Follow the family. US 10,921,134 B2 (continuation, granted 2021‑02‑16) shares the disclosure and has a later expiration; EP 3436781 B1 covers the same subject matter. Any invalidity or FTO analysis should cover the whole family before committing to a position.
6. Preserve the record on admissions. Admissions (A)–(D) should be captured verbatim from the granted specification (identical to the published application) and cited as § 102(b) prior art under MPEP § 2129 in any invalidity contention.

---

10. Confidence and caveats

• Confidence in the verified reference contents: High for US 6,134,501, US 5,902,350 and US 5,910,177 (full-text and/or PDF content retrieved and quoted). Low/unverified for US 4,907,160, 5,928,308, 6,018,697, 6,188,697 (disputed), 6,249,740, 6,269,305, 6,278,943, 6,292,752, 6,385,536, 6,405,128, 6,405,131, EP 0 965 970 A1, EP 1 096 229 A1 — no subject matter attributed.
• Confidence in the conclusion that no verified combination reaches claim 1: Moderate-to-high as to the verified subset; not a certification as to the full cited art or the field.
• Confidence in the "patent's own admissions" theory: High — the quoted passages appear in the authoritative full text supplied.
• Claim-text confidence: claim 1 remains aggregator-sourced and materially narrower than the abstract; obtain the certified claims before filing anything that depends on claim 1's exact wording.
• Nothing here is a legal opinion on validity or infringement, and no PTAB or court has ever construed these claims (the patent is completely untested — 0 IPR/PGR, no litigation identified).

Sources: https://patents.google.com/patent/US10234294B2/en · https://patents.google.com/patent/US6134501 · https://patents.google.com/patent/US5902350 · https://patents.google.com/patent/US5910177 · https://portal.unifiedpatents.com/patents/patent/US-6134501-A · https://insight.rpxcorp.com/patent/US6134501A · https://patentimages.storage.googleapis.com/54/7e/37/a30302199cbf73/US5902350.pdf · https://portal.unifiedpatents.com/patents/patent/US-10234294-B2 · EP 2 175 385 A1 (http://data.epo.org/gpi/EP2175385A1)