US 11391587

I'll verify the patent details and check for any 2026 litigation activity.

Let me check for any litigation or post-grant activity.

US Patent 11,391,587 — Verified Summary

Search scope note: I ran targeted searches for the exact number 11,391,587 / 11391587 across patent databases, PTAB filings, and district-court/CAFC docket material. I found no litigation, IPR/PGR, or Federal Circuit appeal involving this specific patent. My tools do not give direct authenticated access to USPTO PatentCenter or the CAFC docket, so treat that as "no activity surfaced," not as a certified all-clear.

Bibliographic Data

Cited prior art (examiner-cited): US 8,706,394 B2 (Gray & Co.); US 2011/0098915 A1 (Disatnik); US 2014/0278052 A1 (Caliper); US 9,494,937 B2 (Verizon Telematics); US 2018/0004214 A1 (GM). Non-patent literature: Behrisch et al., SUMO – Simulation of Urban Mobility (2011).

Abstract (verbatim)

"The disclosure relates to assessing impact of blockages on an autonomous vehicle transportation service. An example method may include identifying a plurality of starting and destination location pairs within a service area; running a first plurality of simulations by determining a first route for each of the plurality of pairs without including a first blockage in the service area; determining a first summary for each determined first route identifying a length of time to complete that determined route; running a second plurality of simulations by determining a second route for each of the plurality of pairs and including the first blockage in the service area; determining a second summary for each determined second route identifying a length of time to complete that determined route; and comparing the determined first summaries with the determined second summaries in order to assess an impact of the first blockage on the service."

Plain-Language Overview of the Independent Claim

Only claim 1 is independent. All of claims 2–20 depend (directly or indirectly) from it.

Claim 1 — A/B simulation "before and after" method for scoring a blockage's service impact. One or more processors do four things:

1. Identify origin–destination (OD) pairs within a service area (spec says ~5,000–10,000 pairs, drawn from historical trips or synthesized).
2. Run simulation set #1 ("baseline") — compute a route for every pair without a designated first blockage in the service area.
3. Summarize set #1 — for each route, produce a summary including the route's expected completion time.
4. Run simulation set #2 — recompute a route for the same pairs, this time with the first blockage present.
5. Summarize set #2 — same form of summary (completion time), then compare set #1 vs. set #2 summaries to quantify the blockage's impact in the service area.

Note a drafting nuance: the abstract says "impact of the first blockage on the service," while granted claim 1 recites "impact of the first blockage in the service area." The specification describes the output as a proxy for how many users are affected, plus aggregate deltas in time, distance, and maneuver counts, used to prioritize engineering/scouting/remapping work.

Dependent Claim Themes (2–20)

• Blockage definition (2–3): the blockage is an area the service's vehicles are forbidden to drive through; it may be time-based (effective only during specified periods).
• How OD pairs are chosen (4–8): prior trips from a past window (4); a start from one trip paired with a destination from a different trip (5), optionally time-stamped to a high-volume period (6) or a low-volume period (7); or a "fuzzed"/perturbed pair derived from a prior pair (8).
• Temporal assignment (9): each OD pair carries a start time used by the routing system.
• Summary contents (10–16): route length/distance (10); counts of a maneuver type (11) — unprotected turn (12), lane change (13), or blocking left turn (14); derivative comparison of maneuver-count differences (15); a forward-progress metric measuring how often the route advances toward the destination (16).
• Simulation/analysis variants (17–20): baseline includes a second blockage (17); comparison outputs a count of impacted OD pairs (18), total added time (19), or total added distance (20).

Caveats

• No litigation/PTAB/CAFC record found for this patent as of this search. Verify independently via USPTO PatentCenter and PACER/CAFC dockets if the question is dispositive.
• The "adjusted expiration 2041-01-03" figure is an automated estimate on a public patent aggregator, not an official PTA/terminal-disclaimer determination.
• The patent text itself contains minor typographical artifacts (e.g., "starring and destination location pairs," a stray "planning system 172" vs. "168," and "summaries 1020 and 1210" where "1220" is meant). These are transcription/OCR-level errors in the source document and are reproduced here rather than corrected, per the instruction to interpret identifiers literally.

I'll research litigation involving US patent 11391587 across patent litigation databases and court records.

Let me search targeted litigation databases for this specific patent.

Let me check specialized litigation databases directly.

Litigation Search Results: US Patent 11,391,587

Bottom Line

No litigation, PTAB proceeding, or other contested proceeding involving US Patent 11,391,587 was found.

I searched multiple litigation/patent sources for the exact number 11391587 / 11,391,587 (and treated it literally, not conflating it with other Waymo patents):

• Google Patents / USPTO patent page for US11391587B1 (patents.google.com/patent/US11391587B1)
• Unified Patents litigation resources (portal.unifiedpatents.com) — indexed cases
• Docket Alarm / CourtListener / Justia dockets (PACER-based aggregations)
• General web/CAFC-oriented searches pairing the number with "litigation," "infringement," "defendant," "ITC," "IPR/PTAB"

Every hit that contained the string "11391587" was either:

1. The patent's own bibliographic page (title: "Assessing the impact of blockages on autonomous vehicle services," assignee Waymo LLC, inventor Ajay Joshi, filed 2019-04-24, granted 2022-07-19), or
2. A coincidental number match — e.g., a Chinese trademark registration no. 11391587 ("瑞生活 RUILIFE"), a German plumbing catalog article no. 11391587, and an unrelated Netlist '087 patent (a different patent entirely). None of these relate to US 11,391,587.

I found no district court case, ITC Section 337 investigation, PTAB IPR/PGR, or CAFC appeal naming this patent.

Related Observations (context, not this patent)

To be explicit, the Waymo cases that surface in searches involve different patents and are not litigation over 11,391,587:

Important Caveats

• Confidence: My searches consistently returned no litigation for this specific number, which supports a "none found" conclusion with reasonable confidence. However, I cannot guarantee completeness — many complaints are unindexed or sealed, and this search was web-source-dependent.
• Verify directly before relying on this: The authoritative checks would be:
  • Unified Patents litigation portal: https://portal.unifiedpatents.com/litigation/caselist
  • PACER Case Locator (search "11,391,587" as patent)
  • USPTO PTAB API / Patent Center "Litigation" tab for US11391587B1
  • Docket Alarm / Lex Machina patent-number search
• The patent's legal status per Google Patents is Active, adjusted expiration 2041-01-03, family ID 82385009, with a continuation US11788854B1 (filed 2022-06-15).

If you need a litigation record on this patent, none is currently known based on the sources searched as of this date.

Proceedings overview

Zero AIA trial proceedings. The canonical structured record supplied for this analysis — the "PTAB proceedings on file" block sourced from the USPTO Open Data Portal (ODP) — returns no IPR, no PGR, and no CBM naming US 11,391,587 B1. My independent web searches for the patent number paired with "IPR," "PTAB," "petition," and "inter partes review" surfaced nothing for this patent; every hit was either the patent's own bibliographic page or a coincidental number match (see the earlier sections of this analysis). Because there are no proceedings, there are no per-proceeding entries below — I am not going to manufacture case numbers, panels, or dispositions that do not exist.

Breakdown by status: active 0 / claims invalidated 0 / claims sustained 0 / settled 0 / institution denied 0.

Defensive posture this gives a defendant: the patent is completely untested at the PTAB. All 20 claims — including independent claim 1 — stand exactly as granted on 2022-07-19, with the full original claim scope intact and no prosecution-history narrowing from an IPR (no certificate of correction, no substitute claims, no disclaimer-by-amendment). This is not the same as "hardened." It means the validity question is open, not resolved: nobody has run the prior art through an Article I tribunal, so there is no FWD to cite, no estoppel to rely on, and no adverse claim-construction ruling to work with. A defendant's IPR option is legally alive (more so than for a patent that has already survived one) but is now far harder to exercise in practice than it was in 2022 — see the risk factors below.

No proceedings to report

Because the ODP-derived list is empty, the following requested elements are not applicable and should not be treated as omissions:

Statutory windows — what is now foreclosed

Even if no petition was ever filed, the types of challenge available have narrowed with time. Computed from the grant date of 2022-07-19:

• Post-Grant Review (§ 321): CLOSED. A PGR petition had to be filed within nine months of grant, i.e., by 2023-04-19 (35 U.S.C. § 321(c)). That window shut roughly three years before today's date of 2026-09-29. Note this patent is a B1 (no pre-grant publication), so the § 122(b) nine-month-from-publication alternative never ran separately.
• Covered Business Method Review: CLOSED, twice over. CBM review under AIA § 18 was unavailable for petitions filed on or after 2020-09-16 (sunset), and independent of that, this patent claims a navigation/simulation method, not the data processing used in the practice, administration, or management of a financial product or service. § 18(d)(1).
• Inter Partes Review (§ 311): STILL STATUTORILY AVAILABLE, but only against the unexpired claims on § 102/§ 103 grounds built on patents and printed publications — § 311(b) excludes system art, on-sale art, public-use art, and § 101/§ 112 challenges. There is no age limit and no deadline unless and until a § 315 bar is triggered (see estoppel section).

Strategic summary

Which claims are CANCELED vs. SUSTAINED vs. UNTESTED. This is the cleanest part of the picture: none are canceled; none are sustained; all 20 are untested. Independent claim 1 retains the full breadth it had at issuance — the A/B simulation method reciting identification of OD pairs, a first (no-blockage) simulation set, a first summary reciting route completion time, a second (with-blockage) simulation set, a second summary reciting route completion time, and comparison of the two summary sets to assess the blockage's impact in the service area. Dependent claims 2–20 retain their fallback subject matter as well: blockage-as-forbidden-area (2), time-based blockages (3), historical-trip and cross-trip OD pairing (4, 5), high-volume and low-volume traffic time-stamping (6, 7), "fuzzed" pairs (8), per-pair start times (9), route-length summaries (10), maneuver-type counts and the unprotected-turn/lane-change/blocking-left-turn species (11–14), maneuver-count deltas (15), forward-progress metrics (16), a baseline that already includes a second blockage (17), and the impact outputs — count of affected pairs (18), total added time (19), total added distance (20).

Practically, that matters because a defendant cannot point to a canceled claim to end the dispute. Claim 1 is the live target, and it is broad — it recites comparison of simulation summaries, not any particular routing algorithm, cost function, or blockage model, and the only summary content the independent claim requires is "a length of time to complete that determined route." The added content in claims 10–16 (distance, maneuver counts, forward progress) does not narrow claim 1. If you are clearing or designing around this patent, the design-around sits squarely on the claim 1 steps themselves (see estoppel/availability below), not on the dependent-claim features.

Estoppel landscape. There is no § 315(e)(2) estoppel against anyone, because no IPR reached a final written decision. That cuts both ways: your prior-art grounds are all still available in district court, and the patent owner cannot argue that you already had your bite at the apple. Conversely, you cannot borrow someone else's FWD. If you do file an IPR now, the standard § 315 bars become the operative constraint:

• § 315(b): a one-year clock from service of a complaint alleging infringement of this patent. If Waymo (or any future assignee) served you more than a year ago, the IPR door is shut — check your service date now, because it is the single most common reason this option disappears.
• § 315(a)(1): filing a declaratory-judgment action of invalidity before the IPR bars institution.
• § 315(e)(1)/(e)(2): if you file and get an FWD, you (and privies/RPIs) are estopped on any ground raised or reasonably could have raised — but note the Federal Circuit's holding in Ingenico Inc. v. IOENGINE, LLC, 136 F.4th 1354 (Fed. Cir. 2025), that system prior art is not subject to IPR estoppel because it cannot be raised in an IPR at all. Any product/system-art theory you preserve stays alive even after an FWD.

Pattern signals. There is no petitioner pattern to report — no repeat filer, no defensive aggregator (no Unified Patents or RPX-style entity) in the chain, and no PTAB appeal history because there is no PTAB history. Two contextual factors are worth flagging, though:

1. The family is not a single patent. US 17/840,735 issued as US 11,788,854 B1 on 2023-10-17 as a continuation of the '587, sharing the same specification, title, inventor, and 2019-04-24 priority date. I have not read the '854 claims, so I will not characterize their scope — but a continuation of this specification filed in June 2022 almost certainly claims overlapping subject matter from the same disclosure. Any invalidity or design-around analysis that addresses only the '587 is incomplete: knocking out the '587 alone would not necessarily clear the family, and a defendant who files an IPR against the '587 without addressing the '854 risks paying for a proceeding that leaves an equivalent assertion on the table. My searches surfaced no PTAB activity against the '854 either.
2. The 2025–2026 institution climate materially disadvantages a would-be petitioner. Acting/confirmed Director practice changed sharply: the Biden-era Fintiv guidance was rescinded on 2025-02-28; a bifurcated institution regime separating discretionary denial from the merits followed on 2025-03-26; the "settled expectations" factor elevated patent age as a standalone denial basis; and Director Squires, confirmed 2025-09-18, asserted personal control over all IPR/PGR institution decisions in an 2025-10-17 memorandum. Commentators documented near-zero early institution rates under that regime, with many denials issued without reasoning — which also strips petitioners of a hook for rehearing. See Kilpatrick's 2025-11-24 practitioner alert (https://ktslaw.com/en/Pdf?page=%7B7CBEEF5C-EEC7-4353-A251-3501EDE5156D%7D&pdf=1) and Irwin IP's "PTAB Year in Review: A Tumultuous 2025" (https://irwinip.com/wp-content/uploads/2025/12/PTAB-2025-Review-Article-4926-2114-7521.1.pdf). Caveat: these sources describe a fast-moving policy environment, and I have no authenticated statistics specific to this patent's technology center.

One nuance to flag rather than resolve: the "settled expectations" factor has been discussed in terms of patents in force more than roughly six years. This patent issued 2022-07-19, so it is about 4 years 2 months past issuance as of 2026-09-29, but about 7 years 5 months past its 2019-04-24 filing/priority date. Whether the clock runs from issuance or from filing is not something the public guidance resolves, and I will not guess. Either way, the general 2025–2026 direction of travel is against institution.

Recommended next steps

1. Verify the null result directly — do not rely solely on this analysis. The authoritative checks, in priority order:

   • PTAB E2E / Patent Trial & Appeal Case Tracking System: https://ptab.uspto.gov/ — search by patent number 11,391,587; this is the system of record for AIA trials.
   • USPTO PTAB Decisions (FWDs, institution decisions, Director decisions): https://www.uspto.gov/patents/ptab/decisions
   • USPTO Open Data Portal (ODP) API — the same source as the structured block used here; query for proceedings on application 16/393,567 and 17/840,735.
   • CourtListener / RECAP for any CAFC appeal tied to a PTAB case number: https://www.courtlistener.com/
   • Mirror the same queries for US 11,788,854 B1 (application 17/840,735).

2. If you are a defendant and the patent has not been to the PTAB, say exactly that in your invalidity narrative. There is no FWD to cite and no claim to quote as canceled — do not represent otherwise. The correct defensive framing is: all 20 claims are as-granted and untested; claim 1 is broad and recites only a time-based summary; the surviving art grounds are fully available because no § 315(e)(2) estoppel has attached.

3. Calendar the § 315(b) date immediately. Compute one year from the date you were served with a complaint alleging infringement of the '587 (and separately the '854). If that date has passed, an IPR is unavailable as a matter of law and your invalidity case must be made in district court — where you have the advantage of the presumption of validity / clear-and-convincing standard and the full menu of § 101, § 112, and system-art grounds that § 311(b) excludes from IPR.

4. If you intend to file, move with an eye on the discretionary regime, not just the merits. Prepare the discretionary-denial record alongside the merits: document why the petition is not a follow-on, why there was no earlier knowledge of the patent, and stipulate carefully. Under the current framework, a merits-strong petition can still be denied at the discretionary gate before a panel ever reaches the art, and unexplained denials leave little room for rehearing.

5. Do not treat the absence of IPRs as evidence of validity. The absence of PTAB activity on a 2022-issued patent is weak evidence of anything. Well-asserted patents attract IPRs; this one appears not to have been asserted in any indexed litigation at all (see the litigation section of this analysis), which is the more likely explanation for the empty PTAB docket than any strength in the claims.

Caveats and confidence. The "no proceedings" conclusion rests on (a) the ODP-derived structured block, which is canonical for institution-level data, and (b) independent web searches that returned nothing. My tools do not provide authenticated, live write-access to PTAB E2E or the ODP API, so this is a well-corroborated absence of evidence, not a certified negative. Recent petitions (particularly those filed within the last few months) may not yet be indexed in the sources I can reach. Confirm via the links in step 1 before treating this as dispositive for a filing deadline or a litigation position. Everything else above — claim scope, statutory windows, and estoppel rules — is grounded in the patent text and the statutory provisions cited.

I'll reconstruct the assignment chain, starting with the source record and then checking the USPTO Assignment Center index and SEC/company-status sources.

Let me try to identify the specific recorded assignment (reel/frame and correspondent) and the inventor's affiliation.

Assignment & Ownership Reconstruction — US 11,391,587 B1

Scope note: This section builds on the verified bibliographic and litigation sections above. The patent text (patents.google.com/patent/US11391587B1) and its embedded "Legal Events / Reassignment" table are treated as authoritative. I attempted to reach the underlying Assignment Center record (reel/frame + correspondent) through public indexes; I could not retrieve the reel/frame number or the recording correspondent from the sources available to me, so those fields are marked as unretrieved rather than guessed. Nothing in this section contradicts the prior sections; the earlier "no litigation" finding is consistent with the NPE signals below.

Inventors

Unusual-pattern check — inventor departure: No evidence found of the inventor departing the assignee, and no successor inventor-assignment record exists. Ajay Joshi continues to appear as an inventor on later Waymo filings (e.g., grants in 2024 and 2025), which is the opposite of the "inventor leaves within 12 months, portfolio later fire-sold" precursor pattern. Signal: not present.

Original assignee

• Entity on the issued patent: Waymo LLC (Mountain View, CA; Delaware-organized LLC; listed as original and current assignee).
• Primary line of business: Commercial autonomous-vehicle ride-hailing and delivery (Waymo One / Waymo Via), an Alphabet Inc. subsidiary.
• Does it ship a product embodying the claims? Yes — materially. The claimed method is an internal engineering-operations method (A/B simulation of OD pairs with and without a map "blockage," summarized by time/distance/maneuver counts to prioritize blockage triage, scouting, and remapping). Waymo operates the production AV service the simulations model, and the specification's own hardware/software stack (routing system 170, storage system 450 software module) is Waymo's. This is a process patent practiced inside a large operating company, not a licensing vehicle.
• Current status: Operating. Waymo LLC remains a large-entity assignee of record (the 2026-03-09 maintenance-fee reminder in the patent's legal events shows the patent is being actively maintained in the name of a large entity). No bankruptcy, dissolution, or acquisition of the assignee occurred. Google's 2017 reorganization (Waymo Holding Inc. → Waymo LLC, USPTO assignment 42108/21, executed 2017-03-22, recorded 2017-03-28, correspondent Lerner David Littenberg Krumholz & Mentlik LLP) predates this 2019-filed application and does not appear in this patent's chain — a point worth stating explicitly, because that 2017 reorg assignment sweeps up hundreds of earlier Waymo/Google AV applications and could be misread as covering this one.

Assignment timeline

Chronological list of every recorded conveyance surfaced for US 11,391,587 / application 16/393,567:

• 2019-04-24 (filing) — no execution date exposed / recorded 2019-04-29 — Reel not retrieved (Google Patents exposes the reassignment event but not the reel/frame in the source text)

  • Conveyance: Assignment of Assignors' Interest
  • Assignor: JOSHI, AJAY (sole inventor)
  • Assignee: WAYMO LLC
  • Correspondent: Not retrieved. Public sources I could reach do not expose the correspondent of record for this specific 2019 recording. (For context only — not this patent — Waymo's recorded assignments elsewhere show correspondents including Botos Churchill IP Law LLP, New Providence NJ, on assignment 60151/161, and Lerner David Littenberg Krumholz & Mentlik LLP, Westfield NJ, on assignment 42108/21. Waymo clearly uses multiple outside IP firms, so neither can be attributed to this record without direct verification.)
  • Context: Standard employee/inventor-to-employer perfecting assignment, filed alongside the application (5 days from filing to recording). Not an acquisition, fire-sale, securitization, or reorg.

• 2022-06-15 (filing of continuation) — no separate assignment recorded — US 17/840,735 → US 11,788,854 B1 (granted 2023-10-17), same title, same sole inventor, filed by/assigned to Waymo LLC. Family ID 82385009.

  • Conveyance: none recorded (continuation filed directly by the assignee; no new inventor assignment surfaced).
  • Context: Internal continuation practice; no change in ownership.

• Post-issuance (2022-07-19 → present): No assignments recorded. Ownership has not moved since the 2019 inventor assignment.

Finding: The chain is a single-link chain to an operating company. The only recorded conveyance is the inventor→Waymo LLC assignment; there are no transfers to holding companies, no security agreements, no releases, and no changes of name recorded against this patent. This is itself the material finding — the original assignee still owns the patent.

Timeline diagram

timeline
    title Ownership of US 11391587
    2019 : Application filed by Waymo LLC
         : Inventor assignment recorded Apr 29
    2022 : Patent granted Jul 19
         : Continuation filed Jun 15
    2023 : Continuation granted Oct 17

NPE / troll-pattern signals

Verdict

Operating-company assertion — with one explicit caveat.

Justification: the only assignment of record is the inventor Ajay Joshi → Waymo LLC, recorded 2019-04-29 (five days after the 2019-04-24 filing), and nothing has been recorded in the seven years since; the patent was granted to Waymo LLC on 2022-07-19 and continued into US 11,788,854 B1 by the same assignee. The current assignee is a large entity that ships the AV ride-hailing service embodying the claimed blockage-impact simulation method, and every downstream transfer that would signal an NPE (shell-entity hop, cascading LLCs, post-grant security interest, pre-suit transfer) is affirmatively absent from the record. The caveat: the taxonomy's second criterion for this verdict — "is suing actual competitors" — is not met, because no assertion of this patent has been found; if the taxonomy requires an actual assertion, the strictly accurate label is "operating-company owned, never asserted," not "assertion."

Verify at the USPTO Assignment Center: https://assignmentcenter.uspto.gov/ (search patent number 11391587), mirrored index: https://assignment.uspto.gov/patent/index.html; patent record: https://patents.google.com/patent/US11391587B1.

Two diligence items to close before relying on this: (1) read the reel/frame and correspondent off the 2019-04-29 recording, since neither is exposed by the sources I could reach; and (2) confirm no unrecorded agreements exist (the recorded chain is silent, but recordation is not compulsory, and an unrecorded license or covenant would not appear at Assignment Center).

I'll verify the patent record and then research each cited reference.

Prior-Art Analysis — US Patent 11,391,587 B1

Step 1: USPTO record for patent number 11391587 (verified literally)

I searched for the exact string 11391587 / 11,391,587 and did not substitute any similar number. The record that returns is:

Source: https://patents.google.com/patent/US11391587B1/en

No confusion to flag: search hits containing the digits "11391587" that are not this patent (a Chinese trademark "瑞生活 RUILIFE," a German plumbing catalog part number, a Netlist "’087" patent) were discarded. The patent-appropriate hits all resolve to the Waymo document above.

Prior-art date. Because the application was filed on 2019-04-24 (post-AIA), the critical date is 2019-04-24. Every reference below predates it and is therefore available as prior art under 35 U.S.C. § 102(a)(1)/(a)(2).

Step 2: The five examiner-cited patent references + one NPL reference

The Google Patents "Citations (5)" and "Non-Patent Citations (1)" blocks for US 11,391,587 list exactly these six references. All are examiner-cited (formal citations of record), not applicant-submitted. I researched each one.

Reference 1 — US 8,706,394 B2

• Full citation: Trepagnier, P., Nagel, J., Dooner, M., Traft, N., Drakunov, S., Dewenter, M., Kinney, P., & Lee, A., Control and systems for autonomously driven vehicles, U.S. Patent 8,706,394 B2.
• Dates: Filed 2013-04-01 (Appl. 13/854,630); granted 2014-04-22; earliest priority 2008-10-24 (chain through 12/289,325 → 8,126,642). Prior publication US 2013/0274986 A1 (2013-10-17).
• Assignee: Gray & Company, Inc. (Metairie, LA).
• Description: An autonomous-vehicle control platform. It discloses: a processor receiving position signals and generating an updated travel path; LIDAR/laser obstacle detection; a route-planning algorithm that determines a route between two waypoints based on recorded traffic patterns, recorded times to travel between waypoints, history of congestion areas, and real-time congestion reports (see its claims 7–8); an A* heuristic search using time-based waypoint cost metrics; and a variable-structure observer that predicts obstacle positions by running a model forward in time (i.e., a simulation) when sensor data is lost.
• Claims of ’587 it potentially touches: The simulation + routing concepts recited in claims 1 and 17, and the time-based routing aspect underlying claims 6, 7, 9, 19, 20.
• § 102 assessment: Not anticipatory. ’394 discloses a single routing computation and an obstacle-prediction simulation, but it does not disclose the claim-1 core: (a) a plurality of starting/destination pairs, (b) running a first simulation without a designated blockage and a second simulation with it, (c) summaries capturing route completion time for each set, or (d) comparing the two summary sets to score the blockage's impact. At most this is a § 103 secondary reference.

Reference 2 — US 2011/0098915 A1

• Full citation: Disatnik, I., Device, system, and method of dynamic route guidance, U.S. Patent Application Publication 2011/0098915 A1.

• Dates: Published 2011-04-28; priority/filing 2009-10-28.

• Description: Dynamic re-routing for a navigation device (PND, smartphone, in-vehicle unit). The routing module continuously recomputes an optimal route and replaces it when triggered by real-time updates. Critically, its paragraph [0024] expressly lists triggers including:

  "data indicating a blockage in a road segment of the optimal route; data indicating removal of a blockage in a road segment of an alternate route …"

  It also compares replacement routes on estimated travel time and number of road segments (paras. [0021]–[0022]).

• Claims of ’587 it potentially touches: The "blockage" concept of claims 1, 2; the comparison metrics of claims 1, 10, 19, 20; and time-component aspects of claim 3.

• § 102 assessment: Not anticipatory, but this is the closest cited reference on the blockage limitation. It discloses a blockage affecting route choice and a time/distance comparison between a prior route and a replacement route. It does not disclose: running pre-blockage and post-blockage simulation sets across a plurality of OD pairs, generating per-route summaries, or comparing summary sets to assess service-level impact. It operates on one device/one route at a time, so it defeats the "plurality of pairs / service-area impact" limitations. § 103 reference only.

Reference 3 — US 2014/0278052 A1 (issued as US 9,964,414 B2)

• Full citation: Slavin, H., Yang, Q., Morgan, D., Rabinowicz, A., Brandon, J., & Balakrishna, R., Lane-level vehicle navigation for vehicle routing and traffic management, U.S. Patent Application Publication 2014/0278052 A1 (later granted as U.S. Patent 9,964,414 B2 on 2018-05-08).
• Dates: Published 2014-09-18; filed 2014-03-14 (Appl. 14/210,673); provisional 61/789,019 filed 2013-03-15.
• Assignee: Caliper Corporation (Newton, MA).
• Description: A lane-level routing and navigation apparatus with (a) a simulation module performing microsimulation of individual vehicles in a traffic stream, (b) a link-level optimizer that computes candidate paths ("hammocks") from an origin to a destination, and (c) a lane-level optimizer that evaluates those candidates and outputs recommended lane-level maneuvers. It expressly discusses evaluating multiple route alternatives simultaneously and simulating "a work zone [that] has blocked the right lane," including where/how a driver should maneuver around the blockage. It describes regional-scale microsimulation (Phoenix, 500 sq mi) running faster than real time.
• Claims of ’587 it potentially touches: The simulation + routing of OD pairs of claim 1; the maneuver type/count summaries of claims 11–15 (it explicitly models lane changes and turning maneuvers); and the distance/time comparison of claims 10, 19, 20.
• § 102 assessment: Not anticipatory, but this is the cited reference that comes closest to the simulation-based route-evaluation architecture. However, it simulates to produce route guidance and to manage traffic, not to run an A/B comparison isolating the incremental impact of a given blockage, and it does not disclose generating a "summary" per route with completion time for two parallel simulation sets and comparing the summary sets. § 103 reference only.

Reference 4 — US 9,494,937 B2

• Full citation: Method and system for drone deliveries to vehicles in route, U.S. Patent 9,494,937 B2.
• Dates: Filed 2014-06-20; granted/issued 2016-11-15; published as US 2015/0370251 A1 (2015-12-24).
• Assignee: Verizon Telematics Inc. (later HTI, IP, L.L.C.).
• Description: A centralized drone-routing system. A server exchanges vehicle and drone information; a computing device identifies multiple pickup locations, scores them by travel-time/wait-time priority score, and updates the drone route. It discusses dynamic routing problems ("information evolves over time"), cost functions with constraints, and heuristics-based routing for multiple vehicles.
• Claims of ’587 it potentially touches: Weakly, the route-determination and time-cost comparison aspects of claims 1, 19.
• § 102 assessment: Not anticipatory — essentially non-analogous/superfluous. It is a drone-delivery logistical routing reference with no OD-pair population, no blockage construct, no baseline-vs-blockage simulation pairing, and no service-impact summary generation. It appears to have been cited as general art for route determination with time-cost optimization. Worth noting for completeness, but it is the weakest of the five.

Reference 5 — US 2018/0004214 A1

• Full citation: Vehicle remote assistance mode, U.S. Patent Application Publication 2018/0004214 A1.
• Dates: Published 2018-01-04; listed priority 2017-09-15.
• Assignee: GM Global Technology Operations LLC.
• Description: Teleoperation/remote-assistance for autonomous vehicles — a remote operator or assistance center intervenes when the autonomous vehicle cannot resolve a situation.
• Claims of ’587 it potentially touches: Marginal; possibly the general "service operations for autonomous vehicles" backdrop of claim 1.
• § 102 assessment: Not anticipatory. Disclosure caveat: I was unable to open the full text of this specific publication within the search budget (my targeted query for it returned the step-limit message). Based on the title, assignee, and the context in which the examiner placed it, it appears to have been cited for background on autonomous-vehicle service operations, not for any element of the claim-1 comparison method. I flag this as the one reference whose internals I could not independently confirm here — treat its characterization as low-confidence and verify against the actual document.

Reference 6 — Non-Patent Literature (SUMO)

• Full citation: Behrisch, M., Bieker, L., Erdmann, J., & Krajzewicz, D., SUMO — Simulation of Urban Mobility: An Overview, Institute of Transportation Systems / DLR, 2011 (presented at SIMUL 2011, Barcelona).
• Dates: 2011 (pre-AIA; well before the 2019-04-24 critical date).
• Description: SUMO is an open-source, microscopic, multi-modal traffic simulation package. It simulates individual vehicles moving through a road network (an OD/route-based traffic model), supports time-dependent demand and network definitions, and is used for evaluating transportation scenarios such as road closures/network changes.
• Claims of ’587 it potentially touches: The general "running a plurality of simulations … determining a route for each pair" concept of claim 1.
• § 102 assessment: Not anticipatory, but it is the strongest § 102 "tool" among the six. SUMO discloses the simulation infrastructure for the recited steps — simulating many vehicles/routes over a road network and evaluating a scenario with a network modification. It does not, however, disclose the specific claim-1 method as an integrated process: identifying a plurality of starting-and-destination location pairs within a service area drawn from historical trips of an autonomous vehicle service, running paired first/second simulation sets differing only by the designated blockage, and determining and comparing "summaries" including route completion time to assess the blockage's impact. As a printed publication it can only anticipate if it disclosed every element arranged as claimed; it does not. Combined with Reference 3, it is a potent § 103 combination.

Step 3: Bottom-line § 102 conclusion

No cited reference anticipates any claim of US 11,391,587 under 35 U.S.C. § 102, individually or in the record as presented.

On anticipatory effect specifically: None of the five patents/applications or the SUMO paper discloses, in a single reference, the claim-1 combination of:

1. a plurality of starting/destination pairs within a service area,
2. a first simulation set routing every pair without a first blockage,
3. a first summary per route including completion time,
4. a second simulation set routing the same pairs with that blockage,
5. a second summary per route including completion time, and
6. comparing the first and second summary sets to assess the blockage's impact.

The nearest references are Disatnik (’915) for the "blockage" term and time/distance comparison, Caliper (’052) for simulation-based route evaluation across OD paths and maneuver modeling, and SUMO for the simulation substrate. Each covers a fragment; none covers the A/B-per-OD-pair summary comparison that constitutes the novelty.

Where the real § 103 exposure lies: A combination of Caliper ’052 + SUMO, optionally with Disatnik ’915 for the blockage/time-comparison teaching, would be the examiner's natural obviousness case. The dependent claims most exposed are:

• claim 2/3 (blockage = forbidden area; time-based) — taught/obvious from Disatnik ’915 [0024];
• claim 10 / 19 / 20 (length, added time, added distance) — taught by Disatnik ’915 [0021]–[0022] and Caliper ’052;
• claims 11–14 (maneuver counts: unprotected turn, lane change, blocking left turn) — actually the most defensible dependents, since none of the cited art generates per-route maneuver-count summaries the way ’587 does; Caliper models lane changes/turns but not as a comparison metric;
• claim 16 (forward-progress metric) — not shown in any cited reference; strong dependent.

Claims drafting note worth flagging: the ’587 record's later continuation US 11,788,854 B1 (from US 17/840,735) issued with the same title, so the family's claim scope was pursued twice — the cited-art set above is the art of record for the ’587 grant.

Confidence and caveats

• High confidence on the identity and bibliographic data of US 11,391,587 and on references 1–4 and 6, all of which I retrieved directly.
• Lower confidence on reference 5 (US 2018/0004214 A1): I could not retrieve its full text this session; its description is inferred from title/assignee/examiner placement. Verify before relying on it.
• No result auto-corrected: irrelevant "11391587" hits were excluded rather than substituted, per instruction.
• "Potentially anticipates" is used strictly: I have marked every reference as non-anticipatory because, on the record available, none discloses all claim-1 elements. If you want, I can produce a claim-by-claim § 103 obviousness chart (element → reference → motivation to combine) for the Caliper + SUMO + Disatnik combination.

I'll research the cited prior art references to build an accurate § 103 analysis.

Let me research the remaining references.

Obviousness Analysis — US 11,391,587 under 35 U.S.C. § 103

Note on scope and sourcing. Per the task, I am using only the references appearing in the "Prior Art" area of the patent page: the five examiner-cited documents (US 8,706,394; US 2011/0098915; US 2014/0278052; US 9,494,937; US 2018/0004214), the single non-patent citation (Behrisch et al., SUMO, 2011), and the one family-cited document (Nvidia US 11,474,519). I did not retrieve the SUMO paper text or the Nvidia patent text (my searches hit the step limit), so those two are characterized from title/general knowledge and flagged as unverified. The four "Cited By" items on the page are all post-2019 and therefore are not prior art — I do not use them below.

1. The legal framework applied

A § 103 challenge requires the fact-finder to resolve the Graham v. John Deere, 383 U.S. 1 (1966), factors: (1) scope and content of the prior art; (2) the level of ordinary skill in the art; (3) the differences between the prior art and the claims; and (4) objective indicia of non-obviousness. Under KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), the claims must be assessed with "an expansive and flexible approach," and a combination is obvious where the improvement is a predictable use of prior-art elements according to known methods, where a known technique is applied to a known device ready for improvement to yield predictable results, or where the combination reflects "obvious to try." The patent's subject matter — A/B simulation of a routing engine with and without a network restriction — sits squarely in the predictable, engineering-heavy side of that spectrum.

I apply a lower-than-average skill level argument cautiously; a realistic PHOSITA here is an engineer/scientist with a graduate degree or 2–5 years' experience in traffic microsimulation, intelligent transportation systems, and AV routing/mapping.

2. What each reference teaches vs. claim 1

Claim 1 is a six-step method. Mapping:

Conclusion as to claim 1: every element is disclosed, but scattered across references. No single reference appears to disclose the whole claim (I found none that explicitly runs the same O-D set twice — once baseline, once with a blockage — and compares the two summaries). So the § 103 question is whether the combination is proper.

3. Proposed grounds of rejection

Ground 1 (primary): Caliper + Disatnik + GM, optionally with SUMO

• Caliper supplies the simulation-to-route architecture, the O-D/candidate-path framework, time-based route metrics, and the explicit traffic-planning use case (Caliper discloses using microsimulation for planning, and even poses a work-zone lane blockage question).
• Disatnik supplies the blockage as a routing input and, critically, the before/after time comparison that is the heart of claim 1's "comparing" step — including comparison of a previously-calculated optimal route against a recalculated alternate route when a blockage appears or is removed.
• GM US 2018/0004214 supplies the "blockage = road segment the vehicle is not permitted to drive through" implementation (blacklisted road segments excluded from the road-map graph, with route recalculation) — which maps directly onto granted claim 2 (and the specification's definition of "blockage").
• SUMO supplies an independent, well-known, open-source traffic-microsimulation engine that can be pointed at a road network and trip demand, i.e., the "simulation" substrate.

Claim 1 differs from this combination only in aggregating per-trip before/after comparisons across the plurality of O-D pairs. That is not an inventive leap; it is the routine application of statistical/analytical aggregation to a known A/B workflow, i.e., a "predictable variation" under KSR. Indeed, none of the dependent claims 17–20 even purport to claim a specific algorithm — they recite summing/counting results.

Ground 2: Gray + Disatnik + GM (AV-specific primary)

Gray (US 8,706,394) is an autonomous-vehicle control/routing patent, which better matches the "autonomous vehicle transportation service" and "service area" framing of the preamble and narrows the field-of-use gap. Gray discloses:

• an onboard route-planning module computing a route between waypoints;
• an A* search whose cost g(x) is observed average travel time and whose heuristic h(x) is straight-line distance to the goal;
• a learning component that records travel times, marks "new areas of congestion," and discredits old observations;
• generating multiple candidate paths and selecting by score ("shortest time, shortest distance").

Combine with Disatnik (blockage + before/after time comparison) and GM (blacklisting/blockage definition). Gray's use of a goal-distance ("as the crow flies") heuristic is also highly relevant to dependent claim 16 (forward-progress metric).

Ground 3 (broader): Any of the above + SUMO as the simulation engine, with Verizon US 9,494,937 as a secondary teaching that routes between locations are selected/updated using travel-time and wait-time priority scores (i.e., time-based route scoring across a network). Verizon is the weakest reference here and would be cumulative at best.

4. Why a PHOSITA would have been motivated to combine (the KSR rationales)

1. Same field, same problem. All five references are classified in navigation/routing/traffic-control art (G01C21/34 et seq., G08G1/x). The examiner's citation of all five confirms they are analogous art. Disatnik, GM, and Caliper all address route computation in the presence of disruptions (construction, closed roads, blocked lanes, traffic jams) — the exact problem the patent addresses.
2. Known technique (simulation) applied to a known device (a routing engine) to yield predictable results. Traffic microsimulation was mature and standardized by 2011 (SUMO); using it to evaluate a routing engine's output before fielding it is a conventional validation technique. See KSR ("a technique [that] has been used to improve one device, and a person of ordinary skill … would recognize that it would improve similar devices in the same way").
3. Design rationale / predictable variation. Once Disatnik teaches comparing an alternate route's travel time against the original route's travel time when a blockage is added, running that comparison over a set of trips and summing the deltas is an obvious aggregation — a mere change in scale, not in kind.
4. Field-of-use limitation is not patentable weight. The preamble's "autonomous vehicle transportation service" is an intended-use/field limitation. Under KSR, "the mere substitution of one [known element] for another known element" or the recitation of a particular field of use does not confer patentability where the body of the claim is otherwise met. GM and Gray both supply the AV context if needed.
5. "Obvious to try" for the blockage design space. Claim 3's time-based blockage and claims 6/7's high-/low-volume traffic assignment are design choices directly suggested by Disatnik's disclosure of using "average driving speeds per specific time periods, day of week and/or time of day" and by SUMO's time-varying demand modeling.

5. Dependent-claim assessment (2–20)

Likely obvious (weak):

• 2 — Blockage as a no-drive area: GM's blacklisted road segments excluded from the map graph. Very strong rejection.
• 4, 5, 8, 9 — Sourcing O-D pairs from historical trips, pairing start-of-one with destination-of-another, "using a prior starting/destination pair," and assigning a start time: the spec itself notes historical trips are stored; Caliper and SUMO both require trip demand/O-D matrices; Disatnik uses historical speeds by time of day. Note: the granted claim 8 is far narrower than the specification's "S2 cell" fuzzing embodiment — granted claim 8 recites only "selecting a starting location and a destination location using a prior starting and destination location pair," which is broad and easily met.
• 10, 11, 13, 15 — Route length/distance and maneuver counts, including lane-change counts. Caliper is directly on point: its lane-level optimizer "determines recommended lane-level maneuvers along the candidate paths," and it discusses lane changes and work zones expressly. Gray scores candidate paths by "shortest distance."
• 17 — Baseline includes a second blockage: an obvious alternative; the spec itself concedes long-term blockages are commonly present in the baseline.
• 18, 19, 20 — Counting impacted pairs, total added time, total added distance: pure aggregation of the claim-1 comparison; Disatnik compares travel time and number of road segments; none of these recites a non-routine algorithm.

Stronger (candidates to survive, absent extra art):

• 3 — Time-based blockage ("effective only during a particular period"). Disatnik discloses time-of-day-dependent speeds, but I did not confirm a reference that teaches a blockage with an effectiveness window (e.g., "school letting out"). The specification's own example is school-dismissal. If no reference teaches a temporally-qualified restriction, this claim is the best non-obviousness candidate — though it is a thin, easily-designed-around limitation.
• 14 — "Blocking left turn" as a maneuver type (no dedicated left-turn lane; vehicle must hold traffic). This is an AV-specific maneuver taxonomy. I found no cited reference that enumerates this maneuver class explicitly. Like claim 3, this is narrow and species-like; a secondary reference on unprotected/blocking turns would likely close the gap.
• 16 — Forward-progress metric ("how often the route progresses toward the destination"). Gray's A* heuristic uses "straight-line distance to the goal checkpoint" — conceptually the same "as-the-crow-flies progress" notion — but Gray discloses it as a search heuristic, not as a per-route reported metric counting frequency of forward progress. This is the single dependent claim where the mapping is closest without being exact, and the strongest realistic non-obviousness argument. It would still be vulnerable to a "predictable metric of route quality" rationale, especially given Gray's explicit discussion of scoring paths.

No dependent claim recites a non-conventional technical implementation (no model architecture, no training method, no network topology). Claims 18–20 in particular are drafted as results ("determining a number of … pairs impacted"), which invites a § 103 rejection as mere aggregation and a § 101 rejection as a mental-process/mathematical-concept risk.

6. Anticipated rebuttals and their strength

• "References were before the examiner, yet the patent issued." This is the most serious obstacle to any § 103 attack. The five documents are listed as examiner-cited art, and the patent granted 2022-07-19. That creates at least a prima facie inference that the examiner already considered the combination and found something patentable (most likely the specific A/B-baseline-plus-blockage-over-many-O-D-pairs workflow). A petitioner would want new art the examiner did not have (e.g., a domestic AV-fleet simulation patent, the Nvidia family, or a SUMO-based planning paper) to avoid the "already considered" problem. I do not have the file wrapper, so I cannot verify whether the examiner articulated a specific reason for allowance.
• Argument from the KSR dissent / "bodily incorporation" and "teaching away." Patentee might argue Disatnik is about real-time single-vehicle rerouting and Caliper about lane-level guidance, so combining them is hindsight. This is weak: both are route-computation systems in the same art, and KSR expressly rejects a rigid "teaching, suggestion, or motivation" test; the motivation here is the shared goal of maintaining routing quality when the network changes.
• Unexpected results / secondary considerations. The prior sections of this analysis found no litigation and no objective indicia in the record (no commercial-success evidence with a nexus, no long-felt need evidence). The specification presents no comparative data showing unexpected results. So this prong currently favors the challenger — again subject to the "examiner already considered" caveat.
• "Assessing impact ... in the service area" as a mental/organizational step. This informs § 101 more than § 103 (the workflow of comparing results is arguably a mental process/collection of data), but a § 103 challenger can note that the claimed advances over Disatnik's per-route comparison are aggregation and reporting, which KSR treats as predictable.

7. Bottom line

• Claim 1 is vulnerable. No single reference anticipates it, but Caliper (simulation + O-D candidate paths + time metrics) in view of Disatnik (blockage-triggered reroute + before/after travel-time comparison) and GM (blockage = blacklisted no-drive road segments + route recalculation), layered on SUMO as the simulation substrate, renders the six-step method an obvious combination under KSR's "predictable use of prior-art elements according to known methods" and "known technique applied to a known device." The only meaningful gap — aggregating across the plurality of O-D pairs — is a routine, predictable variation.
• Most dependents fall with claim 1 (2, 4–13, 15, 17–20), several quite cleanly (claim 2 via GM; claims 11/13 via Caliper).
• Claims 3, 14, and especially 16 are the best candidates to survive on the cited art alone, because I could not map a temporally-qualified blockage, a blocking-left-turn maneuver taxonomy, or a reported frequency-of-forward-progress metric onto the references. Each, however, is narrow and likely readily invalidated with one additional reference.
• Critical caveat: these are the very references the examiner had, and the patent issued anyway. An obviousness case built only on this set must rebut the inference that the examiner already rejected the combination. My conclusion is a technical vulnerability assessment, not a legal conclusion of invalidity.

Confidence and limitations

• High confidence on the contents of US 8,706,394 (Gray), US 2011/0098915 (Disatnik), US 2014/0278052 (Caliper), and US 2018/0004214 (GM) — I retrieved their abstracts/key passages and quoted them above.
• Medium confidence on US 9,494,937 (Verizon) — retrieved abstract/key portions; it is cumulative art.
• Low/unverified on the SUMO (Behrisch 2011) paper and Nvidia US 11,474,519 B2 — I could not retrieve their text before hitting the search step limit. I treat SUMO as the standard open-source microscopic traffic simulator it is generally known to be, and I flag Nvidia US 11,474,519 (priority 2018-02-26) as potential additional 102(a)(2)/103 art whose substantive disclosure I have not verified. Do not rely on the SUMO or Nvidia characterizations without confirming the source documents.
• I was unable to inspect the USPTO file wrapper / office actions for this patent, so the "examiner already considered this combination" inference rests on the Google Patents citation labels alone.