US 10655981

I'll search for this specific patent number across the sources you mentioned.

Let me check for any CAFC litigation or USPTO proceedings involving this patent.

One more targeted check on dockets and any post-grant proceedings.

Summary — US Patent No. 10,655,981 B2

Important sourcing note: I searched Google Patents (full text supplied), Espacenet, Justia, and FreePatentsOnline. I did not get direct access to USPTO Patent Public Search / PatentCenter, PACER, or the CAFC docket system; my search for litigation was a general web search. See the litigation note below.

Bibliographic data (consistent across all sources found):

Abstract (as published):

"A method is provided for updating parking area information in a navigation system for a vehicle. The method includes the following steps: detecting a parking area by way of data received at least one device on at least one vehicle; determining information relating to the parking area which at least partially relates to characterizing characteristics of the parking area; analyzing the determined information in relation to an adequate characterization of the parking area; comparing the determined information with data of at least one database; and updating the database in the event of a deviation between the determined information and the data of the database."

---

Plain-language overview of the independent claims

Claim 1 — Method. A vehicle-based method for keeping a parking database current, with six steps:

1. Recognize a parking space from sensor data captured by sensor(s) on one or more vehicles, where the sensor data reflects the vehicle's surroundings.
2. Determine parking-space information at least in part from that captured sensor data, where the information characterizes properties of the parking space — expressly "independent of whether the parking space is occupied."
3. Analyze the determined information to decide whether it sufficiently characterizes the parking space (i.e., is there enough to describe it?).
4. Compare the determined information against data in at least one database.
5. Update the database if the determined information deviates from the stored data.
6. Navigate vehicles to parking spaces, each via a respective navigation system, using the updated database.

The gist: rather than only recording where a parking space is, the system uses on-vehicle sensors (cameras, ultrasonic, infrared) to capture what kind of parking space it is (e.g., residents'-only, time-restricted), checks completeness, compares to a map/POI database, and updates it when reality differs — then feeds the corrected data back into navigation.

Claim 10 — Navigation system. A system comprising: a vehicle-side recognition unit that recognizes a parking space from on-vehicle sensor data reflecting surroundings; a determination unit that derives parking-space information at least partly from that sensor data (again characterizing properties independent of occupancy); an analysis unit for sufficient-characterization analysis; a comparison unit for comparing against database data; and an update unit that updates the database — the system then uses the updated database to provide vehicle navigation to parking spaces.

Claim 12 — Computer program product. A non-transitory computer-readable medium storing program code segments that, when executed, carry out the same sequence of acts as claim 1 (recognize → determine → analyze → compare → update on deviation → navigate using the updated database).

Notable dependent claims: Claim 2 — information determined using user inputs; Claims 3/4 — parking space recognized only when a parking process is recognized; Claim 5 — a prompt is output to the user when the analysis result is negative (insufficient characterization); Claims 6/7 — the prompt is produced by a vehicle component or a central/backend component; Claim 8 — the database comparison occurs before the sufficiency analysis; Claim 9 — the recognition data is captured by at least one vehicle sensor.

---

Litigation / CAFC 2026 dockets

I found no evidence of any CAFC appeal, district court case, ITC investigation, IPR, PGR, or other post-grant proceeding involving US 10,655,981. My searches for the patent number combined with "CAFC," "Federal Circuit," "litigation," "IPR," and "docket" returned only unrelated matters (e.g., patents with superficially similar numbers such as 10,155,981 in 10x Genomics v. Parse Biosciences, and 10,526,617 in Regenxbio v. Sarepta). Because I could not query PACER or the CAFC docket database directly, treat this as "no publicly indexed proceedings found," not as a definitive negative. If you need certainty, the authoritative checks would be the CAFC docket (cafc.uscourts.gov) and PACER/RECAP by patent number.

---

Points of uncertainty / flags

• Prosecution history inference (not certain): The granted claim 1's final two limitations — "independent of whether the parking space is occupied" and "navigating one or more vehicles to the parking spaces … using the updated database" — do not appear in the specification's own restatement of the method, which suggests they were added during prosecution. The file history shows a non-final rejection (Mar. 19, 2019), a response (Jul. 26, 2019), a final rejection (Sep. 10, 2019), then a notice of allowance (Jan. 13, 2020). I have not reviewed the actual office actions or amendments, so I cannot state what arguments or amendments were made.
• Expiration date: Google Patents lists "adjusted expiration 2036-05-16" (a PTA-based figure). Separately, the corresponding PCT-based 20-year term from the Feb. 1, 2016 PCT filing would be Feb. 1, 2036, which is consistent; but I have not verified the PTA certificate.
• No reissue, reexamination, or terminal disclaimer was found, but this is based on the sources searched, not an official USPTO register check.
• Bibliographic details (inventor cities, examiner, classifications) come from Justia/Google Patents aggregation rather than a direct USPTO record.

I'll run targeted searches for litigation involving this specific patent.

No relevant hits yet — the results returned are for the different patent 10,155,981 (10x Genomics). Let me run more targeted searches.

Let me try a few more angles — the patent title in litigation, and a direct portal/patent-page check.

Litigation search results for US Patent No. 10,655,981

Bottom line

I found no litigation — and no post-grant proceeding — involving U.S. Patent No. 10,655,981. Searching the patent number, its title, and its assignee in combination produced no district court case, no CAFC appeal, no ITC investigation, no IPR/PGR, and no reexamination or reissue.

I want to be precise about the strength of that conclusion: it is "nothing found in the sources I could reach," not a registry-verified negative. I could not query PACER, the CAFC docket, or the Unified Patents litigation caselist directly, so a low-profile case (e.g., a DJ action, a case where the patent is one of several in a large portfolio assertion, or a recently filed case not yet indexed) could exist without appearing in my results.

---

What I searched, and what came back

Sources reachable in this session included Google Patents, FreePatentsOnline, CourtListener/RECAP pages surfaced via search, Unified Patents portal pages and IPR filings surfaced via search, and various briefs/pleadings. My searches for the Unified Patents caselist (https://portal.unifiedpatents.com/litigation/caselist) did not return a case-list entry keyed to 10,655,981.

---

Important disambiguation — near-miss numbers that are not this patent

The searches repeatedly surfaced a different patent with a similar number, and I flag it explicitly so it is not mistaken for a hit:

• US 10,155,981 — the "’981 patent" in 10x Genomics, Inc. v. Parse Biosciences, Inc., C.A. No. 1:22-cv-01117 (D. Del., filed Aug. 24, 2022, Judge Maryellen Noreika). That case asserted U.S. Patents 10,155,981; 10,697,013; 10,240,197; 10,150,995; 10,619,207; and 10,738,357 (single-cell sequencing). It was followed by IPRs IPR2023-00876 (’981), IPR2023-00955 (’197), and IPR2023-00958 (’013), with the Federal Circuit affirming the Board's obviousness holdings in 10x Genomics, Inc. v. Parse Biosciences, Inc. (nonprecedential, opinion dated Aug. 19, 2026). This is a different patent number, a different technology, and different parties — it has nothing to do with BMW or parking.
• Also returned but irrelevant: U.S. Patent 8,218,481 (IPR2016-00981, Apple v. Evolved Wireless) and unrelated "’981 patents" in Stryker v. Orthophoenix and elsewhere.

There is also an unrelated "581" matter — BMW AG v. Onesta IP, LLC, Case No. 6:25-cv-00581 (W.D. Tex.), with a related Federal Circuit appeal and filings in March 2026 — but that case does not involve 10,655,981, and the "581" is a docket number, not this patent.

---

Why "no litigation" is plausible here

Consistent with the earlier summary section, 10,655,981 is held by Bayerische Motoren Werke Aktiengesellschaft (BMW) as both original and current assignee. BMW is an operating company that uses its patents primarily defensively; its patents are far more commonly cited as prior art and used in prosecution than asserted offensively. The patent is also relatively recent (issued May 19, 2020) and in an area (crowd-sourced parking-data updating) where BMW itself is not a licensor in the way a PAE would be. None of that is proof of a negative, but it is consistent with what the searches returned.

---

Where to verify authoritatively

Because I could not reach these systems directly, if you need a definitive answer, check:

1. Unified Patents litigation caselist — https://portal.unifiedpatents.com/litigation/caselist (filter by patent number).
2. CourtListener / RECAP — https://www.courtlistener.com (search dockets for "10,655,981").
3. PACER Case Locator — https://pcl.uscourts.gov (search by patent number across all districts).
4. CAFC docket / opinions — https://cafc.uscourts.gov and the Fed Circuit Blog (fedcircuitblog.com) for any appeal.
5. PTAB — https://ptacts.uspto.gov (search for IPR/PGR/DER petitions naming the patent).
6. USPTO Patent Public Search / PatentCenter for any reissue, reexamination, or adverse-assignment record.
7. Docket Alarm (docketalarm.com) and RPX Insight (insight.rpxcorp.com) as commercial cross-checks.

---

Disclosure

If your question also covers licensing or non-litigated assertions (e.g., a demand letter, a standards-pool commitment, or a covenant), I found no such record either — but that is outside what the searches above can conclusively establish, and I do not want to represent a negative finding there as verified.

Summary of the answer: No known litigation involving US Patent 10,655,981 in any forum I could reach; no case number, parties, jurisdiction, or status to report because no case was found.

I'll verify the canonical "no proceedings" signal against live sources before writing the analysis.

Let me run a couple of targeted confirmations — a direct patent-number search and a check for any litigation/PTAB pairings.

Let me run two final verification searches — one against the patent-analytics portals and one for any district-court assertion that might have spawned an IPR.

PTAB / AIA Trial Proceedings — US 10,655,981 B2

Verification basis: The structured "PTAB proceedings on file" block supplied with this task (USPTO Open Data Portal ingest) is the canonical source and returns zero AIA trial proceedings. I independently ran six targeted live searches (Google Patents, Espacenet, FreePatentsOnline, Justia, PTAB petition documents via ptacts.uspto.gov, and docket-analytics portals) to look for a proceeding the ODP had not yet indexed. No IPR, PGR, or CBM naming US 10,655,981 as the challenged patent was found.

---

Proceedings overview

Total AIA trial proceedings on file for US 10,655,981: 0. Breakdown by status: 0 active, 0 with claims invalidated, 0 with claims sustained, 0 settled, 0 with institution denied. There is no petitioner, no panel, no institution decision, and no Final Written Decision because no petition has ever been filed.

The bottom-line defensive posture: this patent is PTAB-untested, not PTAB-hardened. All 12 claims stand exactly as issued on 2020-05-19, with the full presumption of validity intact. That cuts both ways for a defendant:

• In your favor: no petitioner has ever been estopped, no claim has been canceled or narrowed by the Board, and there is no adverse Board construction of any term. You carry zero § 315(e)(2) estoppel baggage and can raise any § 102/§ 103 ground you can support — including art the examiner already considered — in district court or in a first-filed IPR.
• Against you: there is also no roadmap. Nobody has done the work of mapping claim 1's added limitations ("independent of whether the parking space is occupied" and the "navigating … using the updated database" step) onto the crowded BMW-parking / connected-vehicle prior art. You would be the first petitioner, absorbing all the § 325(d) and § 112 risk.

Important framing correction: the two illustrative postures in the task prompt ("the patent has survived two IPRs and is hardened" / "claims 1-5 have been canceled — the troll has no case") both presume proceedings exist. Neither applies here. If anyone hands you a "claims 1–5 are canceled" narrative for this patent number, that is a false claim and should be flagged as fabricated.

---

Proceedings on file

None. No proceeding entries to report, and per the task constraints I will not invent proceeding numbers, panels, or dispositions.

Confusion trap — three different patents your search will surface (all NOT this patent)

Because 10,655,981 is numerically adjacent to heavily-litigated patents, any patent-number search here produces loud false positives. Do not let these contaminate your analysis:

The BMW docket also contains genuine PTAB activity on other patents in the navigation space (e.g., NorthStar Systems LLC v. Bayerische Motoren Werke AG, No. 2:22-cv-00496 (E.D. Tex.), and BMW of North America LLC v. NorthStar Systems LLC, No. 6:23-cv-00456 (W.D. Tex.), which produced an IPR petition in which BMW is petitioner and an RPI). None of those challenges US 10,655,981.

---

Strategic summary

Claim status: everything is live and untested

There is no inter partes review certificate, no PGR, no CBM, and — based on the sources I searched — no reissue or ex parte reexamination. Unlike a patent that has been through the Board, no claim of US 10,655,981 has been narrowed, disclaimed, or canceled since issuance.

Estoppel landscape: a blank slate

Because no IPR/PGR has ever been instituted, § 315(e)(2) estoppel attaches to nobody. There is no petitioner, no real party in interest, and no privy carrying forward a bar into litigation. Practical consequences for a defendant asserting invalidity today:

• No ground is foreclosed by statute. You may raise any § 102 or § 103 combination, before the Board or the court, on any reference — including the art the examiner expressly considered.
• The § 325(d) discretion risk is real, though. The Office is now more willing to deny petitions built on "the same or substantially the same art or arguments previously presented to the Office" (Advanced Bionics). The heavy overlap with the cited references below makes a fresh-art-first petition strategy advisable.
• Watch the § 315(b) clock. If you have been served with a complaint alleging infringement of this patent, a petition must be filed within one year of service. Missing it permanently bars IPR. This is the single most time-critical item on the list.
• Watch for a Fintiv/Sotera issue. If parallel litigation exists, a Sotera-style stipulation (agreeing not to re-assert in court the grounds raised in the petition) materially improves the odds of institution.

Pattern signals

• No serial petitioner. No petitioner has filed against this patent once, let alone repeatedly.
• No patent-owner appellate aggression. BMW has no PTAB appeal history on this patent to read as a signal of how hard it will fight — it has simply never needed to.
• No defensive aggregator in the chain. Unified Patents and similar entities have been active in vehicle-navigation art generally (e.g., Unified Patents LLC v. NavBlazer LLC, IPR2020-00983), but I found no indication that Unified or any comparable aggregator has targeted US 10,655,981. The patent is a BMW portfolio asset, not an NPE assertion vehicle — which is itself consistent with the "no IPR" result. Well-asserted, NPE-held patents attract IPRs; internally held, never-asserted portfolio patents typically do not.

The absence of PTAB activity is the finding. For a defendant, it means the patent is a genuine unknown rather than a known-bad or known-good asset: it has neither been hardened by surviving an IPR nor weakened by one. It is, however, a 2015-priority BMW continuation in a very crowded parking-POI art space, and the Board has never been asked whether that art renders claim 1 obvious once the two post-prosecution limitations (independent of whether the parking space is occupied and navigating … using the updated database) are given their natural scope.

---

Recommended next steps

1. If you are a defendant facing assertion — the clock is everything.

• Confirm whether the complaint has actually been served on you or a privy. If it alleges infringement of US 10,655,981, your § 315(b) one-year window is running. No PTAB route exists after it closes.
• There is no FWD to link to and no disposition to quote — do not represent otherwise in any brief or demand response. If opposing counsel or a demand letter tells you claims 1–5 are canceled, that is false; the correct statement is that all 12 claims are live and untested.

2. Build the prior-art map yourself — nobody has done it. The natural starting set is the art already in the file wrapper, all of which the examiner of record considered and which therefore carries Advanced Bionics / § 325(d) risk if reused alone:

• WO 2010/081547 A1 and US 8,890,715 B2 (TomTom / Geelen) — "Navigation apparatus, server apparatus and method of collecting parking location information." Closest in concept.
• US 2014/0340242 A1 (BMW/Belzner) — "Method for Providing Parking Information on Free Parking Spaces."
• DE 10 2013 203 909 A1 (BMW) — "Method and mobile terminal for recognizing a parking space."
• DE 10 2009 028 024 A1 (Bosch) — sensor-recognized place information compared against reference data.
• US 8,063,797 B1 (ParkMe/Sonnabend) — "Parking information collection system and method."
• US 9,558,664 B1 (Here Global) — parking availability via vehicle trajectory.
• BMW's own sibling publication US 2017/0369078 A1 / US 9,959,759 B2.
• Any post-2015-03-09 art cannot be prior art; anchor your search to the 2015-03-09 DE priority date and be prepared to test whether that priority is actually supported.

3. Consider a § 112 written-description / priority attack as a second front. As flagged in the earlier section, the "independent of whether the parking space is occupied" and "navigating … using the updated database" limitations do not appear in the specification's own restatement of the claimed method, and the file history shows the claim set was substantially amended between the 2019-03-19 non-final rejection and the 2020-01-13 notice of allowance. Whether the German priority application (DE 10 2015 204 169.7, filed 2015-03-09) supports those limitations is an open question I have not resolved — it requires reading the DE priority document, which I have not done. If it does not, you may be able to push the effective filing date later and open up intervening art. Treat this as a hypothesis to verify, not a conclusion.

4. Set monitoring. Because there is nothing to track yet, put a docket/patent-number watch on US 10,655,981 (USPTO PTAB E2E, PACER/RECAP, and CourtListener) so that the first petition, any district-court assertion, or any CAFC appeal is captured the day it appears. The right authoritative checks, which I could not execute directly and which should be run before you rely on this memo for a filing decision, are:

• USPTO PTAB E2E / Patent Trial and Appeal Board EndPoint: https://ptacts.uspto.gov/ptacts/
• USPTO Patent Public Search / PatentCenter by application number 15/699,332
• PACER or CourtListener by patent number
• CAFC docket (cafc.uscourts.gov) by patent number

---

Caveats on this memo. My "no proceedings" conclusion rests on (a) the canonical ODP-derived proceedings block, which is empty, and (b) six live web searches that surfaced only the soundalike confusions catalogued above. I could not query the PTAB E2E API, PACER, or the CAFC docket directly, so this should be read as "no publicly indexed or ODP-registered proceeding found," not as an official register negative. I have not fabricated any proceeding number, panel, institution date, FWD, settlement term, or appeal disposition, because none exists to report.

I'll start by pulling the recorded assignment data and cross-referencing the entities in this chain.

Let me verify the recorded assignment details and check whether any post-issuance transfers exist.

Let me confirm the assignment reel/correspondent detail and check for any asserter-directory or litigation activity tied to the patent.

I could not query the USPTO Patent Assignment Center (assignmentcenter.uspto.gov / assignment.uspto.gov) directly, so the assignment record below is reconstructed from the authoritative full text you supplied (Google Patents legal events, which mirror the USPTO assignment index for this patent) plus FreePatentsOnline / Google Patents bibliographic data. I flag where that leaves uncertainty rather than fabricating reel/frame detail.

Inventors

Employer determination: All six are listed only as BMW assignors in the single recorded assignment; there is no evidence any of them is employed elsewhere. The assignor execution dates (see below) cluster within days of the Sept. 8, 2017 US filing, which is the ordinary pattern for a continuation filing — not a departure signal.

Unusual patterns: None found. There is no evidence of inventors departing BMW within 12 months of filing, no co-assignment to a third party, and no inventor-to-NPE assignments. I cannot independently verify current employment status of the inventors, so treat "still at BMW" as undetermined rather than confirmed.

Original assignee

Bayerische Motoren Werke Aktiengesellschaft (BMW AG) — Petuelring 130, 80809 Munich, Germany. Original and current assignee of record.

• Primary line of business: Automotive OEM (vehicles, motorcycles, mobility/navigation services). Publicly traded (Frankfurt: BMW); the entity appears in German commercial register (Amtsgericht München HRB 42243).
• Did they ship a product embodying the claims? Highly likely on the evidence of the patent itself: the claimed vehicle-side recognition unit, navigation database update, and "navigating … using the updated database" are squarely within BMW's in-car navigation and parking-assistance product lines. BMW is an operating company that practices in this field. I have not found a specific product/feature named as the commercial embodiment, so this is an inference, not a documented finding.
• Current status: Operating. No acquisition, dissolution, or bankruptcy of the assignee. BMW is also an active defensive patent litigant — e.g., it appeared as IPR petitioner (BMW of North America, LLC and Bayerische Motoren Werke AG v. Theodore & Associates, LLC, IPR2017-01380, U.S. Pat. No. 9,045,163) — i.e., it defends itself against asserters rather than acting as one.

Assignment timeline

Chronological list of every recorded assignment:

• Executed 2017-08-09 to 2017-09-11 / recorded 2017-09-15 — Reel 043605/0097
  • Conveyance: Assignment (assignment of assignors' interest)
  • Assignors: FALK, ANDREAS; BREINING, CHRISTINA; CONNETTE, CHRISTIAN; GRÜNEWEDEL, SEBASTIAN; KISSLINGER, BERNHARD; SOBOTTA, BENJAMIN
  • Assignee: Bayerische Motoren Werke Aktiengesellschaft
  • Correspondent: Not surfaced in the sources accessible to me. The prosecution attorney/agent firm of record is Crowell & Moring LLP (per FreePatentsOnline "Attorney, Agent or Firm"), which is the likely recording correspondent — but I could not verify the assignment-record correspondent field directly, so I am not asserting it as fact.
  • Context: Original inventor-to-employer assignment on filing of the US continuation — a routine first assignment, not a fire-sale, reorg, or securitization.

No other assignments are indexed. Google Patents' legal-events/transaction history for US 10,655,981 shows only (a) the Sept. 15, 2017 assignment above, (b) the Dec. 28, 2017 pre-grant publication, (c) the May 19, 2020 grant, and (d) a Nov. 9, 2023 4th-year maintenance-fee payment. There is no post-issuance transfer to any LLC, NPE, or aggregator.

If the Assignment Center record differs (e.g., an assignment I could not see), this section would need correction — but the authoritative text you supplied does not show one, and my independent searches returned none.

Timeline diagram

timeline
    title Ownership of US 10655981
    2015 : Priority filed by BMW in Germany
    2016 : PCT application filed
    2017 : US continuation filed
         : Inventors assign to BMW
    2020 : Patent issued to BMW
    2023 : Maintenance fee paid

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. The only recorded assignment (reel 043605/0097, 2017) runs from individual inventors to BMW AG, an operating auto OEM. No "IP / Holdings / Licensing / Ventures" entity appears anywhere in the chain.

2. Known asserter in the chain — Not present. Neither the original nor any current assignee matches an NPE list. BMW AG is an operating company and, on the public record, appears on the defensive side (IPR petitioner against Theodore & Associates, IPR2017-01380).

3. Repeat correspondent across the chain — Unclear / not assessable. There is only one link in the chain, so recurrence cannot be tested. Prosecution correspondent of record is Crowell & Moring LLP; I could not confirm the assignment-record correspondent, so I cannot run a recurrence check. No NPE-list appearance by any attorney has surfaced.

4. Cascading transfers — Not present. No consecutive assignments; the chain is a single link ending at the original assignee.

5. Pre-litigation transfer — Not present. No infringement suit naming US 10,655,981 was found, and no assignment occurred within six months of any suit (the only assignment predates issuance by ~2.5 years).

6. Bankruptcy fire-sale — Not present. No bankruptcy of BMW; the patent was not sold in any proceeding.

7. Privateering — Not present. No transfer to an NPE; BMW retains the patent and practices in the field.

8. Defensive aggregator (anti-NPE) — Not present. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN.

Verdict

Insufficient data (per the defined category "no records, or only the original assignment").

The chain consists of exactly one recorded link — the Feb./Sept. 2017 inventor-to-BMW assignment at reel 043605/0097, recorded 2017-09-15 — with no post-issuance transfers, no assertion activity, and no NPE-adjacent party anywhere. All eight troll-pattern signals are either not present or not assessable (signal 3, due to the single-link chain). This is the profile of an operating company that retained its own patent, not an NPE situation; the verdict is "insufficient data" only because the assignment record contains nothing beyond the original filing assignment to analyze.

Verification link: USPTO Assignment Center search for US 10,655,981 — https://assignmentcenter.uspto.gov/ (or https://assignment.uspto.gov/patent/index.html). Note: I was unable to open these directly and am reporting the record reconstructed from the authoritative patent text and public aggregates; confirm reel 043605/0097 and the abs

I'll start by confirming the patent record, then work through the cited prior art.

Patent record confirmed. Now let me pull details on the most substantive cited references.

Scope note

I confirmed US 10,655,981 B2 against Google Patents (full text supplied), Espacenet, and FreePatentsOnline (URLs below). I could not reach USPTO Patent Public Search / PatentCenter directly — my searches were general web searches, so the citation roster below is taken from the patent's own "References Cited" tables (Google Patents full text + FPO). That is authoritative for what is on the face of the patent, which is what this task asks for. Litigation/PACER status was covered in the earlier section and is not repeated here.

Sources used:

• https://patents.google.com/patent/[US10655981B2](/patent/US10655981B2)/en
• https://worldwide.espacenet.com/publicationDetails/biblio?CC=US&NR=[10655981B2](/patent/10655981B2)&KC=B&FT=D
• https://www.freepatentsonline.com/[10655981](/patent/10655981).html
• https://patents.justia.com/patent/10655981
• https://patentimages.storage.googleapis.com/62/73/be/daae79d9af04ab/US10655981.pdf

---

1. Critical dates that govern § 102 here

Because the US case was filed on/after 16 March 2013, it is an AIA patent. Its effective filing date is 9 March 2015. Therefore:

• § 102(a)(1) art = anything patented / described in a printed publication / in public use / on sale before 9 March 2015.
• § 102(a)(2) art = only US patents, US patent-application publications, and PCT applications designating the US, effectively filed before 9 March 2015. A German or Japanese application publication is not § 102(a)(2) art; it is only art if it was published before the critical date (§ 102(a)(1)).

This matters a great deal here, because a large block of the cited references were published after 9 March 2015 (some were even filed after it). Those references cannot be § 102 prior art against claim 1 as written, regardless of why they appear on the face of the patent. Two possible explanations, both worth flagging: (a) they are third-party IDS submissions the examiner listed but did not rely on (the Google Patents star/† markers distinguish "cited by examiner" from "cited by third party"), and/or (b) they were cited against the later-filed JP/CN/DE family members and swept into the family record. If your purpose is an invalidity analysis, the post-critical-date references should be screened out on day one.

---

2. Important methodological caution

A reference appearing in the "References Cited" list is not an anticipation finding. Under § 102 a single reference must disclose every element of the claim, arranged as in the claim. Claim 1 here has six elements, and two of them are narrow:

• "said information at least partly characterizing properties of the parking space independent of whether the parking space is occupied" (emphasis added — a deliberate carve-out from occupancy/availability art), and
• "analyzing the determined information in relation to a sufficient characterization of the parking space" (a completeness test on the record being built).

None of the cited references appears to disclose both of those in a single teaching, so on the face of the cited art, no reference anticipates claim 1. The real exposure is § 103. I therefore report below the claims each reference implicates, and label anticipation candidacy only where a reference comes genuinely close.

---

3. The Tier-1 references — published before 9 March 2015 (capable of § 102(a)(1))

These are the only cited references that can anticipate. Ordered roughly by closeness to claim 1.

3.1 US 8,890,715 B2 — TomTom International B.V. — Navigation apparatus, server apparatus and method of collecting parking location information

Priority 2009-01-14; published/issued 2014-11-18. (Same family: WO 2010/081547 A1, pub. 2010-07-22; US 2012/0098677 A1, pub. 2012-04-26; EP 2 387 697 B1.)
Disclosure: A navigation apparatus with a "parking location determination module" that infers a candidate parking location when the GPS location is unchanged for a predetermined period (e.g., 2 min–2 h); the device may solicit user input as to whether the candidate parking location is an actual parking location; the identity of the candidate is transmitted to a server which maintains a database of known parking locations and "queries the database of known parking locations to determine if the candidate parking location corresponds to a known parking location"; the server enriches/reports parking information back to the navigation device for route/navigation use.
Claims implicated: Claim 1 (recognise from device data → determine → compare against a database → update → navigate), Claim 2 (user input used to determine/confirm the information), Claims 3/4 (recognition triggered by a stationary/parking event — analogously to "parking process recognised"), Claim 9 (data captured at a vehicle), Claim 10 (server + navigation apparatus architecture).
§ 102 assessment: Closest single reference on the compare-against-database and user-confirmation elements. Weakness: it keys on dwell/location and occupancy probability, i.e., precisely the occupancy dimension claim 1 excludes. It is a strong § 103 primary reference, not a clean § 102 anticipation of claim 1.

3.2 US 8,063,797 B1 — ParkMe LLC — Parking information collection system and method

Priority 2010-07-31; issued 2011-11-22. (Pub. counterpart US 2012/0062395 A1.)
Disclosure: Vehicle-mounted (and pedestrian hand-held) cameras with object recognition/OCR that read parking signs and parking rules street-by-street; the captured sign/rule data is geocoded and uploaded to a backend server that builds/maintains a parking database (rules by street, time of day, day of week, residential vs. commercial, meters, bans). When a parked sign appears obscured, the system can notify the backend server to create a request to have a person verify whether a sign was hidden — a verification loop. The database is then queried by drivers via a GPS app/nav device to locate parking.
Claims implicated: Claim 1 (vehicle sensor captures surroundings → derive characterising properties → backend database → update → navigate), Claim 2, Claim 9, Claim 10, Claim 12.
§ 102 assessment: Best match on the "characterising properties of the parking space" half of claim 1 — it literally reads residential/commercial parking rules and signs off the street into a backend database, which is the patent's core example (residents' parking space, time restrictions). Weaknesses for anticipation: the structured "analyze … in relation to a sufficient characterization" step is not disclosed as such, and the update is largely inventory/collection rather than a deviation-triggered update. Strong § 103 reference; arguably the most damaging for claim 1's characterising-properties element.

3.3 US 2014/0340242 A1 — Bayerische Motoren Werke AG — Method for Providing Parking Information on Free Parking Spaces

Priority 2012-02-01; published 2014-11-20. (Granted as US 9,652,986 B2.) Inventors Kates, Belzner.
Disclosure: Vehicle cameras / optical and non-optical sensors detect available free parking spaces; a central computer maintains a knowledge/historical database (which spaces are usable "valid parking spaces", average free count by block and time); historical data + current vehicle-reported data are fused (Bayes/Erlang) into a probability distribution; result feeds route guidance; user (app) inputs of free spaces are also contemplated and "always added to the historical data".
Claims implicated: Claim 1 (sensor → information → database → update → navigate), Claim 2 (user inputs), Claim 4 (data gathered on a manoeuvring/parking event), Claim 9, Claim 10, Claim 12.
§ 102 assessment: Note this is the same applicant's own earlier application — it is not citable as § 102(a)(2) secret art against its own successor-in-interest's later patent only if the inventorship/common-ownership conditions of § 102(b)(2)(C) are met (frequent in BMW's portfolio); it is citable as § 102(a)(1) art because it published 2014-11-20. Its focus is availability/occupancy — the very thing claim 1 disclaims — so it attacks claim 1 only via claims that don't contain the carve-out, and it is best used for the "central computer + historical database + sensor detection" architecture.

3.4 DE 10 2009 028 024 A1 — Robert Bosch GmbH — Parking guiding system … sensor recognizing place, where information about place is compared with vehicle related data

Priority 2009-07-27; published 2011-02-03.
Disclosure: A parking guidance system for a navigation device in which a sensor recognises a (free) place and the information about the place is compared with vehicle-related data (e.g., is the space big enough for this vehicle).
Claims implicated: Claim 1 (recognise → determine → compare), Claim 9, Claim 10.
§ 102 assessment: Its "compare determined information against stored data" is relevant, but it compares against vehicle dimensions, not against a parking database entry with an update-on-deviation — so it does not reach claim 1's update element.

3.5 US 2011/0224899 A1 — Telenav, Inc. — Navigation system with parking space locator mechanism and method of operation thereof

Priority 2010-03-12; published 2011-09-15.
Claims implicated: Claim 1, Claim 10 (navigation-side parking locator).

3.6 US 2015/0009047 A1 — Mordechai Ashkenazi — Method and apparatus for vehicle parking spaces management using image processing

Priority 2013-07-04; published 2015-01-08 (note: still before the 2015-03-09 critical date).
Claims implicated: Claim 1 (image processing to identify/manage parking spaces), Claim 9, Claim 10.

3.7 DE 10 2013 203 909 A1 — Bayerische Motoren Werke AG — Method and mobile terminal for recognizing a parking space

Priority 2013-03-07; published 2014-09-11.
Claims implicated: Claim 1 (parking-space recognition step), Claim 9, Claim 10. Same-applicant art; relevant to the recognition unit, but silent on database update.

3.8 US 2014/0036076 A1 — Steven David Nerayoff — Method for Controlling Vehicle Use of Parking Spaces by Use of Cameras

Priority 2012-08-06; published 2014-02-06.
Claims implicated: Claim 1 (camera-based parking-space determination), Claim 9.

3.9 DE 10 2011 086 268 A1 — Robert Bosch GmbH — Method for detecting a free parking space

Priority 2011-11-14; published 2013-05-16.
Claims implicated: Claim 1 (sensor detection of a parking space), Claim 9.

3.10 DE 10 2010 033 215 A1 — Valeo Schalter und Sensoren GmbH — Method for supporting parking in a parking garage … occupancy status recognition system

Priority 2010-08-03; published 2012-02-09.
Claims implicated: Claim 1, Claim 9, Claim 10 (occupancy-status recognition; note the occupancy focus cuts against the claim-1 carve-out).

3.11 Ford pair — DE 10 2011 003 772 A1 / US 2012/0200430 A1 — Method and device for assisting a driver in finding a parking spot

Priority 2011-02-08; both published 2012-08-09.
Claims implicated: Claim 1, Claim 9 (parking-spot detection and assistance).

3.12 US 2014/0285361 A1 — Audi AG — Personalized parking assistant

Priority 2013-03-25; published 2014-09-25.
Claims implicated: Claim 5/11 territory (driver prompting/personalisation), Claim 1.

3.13 US 2012/0323643 A1 — Premier Parking LLC — Parking management systems and methods

Priority 2011-03-24; published 2012-12-20. Claims implicated: Claim 1, Claim 10.

3.14 DE 10 2011 003 772 / US 2013/0325342 A1 — Apple Inc. — Navigation application with adaptive instruction text

Priority 2012-06-05; published 2013-12-05. Peripheral — touches the navigating … using element of claim 1 only.

3.15 US 2014/0309934 A1 — Flextronics AP, LLC — Automatic Alert Sent to User Based on Host Location Information

Priority 2013-04-15; published 2014-10-16. Peripheral (location-triggered alerts, HMI).

---

4. Tier-2 references — publicly available only after 9 March 2015, but effectively filed before it (possible § 102(a)(2) art only)

These are citable only if the cited document is a US patent, US pre-grant publication, or a PCT designating the US with an effective filing date before 2015-03-09. For each, the "priority date" shown may be a foreign priority that does not count for § 102(a)(2) purposes — I flag that where relevant, and I could not verify the underlying US filing dates from the sources I reached.

---

5. Tier-3 references — post-date the critical date entirely (not § 102 art against claim 1)

I flagged these explicitly because they constitute roughly a third of the cited roster. Each of these was filed and published after 9 March 2015, so it cannot be § 102(a)(1) or § 102(a)(2) art:

Even if one were minded to use them, the Conduent/PARC cluster is directed to enforcement-officer logistics, and the 2017 Uber/IBM/HERE references to probe-data modelling and risk routing — none touches the "sufficient characterization of the parking space property" analysis of claim 1.

A specific flag: US 9,558,664 B1 (HERE) is often conflated with the "vehicle trajectory → parking availability" family. It was filed 2015-08-13 — five months after the critical date — so it is not prior art here, notwithstanding that it is an examiner-cited reference.

---

6. References cited against the family (JP/CN/DE) — "Family Cites Families"

These appear in the family record (Google Patents "Family Cites Families" list) rather than necessarily in the US file wrapper. I list them because several are pre-critical-date and are therefore genuine § 102(a)(1) printed publications that a full prior-art search should include:

Two of these deserve attention as potential anticipation candidates, subject to full-text review:

• US 8,606,499 B2 (Navteq) — "determining parking area location information" is squarely in the parking-area-database space and predates the critical date. Worth pulling in full.
• US 2012/0056758 A1 (Delphi) — connected-vehicle parking-spot locator with a database, i.e., the multi-vehicle collection architecture of claim 1.

---

7. Bottom line

1. No cited reference appears to anticipate claim 1. Anticipation requires all six elements — including the "characterizing properties … independent of whether the parking space is occupied" clause and the "sufficient characterization" analysis — in one reference. The closest single teachings (TomTom US 8,890,715; ParkMe US 8,063,797; BMW US 2014/0340242) each supply a proper subset and each is oriented toward occupancy/availability, which claim 1 expressly excludes.
2. The strongest invalidity theory is obviousness (§ 103), with the most natural combination being:

• ParkMe US 8,063,797 B1 (vehicle camera + OCR of parking signs/rules → backend parking database → nav query) as to the "determining characterising properties" and "database" elements; combined with
• TomTom US 8,890,715 B2 (candidate-parking-location determination → query database of known parking locations → solicit user confirmation → server-side enrichment) as to the compare/update/user-input elements; and optionally
• BMW US 2014/0340242 A1 (central computer, historical + current data fusion, route guidance) as to the update-and-navigate elements.
  That combination maps onto claims 1, 2, 9, 10 and 12; TomTom's user-confirmation also bears on claims 5–7.

1. Claim-by-claim exposure of the field generally: claims 9 and 10 are the broadest and most exposed (a vehicle sensor plus five generic functional units); claims 5–7 (prompt-on-negative-analysis) map to TomTom's user solicitation and ParkMe's obstruction-verification request; claim 8 (compare before analyze) is a step-ordering claim with thin support.
2. Screen the roster before doing any § 102 work. Approximately 15 of the ~45 cited references — including the eight Conduent/PARC parking-enforcement references, HERE US 9,558,664, Sparkcity US 2016/0180712, the 2017 IBM/Uber/HERE applications, and DE 10 2014 209 453 A1 — cannot be § 102 prior art against claim 1 because they post-date the 9 March 2015 critical date. Their presence in the list is most consistent with third-party IDS submissions and/or citations imported from the JP/CN/DE family prosecutions.

---

8. Caveats and flags (consistency with the earlier section)

• Source limitation (unchanged from the earlier section): I still could not query USPTO Patent Public Search / PatentCenter directly, so I cannot confirm (a) the examiner's star vs. dagger designation for each reference, (b) whether any of these were relied upon in the 2019 non-final/final rejections, or (c) the actual US filing dates behind the foreign-priority figures. Items (b) and (c) matter a lot for the Tier-2 table in § 4.
• New flag, not in the earlier section: the citation roster contains a substantial block of references that post-date the priority date. This is a fact about the record, not an inference about prosecution strategy. It should be reconciled before any invalidity opinion relies on this list.
• New flag — self-citation: US 2014/0340242 A1 and DE 10 2013 203 909 A1 are BMW's own earlier applications. Whether they are citable at all depends on the § 102(b)(2)(C) common-ownership/common-inventorship analysis, which I have not performed.
• Flag re: the earlier section's prosecution inference — it remains correct that I have not read the office actions, so the observation that the "independent of whether the parking space is occupied" and "navigating … using the updated database" limitations look added during prosecution remains an inference from the face of the documents, not a verified fact.

I'll verify the key prior-art disclosures before building the § 103 combinations, so the motivation-to-combine analysis rests on what each reference actually teaches.

Obviousness Analysis — US 10,655,981 B2 under 35 U.S.C. § 103

Builds on the Prior Art section; does not repeat its citation roster. All claim language quoted from the granted claims as reproduced in the authoritative full text. Where a reference's disclosure is material to the combination, I re-verified it against live sources this session (URLs cited inline) and the live results confirmed the Prior Art section's characterizations — no contradictions found.

---

0. Framework and the one fact that shapes everything

Two threshold points drive the whole analysis:

1. The patent's own background is an Applicant Admitted Prior Art (AAPA) problem statement. The specification states: "Usually, this information is static and, at best, updated after a system update. However, this information can also be captured and maintained manually. On account of manual activities, POI information is not always available in a current form…" and frames the object as "updat[ing] parking space information in a navigation system… as quickly and easily as possible, but nevertheless in a reliable manner." Under KSR, this admission supplies the design incentive for the entire claimed architecture. An obviousness case need not work hard to establish "why combine" when the patent itself names the problem the combination solves.

2. The closest art is directed at the same problem in the same terms. TomTom's EP 2 387 697 B1 specification (family of US 8,890,715 B2) states verbatim: "new parking opportunities are created over a time and so do not feature in the database of POIs until the provider of the database of POIs decides to release an update of the database of POIs. Sometimes, the delay between releases can be considerable resulting in the PND having access to an incomplete and/or inaccurate database of parking information." That is the '981 patent's stated problem, stated by a third party, ~5 years before the priority date. This is a § 103 motivation anchor, not merely a § 102 disclosure.

---

1. Element-by-element mapping of claim 1

Claim 1 (method) has six elements. Here is what each requires, and what the top three references supply:

No single reference contains all six + the two narrow limitations. That is why the case is a § 103 case, not a § 102 case — consistent with the Prior Art section's bottom line.

---

2. Combination A — ParkMe as primary, TomTom as secondary (the lead theory)

2.1 The combination

Primary: US 8,063,797 B1 (ParkMe). Teaches elements (a), (b), (c), (d), (e), (f) — in particular the element the examiner most likely allowed over: characterizing properties of a parking space from on-vehicle sensor data, occupancy-blind (the OCR'd sign/rule record: residents'-only, time restrictions, meters). This is literally the '981 patent's own worked example — the spec's example of a determined characterizing property is "a parking sign for a residents' parking space."

Secondary: US 8,890,715 B2 (TomTom). Teaches the deviation-triggered compare-and-verify loop: candidate location → query the database of known parking locations → if not a known location, prompt the user → server-side enrichment → return to the navigation device. This supplies (c)'s completeness test as an operational step and claims 2, 5, 6, 8.

Optional tertiary: US 2014/0340242 A1 / US 9,652,986 B2 (BMW). Supplies the central-computer + historical-database + current-data fusion + route-guidance back end, plus the maneuver-detection trigger (claims 3/4) — detected via "the start-up of the vehicle engine, detection of the current location, as well as the evaluation of steering movements," which is the specification's own definition of a recognized parking process ("a reverse drive with extreme steering movement").

2.2 Motivation to combine — articulated in KSR/MPEP terms

1. Same field of endeavor, same problem, same solution direction. All three are vehicle-navigation / parking-assistance systems. ParkMe and TomTom both expressly target the currency of parking POI data; TomTom's own specification names the "considerable delay between database releases" problem that the '981 patent identifies. Combining a crowd-sourced parking-characterization pipeline (ParkMe) with a candidate-verification-and-user-prompt loop (TomTom) is the application of one known technique to improve the other in the same way, for the same purpose. MPEP 2144.03, Example C; KSR (predictable use of prior-art elements).

2. The references themselves supply the suggestion. ParkMe discloses that when a sign is unresolvable/obscured, the system may notify the backend to have a person verify; TomTom discloses prompting the driver to confirm a candidate. One of ordinary skill, seeking to raise the completeness/quality of the characterizing record (ParkMe's stated quality-control goal), would look to a driver-confirmation mechanism — which TomTom supplies, in the same HMI, on the same device.

3. Design incentive / market forces. TomTom's spec: release delays produce "an incomplete and/or inaccurate database." ParkMe: compensation auto-adjusts where "data is becoming stale." Both articulate a commercial driver for continuous, automated database refresh from the vehicle fleet — which is precisely the claimed architecture. KSR design-incentive rationale.

4. Predictable results, no new mechanism required. Every step is a known data-processing operation: (i) OCR/object recognition on camera frames; (ii) look-up in an indexed geocoded record (ParkMe's linked-list street/block/POI structure); (iii) a field-level mismatch → write (an ordinary UPSERT); (iv) a query to serve navigation. The patent adds no new sensor, algorithm, or hardware. Its asserted advantages (new spaces incorporated quickly; stale spaces removed; changed attributes updated) are the natural and predictable consequences of automating what the AAPA says was manual.

5. Reasonable expectation of success. ParkMe had already demonstrated reading ~250 parking signs per mile from a moving vehicle at speed with threshold-match resolution. TomTom had already demonstrated server-side aggregation of parking locations from a fleet. A POSITA would have expected the combination to work.

6. "Obvious to try" / design choice on which properties to characterize. The specific properties recited — residents'-only, time restrictions, meters — are dictated by the target application (the POI record), not by any technical insight. ParkMe's POI type field ("type #32 = valet parking sign") and rule attributes make the selection a mere design choice among a finite set of known parking attributes.

2.3 Rebuttal to the likely non-obviousness arguments

(i) "The prior art is occupancy-focused; claim 1 requires occupancy-independent characterization."
This is the patent's strongest-looking argument and it fails on the face of ParkMe. ParkMe's characterizing data (signs, rules, meters, time-of-day, residential/commercial) is not occupancy data at all — it is static-and-slowly-changing property data. The negative limitation is met. Moreover, the '981 patent does not exclude occupancy data; it requires the characterizing to be independent of occupancy — i.e., performed whether or not a space is occupied. ParkMe reads signs from a moving vehicle irrespective of occupancy; BMW '242 detects spaces while "moving along the traveled street." Element (b) is squarely met.

(ii) "The 'sufficient characterization' analysis is a mental/subjective judgment; the art doesn't teach it."
Attractive but wrong-headed for the patentee. In re Comiskey-type arguments about mental steps cut against the claim, not for it (§ 101/§ 112 concerns). And the art does teach a machine-implemented completeness test: ParkMe's threshold-match percentage test ("better than a threshold percentage match of what is expected based on patterns of side of street") is a literal sufficiency threshold — the record is either sufficient to positively identify the sign or it is not, and the branch determines whether a verification request issues. TomTom's "candidate not a known candidate parking location → solicit user input" is the same logical structure. Under In re Keller, 642 F.2d 413 (CCPA 1981), a combination claim may be obvious even if each reference's purpose is not identical to the claim's; one cannot show non-obviousness by attacking references individually.

(iii) "The references show no reason to add attribute-characterization to a location-only DB."
But TomTom's stated object is "a database of parking locations to be created and/or maintained so as to provide as complete repository of parking locations as possible for any number of parking related services." "As complete as possible" is the motivation to enrich beyond bare location. ParkMe supplies the enrichment payload.

(iv) Ordering (claim 8). Claim 8 recites compare before analyze. The '981 specification concedes this is an alternative: "the step of comparing with data in a database may be effectuated prior to the analysis of the determined data." TomTom performs the DB query first (is the candidate known?), then solicits user input only on a miss — i.e., compare-then-complete. Claim 8 is therefore closely approached, and in any event reordering steps in a data pipeline absent a showing of a new/unexpected technical result is an obvious variation. In re Angstadt; MPEP 2144.04(III).

Secondary considerations. I found no evidence of unexpected results, commercial success with nexus, long-felt-but-unsolved need, or industry praise. The patent's own listed advantages (fast incorporation of new spaces, removal of obsolete ones, propagation of changed properties to the backend) are the expected benefits of automating the AAPA-described manual process — which under KSR is not a rebuttal of obviousness but a confirmation of it. Flagging this honestly: absence of evidence is not proof of absence, but nothing in the record I could reach supports a secondary-considerations argument.

---

3. Combination B — TomTom as primary (alternative / independent footing)

Primary US 8,890,715 B2 (TomTom) for the aggregate-select-prompt-update architecture + US 9,759,569 B2 (TomTom, "Apparatus and method for determining parking information") for server-side determination of parking areas from a plurality of navigation devices (this is the multi-vehicle collection architecture the '981 claim 1's "at least one vehicle" and the spec's FIG. 5 "customer feedback" embodiment rely on) + ParkMe for the property-characterization payload + US 8,606,499 B2 (Navteq/Nokia) for the deviation-triggered database write.

Navteq US 8,606,499 B2 is verified as disclosing the closest thing to element (e) in a trace-based system: "If the system 100 determines that the parking area is not already stored within the one or more databases, the system 100 causes… an addition of the location-based information associated with the parking area to the one or more databases" — i.e., compare-to-DB → add-if-absent, with a corroboration threshold ("valid if more than three independent GPS traces recently (e.g., in the last six months)"). That corroboration threshold independently reads on the '981 FIG. 5 embodiment's requirement to "check… whether the probability that the information which is available is applicable and sufficient to justify an update of the database" and on the cross-vehicle-deviation check in the spec.

Motivation (Combination B): TomTom's own specification frames the inadequacy of single-vehicle storage ("WO2007083997 discloses the storing of frequently used parking positions in a single vehicle, this information not being shared with other users") as the problem to be solved, and expressly seeks a "low-cost and minimally intrusive" fleet-sourced DB. Navteq and ParkMe both supply fleet-sourcing with attribute/confidence handling. The combination is the predictable union of three systems that each already send vehicle-observed parking data to a remote DB and return it to navigation devices.

---

4. Combination C — Generic-architecture attack on claims 9, 10, 12

Claims 9, 10, and 12 are the broadest claims in the patent and the most exposed.

• Claim 9 adds only "data used to recognize the parking space are captured by at least one sensor of a vehicle" — admitted in the spec, disclosed by ParkMe (camera), TomTom (GPS/location receiver), BMW '242 (camera/radar), Delphi US 2012/0056758, Bosch DE 10 2011 086 268 A1.
• Claim 10 is a five-means-plus-function-style recitation: recognition unit / determination unit / analysis unit / comparison unit / update unit. Each unit maps 1:1 onto a generic programmed processor disclosed in the art. TomTom US 8,890,715 claim 1 recites a "processing resource… supporting a parking location determination module and an inference engine operably coupled to a database" — i.e., structurally the same architecture. ParkMe US 8,063,797 claim 12 recites "one or more memories for storing program code; one or more communication links to a parking information database; and one or more processors." Delphi US 2012/0056758 recites a remote database, a communication link receiving parking location information from connected vehicles and providing it back, and a processor — a functionally identical three-unit core. Under § 103, distributing those known functions across named software modules is an obvious implementation choice. MPEP 2144.04; In re Kuhle.
• Claim 12 (CRM) adds nothing beyond claim 1 plus "non-transitory computer readable medium" — squarely obvious given ParkMe's downloadable app / collection application and TomTom's software-updatable navigation device.

Motivation (Combination C): Mere automation/computer-implementation of known data-collection and database-query steps. KSR: "if a technique 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, using the technique is obvious."

---

5. Claim-by-claim vulnerability map

Weighted conclusion: the independent claims (1, 10, 12) are the harder targets because of element (c)'s "sufficient characterization" phrasing and element (b)'s occupancy carve-out. The dependent claims — especially 9, and the claim 3/4 parking-process trigger and claim 2 user-input feature — fall away almost mechanically once a viable primary reference is established for claim 1. This is the classic profile of a patent whose apparent strength sits entirely in one independent claim.

---

6. Why this is a § 103 and not a § 102 case — restated briefly

Anticipation requires one reference disclosing all elements arranged as claimed. ParkMe is the only reference that comes near, and it does not expressly disclose a deviation-triggered update as such (its update is inventory-driven and its quality-control branch is a human-verification request, not an automatic write-back on mismatch). TomTom keys on dwell/candidate location and is silent on property characterization. BMW '242 is availability/occupancy-focused and silent on signed-rule capture. None of the three alone reads on claim 1 with the "sufficient characterization" step as an explicit machine test. The combination, however, closes every gap with unrebutted KSR rationales.

---

7. Material caveats and strategic flags

7.1 The examiner-cited-art problem (most important flag on this analysis). ParkMe US 8,063,797, TomTom US 8,890,715 / WO 2010/081547, BMW US 2014/0340242, and Bosch DE 10 2009 028 024 all appear on the face of US 10,655,981 as cited references and were therefore available to the examiner of record before the 2020-01-13 notice of allowance. The patent nonetheless issued. This means:

• A § 103 theory built on exactly ParkMe + TomTom + BMW '242 faces § 325(d) discretionary-denial risk if presented to the PTAB (Advanced Bionics), because it uses "the same or substantially the same art or arguments previously presented to the Office."
• The theory is only as strong as the gap it explains the examiner did not address. The likely allowance hook was the "analyzing… in relation to a sufficient characterization" step, not the reference combination per se. A petitioner should therefore lead with that element and with non-cumulative art — Navteq US 8,606,499, Delphi US 2012/0056758, JP 2015-007557, US 9,759,569, and any foreign-language art from the JP/CN/DE family prosecutions — rather than re-running the same trio.
• I have not read the office actions, so I cannot say what arguments or amendments were made or which reference was relied upon. This is the single most important unresolved item for a real § 103 opinion.

7.2 Priority / § 112 opening (flagged in earlier sections — restated because it changes the § 103 art set). The final two claim-1 limitations ("independent of whether the parking space is occupied" and "navigating… using the updated database") are absent from the specification's own restatement of the method. If they are not supported by DE 10 2015 204 169.7 (filed 2015-03-09), the effective filing date moves later, and the following currently-non-art references become available as § 102/§ 103 art: US 9,558,664 B1 (HERE, filed 2015-08-13), US 2016/0180712 A1 (Sparkcity, 2015-08-27), and others from the Tier-3 table. That is a hypothesis requiring the DE priority document — I have not read it. Do not assert it as established.

7.3 Common-ownership nuance. US 2014/0340242 A1 and DE 10 2013 203 909 A1 are BMW's own earlier applications. Their use as § 102(a)(2) art may be defeated by the § 102(b)(2)(C) common-ownership exception, but they remain available as § 102(a)(1) printed publications (published 2014-11-20 and 2014-09-11, both before the 2015-03-09 critical date). For § 103 purposes, that is sufficient — and their value as motivation evidence ("the same applicant recognized this architecture in the same art") is unaffected.

7.4 Claim-construction dependence. The "sufficient characterization" limitation has no absolute standard; its scope depends on construction. Under the specification, sufficiency = having the fields on a stored required-information list (position + use conditions). Under that construction the limitation is broad and the art reads on it. If a defendant's expert applies a narrower construction, the analysis changes. This should be pinned down before any opinion is finalized.

7.5 Source limits. I re-verified ParkMe, TomTom, BMW '242/'986, Navteq '499, and Delphi '0056758 against live full-text/claim sources (freepatentsonline, RPX Insight, Google Patents PDFs) this session. I could not reach USPTO Patent Public Search, PTAB E2E, or PACER directly. The star/dagger designations, the examiner's actual reliance in the 2019 rejections, and the true US filing dates behind the foreign-priority figures in the Tier-2 table remain unverified, and item (b) of that list (reliance) materially affects the § 325(d) analysis in § 7.1.

---

Bottom line

The strongest § 103 case against US 10,655,981 is:

US 8,063,797 B1 (ParkMe) — on-vehicle camera/machine-vision capture of parking signs and rules into a geocoded, backend parking information database, with a threshold-match sufficiency test and a human-verification branch, queried through in-dash navigation —
in view of US 8,890,715 B2 (TomTom) — candidate-parking-location determination, query of the database of known parking locations, driver solicitation upon non-confirmation, and server-side maintenance of the parking-location database —
and optionally US 2014/0340242 A1 (BMW) — central computer, historical + current data fusion, maneuver-detection trigger, and route guidance to the resulting parking information.

That combination maps onto claims 1, 2, 5, 6, 7, 8, 9, 10, 11, and 12; adding BMW's steering/maneuver detection maps onto claims 3 and 4. Every KSR rationale is available and each is supported by the references' own texts: same field and same stated problem (TomTom's "considerable delay between releases"; ParkMe's "data is becoming stale"), explicit suggestion in the references (ParkMe's verification request; TomTom's solicitation step), a design incentive (fleet-sourced DB currency), predictable use of known techniques (OCR + DB lookup + conditional write + navigation query), and reasonable expectation of success (both systems were commercially deployed). The patentee's best counter — that the cited art is occupancy-focused — is defeated by ParkMe, which characterizes parking properties irrespective of occupancy.

The principal weakness of this theory is not technical; it is procedural (§ 325(d), because the three lead references were before the examiner). Any serious challenge should be built on that gap plus fresher art, not on a re-run of the examiner's own citations.