patent.fyi
Applications  /  US 2026/0356385
United States Patent Application Publication

US 2026/0356385 A1

Published
Nov 12, 2026
Filed
Aug 20, 2025
Earliest date on the face
May 12, 2025
Classification
B60W 50/10
First published here
Nov 20, 2026

Generated automatically. Not reviewed by a person. Not legal advice or an opinion.

Section 1 · machine-generated

What if?

Questions about how individual parts of claim 1 might be changed. A question may describe something different from what the claim language recites, or a version of it; no view is offered on which.

1.1Timed maneuver sequence

What ifWhat if the processor detects an ordered sequence of driver actions using discrete states?
Claim language“determine a behavior confidence value”

A finite-state machine starts a five-second timer when the turn signal activates below 10 km/h. It records a parking maneuver when the brake pedal is pressed and the gear selector enters reverse before the timer expires.

1.1Boolean driver action rule

What ifWhat if the processor combines driver actions with a Boolean rule that uses no per-characteristic weights?
Claim language“a weight associated with the behavior characteristic”

The processor converts wheel speed, brake pedal position, and gear selection into Boolean flags. An AND gate sets the behavior flag when wheel speed is below 3 km/h, the brake pedal is depressed, and reverse is selected.

1.2Direct parking geometry checks

What ifWhat if the processor determines parking suitability directly from measured clearances?
Claim language“determine a parking situation confidence value”

Side-facing range sensors measure the gap between neighboring vehicles. The processor marks the gap as usable when its measured length exceeds vehicle length by 1 m and its measured width exceeds vehicle width by 0.5 m.

1.3Parking path feasibility flag

What ifWhat if the processor returns a feasible or infeasible flag from a parking path check?
Claim language“determine a parking situation characteristic score”

The processor projects a vehicle footprint along candidate reverse paths generated within the steering angle limits. It sets the feasibility flag when a path reaches the target pose and maintains at least 0.25 m of clearance from occupied cells throughout the maneuver.

1.4Button initiated parking offer

What ifWhat if a dedicated parking button supplies the driver-intent trigger for the offer?
Claim language“being greater than a first predetermined threshold”

The processor registers a button press after the switch remains closed for 30 ms. It displays the parking assistance offer when it registers the press and the parking situation characteristic score exceeds the second threshold.

Section 2 · machine-generated

Variations

Further arrangements of the same subject matter, described here as of the publication date at the foot of this page.

Other field of use

Docking assistance for warehouse lift trucks

A lift truck controller reads traction speed, steering angle, and direction selector state to derive operator behavior confidence values. A forward range sensor and a camera supply separate confidence values for bay clearance, alignment with floor markers, and distance to the loading bay. The controller combines weighted partial scores within each group and displays an automatic docking offer when both group scores exceed their respective thresholds.

Component substitution

Weighted median for group score aggregation

The processor replaces sum-based aggregation with a weighted median for each of the two characteristic groups. Within each group, it sorts the confidence values and forms a partial score for each characteristic by dividing its weight by the total group weight. It accumulates these partial scores in confidence order and uses the first confidence value at which the cumulative score reaches one half as the group score. The display presents a parking assistance offer when both group scores exceed their respective thresholds.

Parameter or scale variation

Timed confirmation at low travel speeds

The controller updates the weighted behavior and parking situation scores at 10 Hz while vehicle speed is below 8 km/h. Both scores use a zero-to-one scale, and the controller counts consecutive updates above 0.70 for behavior and 0.80 for the parking situation. A failed comparison or a speed at or above 8 km/h resets the count. The controller generates the parking assistance offer after eight consecutive qualifying updates.

Separate situation scores for two sides

The processor evaluates up to two candidate spaces on each side of the vehicle and computes one shared behavior score from driver inputs. For each candidate, it derives confidence values for opening length, lateral clearance, and distance to the space entrance, then aggregates their weighted partial scores into a candidate-specific situation score. When the behavior score exceeds 0.70 and at least one candidate scores above 0.80 on a zero-to-one scale, the display places the assistance offer beside the highest-scoring candidate.

Combination with existing systems

Parking offers using route guidance data

A route guidance unit supplies the parking controller with distance to the destination and the mapped locations of parking entrances. The controller assigns separate weighted situation confidence values for destination proximity and proximity to an entrance, and combines them with a space clearance confidence value from local sensors. It forms the behavior score separately from speed and steering data. The route guidance display presents the parking assistance offer when both scores exceed their respective thresholds.

Automated or data-driven variant

Trained models for individual confidence values

A development system trains a separate logistic regression model for each behavior and parking situation characteristic using recorded approaches labeled by whether the driver begins a parking maneuver. Each model takes measurements for one characteristic, such as deceleration over a two-second window or lateral clearance beside the vehicle, and returns a confidence value between zero and one. The vehicle processor applies stored weights to the model outputs and aggregates the partial scores separately for behavior and parking situation. It displays the offer when both aggregate scores exceed their respective thresholds.

Specific version of a part

Weighted driver input signals

The processor assigns confidence values from 0 to 1 to low wheel speed, brake pedal application, and turn signal activation using lookup tables. It multiplies those values by weights of 0.40, 0.35, and 0.25, respectively, and sums the partial scores.

Weighted parking gap measurements

The processor maps measured gap length, usable width, and curb alignment to separate intention-to-park confidence values between 0 and 1 using calibrated lookup tables. It multiplies those values by weights of 0.50, 0.30, and 0.20, respectively, to produce partial parking situation scores.

Sum of weighted gap scores

The processor adds the partial scores for gap length, usable width, and curb alignment at each 100 ms update. It stores the sum as a parking situation characteristic score between 0 and 1 when the weights sum to 1.

Dashboard parking assistance prompt

The processor displays a parking assistance offer when the behavior characteristic score exceeds 0.70 and the parking situation characteristic score exceeds 0.80. The dashboard presents accept and dismiss controls, and the processor sends a request to the parking controller when the driver selects accept.

Section 3 · quoted from USPTO records

Other work

Passages reproduced word for word from two earlier-published U.S. patent documents, placed beside the language of claim 1. No characterization is offered. Each passage is marked A or B for the document it comes from; ¶ numbers are the paragraph numbers in the USPTO’s published full text of that document.

Document AUS 2006/0190259 A1
Published
Aug 24, 2006
Filed
Feb 16, 2006
Class
G10L 15/14
Document BUS 2015/0266486 A1
Published
Sep 24, 2015
Filed
Mar 11, 2015
Class
B60W 40/09
Claim 1
Document A
Document B
1.1an electronic processor, the electronic processor configured to: for each behavior characteristic of a plurality of behavior characteristics, determine a behavior confidence value, wherein the behavior confidence value indicates a likelihood of intention to park based on the behavior characteristic; and determine a partial behavior characteristic score based on the behavior confidence value and a weight associated with the behavior characteristic; determine a behavior characteristic score based on the partial behavior characteristic scores;
The method includes the operations of obtaining frequency features of a received speech signal for the respective frames having a predetermined length, calculating a keyword model-based likelihood and a filler model-based likelihood for each of the frame, calculating a confidence score based on the two types of likelihoods, and deciding whether the received speech signal corresponds to a keyword or a non-keyword based on the confidence scores.
A · US 2006/0190259 A1 · abstract
1.2for each parking situation characteristic of a plurality of parking situation characteristics, determine a parking situation confidence value, wherein the parking situation confidence value indicates a likelihood of intention to park based on the parking situation characteristic; and determine a partial parking situation characteristic score based on the parking situation confidence value and a weight associated with the parking situation characteristic;
The processing unit 9 may be configured to raise a level of overtaking intention in case the measured lateral position p indicates that the lateral position of the host vehicle 1 is closer a centre line 21 of the road than a predetermined threshold distance.
B · US 2015/0266486 A1 · ¶ 71
1.3determine a parking situation characteristic score based on the partial parking situation characteristic scores; and
1.4in response to the behavior characteristic score being greater than a first predetermined threshold and the parking situation characteristic score being greater than a second predetermined threshold, generate a notification regarding an offer to automatically assist parking the vehicle.
Further, the forward collision warning system 2 is configured to delay issuance of the warning in case the estimated level of overtaking intention is above a predetermined threshold level.
B · US 2015/0266486 A1 · ¶ 81

An empty cell means only that no passage was reproduced there.

Statements in these documents
Even the words of the same subword provide difference pieces information on a frame basis, but uniform calculation is performed without consideration of the difference.
A · US 2006/0190259 A1 · ¶ 6
According to another aspect of the present invention, the Equal Error Rate (EER) of a speech recognizer is reduced by correcting calculation results on confidence levels with weights assigned to respective frames and, therefore, reduce a false response rate.
A · US 2006/0190259 A1 · ¶ 9
Therefore, in view of the above, there is a need for methods and arrangements providing an improved estimation of a level of overtaking intention of a vehicle driver.
B · US 2015/0266486 A1 · ¶ 5
Since the level of overtaking intention is estimated on the basis of the gaze direction, an arrangement is provided with an improved capability to estimate a level of overtaking intention of a driver of a vehicle hosting the arrangement.
B · US 2015/0266486 A1 · ¶ 8
Cite aspatent.fyi, US 2026/0356385 A1, claim 1, first published 2026-11-20T01:05:31-05:00, https://patent.fyi/US20260356385/, SHA-256 2b0d3588161a553158c258d19fa872e0d0eb50c9d6e0bf10881dc51c9876a041
First published
Permanent addresshttps://patent.fyi/US20260356385/
Content fingerprint (SHA-256)2b0d3588161a553158c258d19fa872e0d0eb50c9d6e0bf10881dc51c9876a041
Methodv2026-11-18.2 · automated, not reviewed by a person
Correctionscorrections@patent.fyi