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.
No entries.
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
Branch detection on airport service roads
A forward camera on an airport service vehicle tracks painted lane boundaries at junctions between service roads and loading bays. A controller uses steering torque and lateral motion to detect a driver-initiated lane change. The controller classifies the change as intended when the vehicle trajectory aligns with a detected branch and as unintended when the vehicle crosses a boundary without following a branch. It suppresses the lane departure tone for an intended change and issues the tone for an unintended change.
Component substitution
Lidar intensity sensing for lane branches
A scanning lidar replaces a camera as the surroundings sensor and supplies point coordinates and return intensities. The processor isolates road surface points, normalizes their intensities by range, and fits curves to clusters of bright returns from painted markings. When steering torque and lateral motion indicate a driver-initiated lane change, the processor checks whether a fitted marking splits into two curves and whether the projected path follows one branch. It classifies the change as intended when both conditions hold and triggers a steering wheel vibration for an unintended change.
Parameter or scale variation
Timed branch confirmation at low speed
A controller operates from 5 to 30 kilometers per hour and processes front and rear camera images at 20 frames per second. It tracks roadway markings up to 20 meters ahead and 8 meters behind the vehicle and confirms a bifurcation when two marking segments diverge by at least 0.5 meters and remain visible for five consecutive frames. After steering torque and lateral motion indicate a lane change, the controller classifies it as intended when the projected path follows a confirmed branch within 1.5 seconds of detection. It raises a lane departure alert when the interval expires without a branch match.
Combination with existing systems
Lane centering at roadway marking splits
A forward camera feeds a roadway marking tracker whose output includes both branch geometry and candidate lane centerlines. Steering torque and lateral motion indicate a driver-initiated lane change, and the controller classifies the change as intended when the projected path aligns with a detected branch. For an intended change, the lane centering controller reduces steering assistance during the crossing and uses the selected branch centerline once both of its boundaries are stable. For an unintended change, it retains the current centerline as its reference and issues a lane departure alert.
Adaptive cruise control at lane splits
A camera processor tracks roadway marking bifurcations and sends branch geometry to an adaptive cruise controller that receives range and relative speed from forward radar. Steering torque and lateral motion indicate a driver-initiated lane change, which the processor classifies as intended when the projected path follows a detected branch. For an intended change, the cruise controller selects the nearest radar track ahead within the selected branch and calculates acceleration from its range and relative speed. For an unintended change, the controller keeps its current lane as the target corridor and requests a lane departure alert.
Automated or data-driven variant
Learned marking segmentation and branch tracking
A convolutional neural network trained on annotated road images produces a per-pixel roadway marking probability map for each camera frame. The processor projects the marking pixels onto the road plane, connects them into curves, and tracks junctions where one curve divides into two. When steering torque and lateral motion indicate a lane change, the processor compares the projected vehicle path with the branches and classifies the change as intended if the path aligns with one. An unintended classification activates a seat vibration on the side of the crossing.
Cite aspatent.fyi, US 2026/0356368 A1, claim 1, first published 2026-11-20T00:27:46-05:00, https://patent.fyi/US20260356368/, SHA-256 55469b851b4f837f9e66e3858fd8beda0e74847267a1763a8f0cf60aa9696417