1.1Constraint mapping via lidar
A lidar sensor scans the work area to generate a point cloud. The point cloud is processed to identify physical boundaries and zones. These zones are converted into digital constraint levels for the control system.
Generated automatically. Not reviewed by a person. Not legal advice or an opinion.
Questions about individual parts of claim 1: what if that part was like this, and not that.
A lidar sensor scans the work area to generate a point cloud. The point cloud is processed to identify physical boundaries and zones. These zones are converted into digital constraint levels for the control system.
A machine learning model is trained on historical traverse data and terrain maps. The model predicts future engageable surface portions based on the current trajectory. These predictions include estimated constraint levels for the upcoming path.
An operator selects the desired operating mode on a touchscreen display. The display shows visual representations of the work area and constraint zones. The selected mode is transmitted to the control unit as a digital signal.
Proportional hydraulic valves control the flow of hydraulic fluid to the actuator cylinders. The control unit sends electrical signals to the valves based on the predicted constraint levels. The valves adjust the cylinder extension to engage or disengage the implement.
A gps receiver determines the current position of the work machine. The position is overlaid on a digital map containing constraint data. The system calculates the upcoming engageable surface portions and their constraint levels based on the map.
Further arrangements of the same subject matter, described here as of the publication date at the foot of this page.
An autonomous underwater vehicle carries a sonar array defining a working length and width. The vehicle continuously predicts seabed surface portions to be traversed and their constraint levels, such as obstacle density or sediment type. The vehicle automatically adjusts the sonar array's engagement depth based on the predicted constraints and the operating mode.
An autonomous floor cleaning robot carries a brush assembly defining a working length and width. The robot continuously predicts floor surface portions to be traversed and their constraint levels, such as obstacle proximity or floor type. The robot automatically adjusts the brush assembly's engagement pressure based on the predicted constraints and the operating mode.
A laser rangefinder replaces the standard radar sensor for detecting engagement constraints. The rangefinder continuously scans the work area ahead of the implement assembly. The control system uses the rangefinder data to predict engageable surface portions and their constraint levels, triggering automatic actuation of the implement assembly.
The work implement assembly has a working length of 0.5 meters and a working width of 0.2 meters. The system predicts engageable surface portions at a frequency of 100 hertz. The implement assembly actuates within 50 milliseconds based on the predicted constraints and the operating mode.
The arrangement integrates with a global navigation satellite system receiver. The receiver provides continuous position data of the work machine. The control system uses this position data to map engagement constraints and predict engageable surface portions, automatically actuating the implement assembly accordingly.
A trained machine learning model predicts constraint levels for engageable surface portions. The model processes sensor data from the work machine and historical work area data. The control system uses the model's predictions to determine the operating mode and automatically actuate the implement assembly.
patent.fyi, US 12,796,585 B2, claim 1, first published 2026-11-18T13:13:23-05:00, https://patent.fyi/US12796585/, SHA-256 698a21d4b865c23dc28231ebabc56a2bc5dce1d884a9ffb494ee776547080717