/Patent Tracker
SamsungGranted patentCars & transportInterest 6/10

Samsung patents AI system for real-time autonomous parking

Samsung has been granted a patent for a method that uses a convolutional neural network to plan parking paths in real time. The system processes map data, obstacle locations, and the car's current position to predict the best route to a destination.

Why it mattersThis technology could improve how self-driving cars handle tight parking spaces by continuously adjusting their path based on immediate surroundings.

In the filing’s own words

A device and method with continuous real-time autonomous parking planning and control are disclosed. A method of training a path distribution estimation model includes: obtaining initial information including map information, departure information of an autonomous driving device, and destination information of the autonomous driving device; obtaining paths through which the autonomous driving device is movable by inputting the initial information to the path distribution estimation model which predicts the paths; and training the path distribution estimation model to output a path distribution corresponding to the paths based on the initial information.

What exactly they claim (claim 1)

1. A method comprising: obtaining, by a processor, initial information comprising map information including obstacle information, departure information of an autonomous driving device, and destination information of the autonomous driving device; obtaining, by the processor, paths through which the autonomous driving device is movable in real time, by inputting the initial information to the path distribution estimation model which outputs a path distribution corresponding to the paths based on a sampling probability distribution of an optimized path; and training, by the processor, the path distribution estimation model to output the path distribution corresponding to the paths, which includes an optimal committed trajectory targeting a destination location based on the initial information wherein the path distribution estimation model comprises a convolutional neural network (CNN) model, wherein the path distribution estimation model is configured to input the initial information in a form of a 5-channel grid map, which comprising (i) an unknown grid map associated with a trajectory of the autonomous driving device, (ii) an occupancy grid map in which a path may be generated, (iii) tree route information including an information on the path through which the autonomous driving device is to move, (iv) the destination information of the autonomous driving device, and (v) a committed trajectory information that the autonomous driving device is currently following, and wherein the path distribution estimation model is configured to output the path distribution in a form of a 4-channel grid map, which comprising (i, ii) a respective path distribution of the paths output by the path distribution estimation model based on the initial information of the unknown grid map and the occupancy grid map, (iii) sine information of departure posture information of the autonomous driving device, and (iv) cosine information of the departure posture information of the autonomous driving device as a recurrent result.

Facts

Company
Samsung (SAMSUNG ELECTRONICS CO., LTD.)
Official title
Device and method with continuous real-time autonomous parking planning and control
Patent number
US12728836B2
Application number
18332146
Filed
Jun 9, 2023
Granted
Sep 8, 2026
Inventors
4
Claims
13

Classified under

B60W 30/06 Operations & transportB60W 60/0011 Operations & transportG06N 3/0464 Physics & computing

More from Samsung in Cars & transport

Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.