IP Library › Granted Patent US 11,427,146
Granted Patent B2
US 11,427,146 · App. 16/508,084 · Granted Aug 30, 2022

Collision protection

Inventor: Zhiling Huang (Fremont, CA)
Assignee: Pony AI Inc.
B60R21/0134G01C21/3697G06V20/58B60R2021/01345
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 11,427,146
App. No.
16/508,084
Granted
Aug 30, 2022
Kind
B2
Abstract

Systems, methods, and non-transitory computer-readable media are provided for collision prevention. In some embodiments, a trajectory of a vehicle with respect to an object can be determined using a trajectory prediction model and sensor data associated with the vehicle and the object. Based on the trajectory, a collision event involving the vehicle and the object can be determined and one or more collision protection devices can be activated.

Claims (54)

1. A collision protection system comprising:

a plurality of collision protection devices operationally coupled to a first object;

a plurality of sensors operationally coupled to the first object;

a memory storing executable instructions; and

one or more processors in communication with the plurality of collision protection devices, the plurality of sensors, and the memory, the one or more processors programmed by the executable instructions to perform a method comprising:

receiving sensor data collected by the plurality of sensors;

determining, based on the sensor data, a first trajectory of the first object with respect to a second object using a trajectory prediction model implemented based on a neural network, wherein the neural network applies one or more transformations to the sensor data, through neural layers of the neural network, to determine the first trajectory;

determining, based on the first trajectory, a collision event to occur between the first object and the second object at a contact location of the first object;

activating, prior to the collision event, at a first time, a first collision protection device based on a type and a size of the second object, the first collision protection device being at least partially on a first exterior of the first object corresponding to the contact location, wherein the activating comprises:

determining a size of the first collision protection device based on a particular segment of the second object that directly contacts the first object during the collision event, wherein a remainder of the second object besides the particular segment does not directly contact the first object during the collision event; and

determining the first time relative to the collision event based on a duration to activate the first collision protection device;

determining that the collision event exceeds coverage of the first collision protection device; and

activating, at a second time, a second collision protection device, the second collision protection device being at least partially on a second exterior of the first object adjacent to the first exterior.

2. The system of claim 1 , wherein the plurality of sensors comprises at least one of a Light Detection and Ranging (Lidar) sensor, a Radio Detection and Ranging (Radar) sensor, an imaging device, an ultrasonic sensor, a vehicle telemetry sensor, an inertial measurement unit (IMU), or a global positioning system (GPS) sensor.

3. The system of claim 1 , wherein the first object is a first motor vehicle, wherein the second object is a second motor vehicle, and wherein the contact location is at a rear location of the first motor vehicle.

4. The system of claim 1 , wherein the trajectory prediction model is based on a machine learning model.

5. The system of claim 4 , wherein the machine learning model comprises at least one of a dynamical model, a maneuver-based model, an interaction aware model, or a combination thereof.

6. The system of claim 1 , wherein determining the first trajectory of the first object comprises:

determining the first trajectory of the first object in a reference frame of the first object using the trajectory prediction model based on the sensor data.

7. The system of claim 1 , wherein determining the first trajectory of the first object comprises:

determining the first trajectory of the first object in an absolute reference frame;

determining a second trajectory of the second object in the absolute reference frame; and

determining the first trajectory of the first object with respect to the second trajectory of the second object in the absolute reference frame.

8. The system of claim 1 , wherein the first collision protection device covers at least partially the contact location when activated.

9. The system of claim 1 , wherein the plurality of collision protection devices comprises at least one of an inflatable device, a thrusting device, or an object protector.

10. The system of claim 1 , wherein activating the first collision protection device comprises:

activating, at a first activation time, a first activation mechanism of the first collision protection device; and

activating, at a second activation time, a second activation mechanism of the first collision protection device, wherein the second activation time is after the first activation time.

11. The system of claim 1 , wherein the first collision protection device is activated within a first duration, and wherein the second collision protection device is activated within a second duration.

12. The system of claim 1 , wherein the first time to activate the first collision protection device and the second time to activate the second collision protection device are identical.

13. The system of claim 1 , wherein the first time to activate the first collision protection device is prior to the second time to activate the second collision protection device.

14. The system of claim 1 , wherein the one or more processors are further programmed by the executable instructions to perform:

determining the first time to activate the first collision protection device based on:

one or more properties of the first collision protection device;

one or more properties of the collision event;

one or more properties of the first object;

one or more properties of the second object; and

the contact location.

15. The system of claim 1 , wherein the one or more processors are further programmed by the executable instructions to perform:

selecting a collision protection device from the plurality of collision protection devices based on the contact location.

16. The system of claim 1 , wherein the one or more processors are further programmed by the executable instructions to perform:

selecting the first collision protection device from the plurality of collision protection devices based on:

one or more properties of the collision event;

one or more properties of the first object;

one or more properties of the second object; and

the contact location.

17. The system of claim 1 , wherein the one or more processors are further programmed by the executable instructions to perform:

determining, based on the sensor data, a desirable first trajectory of the first object using the trajectory prediction model; and

causing the first trajectory of the first object to change based on the desirable first trajectory of the first object.

18. The system of claim 1 , wherein the one or more processors are further programmed by the executable instructions to perform:

transmitting information relating to the collision event to a service provider; and

causing the service provider to replace the first collision protection device.

19. The system of claim 1 , wherein the neural network comprises at least one of a normalization layer, a convolutional layer, a soft sign layer, a rectified linear layer, a concatenation layer, a pooling layer, a recurrent layer, an inception-like layer, or a combination thereof.

20. The system of claim 1 , wherein, in response to the second object being a person, the activating comprises setting a rate of expansion of the first collision protection device to below a threshold rate.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Sep 6, 2019
From: HUANG, ZHILING
To: PONY AI INC. C/O OSIRIS INTERNATIONAL CAYMAN LIMITED
Reel/Frame 050297/0968 →
Continuity (1)
Related Publication 20210009060A1 · Jan 14, 2021