Granted Patent B2

US 12,485,909 · App. 18/006,005 · Granted Dec 2, 2025

Inventors: Peter Sautter (Lauffen, DE); Georg Mayer (Ditzingen, DE); Ulrich Kersken (Diekholzen, DE); Folko Flehmig (Stuttgart, DE); Christian Connette (Leonberg, DE)

Assignee: Robert Bosch GmbH

B60W50/029B60W50/0097B60W2050/0292

Abstract

Claims (91)

1 . A computer-implemented method for determining a trajectory for controlling a vehicle, the vehicle being equipped with a sensor system configured to sense an environment of the vehicle and with a vehicle computer configured to process sensor data and control the vehicle, the method comprising:

receiving sensor data, generated by the sensor system, in a control module of the vehicle computer;

inputting the sensor data into a safety algorithm configured to detect safety-relevant objects based on the sensor data;

inputting the sensor data into a comfort algorithm configured to detect comfort-relevant objects based on the sensor data, the comfort algorithm different from the safety algorithm, and the safety-relevant objects different from the comfort-relevant objects;

estimating future states of the detected safety-relevant objects using an environment model representing the environment of the vehicle;

estimating future states of the detected comfort-relevant objects using the environment model;

storing and tracking the detected safety-relevant objects and the comfort-relevant objects over time in the environment model;

applying safety rules to the environment model to calculate a safety trajectory based on the estimated future states of the detected safety-relevant objects;

applying comfort rules to the environment model to calculate a comfort trajectory based on the estimated future states of the detected comfort-relevant objects, the comfort rules different from the safety rules, and the comfort trajectory different from the safety trajectory;

checking whether the calculated comfort trajectory satisfies the safety rules;

using the calculated comfort trajectory to control the vehicle when the calculated comfort trajectory satisfies the safety rules and the comfort rules; and

using the calculated safety trajectory to control the vehicle when the calculated comfort trajectory does not satisfy the safety rules.

2 . The method according to claim 1 , wherein:

further outputs of the safety algorithm are input into the comfort algorithm; and

the comfort algorithm is configured to detect the comfort-relevant objects further based on the outputs of the safety algorithm.

3 . The method according to claim 1 , wherein the comfort algorithm has been trained by machine learning to detect at least one of the comfort-relevant objects and the safety-relevant objects based on the sensor data.

4 . The method according to claim 3 , wherein the comfort algorithm is based on an artificial neural network.

5 . The method according to claim 1 , wherein:

a further safety trajectory, configured to transfer the vehicle into a safe state, is calculated based on further safety rules based on the estimated future states of the detected safety-relevant objects and the comfort-relevant objects; and

the further safety trajectory is used to control the vehicle when the calculating of the safety trajectory fails.

6 . The method according to claim 1 , further comprising:

receiving the sensor data in a further control module of the vehicle computer, the further control module being supplied with electrical power independently of the control module;

inputting the sensor data into the safety algorithm using the further control module;

estimating, using the further control module, further future states of the detected safety-relevant objects and comfort-relevant objects using the environment model;

calculating, using the further control module, a further safety trajectory for transferring the vehicle into a safe state taking into account further safety rules based on the estimated further future states of the detected safety-relevant objects and comfort-relevant objects;

checking, using the further control module, whether the control module is operational; and

using the further safety trajectory to control the vehicle using the further control module in response to the control module not being operational.

7 . The method according to claim 1 , wherein the method is performed by a computer program comprising instructions that are executed by the vehicle computer.

8 . The method according to claim 7 , wherein the computer program is stored on a non-transitory computer-readable medium.

9 . The method according to claim 1 , wherein:

the detected comfort-relevant objects include a puddle, and

the detected safety-relevant objects include a vehicle and a road marking.

10 . The method according to claim 9 , wherein:

the calculated comfort trajectory circumvents the puddle, and

the calculated safety trajectory passes through the puddle.

11 . The method according to claim 10 , wherein:

the vehicle will meet an oncoming vehicle while circumventing the puddle and, therefore, the calculated comfort trajectory is determined not to satisfy the safety rules.

12 . The method according to claim 1 , wherein:

the safety rules include a first specified distance, a first specified orientation, and/or a first specified relative speed of the vehicle with respect to the detected safety-relevant objects; and

the comfort rules include a second specified distance, a second specified orientation, and/or a second specified relative speed of the vehicle with respect to the detected comfort-relevant objects.

13 . A vehicle computer for determining a trajectory for controlling a vehicle, the vehicle computer being configured to:

receive sensor data generated by a sensor system of the vehicle;

input the sensor data into a safety algorithm configured to detect safety-relevant objects based on the sensor data;

input the sensor data into a comfort algorithm configured to detect comfort-relevant objects based on the sensor data, the comfort algorithm different from the safety algorithm, and the safety-relevant objects different from the comfort-relevant objects;

estimate future states of the detected safety-relevant objects using an environment model representing an environment of the vehicle;

estimate future states of the detected comfort-relevant objects using the environment model;

store and track the detected safety-relevant objects and the comfort-relevant objects over time in the environment model;

apply safety rules to the environment model to calculate a safety trajectory based on the estimated future states of the detected safety-relevant objects;

apply comfort rules to the environment model to calculate a comfort trajectory based on the estimated future states of the detected comfort-relevant objects, the comfort rules different from the safety rules, and the comfort trajectory different from the safety trajectory;

check whether the calculated comfort trajectory satisfies the safety rules;

use the calculated comfort trajectory to control the vehicle in response to the calculated comfort trajectory satisfying the safety rules; and

use the calculated safety trajectory to control the vehicle in response to the calculated comfort trajectory not satisfying the safety rules.

14 . The vehicle computer according to claim 13 comprising:

a control module configured to (i) receive the sensor data, (ii) input the sensor data into the safety algorithm, (iii) input the sensor data into the comfort algorithm, (iv) estimate the future states, (v) calculate the safety trajectory, (vi) check whether the comfort trajectory satisfies the safety rules, (vii) use the comfort trajectory to control the vehicle, and (viii) use the safety trajectory to control the vehicle; and

at least one further control module supplied with electrical power independently of the control module, the at least one further control module configured to:

receive the sensor data;

inputting the sensor data into the safety algorithm;

estimate further future states of the detected safety-relevant objects using the environment model;

estimate further future states of the detected comfort-relevant objects using the environment model;

calculate a further safety trajectory for transferring the vehicle into a safe state based on further safety rules based on the estimated further future states of the detected safety-relevant objects and the estimated further future states of the detected comfort-relevant objects;

checking, using the further control module, whether the control module is operational; and

using the further safety trajectory to control the vehicle using the further control module in response to the control module not being operational.

15 . The vehicle computer according to claim 14 , wherein:

the control module comprises software modules configured to perform operations thereof;

the at least one further control module comprises software modules configured to perform operations thereof; and

the software modules of the at least one further control module are at least partially copies of the software modules of the control module.

16 . The vehicle computer according to claim 13 , comprising:

a safety level;

a comfort level; and

a redundancy level,

wherein each of the safety level, the comfort level, and the redundancy level are separate domains of the vehicle computer including corresponding software modules.

17 . The vehicle computer according to claim 16 , wherein:

the environment model is located in the safety level and includes a secure subgraph,

data of the secure subgraph is used both in the safety level and in the redundancy level, and

a safety access module is located in the safety level and is configured to control access to the data of the secure subgraph in the safety level.

18 . The vehicle computer according to claim 17 , wherein:

a safety prediction module is located in the safety level and is configured to receive the data of the secure subgraph from the safety access module, and

the safety prediction module is configured to estimate future states of the environment model.

19 . A vehicle system comprising:

a sensor system configured to sense an environment of a vehicle and to generate corresponding sensor data; and

a vehicle computer configured to receive the sensor data of the sensor system and to control the vehicle, the vehicle computer further configured to:

input the sensor data into a safety algorithm configured to detect safety-relevant objects based on the sensor data;

input the sensor data into a comfort algorithm configured to detect comfort-relevant objects based on the sensor data, the comfort algorithm different from the safety algorithm, and the safety-relevant objects different from the comfort-relevant objects;

estimate future states of the detected safety-relevant objects using an environment model representing the environment of the vehicle;

estimate future states of the detected comfort-relevant objects using the environment model;

store and track the detected safety-relevant objects and the comfort-relevant objects over time in the environment model;

apply safety rules to the environment model to calculate a safety trajectory based on the estimated future states of the detected safety-relevant objects;

apply comfort rules to the environment model to calculate a comfort trajectory based on the estimated future states of the detected comfort-relevant objects, the comfort rules different from the safety rules, and the comfort trajectory different from the safety trajectory;

check whether the calculated comfort trajectory satisfies the safety rules;

use the calculated comfort trajectory to control the vehicle in response to the calculated comfort trajectory satisfying the safety rules; and

use the calculated safety trajectory to control the vehicle in response to the calculated comfort trajectory not satisfying the safety rules.

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (30)