Granted Patent B2

US 12,596,015 · App. 18/497,540 · Granted Apr 7, 2026

Inventors: Michael Baus (Bietigheim-Bissingen, DE); Carsten Hasberg (Ilsfeld-Auenstein, DE); Daniel Zaum (Sarstedt, DE)

Assignee: ROBERT BOSCH GMBH

G01C21/3848G08G1/0112

Abstract

Claims (37)

1 . A method for creating a map representation of a road traffic network for navigation of one or more vehicles forming a vehicle fleet, the method comprising the following steps:

continually repeating the following set of steps:

receiving environment sensor data from environment sensors of a plurality of vehicles of the vehicle fleet, using an external computing unit, wherein the environment sensor data reproduce roads of the road traffic network traveled by each of the plurality of vehicles;

carrying out a simultaneous localization and mapping (SLAM) method on the received environment sensor data and ascertaining a SLAM map representation of the road traffic network traveled by the plurality of vehicles, using the external computing unit, the SLAM map representation including features mapped based on the environment sensor data;

receiving differential positioning data of a differential global positioning system (DGPS) of the plurality of vehicles of the vehicle fleet, using the external computing unit;

correlating the received DGPS data of the plurality of vehicles to one another and to features mapped based on the environment sensor data, using the external computing unit;

for each of two or more of the plurality of vehicles, determining a respective pose difference between (i) a SLAM pose corresponding to a respective one of the mapped features of the SLAM map representation and (ii) a pose derived from the DGPS data for the respective vehicle;

averaging, using the external computing unit, the respective pose differences determined for the two or more of the plurality of vehicles to obtain an averaged pose correction; and

modifying, using the external computing unit, one or more of the mapped features of the SLAM map representation based on the averaged pose correction, thereby generating an enriched map representation, wherein the modifying includes shifting a location of a mapped road segment in the SLAM map representation based on the averaged pose correction, and wherein the continual repetition dynamically updates the enriched map representation over time; and

continually providing the updated enriched map representation to one or more vehicles of the vehicle fleet, using the external computing unit, thereby influencing a navigation of the one or more vehicles of the vehicle fleet.

2 . The method according to claim 1 , wherein the repetition is carried out for a predefined distance between successive poses of the map representation.

3 . The method according to claim 1 , wherein the differential position data of the differential global positioning system are based on corrected propagation time information of satellite signals of the global positioning system, and wherein the corrected propagation time information is corrected in relation to propagation time information of predefined reference signals.

4 . The method according to claim 1 , wherein the environment sensor data include: camera data and/or lidar data and/or radar data.

5 . A computing unit configured to create a map representation of a road traffic network for navigation of one or more vehicles forming a vehicle fleet, the computing unit comprising a processor that is programmed to:

continually repeat the following set of steps:

receive environment sensor data from environment sensors of a plurality of vehicles of the vehicle fleet, using an external computing unit, wherein the environment sensor data reproduce roads of the road traffic network traveled by each of the plurality of vehicles;

carry out a simultaneous localization and mapping (SLAM) method on the received environment sensor data and ascertain a SLAM map representation of the road traffic network traveled by the plurality of vehicles, using the external computing unit, the SLAM map representation including features mapped based on the environment sensor data;

receive differential positioning data of a differential global positioning system (DGPS) of the plurality of vehicles of the vehicle fleet, using the external computing unit;

correlate the received DGPS data of the plurality of vehicles to one another and to features mapped based on the environment sensor data, using the external computing unit;

for each of two or more of the plurality of vehicles, determining a respective pose difference between (i) a SLAM pose corresponding to a respective one of the mapped features of the SLAM map representation and (ii) a pose derived from the DGPS data for the respective vehicle;

average, using the external computing unit, the respective pose differences determined for the two or more of the plurality of vehicles to obtain an averaged pose correction; and

modify, using the external computing unit, one or more of the mapped features of the SLAM map representation based on the averaged pose correction, thereby generating an enriched map representation, wherein the modification includes a shift of a location of a mapped road segment in the SLAM map representation based on the averaged pose correction, and wherein the continual repetition dynamically updates the enriched map representation over time; and

continually provide the updated enriched map representation to one or more vehicles of the vehicle fleet, using the external computing unit, thereby influencing a navigation of the one or more vehicles of the vehicle fleet.

6 . A non-transitory computer-readable medium on which is stored a computer program including commands for creating a map representation of a road traffic network for navigation of one or more vehicles forming a vehicle fleet, the commands, when executed by a data processor, causing the data processor to perform the following steps:

continually repeating the following set of steps:

receiving environment sensor data from environment sensors of a plurality of vehicles of the vehicle fleet, using an external computing unit, wherein the environment sensor data reproduce roads of the road traffic network traveled by each of the plurality of vehicles;

carrying out a simultaneous localization and mapping (SLAM) method on the received environment sensor data and ascertaining a SLAM map representation of the road traffic network traveled by the plurality of vehicles, using the external computing unit, the SLAM map representation including features mapped based on the environment sensor data;

receiving differential positioning data of a differential global positioning system (DGPS) of the plurality of vehicles of the vehicle fleet, using the external computing unit;

correlating the received DGPS data of the plurality of vehicles to one another and to features mapped based on the environment sensor data, using the external computing unit;

for each of two or more of the plurality of vehicles, determining a respective pose difference between (i) a SLAM pose corresponding to a respective one of the mapped features of the SLAM map representation and (ii) a pose derived from the DGPS data for the respective vehicle;

averaging, using the external computing unit, the respective pose differences determined for the two or more of the plurality of vehicles to obtain an averaged pose correction; and

modifying, using the external computing unit, one or more of the mapped features of the SLAM map representation based on the averaged pose correction, thereby generating an enriched map representation, wherein the modifying includes shifting a location of a mapped road segment in the SLAM map representation based on the averaged pose correction, and wherein the continual repetition dynamically updates the enriched map representation over time; and

continually providing the updated enriched map representation to one or more vehicles of the vehicle fleet, using the external computing unit, thereby influencing a navigation of the one or more vehicles of the vehicle fleet.

7 . The method according to claim 1 , wherein the respective pose difference determined for each of the two or more of the plurality of vehicles includes a position difference and/or an orientation difference between the SLAM pose and the pose derived from the DGPS data.

8 . The method according to claim 1 , wherein the respective pose difference determined for each of the two or more of the plurality of vehicles includes a lateral difference and/or a longitudinal difference between the SLAM pose and the pose derived from the DGPS data.

9 . The method according to claim 1 , wherein the averaging is weighted based on an accuracy value associated with the DGPS data of each of the vehicles of the plurality of vehicles.

10 . The method according to claim 1 , wherein the correlating includes identifying features of the road traffic network that are commonly traveled by two or more of the plurality of vehicles.

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (10)