Calibration of inertial measurement units of vehicles using localization

1. A method, comprising:

obtaining a first measure of velocity of a vehicle based at least on a plurality of locations determined for the vehicle;

obtaining, based at least on IMU measurements of an inertial measurement unit (IMU) of the vehicle, a second measure of velocity of the vehicle; and

calibrating, based at least on the first measure of velocity and the second measure of velocity, the IMU by, at least, performing one or more of an IMU orientation transformation or an IMU translation transformation, the calibrating being performed based at least on a difference between the first measure of velocity and the second measure of velocity satisfying a threshold that is based at least on one or more physical characteristics of the vehicle.

2. The method of claim 1 , wherein the calibrating comprises:

determining an orientation of the IMU on the vehicle based at least on the first measure of velocity and the second measure of velocity,

wherein the performing the IMU orientation transformation is based at least on the orientation of the IMU.

3. The method of claim 1 , wherein the calibrating comprises:

determining a lever arm vector that corresponds to the IMU based at least on the first measure of velocity and the second measure of velocity, wherein the performing the IMU translation transformation is based at least on the lever arm vector.

4. The method of claim 1 , wherein one or more locations of the plurality of locations are determined, at least, by performing localization based at least on a scanned point cloud corresponding to a region through which the vehicle is traveling and a map point cloud corresponding to the region.

5. The method of claim 1 , wherein the obtaining the first measure of velocity of the vehicle comprises:

for a first point in time, determining a first location of the plurality of locations of the vehicle;

for a second point in time, determining a second location of the plurality of locations of the vehicle; and

determining the first measure of velocity of the vehicle based at least on the first location, the second location, and a measure of length of time between the first point in time and the second point in time.

6. The method of claim 1 , wherein the calibrating occurs with respect to one or more of:

data the vehicle is gathering in real time as the vehicle is driving; or

previously gathered data when the vehicle is not driving.

7. The method of claim 1 , wherein the obtaining the second measure of velocity of the vehicle is further based at least on GPS data.

8. The method of claim 1 , wherein one or more of the first measure of velocity or the second measure of velocity include one or more of: angular velocity or linear velocity.

9. The method of claim 1 , wherein the calibrating includes aggregating multiple velocity measurements.

10. The method of claim 1 , wherein one or more of the first measure of velocity or the second measure of velocity is obtained with respect to driving patterns of the vehicle associated with the vehicle turning.

11. The method of claim 1 , wherein the one or more physical characteristics of the vehicle include one or more of: a size of the vehicle or a turning radius of the vehicle.

12. A system comprising:

one or more processing units to perform operations, the operations comprising:

obtaining a first measure of velocity of a vehicle based at least on a plurality of locations determined for the vehicle;

obtaining, based at least on IMU measurements of an inertial measurement unit (IMU) of the vehicle, a second measure of velocity of the vehicle;

determining an orientation of the IMU on the vehicle based at least on the first measure of velocity and the second measure of velocity, the orientation indicating a rotation between the IMU and a reference point corresponding to the vehicle;

determining a lever arm vector that corresponds to the IMU based at least on the first measure of velocity and the second measure of velocity, the lever arm vector indicating a positional displacement between the IMU and the reference point corresponding to the vehicle; and

performing calibration of the IMU based at least on the orientation of the IMU and the lever arm vector that corresponds to the IMU, the performing of the calibration being based at least on a threshold difference between the first measure of velocity and the second measure of velocity, the threshold difference being based at least on one or more physical characteristics of the vehicle.

13. The system of claim 12 , wherein the obtaining the first measure of velocity of the vehicle comprises:

obtaining a first lidar scan comprising a first scanned point cloud corresponding to a region;

determining a first pose for the vehicle by performing localization based at least on the first scanned point cloud and a map point cloud corresponding to the region;

obtaining a second lidar scan comprising a second scanned point cloud of the region;

determining a second pose for the vehicle by performing localization based at least on the second scanned point cloud and the map point cloud; and

determining the first measure of velocity of the vehicle based at least on the first pose, the second pose, and a measure of length of a time interval associated with the first pose and the second pose.

14. The system of claim 12 , wherein the obtaining the first measure of velocity of the vehicle comprises:

for a first point in time, determining a first location of the plurality of locations of the vehicle;

for a second point in time, determining a second location of the plurality of locations of the vehicle; and

determining the first measure of velocity of the vehicle based at least on the first location, the second location, and a measure of a length of time between the first point in time and the second point in time.

15. The system of claim 12 , wherein the performing the calibration is in response to a difference between the first measure of velocity and the second measure of velocity satisfying the threshold difference.

16. The system of claim 12 , wherein the performing the calibration occurs:

as the vehicle is driving, based at least on data the vehicle is gathering in real time; or while the vehicle is not driving, based at least on previously gathered data.

17. The system of claim 12 , wherein the obtaining the second measure of velocity of the vehicle is further based on Global Positioning System (GPS) data.

18. The system of claim 12 , wherein one or more of the first measure of velocity or the second measure of velocity include one or more of: angular velocity or linear velocity.

19. The system of claim 12 , wherein the performing the calibration comprises:

an IMU orientation transformation representing the rotation between the IMU and the reference point corresponding to the vehicle; and

an IMU translation transformation representing the positional displacement between the IMU and the reference point corresponding to the vehicle.

20. The system of claim 12 , wherein the performing the calibration further comprises aggregating multiple velocity measurements.

21. The system of claim 12 , wherein one or more of the first measure of velocity or the second measure of velocity are obtained with respect to driving patterns of the vehicle associated with the vehicle turning.

22. The system of claim 12 , wherein the one or more physical characteristics of the vehicle include one or more of: a size of the vehicle or a turning radius of the vehicle.

23. A processor comprising:

processing circuitry to cause performance of operations comprising:

obtaining a first measure of velocity of a vehicle based at least on a plurality of locations determined for the vehicle;

obtaining, based at least on IMU measurements of an inertial measurement unit (IMU) of the vehicle, a second measure of velocity of the vehicle; and

calibrating, based at least on the first measure of velocity and the second measure of velocity, the IMU by, at least, performing an IMU translation transformation representing a positional displacement between the IMU and a reference point corresponding to the vehicle, the calibrating being performed based at least on a threshold difference between the first measure of velocity and the second measure of velocity, the threshold difference being based at least on one or more physical characteristics of the vehicle.

24. The processor of claim 23 , wherein the calibrating further comprises:

determining an orientation of the IMU on the vehicle based at least on the first measure of velocity and the second measure of velocity, the orientation indicating a rotation between the IMU and the reference point corresponding to the vehicle; and

performing an IMU orientation transformation based at least on the orientation of the IMU.

25. The processor of claim 23 , wherein the performing the IMU translation transformation is based at least on a lever arm vector indicating the positional displacement between the IMU and the reference point corresponding to the vehicle.

26. The processor of claim 23 , wherein one or more locations of the plurality of locations are determined, at least, by performing localization based at least on a scanned point cloud corresponding to a region through which the vehicle is traveling and a map point cloud corresponding to the region.

27. The processor of claim 23 , wherein the calibrating is performed in response to a difference between the first measure of velocity and the second measure of velocity satisfying the threshold difference.