Systems and methods for monitoring LiDAR sensor health

1. A method for operating an autonomous vehicle, comprising:

obtaining, by a computing device, LiDAR point cloud data generated by a LiDAR system of the autonomous vehicle;

inspecting, by the computing device, the LiDAR point cloud data to infer a health of LiDAR beams;

identifying, by the computing device, bad quality point cloud data based on the inferred health of the LiDAR beams;

removing, by the computing device, the bad quality point cloud data from the LiDAR point cloud data to generate modified LiDAR point cloud data; and

causing, by the computing device, the autonomous vehicle to perform at least one autonomous driving operation or mode change based on the modified LiDAR point cloud data;

wherein the health of the LiDAR beam is inferred by computing a plurality of metrics based on characteristics of a LiDAR point cloud defined by the LiDAR point cloud data; and

wherein a metric of the plurality of metrics is generated by:

determining a first number representing a total number of data points in the LIDAR point cloud that have a first intensity value of a plurality of intensity values, and a second number representing a total number of data points in the LIDAR point cloud that have a different second intensity value of the plurality of intensity values;

combining the first number and the second number to obtain a third number; and

setting a value of the metric equal to the third number.

2. The method according to claim 1 , wherein the health of the LiDAR beam is inferred by further computing a confidence score based on the plurality of metrics.

3. The method according to claim 2 , wherein the health of the LiDAR beam is inferred by further classifying the LiDAR beam as a faulty beam or a good beam based on the confidence score.

4. The method according to claim 3 , wherein the LiDAR beam is classified as a faulty beam when the confidence score is less than a threshold value.

5. The method according to claim 1 , wherein the plurality of metrics comprise at least one of a shape context metric, and a height metric.

6. The method according to claim 1 , wherein the plurality of metrics are determined based on at least one of a total number of data points in the LiDAR point cloud, intensity values of the LiDAR point cloud, and z-coordinate values for data points in the LiDAR point cloud.

7. The method according to claim 1 , wherein another metric of the plurality of metrics is equal to a total number of data points in the LiDAR point cloud.

8. The method according to claim 1 , wherein another metric of the plurality of metrics is generated by:

using a z-coordinate value to compute a distance from each given data point to an average z-coordinate value;

determining a standard deviation of the z-coordinate values for data points in a LiDAR point cloud;

dividing each distance by the standard deviation to obtain a value;

computing an average of the values.

9. A system, comprising:

a processor;

a non-transitory computer-readable storage medium comprising programming instructions that are configured to cause the processor to implement a method for operating an autonomous vehicle, wherein the programming instructions comprise instructions to:

obtain LiDAR point cloud data generated by a LiDAR system of the autonomous vehicle;

inspect the LiDAR point cloud data to infer a health of LiDAR beams;

identify bad quality point cloud data based on the inferred health of the LiDAR beams;

remove the bad quality point cloud data from the LiDAR point cloud data to generate modified LiDAR point cloud data; and

cause the autonomous vehicle to perform at least one autonomous driving operation or mode change based on the modified LiDAR point cloud data;

wherein the health of the LiDAR beam is inferred by computing a plurality of metrics based on characteristics of a LiDAR point cloud defined by the LiDAR point cloud data; and

wherein a metric of the plurality of metrics is generated by:

determining a first number representing a total number of data points in the LIDAR point cloud that have a first intensity value of a plurality of intensity values, and a second number representing a total number of data points in the LIDAR point cloud that have a different second intensity value of the plurality of intensity values;

combining the first number and the second number to obtain a third number; and

setting a value of the metric equal to the third number.

10. The system according to claim 9 , wherein the health of the LiDAR beam is inferred by further computing a confidence score based on the plurality of metrics.

11. The system according to claim 10 , wherein the health of the LiDAR beam is inferred by further classifying the LiDAR beam as a faulty beam or a good beam based on the confidence score.

12. The system according to claim 11 , wherein the LiDAR beam is classified as a faulty beam when the confidence score is less than a threshold value.

13. The system according to claim 9 , wherein the plurality of metrics comprise at least one of an outlier metric, a shape context metric, and a height metric.

14. The system according to claim 9 , wherein the plurality of metrics are determined based on at least one of a total number of data points in the LiDAR point cloud, intensity values of the LiDAR point cloud, and z-coordinate values for data points in the LiDAR point cloud.

15. The system according to claim 9 , wherein another metric of the plurality of metrics is equal to a total number of data points in the LiDAR point cloud.

16. The system according to claim 9 , wherein another metric of the plurality of metrics is generated by:

using a z-coordinate value to compute a distance from each given data point to an average z-coordinate value;

determining a standard deviation of the z-coordinate values for data points in a LiDAR point cloud;

dividing each distance by the standard deviation to obtain a value;

computing an average of the values.