Handling input data errors in an autonomous vehicle

1. A method of handling input data errors in an autonomous vehicle using predictive inputs, the method comprising:

determining an error in input data for a model of a plurality of models of an automation system of the autonomous vehicle;

generating, in response to the error, predicted input data for the model;

generating, by providing the predicted input data to the model instead of the input data, output data for the model, wherein the output data comprises an indication that the output data was generated based on the predicted input data instead of the input data;

providing the output data comprising the indication as input data to an other model;

generating, by the other model and based on the input data, other output data by modifying, in response to receiving the indication, one or more thresholds or one or more confidence scores of the other model used in generating the other output data;

causing the autonomous vehicle to perform one or more driving decisions based on the other output data of the other model.

2. The method of claim 1 , wherein the plurality of models comprises a chain of models for generating the one or more driving decisions.

3. The method of claim 1 , wherein generating the predicted input data comprises generating, based on prior input data to the model, the predicted input data.

4. The method of claim 1 , wherein the input data comprises first sensor data from a first sensor, and wherein generating the predicted input data for the model comprises generating, based on second sensor data from a second sensor, the predicted input data for the model.

5. The method of claim 4 , wherein the second sensor comprises one or more of a different fidelity sensor relative to the first sensor or a different sensor type than the first sensor.

6. The method of claim 1 , wherein the other output data comprises another indication that the other output was based on some portion of the predicted input data.

7. The method of claim 6 , wherein the model comprises a first model, the other model comprises a second model, and the method further comprises providing the other output data to a third model configured to modify one or more other thresholds or one or more other confidence scores of the third model responsive to the other indication being included in the other output data.

8. An apparatus for handling input data errors in an autonomous vehicle using predictive inputs, the apparatus configured to perform steps comprising:

determining an error in input data for a model of a plurality of models of an automation system of the autonomous vehicle;

generating, in response to the error, predicted input data for the model;

generating, by providing the predicted input data to the model instead of the input data, output data for the model, wherein the output data comprises an indication that the output data was generated based on the predicted input data instead of the input data;

providing the output data comprising the indication as input data to an other model;

generating, by the other model and based on the input data, other output data by modifying, in response to receiving the indication, one or more thresholds or one or more confidence scores of the other model used in generating the other output data; and

causing the autonomous vehicle to perform one or more driving decisions based on the other output data of the other model.

9. The apparatus of claim 8 , wherein the plurality of models comprises a chain of models for generating the one or more driving decisions.

10. The apparatus of claim 8 , wherein generating the predicted input data comprises generating, based on prior input data to the model, the predicted input data.

11. The apparatus of claim 8 , wherein the input data comprises first sensor data from a first sensor, and wherein generating the predicted input data for the model comprises generating, based on second sensor data from a second sensor, the predicted input data for the model.

12. The apparatus of claim 11 , wherein the second sensor comprises one or more of a different fidelity sensor relative to the first sensor or a different sensor type than the first sensor.

13. The apparatus of claim 8 , wherein the other output data comprises another indication that the other output was based on some portion of the predicted input data.

14. An autonomous vehicle for handling input data errors in an autonomous vehicle using predictive inputs, the autonomous vehicle comprising an apparatus configured to perform steps comprising:

determining an error in input data for a model of a plurality of models of an automation system of the autonomous vehicle;

generating, in response to the error, predicted input data for the model;

generating, by providing the predicted input data to the model instead of the input data, output data for the model, wherein the output data comprises an indication that the output data was generated based on the predicted input data instead of the input data;

providing the output data comprising the indication as input data to an other model;

generating, by the other model and based on the input data, other output data by modifying, in response to receiving the indication, one or more thresholds or one or more confidence scores of the other model used in generating the other output data; and

causing the autonomous vehicle to perform one or more driving decisions based on the other output data of the other model.

15. The autonomous vehicle of claim 14 , wherein the plurality of models comprises a chain of models for generating the one or more driving decisions.

16. The autonomous vehicle of claim 14 , wherein generating the predicted input data comprises generating, based on prior input data to the model, the predicted input data.

17. The autonomous vehicle of claim 14 , wherein the input data comprises first sensor data from a first sensor, and wherein generating the predicted input data for the model comprises generating, based on second sensor data from a second sensor, the predicted input data for the model.

18. The autonomous vehicle of claim 14 , wherein the other output data comprises another indication that the other output was based on some portion of the predicted input data.

19. A computer program product disposed upon a non-transitory computer readable medium, the computer program product comprising computer program instructions for handling input data errors in an autonomous vehicle using predictive inputs that, when executed, cause a computer system of the autonomous vehicle to carry out the steps of:

determining an error in input data for a model of a plurality of models of an automation system of the autonomous vehicle;

generating, in response to the error, predicted input data for the model;

generating, by providing the predicted input data to the model instead of the input data, output data for the model, wherein the output data comprises an indication that the output data was generated based on the predicted input data instead of the input data;

providing the output data comprising the indication as input data to an other model;

generating, by the other model and based on the input data, other output data by modifying, in response to receiving the indication, one or more thresholds or one or more confidence scores of the other model used in generating the other output data; and

causing the autonomous vehicle to perform one or more driving decisions based on the other output data of the other model.

20. The computer program product of claim 19 , wherein the other output data comprises another indication that the other output was based on some portion of the predicted input data.