Systems and methods for maintaining vehicle state information

1. A method for monitoring a fleet of self-driving vehicles, comprising:

autonomously navigating a self-driving vehicle in an environment;

collecting current state information using at least one sensor of the self-driving vehicle;

comparing the current state information with reference data, wherein the reference data is based on previously-acquired state information;

based on the comparing, determining whether outlier data exists within the current state information; and

in response to determining that outlier data does not exist within the current state information, updating the reference data based on the current state information;

otherwise in response to determining that outlier data does exist:

determining at least one of a type or a severity of the outlier data;

associating the determined at least one of type or severity of the outlier data to an escalation level, wherein the escalation level is associated with one or more user roles;

identifying one or more user devices associated with the user roles;

generating an alert based on the outlier data; and

transmitting a notification to the one or more user devices based on the alert.

2. The method of claim 1 , wherein:

the current state information consists of current vehicle state information pertaining to the self-driving vehicle;

the reference data comprises reference vehicle data; and

the sensor is a vehicle sensor.

3. The method of claim 2 , further comprising:

collecting current environment state information pertaining to the environment; and

comparing the current environment state information with reference environment data,

wherein identifying outlier data comprises identifying outlier data in at least one of the current vehicle state information based on the reference vehicle data and the current environment state information based on the reference environment data.

4. The method of claim 1 , wherein:

the current state information consists of current environment state information pertaining to the environment;

the reference data comprises reference environment data; and

the sensor is an environment sensor.

5. The method of claim 1 , wherein identifying the outlier data comprises:

calculating an anomaly within the current state information based on the reference data using one or more processors;

transmitting a verification message to at least one user device based on the anomaly;

receiving a verification confirmation from the at least one first monitoring device; and

determining the outlier data based on the anomaly and the verification confirmation.

6. The method of claim 1 , further comprising, prior to collecting the current state information:

collecting preliminary state information using the sensor;

comparing the preliminary state information with the reference data;

determining that the preliminary state information does not comprise preliminary outlier data; and

updating the reference data based on the preliminary state information.

7. The method of claim 1 , wherein the reference data comprises one of an average, mean, deviation, dispersion or variation of previously-acquired state information.

8. The method of claim 1 , wherein comparing the current state information with the reference data comprises using pattern recognition methods comprising one or more of k-nearest neighbor algorithms, local outlier factor algorithms, density-based techniques, sub-space and correlation-based outlier detection, classification and regression analysis, support-vector machines, association rules, recurrent neural networks and cluster analysis.

9. A system for monitoring a fleet of self-driving vehicles, comprising:

one or more self-driving vehicles each having at least one sensor for collecting current state information;

a fleet-management system in communication with the one or more self-driving vehicles;

one or more user devices in communication with the fleet-management system; and

a non-transient computer-readable media for storing reference data in communication with the fleet-management system;

wherein each of the one or more self-driving vehicles is configured to collect the current state information and transmit the current state information to the fleet-management system;

wherein the fleet-management system is configured to:

receive the reference data from the non-transient computer-readable media, wherein the reference data is based on previously-acquired state information;

receive the current state information from a self-driving vehicle of the one or more self-driving vehicles;

compare the current state information with the reference data;

based on the comparing, determine whether outlier data exists in the current state information; and

in response to determining that no outlier data exists in the current state information, updating the reference data based on the current state information;

otherwise in response to determining that outlier data does exist:

determine at least one of a type or a severity of the outlier data;

associate the determined at least one of type or severity of the outlier data to an escalation level, wherein the escalation level is associated with one or more user roles;

identify one or more user devices associated with user roles;

generate an alert based on the outlier data; and

transmit a notification to the one or more user devices based on the alert.

10. The system of claim 9 , wherein:

the current state information consists of current vehicle state information pertaining to the self-driving vehicle;

the reference data comprises reference vehicle data; and

the sensor comprises a vehicle sensor.

11. The system of claim 10 wherein:

the sensor further comprises an environment sensor;

the reference data further comprises reference environment data;

the self-driving vehicle is further configured to collect current environment state information pertaining to the environment using the environment sensor;

the fleet-management system is further configured to compare the current environment state information with the reference environment data;

the fleet-management system configured to identify the outlier data comprises the fleet-management system configured to identify the outlier data in at least one of the current vehicle state information based on the reference vehicle data and the current environment state information based on the reference environment data.

12. The system of claim 9 , wherein:

the current state information consists of current environment state information pertaining to the environment;

the reference data comprises reference environment data; and

the sensor is an environment sensor.

13. The system of claim 9 , wherein the fleet-management system configured to identify the outlier data further comprises the fleet-management system configured to:

calculate an anomaly within the current state information based on the reference data;

transmit a verification message to at least one user device selected from the one or more user devices based on the anomaly;

receive a verification confirmation from the at least one first monitoring device; and

determine the outlier data based on the anomaly and the verification confirmation.

14. The system of claim 9 , wherein the reference data comprises one of an average, mean, deviation, dispersion or variation of previously-acquired state information.

15. The system of claim 9 , wherein comparing the current state information metric information with reference data comprises using pattern recognition methods comprising one or more of k-nearest neighbor algorithms, local outlier factor algorithms, density-based techniques, sub-space and correlation-based outlier detection, classification and regression analysis, support-vector machines, association rules, recurrent neural networks and cluster analysis.

16. A method for monitoring a fleet of self-driving vehicles, comprising:

transmitting a respective mission to at least one self-driving vehicle of the fleet using a fleet-management system;

executing the respective mission by autonomously navigating the at least one self-driving vehicle according to the respective mission;

collecting current state information from the at least one self-driving vehicle based on the executing the respective mission;

determining current fleet-performance metric information based on the current state information;

comparing the current fleet-performance metric information with reference data, wherein the reference data is based on previously-acquired state information;

based on the comparing, determining whether outlier data exists in the current fleet-performance metric information;

in response to determining that no outlier data exists in the current fleet-performance metric information, updating the reference data based on current fleet-performance metric information;

otherwise in response to determining that outlier data exists:

determining at least one of a type or a severity of the outlier data;

associating the at least one of type or severity of the outlier data to an escalation level, wherein the escalation level is associated with one or more user roles;

identifying one or more user devices associated with user roles; and

generating an alert based on the outlier data; and

transmitting a notification to the one or more user device based on the alert.

17. The method of claim 16 , wherein identifying the outlier data comprises:

calculating an anomaly within the current state information based on the reference data using one or more processors;

transmitting a verification message to at least one user device based on the anomaly;

receiving a verification confirmation from the at least one user device; and

determining the outlier data based on the anomaly and the verification confirmation.

18. The method of claim 16 , further comprising, prior to transmitting the respective mission to the at least one self-driving vehicle:

transmitting a respective preliminary mission to the at least one self-driving vehicle of the fleet using the fleet-management system;

executing the respective preliminary mission;

collecting preliminary state information from the at least one self-driving vehicle based on the executing the respective preliminary mission;

determining preliminary fleet-performances metric information based on the preliminary state information;

comparing the preliminary fleet-performance metric information with the reference data;

determining that the preliminary fleet-performance metric information does not comprise preliminary outlier data; and

updating the reference data based on the preliminary fleet-performance metric information.

19. The method of claim 16 , wherein the reference data comprises one of an average, mean, deviation, dispersion or variation of previously-acquired state information.

20. The method of claim 16 , wherein comparing the current fleet-performance metric information with the reference data comprises using pattern recognition comprises using pattern recognition methods comprising one or more of k-nearest neighbor algorithms, local outlier factor algorithms, density-based techniques, sub-space and correlation-based outlier detection, classification and regression analysis, support-vector machines, association rules, recurrent neural networks and cluster analysis.