Autonomous vehicle performance optimization system
An autonomous vehicle (AV) performance optimization system can receive vehicle data from human-driven vehicles, and determine a set of performance metrics for determining AV performance in relation to human performance based on the vehicle data. Thereafter, the system can receive AV data from AVs operating or configured for operation throughout a given region. Based on the AV data, the system can determine a performance score for each of the set of performance metrics for the AV.
1. An autonomous vehicle (AV) performance optimization system comprising:
one or more processors; and
one or more memory resources storing instructions that, when executed by the one or more processors, cause the one or more processors to:
in an initial phase, (i) receive vehicle data from human-driven vehicles, and (ii) utilizing the vehicle data, determine a set of performance metrics for determining AV performance in relation to human performance;
after the initial phase, receive AV data from an AV;
based on the AV data, determine a performance score for each of the set of performance metrics for the AV;
based on the performance score for each of the set of performance metrics, determine a number of deficient performance metrics in which the AV does not meet one or more performance thresholds corresponding to at least one of traffic law compliance, ride comfort, or performance etiquette;
generate a configuration package, executable by the AV, comprising a number of control system parameter adjustments for meeting the one or more performance thresholds corresponding to the deficient performance metrics; and
transmit the configuration package to the AV for execution.
2. The AV performance optimization system of claim 1 , wherein the executed instructions further cause the one or more processors to:
construct a set of performance models based on the set of performance metrics;
wherein the executed instructions cause the one or more processors to determine the performance score for each of the set of performance metrics by running the AV data through the set of performance models.
3. The AV performance optimization system of claim 1 , wherein the performance score for each of the set of performance metrics comprises a percentile indication of the AV's performance in relation to human drivers.
4. The AV performance optimization system of claim 1 , wherein the vehicle data and the AV data comprise data from an inertial measurement unit, a LIDAR, and one or more cameras.
5. The AV performance optimization system of claim 4 , wherein the vehicle data and the AV data further comprise location-based data from a global positioning system (GPS) receiver.
6. The AV performance optimization system of claim 1 , wherein execution of the configuration package by the AV causes a control system of the AV to alter a manner in which the control system interprets sensor data from an on-board sensor array to generate control inputs for execution by acceleration, braking, and steering systems of the AV.
7. The AV performance optimization system of claim 1 , wherein the executed instructions cause the one or more processors to classify individual performance metrics of the set of performance metrics as (i) traffic law compliance metrics, (ii) ride comfort metrics, or (iii) etiquette metrics.
8. A non-transitory computer-readable medium storing instructions that, when executed by one or more processors of an autonomous vehicle (AV) performance optimization system, cause the one or more processors to:
in an initial phase, (i) receive vehicle data from human-driven vehicles, and (ii) utilizing the vehicle data, determine a set of performance metrics for determining AV performance in relation to human performance;
after the initial phase, receive AV data from an AV;
based on the AV data, determine a performance score for each of the set of performance metrics for the AV;
based on the performance score for each of the set of performance metrics, determine a number of deficient performance metrics in which the AV does not meet one or more performance thresholds corresponding to at least one of traffic law compliance, ride comfort, or performance etiquette;
generate a configuration package, executable by the AV, comprising a number of control system parameter adjustments for meeting the one or more performance thresholds corresponding to the deficient performance metrics; and
transmit the configuration package to the AV for execution.
9. The non-transitory computer-readable medium of claim 8 , wherein the executed instructions further cause the one or more processors to:
construct a set of performance models based on the set of performance metrics;
wherein the executed instructions cause the one or more processors to determine the performance score for each of the set of performance metrics by running the AV data through the set of performance models.
10. The non-transitory computer-readable medium of claim 8 , wherein the performance score for each of the set of performance metrics comprises a percentile indication of the AV's performance in relation to human drivers.
11. The non-transitory computer-readable medium of claim 8 , wherein the vehicle data and the AV data comprise data from an inertial measurement unit, a LIDAR, and one or more cameras.
12. The non-transitory computer-readable medium of claim 11 , wherein the vehicle data and the AV data further comprise location-based data from a global positioning system (GPS) receiver.
13. The non-transitory computer-readable medium of claim 8 , wherein execution of the configuration package by the AV causes a control system of the AV to alter a manner in which the control system interprets sensor data from an on-board sensor array to generate control inputs for execution by acceleration, braking, and steering systems of the AV.
14. The non-transitory computer-readable medium of claim 8 , wherein the executed instructions cause the one or more processors to classify individual performance metrics of the set of performance metrics as (i) traffic law compliance metrics, (ii) ride comfort metrics, or (iii) etiquette metrics.
15. A computer-implemented method of determining performance of an autonomous vehicle (AV), the method being performed by one or more processors and comprising:
in an initial phase, (i) receiving vehicle data from human-driven vehicles, and (ii) utilizing the vehicle data, determining a set of performance metrics for determining AV performance in relation to human performance;
after the initial phase, receiving AV data from the AV;
based on the AV data, determining a performance score for each of the set of performance metrics for the AV;
based on the performance score for each of the set of performance metrics, determining a number of deficient performance metrics in which the AV does not meet one or more performance thresholds corresponding to at least one of traffic law compliance, ride comfort, or performance etiquette;
generating a configuration package, executable by the AV, comprising a number of control system parameter adjustments for meeting the one or more performance thresholds corresponding to the deficient performance metrics; and
transmitting the configuration package to the AV for execution.
16. The computer-implemented method of claim 15 , further comprising:
constructing a set of performance models based on the set of performance metrics;
wherein determining the performance score for each of the set of performance metrics comprises running the AV data through the set of performance models.
17. The computer-implemented method of claim 15 , wherein execution of the configuration package by the AV causes a control system of the AV to alter a manner in which the control system interprets sensor data from an on-board sensor array to generate control inputs for execution by acceleration, braking, and steering systems of the AV.