System and method for monitoring test data for autonomous operation of an automated vehicle

Abstract

To provide a method.SOLUTION: A method for distributing and analyzing autonomous driving vehicle test data is described. The method includes uploading driving session data from a computer at an operating site to network-attached storage at the operating site. The method also includes uploading the driving session data from the network-attached storage at the operating site to a cloud-based storage location. The method further includes distributing the driving session data and work units from the cloud-based storage location to at least one research site that is separate from the operating site. The method also includes processing the driving session data by at least one research site according to an analysis / processing task associated with the work unit.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] Certain aspects of the present disclosure relate generally to machine learning, and more particularly to systems and methods for monitoring test data from the autonomous operation of a self-driving vehicle. [Background technology]

[0002] Autonomous agents, such as self-driving cars and robots, are advancing rapidly. Self-driving cars rely on various methods to perceive their environment. Unfortunately, the various methods used by self-driving cars to perceive their surroundings are not completely reliable. In addition, because self-driving cars must interact with other vehicles, many significant concerns arise. For example, one significant concern is how to design vehicle control for autonomous vehicles using machine learning.

[0003] Unfortunately, machine learning vehicle control can be ineffective in situations involving complex interactions between vehicles (e.g., a situation where a controlled (ego) vehicle merges into a traffic lane). Machine learning techniques for vehicle control by selecting an appropriate vehicle control behavior for the ego vehicle are desired. For example, selected speed / acceleration / steering angle of the controlled (ego) vehicle can be applied as the vehicle control behavior. The autonomous test vehicle can operate according to the selected vehicle control behavior. Unfortunately, test autonomous vehicles generate a substantial amount of data (e.g., 100 gigabytes (GB)) during a test run. Systems and methods for accessing this test data from different research locations are desired. Summary of the Invention

[0004] A method for autonomous vehicle test data distribution and analysis is described. The method includes uploading driving session data from a driving site computer to a network-attached storage device (hereinafter referred to as network-attached storage) at the driving site. The method also includes uploading the driving session data from the driving site's network-attached storage to a cloud-based storage location. The method further includes distributing the driving session data and the operational units from the cloud-based storage location to at least one research site separate from the driving site. The method also includes processing, by the at least one research site, the driving session data according to analysis / processing tasks associated with the operational units.

[0005] A non-transitory computer-readable medium having recorded thereon program code for autonomous vehicle test data distribution and analysis is described. The program code is executed by a processor. The non-transitory computer-readable medium includes program code for uploading driving session data from a computer at an operating site to network-attached storage at the operating site. The non-transitory computer-readable medium also includes program code for uploading the driving session data from the network-attached storage at the operating site to a cloud-based storage location. The non-transitory computer-readable medium further includes program code for distributing the driving session data and the operational units from the cloud-based storage location to at least one research site separate from the operating site. The non-transitory computer-readable medium also includes program code for processing the driving session data by the at least one research site according to analysis / processing tasks associated with the operational units.

[0006] A system for autonomous vehicle test data distribution and analysis is described. The system includes a driving site comprising network-attached storage and a computer. The computer is configured to upload driving session data from the driving site computer to the network-attached storage in response to insertion of a session data memory module in the test vehicle. The network-attached storage is configured to upload the driving session data from the driving site's network-attached storage to a cloud-based storage location. The system's cloud-based storage location is configured to distribute the driving session data and the work units to research sites separate from the driving site. At least two of the research sites are configured to process the driving session data according to analysis / processing tasks associated with the work units received by each of the two research sites.

[0007] This has outlined broadly the features and technical advantages of the present disclosure in order that the detailed description that follows may be better understood. Additional features and advantages of the present disclosure are described below. Those skilled in the art should realize that they may readily utilize this disclosure as a basis for designing modifications or other structures for carrying out the same purposes of the present disclosure. Those skilled in the art should also realize that such equivalent constructions do not depart from the teachings of the present disclosure as set forth in the appended claims. The novel features believed characteristic of the present disclosure, both as to its organization and method of operation, together with further objects and advantages, will be better understood from the following description when considered in conjunction with the accompanying drawings. It is to be expressly understood, however, that each of the figures is provided for the purpose of illustration and description only and is not intended as a definition of the limits of the present disclosure. [Brief explanation of the drawings]

[0008] The features, nature, and advantages of the present disclosure will become more apparent from the following detailed description when considered in conjunction with the drawings in which like reference characters identify correspondingly throughout. [Figure 1] 1 illustrates an example implementation of designing neural networks using a system-on-chip (SOC) of an autonomous vehicle test data dissemination and analysis system according to aspects of the present disclosure. [Figure 2] FIG. 1 is a block diagram illustrating an example software architecture that may modularize artificial intelligence (AI) capabilities for an autonomous vehicle test data distribution and analysis system, according to aspects of the present disclosure. [Figure 3] FIG. 1 illustrates a hardware implementation for an autonomous vehicle test data distribution and analysis system, according to aspects of the present disclosure. [Figure 4] FIG. 1 illustrates an example ingest process that enables systems and methods for monitoring the status of test data in various memory locations and performing tasks on the test data, according to aspects of the present disclosure. [Figure 5] 1 illustrates an example sensor data image captured by a test autonomous vehicle operating according to a driving stack (e.g., a test vehicle application module), according to aspects of the present disclosure. [Figure 6] 1 is a flowchart illustrating a method for autonomous vehicle test data distribution and analysis according to an aspect of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0009] The detailed description set forth below, in conjunction with the accompanying drawings, is intended as a description of various configurations and is not intended to represent the only configurations in which the concepts described herein may be practiced. The detailed description includes specific details for the purpose of providing a thorough understanding of the various concepts. However, it will be apparent to those skilled in the art that these concepts may be practiced without these specific details. In some instances, well-known structures and components are shown in block diagram form to avoid obscuring such concepts.

[0010] Based on the teachings, one skilled in the art should recognize that the scope of the present disclosure is intended to cover any aspect of the present disclosure, whether implemented independently or in combination with any other aspect of the present disclosure. For example, an apparatus can be implemented using any number of the described aspects, and a method can be practiced using any number of the described aspects. In addition, the scope of the present disclosure is intended to cover such apparatuses or methods that are practiced using other structure, function, or structure and function in addition to or other than the various aspects of the present disclosure described. It should be understood that any aspect of the present disclosure disclosed can be embodied by one or more elements of a claim.

[0011] While particular embodiments are described herein, numerous variations and permutations of these embodiments are within the scope of the present disclosure. While certain benefits and advantages of the preferred embodiments are mentioned, it is not intended that the scope of the present disclosure be limited to particular benefits, uses, or purposes. Rather, the embodiments of the present disclosure are intended to be broadly applicable to different technologies, system configurations, networks, and protocols, some of which are illustrated by way of example in the drawings and description of the preferred embodiments below. The detailed description and drawings are merely illustrative of the present disclosure and do not limit the scope of the present disclosure, which is defined by the appended claims and their equivalents.

[0012] Autonomous agents, such as self-driving cars and robots, are advancing rapidly. Self-driving cars rely on various methods to perceive their environment. Unfortunately, the various methods used by self-driving cars to perceive their surroundings are not completely reliable. In addition, because self-driving cars must interact with other vehicles, many significant concerns arise. For example, one significant concern is how to test the vehicle control of an autonomous vehicle using machine learning.

[0013] Automated vehicle control on major highways is advancing rapidly. These automated vehicles are expected to reduce traffic accidents and improve traffic efficiency. In particular, machine learning techniques for vehicle control by selecting appropriate vehicle control behaviors for the ego vehicle are desirable. For example, selected speed / acceleration / steering angle of the controlled (ego) vehicle can be applied as the vehicle control behavior. Unfortunately, machine learning-based vehicle control can be ineffective in situations involving complex interactions between vehicles (e.g., when the controlled (ego) vehicle merges into a traffic lane).

[0014] Safety is a key concern, especially when building autonomous agents that operate in human environments. For autonomous driving in particular, safety is a significant challenge due to high speeds, rich environments, and complex dynamic interactions with numerous traffic participants, including sensitive road users. Testing and validating machine learning techniques for vehicle control by selecting appropriate vehicle control behaviors for the ego vehicle is desirable. For example, an autonomous test vehicle can operate according to the selected vehicle control behaviors. Unfortunately, the test autonomous vehicle generates a significant amount of data.

[0015] For example, a test autonomous vehicle generates a substantial amount of data (e.g., 100 gigabytes (GB)) during a test run. Transmitting such a large amount of data is difficult and labor-intensive. Furthermore, storing the data in a cloud-based storage device for use by researchers may not be practical due to the substantial time required to download and / or process the data according to the researchers' needs. Therefore, the data should ideally be located close to the researchers. Systems and methods for accessing this test data at different research locations are desired.

[0016] Aspects of the present disclosure are directed to systems and methods for monitoring the status of test data in various memory storage locations and performing tasks on the test data. Aspects of the present disclosure provide a test data pipeline that distributes relevant data from a test autonomous vehicle to a researcher's location for rapid use by the researcher. The testing method begins when the test autonomous vehicle completes a test run and enters a garage. In this example, the sensor data memory module is removed from the test autonomous vehicle and inserted into a computer at the garage.

[0017] In this aspect of the disclosure, sensor data is provided to a cloud-based repository and removed from the source site (e.g., the garage of the test autonomous vehicle). The raw sensor data can be stored in the cloud-based repository. Additionally, any irrelevant information can be removed from the sensor data. However, a complete, unfiltered version of the sensor data remains in the cloud-based repository. The filtered information is then provided to a processing pipeline, which operates as a distributed computing network. This aspect of the disclosure is directed to the ability to collect and rapidly distribute autonomous vehicle test data in a data pipeline so that it is readily available to researchers.

[0018] According to this aspect of the disclosure, the processing pipeline allows each researcher (or research site) to receive sensor data along with a processing unit. That is, the researcher provides processing power to process the sensor data as soon as it is ingested and receives the information. The processing (e.g., operational) units can correspond to specialized analysis or processing tasks. For example, one research site may be responsible for indexing the sensor data and making it compatible with keyword searches. Thus, an assigned operational unit may be to process the sensor data for indexing. Another research site may be responsible for performing machine learning tasks on the sensor data. Once a research site completes an operational unit, the output is distributed to a cloud-based storage location and to other research sites that can use the output.

[0019] FIG. 1 illustrates an example implementation of the above-described systems and methods for an autonomous vehicle test data distribution and analysis system using a system-on-chip (SOC) 100 of a vehicle vision system for an autonomous vehicle 140. The SOC 100 may include a single processor or a multi-core processor (e.g., a central processing unit (CPU) 102) according to certain aspects of the present disclosure. Variables (e.g., neural signals and synaptic weights), system parameters associated with computational devices (e.g., weighted neural networks), delays, frequency bin information, and task information may be stored in memory blocks. The memory blocks may be associated with a neural processing unit (NPU) 108, the CPU 102, a graphics processing unit (GPU) 104, a digital signal processor (DSP) 106, a dedicated memory block 118, or may be distributed across multiple blocks. Instructions executed by the processor (e.g., the CPU 102) may be loaded from a program memory associated with the CPU 102 or from the dedicated memory block 118.

[0020] SOC 100 may also include additional processing blocks configured to perform specific functions, such as GPU 104, DSP 106, and connectivity block 110, which may include Fourth Generation Long Term Evolution (4G LTE) connectivity, unlicensed Wi-Fi connectivity, USB connectivity, Bluetooth® connectivity, etc. Additionally, multimedia processor 112, coupled with display 130, may select warranted vehicle control actions according to display 130 illustrating, for example, a diagram of the vehicle.

[0021] In some aspects, NPU 108 can be implemented in CPU 102, DSP 106, and / or GPU 104. SOC 100 can further include a sensor processor 114, an image signal processor (ISP) 116, and / or a navigation 120, which can include, for example, a global positioning system. SOC 100 can be based on an Advanced Risc Machine (ARM) instruction set, or the like. In other aspects of the present disclosure, SOC 100 can be a server computer in communication with autonomous vehicle 140. In this arrangement, autonomous vehicle 140 can include the processor and other features of SOC 100.

[0022] In this aspect of the disclosure, the instructions loaded into the processor (e.g., CPU 102) or NPU 108 of the autonomous vehicle 140 may include code for uploading driving session data from a driving site computer to network-attached storage at the driving site based on images captured by the sensor processor 114. The instructions loaded into the processor (e.g., CPU 102) may also include code for uploading the driving session data from the driving site's network-attached storage to a cloud-based storage location in response to images captured by the sensor processor 114. The instructions loaded into the processor (e.g., CPU 102) may also include code for distributing the driving session data and the task units from the cloud-based storage location to at least one research site separate from the driving site. The instructions loaded into the processor (e.g., CPU 102) may also include code for processing the driving session data by the at least one research site according to analysis / processing tasks associated with the task units.

[0023] FIG. 2 is a block diagram illustrating a software architecture 200 that can modularize artificial intelligence (AI) functionality for an autonomous vehicle test data distribution and analysis system, according to an embodiment of the present disclosure. Using the architecture, a planning / controller application 202 can be designed such that it can have various processing blocks of the SOC 220 (e.g., the CPU 222, the DSP 224, the GPU 226, and / or the NPU 228) perform supporting computations during runtime operation of the planning / controller application 202. While FIG. 2 describes a software architecture 200 for autonomous vehicle test data distribution and analysis, it should be appreciated that vehicle test data distribution and analysis is not limited to autonomous agents. According to an embodiment of the present disclosure, the vehicle test data distribution and analysis functionality is applicable to any vehicle type.

[0024] The planning / controller application 202 can be configured to invoke functions defined in user space 204 that can, for example, provide vehicle test data distribution and analysis services. The planning / controller application 202 can request compiled program code associated with libraries defined in a test data pipeline application programming interface (API) 206. The test data pipeline API 206 is configured to distribute test sensor data to a cloud-based repository and remove test data from a source site (e.g., a test autonomous vehicle garage) provided to the test data pipeline API 206. In response, the compiled code of the test data analysis API 207 enables each researcher (or research site) to receive sensor data with a processing unit. That is, the researcher receives information and provides processing power to process the sensor data as it is ingested. The processing (e.g., work) unit can handle specialized analysis or processing tasks associated with the test data analysis API 207.

[0025] A runtime engine 208, which may be compiled code of a runtime framework, may further be accessible to the planning / controller application 202. The planning / controller application 202 may cause the runtime engine 208 to take action, for example, for vehicle test data distribution and analysis of sensor data from a test autonomous vehicle. When the ego vehicle encounters a safety situation, the runtime engine 208 may accordingly signal an operating system 210, such as a Linux® kernel 212 running on the SOC 220. FIG. 2 illustrates the Linux® kernel 212 as an example software architecture for autonomous vehicle test data distribution and analysis. However, it should be appreciated that aspects of the present disclosure are not limited to this example software architecture. For example, other kernels may provide a software architecture to support autonomous vehicle test data distribution and analysis functions.

[0026] Operating system 210 can result in computations being performed in CPU 222, DSP 224, GPU 226, NPU 228, or some combination thereof. CPU 222 can be directly accessible by operating system 210, and other processing blocks can be accessed through drivers, such as drivers 214-218, to DSP 224, GPU 226, or NPU 227. In the example shown, the deep neural network can be configured to run on a combination of processing blocks, such as CPU 222 and GPU 226, or can run on NPU 228, if present.

[0027] The increasing complexity of software in autonomous vehicles makes it more difficult to ensure the reliability of these autonomous vehicles. For example, even with improved comprehensive safety measures, the risk of unexpected catastrophic failure remains. Safety is a critical concern, especially when building autonomous agents that operate in human environments. Safety is a significant challenge, especially for autonomous driving, due to high speeds, rich environments, and complex dynamic interactions with numerous traffic participants, including sensitive road users. Testing and validating machine learning techniques for vehicle control by selecting appropriate vehicle control behaviors for the ego vehicle is desirable. For example, an autonomous test vehicle can operate according to the selected vehicle control behavior. Unfortunately, test autonomous vehicles generate a considerable amount of data.

[0028] For example, a test autonomous vehicle generates a substantial amount of data (e.g., 100 gigabytes (GB)) during a test run. Transmitting such a large amount of data is difficult and labor-intensive. Furthermore, storing the data in a cloud-based storage device for use by researchers may be impractical due to the substantial time required to download and / or process the data according to the researchers' needs. Therefore, the data should ideally be located close to the researchers. Systems and methods for accessing this test data at different research locations are desired.

[0029] Aspects of the present disclosure are directed to systems and methods for monitoring the status of test data in various memory storage locations and performing tasks on the test data. Aspects of the present disclosure provide a test data pipeline that distributes relevant data from a test autonomous vehicle to a researcher's location for rapid use by the researcher. The testing method begins when the test autonomous vehicle completes a test run and enters a garage. In this example, the sensor data memory module is removed from the test autonomous vehicle and inserted into a computer at the garage.

[0030] According to this aspect of the disclosure, the processing pipeline allows each researcher (or research site) to receive sensor data along with a processing unit. That is, the researcher provides processing power to process the sensor data as soon as it is ingested and receives the information. The processing (e.g., operational) units can correspond to specialized analysis or processing tasks. For example, one research site may be responsible for indexing the sensor data and making it compatible with keyword searches. Thus, an assigned operational unit may be to process the sensor data for indexing. Another research site may be responsible for performing machine learning tasks on the sensor data. Once a research site completes an operational unit, the output is distributed to a cloud-based storage location and to other research sites that can use the output.

[0031] FIG. 3 illustrates a hardware implementation for an autonomous vehicle test data distribution and analysis system 300 according to an embodiment of the present disclosure. The autonomous vehicle test data distribution and analysis system 300 can be configured to enhance testing of ego vehicles using distributed analysis and processing of driving test session data from an origin site (hereinafter referred to as the origin site). The autonomous vehicle test data distribution and analysis system 300 includes a test agent control system 301, which may be a component of a vehicle, a robotic device, or other non-autonomous device (e.g., a non-autonomous vehicle, a ride-sharing vehicle, etc.). For example, as shown in FIG. 3, the test agent control system 301 is a component of a test autonomous vehicle 350. Aspects of the present disclosure are not limited to the test agent control system 301 being a component of the test autonomous vehicle 350. Other devices, such as a bus, a motorcycle, or other similar non-autonomous vehicles, are also contemplated for implementing the test agent control system 301. In this example, the test autonomous vehicle 350 may be autonomous or semi-autonomous, although other configurations for the test autonomous vehicle 350 are also contemplated.

[0032] Test agent control system 301 can be implemented with an interconnect architecture, generally represented by interconnect 346. Interconnect 346 can include any number of point-to-point interconnects, buses, and / or bridges, depending on the particular application of test agent control system 301 and the overall design constraints. Interconnect 346 can couple together various circuits, including one or more processors and / or hardware modules, represented by sensor module 302, vehicle perception module 310, processor 320, computer-readable medium 322, communication module 324, on-board unit 326, position module 328, movement module 329, planning module 330, and controller module 340. Interconnect 346 can also couple various other circuits, such as timing sources, peripherals, voltage regulators, power management circuitry, etc., which are well known in the art and will not be described further.

[0033] Test agent control system 301 includes a transceiver 342 coupled to sensor module 302, vehicle perception module 310, processor 320, computer-readable medium 322, communication module 324, on-board unit 326, position module 328, movement module 329, planning module 330, and controller module 340. Transceiver 342 is coupled to antenna 344. Transceiver 342 communicates with various other devices over a transmission medium. For example, transceiver 342 can receive commands via transmission from a user or a connected vehicle. In this example, transceiver 342 can send information for vehicle perception module 310 to and receive information from connected vehicles within the vicinity of test autonomous vehicle 350.

[0034] Test agent control system 301 includes processor 320 coupled to computer-readable medium 322. Processor 320 performs processes, including executing software stored on computer-readable medium 322, to provide functionality according to the present disclosure. When executed by processor 320, the software causes test agent control system 301 to perform various functions described for autonomous vehicle test data distribution and analysis for test autonomous vehicle 350 or any of the modules (e.g., 302, 310, 324, 328, 329, 330, and / or 340). Computer-readable medium 322 can also be used to store data processed by processor 320 when executing the software.

[0035] The sensor module 302 may obtain measurements via different sensors, such as a first sensor 306 and a second sensor 304. The first sensor 306 may be a visual sensor for capturing 2D images (e.g., a stereoscopic camera or a red / green / blue (RGB) camera). The second sensor 304 may be a ranging sensor, such as a light detection and ranging (LIDAR) sensor or a radio detection and ranging (RADAR) sensor. Of course, aspects of the present disclosure are not limited to the above sensors, as other types of sensors (e.g., thermal, ultrasonic, and / or laser) are also contemplated for either the first sensor 306 or the second sensor 304.

[0036] Measurements from first sensor 306 and second sensor 304 may be processed by processor 320, sensor module 302, vehicle perception module 310, communication module 324, on-board unit 326, position module 328, movement module 329, planning module 330, and / or controller module 340. In conjunction with computer-readable medium 322, measurements from first sensor 306 and second sensor 304 are processed to implement the functionality described herein. In one configuration, data obtained by first sensor 306 and second sensor 304 may be transmitted to a connected vehicle via transceiver 342. First sensor 306 and second sensor 304 may be coupled to or in communication with test autonomous vehicle 350.

[0037] Location module 328 can determine the location of test autonomous vehicle 350. For example, location module 328 can use a Global Positioning System (GPS) to determine the location of test autonomous vehicle 350. Location module 328 can implement a Dedicated Short Range Communications (DSRC) compliant GPS unit. A DSRC compliant GPS unit includes hardware and software that causes test autonomous vehicle 350 and / or location module 328 to comply with one or more of the following DSRC standards: DSRC standards include EN 12253:2004 Dedicated short range communications - Physical layer using 5.8 GHz microwave (review); EN 12795:2002 Dedicated short range communications (DSRC) - DSRC data link layer: medium access and logical link control (review); EN 12834:2002 Dedicated short range communications - Application layer (review); EN 13372:2004 Dedicated short range communications (DSRC) - DSRC profile for RTTT applications (review); and any derivatives or branches of EN ISO 14906:2004 Electronic toll collection application interface.

[0038] The communications module 324 can facilitate communications via the transceiver 342. For example, the communications module 324 can be configured to provide communications capabilities via different wireless protocols, such as 5G New Radio (NR), Wi-Fi, Long Term Evolution (LTE), 4G, 3G, etc. The communications module 324 can also communicate with other components of the test autonomous vehicle 350 that are not modules of the test agent control system 301. The transceiver 342 can be a communications channel through a network access point 360. The communications channel can include DSRC, LTE, LTE-D2D, mmWave, Wi-Fi (infrastructure mode), Wi-Fi (ad hoc mode), visible light communications, TV white space communications, satellite communications, full-duplex wireless communications, or any other wireless communications protocol, such as those mentioned herein.

[0039] Test agent control system 301 also includes a planning module 330 for planning a route and a controller module 340 for controlling the movement of test autonomous vehicle 350 via a movement module 329 for autonomous operation of test autonomous vehicle 350. In one configuration, controller module 340 can override a user input when the user input is expected (predicted) to cause a collision according to the autonomy level of test autonomous vehicle 350. The modules may be software modules operating in processor 320 and residing / stored on computer-readable medium 322 and / or hardware modules coupled to processor 320, or some combination thereof.

[0040] The National Highway Traffic Safety Administration (NHTSA) defines different "levels" of autonomous vehicles (e.g., Level 0, Level 1, Level 2, Level 3, Level 4, and Level 5). For example, if an autonomous vehicle has a higher level number than another autonomous vehicle (e.g., Level 3 is a higher level number than Level 2 or 1), the autonomous vehicle with the higher level number offers a greater combination and amount of autonomous functionality relative to a vehicle with a lower level number. These different levels of autonomous vehicles are briefly described below.

[0041] Level 0: In a Level 0 vehicle, the set of Advanced Driver Assistance Systems (ADAS) features installed on the vehicle does not provide control of the vehicle, but can issue warnings to the vehicle driver. Level 0 vehicles are neither autonomous nor semi-autonomous.

[0042] Level 1: In a Level 1 vehicle, the driver is ready to assume driving control of the autonomous vehicle at any time. The set of ADAS features installed on the autonomous vehicle can provide autonomous functions such as adaptive cruise control ("ACC"), park assist with automated steering, and lane keeping assist ("LKA") Type II, in any combination.

[0043] Level 2: In a Level 2 vehicle, the driver is responsible for detecting objects and events in the road environment and responding when the set of ADAS features installed on the autonomous vehicle does not respond appropriately (based on the driver's subjective judgment). The set of ADAS features installed on the autonomous vehicle may include acceleration, braking, and steering. In a Level 2 vehicle, the set of ADAS features installed on the autonomous vehicle can immediately cease activation when control is taken over by the driver.

[0044] Level 3: In a Level 3 ADAS vehicle, the driver can safely divert their attention from the driving task within a known and limited environment (such as a highway), but must still be prepared to assume control of the autonomous vehicle when necessary.

[0045] Level 4: In a Level 4 vehicle, the set of ADAS features installed on the autonomous vehicle can control the autonomous vehicle in all environments except for some environments, such as severe weather. The driver of a Level 4 vehicle can enable the automation system (which consists of the set of ADAS features installed on the vehicle) only when it is safe to do so. Once an automated Level 4 vehicle is enabled, driver attention is no longer required for the autonomous vehicle to operate safely and remain stable within acceptable standards.

[0046] Level 5: In a Level 5 vehicle, there is no human intervention other than setting the destination and activating the system. The automated system can drive to any location where driving is legal and can make its own decisions (this can vary based on the jurisdiction in which the vehicle is located).

[0047] A highly autonomous vehicle ("HAV") is an autonomous vehicle that is Level 3 or higher. Thus, in some configurations, test autonomous vehicle 350 is one of a Level 1 autonomous vehicle, a Level 2 autonomous vehicle, a Level 3 autonomous vehicle, a Level 4 autonomous vehicle, a Level 5 autonomous vehicle, and an HAV.

[0048] The vehicle perception module 310 can be in communication with the sensor module 302, the processor 320, the computer-readable medium 322, the communication module 324, the on-board unit 326, the position module 328, the movement module 329, the planning module 330, the controller module 340, and the transceiver 342. In one configuration, the vehicle perception module 310 receives sensor data from the sensor module 302. The sensor module 302 can receive sensor data from the first sensor 306 and the second sensor 304. According to aspects of the present disclosure, the sensor module 302 can perform filtering on the data to remove noise, encode the data, decode the data, fuse the data, extract frames, or perform other functions. In an alternative configuration, the vehicle perception module 310 can receive sensor data directly from the first sensor 306 and the second sensor 304.

[0049] 3 , the vehicle perception module 310 includes a test vehicle application module 312, a drive log telemetry module 314, and a session data memory module 316. The test vehicle application module 312 and the drive log telemetry module 314 may be components of the same or different artificial neural networks, such as a deep convolutional neural network (CNN). The vehicle perception module 310 is not limited to a CNN. The vehicle perception module 310 receives a data stream from the first sensor 306 and / or the second sensor 304. The data stream may include a 2D RGB image from the first sensor 306 and LIDAR data points from the second sensor 304. The data stream may include multiple frames, such as image frames of a scene.

[0050] This configuration of vehicle perception module 310 includes a test vehicle application module 312 (e.g., a driving stack) for operating a test autonomous vehicle 350 during a driving session. The driving session is categorized by stored telemetry information from driving log telemetry module 314 and sensor data from sensor module 302, which is stored in a removable session data memory module 316. In this example, test agent control system 301 and test autonomous vehicle 350 are associated with a site that includes garage machine 370, site network attach and storage (NAS) 380, and site virtual machine 390, which are further illustrated in the ingestion process shown in FIG. 4.

[0051] 4 is a diagram illustrating an ingestion process 400 for enabling systems and methods for monitoring the status of test data in various memory locations and performing tasks on the test data, according to aspects of the present disclosure. In this example, ingestion process 400 begins when test autonomous vehicle 350 completes a test run ("session") and enters a garage. In this example, session data memory module 316 is removed from test autonomous vehicle 350 and inserted into a computer (e.g., garage machine 370).

[0052] In this example, inserting session data memory module 316 into garage machine 370 triggers an upload of session data from garage machine 370 to site NAS 380 at origin site 402. In one configuration, the upload of session data to site NAS 380 triggers a message to a session data queue monitored by site virtual machine 390. In response, site virtual machine 390 uploads the session data from origin site 402 (e.g., driving site) data to cloud-based storage location 430. Additionally, the session data is removed from origin site 402 (e.g., garage machine 370 and / or site NAS 380).

[0053] The raw session data can be stored in cloud-based storage 430. Additionally, any irrelevant information can be removed from the session data. However, a complete, unfiltered version of the session data remains in cloud-based storage 430. The filtered information can be provided to data pipeline 440. Ingestion process 400 serves as a distributed computing network for distribution of the filtered information to research sites (e.g., site 1, site 2, etc.). This aspect of the present disclosure is directed to the ability to collect and rapidly distribute autonomous vehicle test data through a distributed computing network for ready use by researchers at other sites (e.g., site 1, site 2, etc.).

[0054] Ingestion process 400 provides test data distribution of relevant data from test autonomous vehicles to researcher locations for rapid use by researchers. For example, test data distribution to a first research site 410 enables the researcher (or research site) to receive the session data along with a processing unit. For example, distribution of the test data and a processing unit to a second research site 420, in response to ingestion process 400, provides the test information to the researcher along with a processing unit for providing processing power for processing the session data.

[0055] A processing (or work) unit can correspond to a specific analysis or processing task. For example, a first research site 410 can be responsible for indexing session data and making the session data compatible for keyword searches in network-attached storage at a second research site 420. Thus, an assigned work unit can be processing the session data for indexing. The second research site 420 can be responsible for performing machine learning tasks on the session data. Once a research site completes a work unit, its output is delivered to a cloud-based storage location 430 as well as to other research locations that can use the output, such as a data pipeline 440.

[0056] 5 illustrates an example sensor data image 500 captured by a test autonomous vehicle 350 for operating based on a driving stack (e.g., test vehicle application module 312) according to an embodiment of the disclosure. To illustrate the operation of test autonomous vehicle 350, consider a simple scenario illustrated by sensor data image 500 of test autonomous vehicle 350 traveling along a straight section of major highway 502.

[0057] In this example, vehicle perception module 310 determines (e.g., using LIDAR) that the only obstacle ahead of test autonomous vehicle 350 is a leading automobile 510 traveling in the same direction and 100 meters away. Test autonomous vehicle 350 is traveling toward leading automobile 510 at a speed of 10 meters per second (m / s) and is traveling at a speed of 5 meters per second (m / s). 2) which indicates a stopping distance of 10 meters, which increases to 40 meters if the speed is doubled. The controller module 340 can propose a vehicle control action to increase the ego vehicle's speed to 20 m / s. This sensor data obtained during this test driving session, along with telemetry information obtained by the driving log telemetry module 314 (e.g., using an inertial measurement unit (IMU)), is stored in the session data module for subsequent processing according to the ingestion process shown in FIG. 4.

[0058] FIG. 6 is a flowchart illustrating a method for autonomous vehicle test data distribution and analysis according to an embodiment of the disclosure. The method 600 of FIG. 6 begins at block 602, where driving session data is uploaded from a driving site computer to the driving site's network-attached storage. For example, as described in FIG. 4, inserting the session data memory module 316 into the garage machine 370 triggers an upload of the session data from the garage machine 370 to the site NAS 380 of the origin site 402. In block 604, the driving session data is uploaded from the driving site's network-attached storage to a cloud-based storage location. For example, as shown in FIG. 4, the upload of the session data to the site NAS 380 triggers a message to a session data queue monitored by the site virtual machine 390. In response, the site virtual machine 390 uploads the session data from the origin site 402 (e.g., the driving site) to the cloud-based storage location 430. Additionally, the session data is removed from the origin site 402 (eg, garage machine 370 and / or site NAS 380).

[0059] Referring again to FIG. 6 , in block 606, the driving session data and task units from the cloud-based storage location are distributed to at least one research site separate from the driving site. For example, as shown in FIG. 4 , ingestion process 400 provides test data distribution of relevant data from test autonomous vehicles to researcher locations for rapid use by researchers. For example, test data distribution to first research site 410 enables the researcher (or research site) to receive the session data along with a processing unit. For example, distribution of test data and a processing unit to second research site 420, in response to ingestion process 400, provides the test information to the researcher along with a processing unit for providing processing power for processing the session data.

[0060] In block 608, the driving session data is processed by at least one research site according to the analysis / processing task associated with the operational unit. For example, as shown in FIG. 4, a processing (or operational) unit can correspond to a specific analysis or processing task. For example, a first research site 410 can be responsible for indexing the session data and making the session data compatible with keyword searches in network-attached storage at a second research site 420. Thus, an assigned operational unit can be processing the session data for indexing. The second research site 420 can be responsible for performing machine learning tasks on the session data. Once a research site completes a operational unit, its output is distributed to a cloud-based storage location 430 and to other research locations that can use the output, such as a data pipeline 440.

[0061] Method 600 also includes uploading the processed session data from at least one study site to a cloud-based storage location. Method 600 further includes distributing the processed session data from the cloud-based storage location to at least other study sites. Method 600 also includes completing a driving session of the test autonomous vehicle. Method 600 further includes removing the session data memory module from the test autonomous vehicle. Method 600 also includes storing the driving session data from the session data memory module on a computer at the driving site.

[0062] The method further includes monitoring, by a virtual machine at the driving site, a message queue associated with the network-attached storage at the driving site. Method 600 also includes uploading, by the virtual machine at the driving site, the driving session data from the network-attached storage at the driving site to a cloud-based storage location in response to the message queue. In method 600, the operation unit may include performing machine learning on the sensor data of the driving session data. Additionally, the operation unit may be performing a filtering operation on the driving session data.

[0063] Aspects of the present disclosure are directed to systems and methods for monitoring the status of test data in various memory storage locations and performing tasks on the test data. Aspects of the present disclosure provide a test data pipeline that distributes relevant data from test autonomous vehicles to researcher locations for rapid use by the researcher.

[0064] The various operations of the methods described above may be performed by any suitable means capable of performing the corresponding functions. The means may include various hardware and / or software components and / or modules, including, but not limited to, circuits, application specific integrated circuits (ASICs), or processors. Generally, where operations are illustrated in figures, those operations may have corresponding means-plus-function components that are similarly numbered.

[0065] As used herein, the term "determining" encompasses a wide variety of actions. For example, "determining" can include calculating, computing, processing, deriving, investigating, examining (e.g., examining a table, database, or other data structure), ascertaining, and the like. Additionally, "determining" can include receiving (receiving information), accessing (e.g., accessing data in a memory), and the like. Further, "determining" can include resolving, selecting, choosing, establishing, and the like.

[0066] As used herein, a phrase referring to "at least one" of a list of items refers to any combination of those items, including a single element (item). By way of example, "at least one of a, b, or c" is intended to cover a, b, c, ab, ac, bc, and abc.

[0067] The various example logic blocks, modules, and circuits described in connection with this disclosure may be implemented or performed by a processor configured in accordance with the present disclosure, a digital signal processor (DSP), an ASIC, a field programmable gate array (FPGA), or other programmable logic device (PLD), discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A processor may be a microprocessor, but alternatively, a processor may be any commercially available processor, controller, microcontroller, or state machine specially configured as described herein. A processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration.

[0068] The steps of a method or algorithm described in connection with the present disclosure may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module may reside on any form of storage medium known in the art. Some examples of storage media that may be used include random access memory (RAM), read-only memory (ROM), flash memory, erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disk, removable disk, CD-ROM, etc. A software module may comprise a single instruction, or many instructions, and may be distributed over several different code segments, among different programs, or across multiple storage media. A storage medium may be coupled to a processor such that the processor can read information from, and write information to, the storage medium. Alternatively, the storage medium may be integral to the processor.

[0069] The methods disclosed herein comprise one or more steps or actions for achieving the described method. Method steps and / or actions may be interchanged with one another without departing from the scope of the claims. In other words, unless a specific order of steps or actions is specified, the order and / or use of specific steps and / or actions may be modified without departing from the scope of the claims.

[0070] The described functions can be implemented in hardware, software, firmware, or any combination thereof. When implemented in hardware, an example hardware configuration can comprise a processing system in an apparatus. The processing system can be implemented in a bus architecture. The bus can include any number of interconnected buses and bridges, depending on the particular application and overall design constraints of the processing system. The bus can couple various circuits together, including a processor, machine-readable media, and a bus interface. The bus interface can, among other things, connect a network adapter to the processing system via the bus. The network adapter can perform signal processing functions. For certain aspects, a user interface (e.g., keypad, display, mouse, joystick, etc.) can also be connected to the bus. The bus can also couple various other circuits, such as timing sources, peripherals, voltage regulators, power management circuits, etc., which are well known in the art and will not be described further.

[0071] The processor may be responsible for managing a bus and processing, including the execution of software stored on a machine-readable medium. Examples of processors that may be specially configured according to the present disclosure include microprocessors, microcontrollers, DSP processors, and other circuitry capable of executing software. Software should be interpreted broadly to mean instructions, data, or any combination thereof, whether referred to as software, firmware, middleware, microcode, hardware description language, or the like. The machine-readable medium may include, by way of example, RAM, flash memory, ROM, programmable read-only memory (PROM), EPROM, EEPROM, registers, magnetic disk, optical disk, hard drive, or any other suitable storage medium, or any combination thereof. The machine-readable medium may be embodied in a computer program product. The computer program product may include packaging materials.

[0072] In a hardware implementation, the machine-readable medium may be part of a processing system separate from the processor. However, as those skilled in the art will readily recognize, the machine-readable medium, or any portion thereof, may be external to the processing system. By way of example, the machine-readable medium may include a transmission line, a carrier wave modulated with data, and / or a computer product separate from the device, all of which are accessible by the processor through a bus interface. Alternatively, or in addition, the machine-readable medium, or any portion thereof, may be integrated into the processor, such as in the case of a cache and / or specialized register file. While the various components discussed may be described as having a specific location, such as a local component, certain components may also be configured in various ways, such as to be configured as part of a distributed computing system.

[0073] The processing system can be comprised of one or more microprocessors providing processor functionality and external memory providing at least a portion of the machine-readable medium, all linked together to other support circuitry through an external bus architecture. Alternatively, the processing system can include one or more neuromorphic processors for implementing the neuron model and model of the neural system described herein. As another alternative, the processing system can be implemented in an ASIC having the processor, bus interface, user interface, support circuitry, and at least a portion of the machine-readable medium integrated on a single chip, or in one or more FPGAs, PDLs, controllers, state machines, gate logic, discrete hardware components, or any other suitable circuitry or combination of circuitry capable of performing the various functions described throughout this disclosure. Those skilled in the art will recognize how best to implement the described functionality for a processing system depending on the particular application and the overall design constraints imposed on the overall system.

[0074] The machine-readable medium may include multiple software modules. The software modules contain instructions that, when executed by a processor, cause the processing system to perform various functions. The software modules may include a transmitting module and a receiving module. Each software module may reside on a single storage device or may be distributed across multiple storage devices. By way of example, a software module may be loaded into RAM from a hard drive upon a triggering event. During execution of a software module, the processor may load some of the instructions into a cache to speed up access. One or more cache lines may then be loaded into a special-purpose register file for execution by the processor. When referring to the functionality of a software module below, it will be understood that such functionality is implemented by the processor when executing instructions from that software module. Furthermore, it should be recognized that aspects of the present disclosure may result in improved functionality of a processor, computer, machine, or other system implementing such aspects.

[0075] If implemented in software, the functions may be stored on or transmitted as one or more instructions or code on a non-transitory computer-readable medium. Computer-readable media includes both computer storage media and communication media, including any medium that facilitates transfer of a computer program from one place to another. Storage media may be any available medium that can be accessed by a computer. By way of example, and not limitation, such computer-readable media may comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage devices, magnetic disk storage devices or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. Additionally, any connection is properly termed a computer-readable medium. For example, if software is transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared (IR), radio, and microwave, the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included within the definition of medium. As used herein, disk and disc include compact discs (CDs), laser discs, optical discs, digital versatile discs (DVDs), floppy disks, and Blu-ray discs, where disks typically reproduce data magnetically, while discs reproduce data optically with a laser. Thus, in some aspects, computer-readable medium can comprise non-transitory computer-readable medium (e.g., tangible medium). Additionally, for other aspects, computer-readable medium can comprise transitory computer-readable medium (e.g., a signal). Combinations of the above should also be included within the scope of computer-readable medium.

[0076] As such, certain aspects may comprise a computer program product for performing the operations presented herein. For example, such a computer program product may comprise a computer-readable medium storing (and / or encoding) instructions executable by one or more processors to perform the operations described herein. For certain aspects, the computer program product may include packaging materials.

[0077] Furthermore, it should be appreciated that modules and / or other suitable means for performing the methods and techniques described herein can be downloaded and / or obtained by a user terminal and / or, where applicable, by a base station. For example, such a device can be coupled to a server to facilitate the transfer of means for performing the methods described herein. Alternatively, the various methods described herein can be provided via storage means (e.g., RAM, ROM, physical storage media such as a CD or floppy disk, etc.), such that the user terminal and / or base station can obtain the various methods by coupling or providing the storage means to the device. Furthermore, any other suitable technique for providing the methods and techniques described herein to a device can be utilized.

[0078] It is to be understood that the claims are not limited to the precise configuration and components illustrated above. Various modifications, changes and variations may be made in the arrangement, operation and details of the methods and apparatus described above without departing from the scope of the claims.

Claims

1. 1. A method for autonomous vehicle test data dissemination and analysis, comprising: The processor: operating an autonomous driving test vehicle while a test drive is being conducted involving other vehicles; uploading driving session test data from a driving site computer to network-attached storage at the driving site, the driving session test data comprising a driving log, measured telemetry information, captured sensor data, and selected vehicle control actions during the test drive of the autonomous driving test vehicle; uploading the driving session test data from the network-attached storage at the driving site to a cloud-based storage location; filtering the driving session test data from the cloud-based storage location according to a first task unit and a first processing unit assigned to a first study site and a second task unit and a second processing unit assigned to a second study site; distributing the filtered driving session test data and a first operational unit and a first processing task to the first research site and a second operational unit and a second processing task to the second research site via a data distribution pipeline, wherein the first research site and the second research site are geographically separated and separate from the driving site; processing, by the first research site, the filtered driving session test data according to a first processing task associated with the first operational unit, and outputting, using the first processing unit, processed sensor data, the selected vehicle control actions, and the processed driving log and telemetry information to the data dissemination pipeline; automatically validating, by a second research site, the selected vehicle control action using the second processing unit according to a machine learning test validation task associated with a second operational unit according to the processed sensor data and the processed driving log and telemetry information received from the data dissemination pipeline; A method comprising:

2. uploading the processed driving session test data from the first study site to the cloud-based storage location; distributing the processed driving session test data from the cloud-based storage location to the second research site; and The method of claim 1 further comprising:

3. Completing a driving session of an autonomous driving test vehicle; removing a session data memory module from the autonomous driving test vehicle; storing the driving session test data from the session data memory module on the computer at the driving site; The method of claim 1 further comprising:

4. said uploading said driving session test data from said network attached storage further comprising: monitoring, by a virtual machine at the operating site, a message queue associated with the network-attached storage at the operating site; uploading, by the virtual machine at the driving site, the driving session test data from the network-attached storage at the driving site to the cloud-based storage location in response to the message queue; The method of claim 1 , comprising:

5. The method of claim 4 , further comprising deleting, by the virtual machine at the driving site, the driving session test data from the network-attached storage and / or the computer at the driving site.

6. The method of claim 1, wherein a third operational unit comprises a filtering operation of the driving session test data.

7. The method of claim 1 , wherein raw sensor data from the driving sensor data of the driving session test data is stored in the cloud-based storage location, and the raw sensor data is filtered at the first study site.

8. A non-transitory computer-readable medium having recorded thereon program code for autonomous vehicle test data distribution and analysis, the program code being executed by a processor; program code for operating an autonomous driving test vehicle while a test drive is being conducted involving other vehicles; program code for uploading driving session test data from a driving site computer to network-attached storage at the driving site, the driving session test data comprising a driving log, measured telemetry information, captured sensor data, and selected vehicle control actions during the test drive of the autonomous driving test vehicle; program code for uploading the driving session test data from the network-attached storage at the driving site to a cloud-based storage location; program code for filtering the driving session test data from the cloud-based storage location according to a first task unit and a first processing unit assigned to a first study site and a second task unit and a second processing unit assigned to a second study site; program code for distributing the filtered driving session test data and a first operation unit and a first processing task to the first research site and a second operation unit and a second processing task to the second research site via a data distribution pipeline, the first research site and the second research site being geographically separate and distinct from the driving site; program code for processing, by the first research site, the filtered driving session test data according to a first processing task associated with the first operational unit, and outputting, with the first processing unit, processed sensor data, the selected vehicle control action, and the processed driving log and telemetry information to the data dissemination pipeline; program code for automatically authenticating, by a second research site, using the second processing unit according to a machine learning test validation task associated with a second operational unit in accordance with the processed sensor data and the processed driving log and telemetry information received from the data dissemination pipeline; 1. A non-transitory computer-readable medium comprising:

9. program code for uploading the processed session test data from the first research site to the cloud-based storage location; program code for distributing the processed session test data from the cloud-based repository to the second research site; 10. The non-transitory computer-readable medium of claim 8, further comprising:

10. Program code for completing a driving session of an autonomous driving test vehicle; program code for removing a session data memory module from the autonomous driving test vehicle; program code for storing the driving session test data from the session data memory module on the computer at the driving site; 10. The non-transitory computer-readable medium of claim 8, further comprising:

11. The program code for uploading the driving session test data from the network attached storage comprises: program code for monitoring, by a virtual machine at the operating site, a message queue associated with the network-attached storage at the operating site; program code for uploading, by the virtual machine at the driving site, the driving session test data from the network-attached storage at the driving site to the cloud-based storage location in response to the message queue; 10. The non-transitory computer-readable medium of claim 8, comprising:

12. 12. The non-transitory computer-readable medium of claim 11, further comprising program code for deleting, by the virtual machine at the driving site, the driving session test data from the network-attached storage and / or the computer at the driving site.

13. 1. A system for autonomous vehicle test data distribution and analysis, comprising: An autonomous driving test vehicle capable of conducting test drives involving other vehicles; a driving site comprising network-attached storage and a computer, the computer configured to upload driving session test data from the computer at the driving site to the network-attached storage in response to insertion of a session data memory module of the autonomous driving test vehicle, the network-attached storage configured to upload the driving session test data from the network-attached storage at the driving site to a cloud-based storage location, the driving session test data comprising driving logs and measured telemetry information, captured sensor data, and selected vehicle control actions during the test drive of the autonomous driving test vehicle; a data distribution pipeline configured to filter the driving session test data according to a first operation unit and a first processing unit assigned to a first research site and a second operation unit and a second processing unit assigned to a second research site, and to distribute the filtered driving session test data and the first operation unit and first processing task to the first research site and the second operation unit and second processing task to the second research site, wherein the first research site and the second research site are geographically separated and separate from the driving sites; the first research site configured to process the filtered driving session test data according to a first processing task associated with the first operational unit, and output, using the first processing unit, processed sensor data, the vehicle control actions, and processed driving logs and telemetry information to the data dissemination pipeline; the second research site configured to automatically authenticate, using the second processing unit, the selected vehicle control action according to a machine learning test validation task associated with a second operational unit according to the processed sensor data and the processed driving log and telemetry information received from the data dissemination pipeline; A system comprising:

14. The system of claim 13, wherein a third operational unit comprises a filtering operation of the driving session test data.

15. The system of claim 13 , wherein raw sensor data from the driving sensor data of the driving session test data is stored in the cloud-based storage location and filters the raw sensor data of the first study site.

Images