Machine learning-based vehicle dynamic simulator

Abstract

In an example implementation, a method includes accessing first input data representing a trajectory for traversal by an autonomous vehicle and second input data representing a first state of the autonomous vehicle, including a position, a heading and a speed of the autonomous vehicle at a first time. The method also includes estimating, using a machine learning system including a long short-term memory (LSTM) network, a second state of the autonomous vehicle based on the first and second input data, the second state of the autonomous vehicle including an estimated position, heading, and speed of the autonomous vehicle at a second time subsequent to the first time. The method also includes generating first output data representing the second state of the autonomous vehicle.

Description

[0001] Attorney Docket No. 46154-0567WO1 / 12023099

[0002] MACHINE LEARNING-BASED VEHICLE DYNAMIC SIMULATOR

[0003] CLAIM OF PRIORITY

[0004] [1] This application claims priority to U.S. Patent Application No. 63 / 735,653 filed on December 18, 2024, the entire contents of which are hereby incorporated by reference.

[0005] BACKGROUND

[0006] [2] In general, autonomous vehicles are configured to autonomously traverse from one location to another. Further, the behavior of an autonomous vehicle in response to commands can be simulated (e.g., using software) to facilitate testing and improvement of autonomous operations.

[0007] BRIEF DESCRIPTION OF THE FIGURES

[0008] [3] FIG. 1 is an example environment in which a vehicle including one or more components of an autonomous system can be implemented.

[0009] [4] FIG. 2 is a diagram of one or more systems of a vehicle including an autonomous system.

[0010] [5] FIG. 3 is a diagram of components of one or more devices and / or one or more systems of FIGS. 1 and 2.

[0011] [6] FIG. 4A is a diagram of certain components of an autonomous system.

[0012] [7] FIG. 4B is a diagram of an implementation of a neural network.

[0013] [8] FIG. 4C and 4D are a diagram illustrating example operation of a CNN.

[0014] [9] FIG. 5 is a diagram of an example system for simulating the behavior of a vehicle (e.g., an autonomous vehicle).

[0015]

[0010] FIG. 6 is a diagram of an example vehicle dynamic simulator module.

[0016]

[0011] FIG. 7 is a diagram of an example model architecture including one or more LSTM networks.

[0017]

[0012] FIG. 8 is a diagram of an example LSTM network and a corresponding rigid body kinematic module.

[0018]

[0013] FIG. 9 is a flow diagram of an example method of simulating an operation of an autonomous vehicle. Attorney Docket No. 46154-0567WO1 / 12023099

[0019] DETAILED DESCRIPTION

[0020]

[0014] In the following description numerous specific details are setforth in order to provide a thorough understanding of the present disclosure for the purposes of explanation. It will be apparent, however, that the embodiments described by the present disclosure can be practiced without these specific details. In some instances, well-known structures and devices are illustrated in block diagram form in order to avoid unnecessarily obscuring aspects of the present disclosure.

[0021]

[0015] Specific arrangements or orderings of schematic elements, such as those representing systems, devices, modules, instruction blocks, data elements, and / or the like are illustrated in the drawings for ease of description. However, it will be understood by those skilled in the art that the specific ordering or arrangement of the schematic elements in the drawings is not meant to imply that a particular order or sequence of processing, or separation of processes, is required unless explicitly described as such. Further, the inclusion of a schematic element in a drawing is not meant to imply that such element is required in all embodiments or that the features represented by such element may not be included in or combined with other elements in some embodiments unless explicitly described as such.

[0022]

[0016] Further, where connecting elements such as solid or dashed lines or arrows are used in the drawings to illustrate a connection, relationship, or association between or among two or more other schematic elements, the absence of any such connecting elements is not meant to imply that no connection, relationship, or association can exist. In other words, some connections, relationships, or associations between elements are not illustrated in the drawings so as not to obscure the disclosure. In addition, for ease of illustration, a single connecting element can be used to represent multiple connections, relationships or associations between elements. For example, where a connecting element represents communication of signals, data, or instructions (e.g., “software instructions”), it should be understood by those skilled in the art that such element can represent one or multiple signal paths (e.g., a bus), as may be needed, to affect the communication.

[0023]

[0017] Although the terms first, second, third, and / or the like are used to describe various elements, these elements should not be limited by these terms. The Attorney Docket No. 46154-0567WO1 / 12023099 terms first, second, third, and / or the like are used only to distinguish one element from another. For example, a first contact could be termed a second contact and, similarly, a second contact could be termed a first contact without departing from the scope of the described embodiments. The first contact and the second contact are both contacts, but they are not the same contact.

[0024]

[0018] The terminology used in the description of the various described embodiments herein is included for the purpose of describing particular embodiments only and is not intended to be limiting. As used in the description of the various described embodiments and the appended claims, the singular forms “a,” “an” and “the” are intended to include the plural forms as well and can be used interchangeably with “one or more” or “at least one,” unless the context clearly indicates otherwise. It will also be understood that the term “and / or” as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items. It will be further understood that the terms “includes,” “including,” “comprises,” and / or “comprising,” when used in this description specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0025]

[0019] As used herein, the terms “communication” and “communicate” refer to at least one of the reception, receipt, transmission, transfer, provision, and / or the like of information (or information represented by, for example, data, signals, messages, instructions, commands, and / or the like). For one unit (e.g., a device, a system, a component of a device or system, combinations thereof, and / or the like) to be in communication with another unit means that the one unit is able to directly or indirectly receive information from and / or send (e.g., transmit) information to the other unit. This may refer to a direct or indirect connection that is wired and / or wireless in nature. Additionally, two units may be in communication with each other even though the information transmitted may be modified, processed, relayed, and / or routed between the first and second unit. For example, a first unit may be in communication with a second unit even though the first unit passively receives information and does not actively transmit information to the second unit. As another example, a first unit may be in Attorney Docket No. 46154-0567WO1 / 12023099 communication with a second unit if at least one intermediary unit (e.g., a third unit located between the first unit and the second unit) processes information received from the first unit and transmits the processed information to the second unit. In some embodiments, a message may refer to a network packet (e.g., a data packet and / or the like) that includes data.

[0026]

[0020] As used herein, the term “if” is, optionally, construed to mean “when”, “upon”, “in response to determining,” “in response to detecting,” and / or the like, depending on the context. Similarly, the phrase “if it is determined” or “if [a stated condition or event] is detected” is, optionally, construed to mean “upon determining,” “in response to determining,” “upon detecting [the stated condition or event],” “in response to detecting [the stated condition or event],” and / or the like, depending on the context. Also, as used herein, the terms “has”, “have”, “having”, or the like are intended to be open-ended terms. Further, the phrase “based on” is intended to mean “based at least partially on” unless explicitly stated otherwise.

[0027]

[0021] Reference will now be made in detail to embodiments, examples of which are illustrated in the accompanying drawings. In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of the various described embodiments. However, it will be apparent to one of ordinary skill in the art that the various described embodiments can be practiced without these specific details. In other instances, well-known methods, procedures, components, circuits, and networks have not been described in detail so as not to unnecessarily obscure aspects of the embodiments.

[0028] General Overview

[0029]

[0022] In general, machine learning can be used to simulate vehicle dynamics, such as simulating the movement and position of a vehicle in response to operating commands.

[0030]

[0023] In an example implementation, a neural network is trained to predict the acceleration and a yaw rate of a vehicle based on an initial state of the vehicle and a commanded trajectory. The predicted acceleration of the vehicle is modified based on the physical constraints of the vehicle, and is used to predict Attorney Docket No. 46154-0567WO1 / 12023099 the change in the vehicle’s state. Successive changes to the vehicle’s state can be used to simulate the vehicle’s trajectory over time.

[0031]

[0024] The embodiments described herein can provide various technical benefits.

[0032]

[0025] As an example, embodiments of the machine learning-based vehicle dynamic simulator can be used to simulate a vehicle’s dynamics more accurately than might otherwise be possible using existing techniques.

[0033]

[0026] In general, existing vehicle dynamic simulators may use complex physics-based models to represent the kinematics of various aspects of a vehicle over time. For instance, the models can predict the behavior of a vehicle’s suspension and wheels / tires over time and in response to specific operating commands, the effects of environment conditions on the vehicle (e.g., wind, temperature, rain, etc.), and any other kinematic aspects of a vehicle’s operation. Further, the output of these models can be used to predict changes to the vehicle’s position, speed, acceleration, and heading over time. However, these models may be vehicle-specific (e.g., due to the specific components and configurations of each vehicle), and be difficult to implement in practice (e.g., due to the variation and complexity of accurately modeling each of the aspects of a vehicle’s operation and integrating the models together). Further, although the models may be simplified in order to ease implementation, the output of the models may be biased and / or less accurate due to the simplification.

[0034]

[0027] In contrast, aspects of the vehicle dynamic simulator described herein can include one or more neural networks that are trained using data sets collected from test vehicles in a wide range of contexts (e.g., different locations, operating conditions, vehicle maneuvers, etc.), and can be used to predict a vehicle’s dynamics without explicit parameter modeling. Accordingly, the neural network can be trained based on the data sets to account for a wide array of characteristics of the vehicle’s operation, even if such characteristics are not expressly identified by a human or expressly represented in a specialized kinematic model.

[0035]

[0028] Further, aspects of the vehicle dynamic simulator can include one or more long short-term memory (LSTM) networks that selectively regulate the flow of information through the network, such that information is maintained and Attorney Docket No. 46154-0567WO1 / 12023099 updated in memory over time. This allows the network to retain information over successive states that may be beneficial in modeling a vehicle’s dynamics more accurately over time (e.g., latent variables or other information that may represent changes of the characteristics of the vehicle and the context of its operation over time).

[0036]

[0029] In some implementations, the vehicle dynamic simulator can be used to facilitate the development of autonomous vehicle operations by allowing users to simulate the behavior of an autonomous vehicle in response to particular commands, without requiring that that test be performed on a physically deployed vehicle. Thus, the autonomous vehicle can be iteratively improved in a safer and more resource efficient manner. For example, a user can modify computer code that controls the operation of an autonomous vehicle, and test the effects of the modifications virtually using the vehicle dynamic simulator, rather than conducting the tests on a physically deployed vehicle. This reduces the risk of damage to the vehicle and the environment, as well as the risk of injury to any passengers of the vehicle or bystanders. Further, this reduces the consumption of physical resources needed to run a test vehicle, such as power and fuel, and reduces physical wear on the vehicle. Further, this allows a larger number of tests to be performed than might otherwise be possible using a physically deployed vehicle.

[0037]

[0030] Referring now to FIG. 1 , illustrated is example environment 100 in which vehicles that include autonomous systems, as well as vehicles that do not, are operated. As illustrated, environment 100 includes vehicles 102a-102n, objects 104a-104n, routes 106a-106n, area 108, vehicle-to-infrastructure (V2I) device 110, network 112, remote autonomous vehicle (AV) system 114, fleet management system 116, and V2I system 118. Vehicles 102a-102n, vehicle- to-infrastructure (V2I) device 110, network 112, autonomous vehicle (AV) system 114, fleet management system 1 16, and V2I system 118 interconnect (e.g., establish a connection to communicate and / or the like) via wired connections, wireless connections, or a combination of wired or wireless connections. In some embodiments, objects 104a-104n interconnect with at least one of vehicles 102a-102n, vehicle-to-infrastructure (V2I) device 110, network 112, autonomous vehicle (AV) system 114, fleet management system Attorney Docket No. 46154-0567WO1 / 12023099

[0038] 116, and V2I system 118 via wired connections, wireless connections, or a combination of wired or wireless connections.

[0039]

[0031] Vehicles 102a-102n (referred to individually as vehicle 102 and collectively as vehicles 102) include at least one device configured to transport goods and / or people. In some embodiments, vehicles 102 are configured to be in communication with V2I device 110, remote AV system 114, fleet management system 116, and / or V21 system 118 via network 112. In some embodiments, vehicles 102 include cars, buses, trucks, trains, and / or the like. In some embodiments, vehicles 102 are the same as, or similar to, vehicles 200, described herein (see FIG. 2). In some embodiments, a vehicle 200 of a set of vehicles 200 is associated with an autonomous fleet manager. In some embodiments, vehicles 102 travel along respective routes 106a-106n (referred to individually as route 106 and collectively as routes 106), as described herein. In some embodiments, one or more vehicles 102 include an autonomous system (e.g., an autonomous system that is the same as or similar to autonomous system 202).

[0040]

[0032] Objects 104a-104n (referred to individually as object 104 and collectively as objects 104) include, for example, at least one vehicle, at least one pedestrian, at least one cyclist, at least one structure (e.g., a building, a sign, a fire hydrant, etc.), and / or the like. Each object 104 is stationary (e.g., located at a fixed location for a period of time) or mobile (e.g., having a velocity and associated with at least one trajectory). In some embodiments, objects 104 are associated with corresponding locations in area 108.

[0041]

[0033] Routes 106a-106n (referred to individually as route 106 and collectively as routes 106) are each associated with (e.g., prescribe) a sequence of actions (also known as a trajectory) connecting states along which an AV can navigate. Each route 106 starts at an initial state (e.g., a state that corresponds to a first spatiotemporal location, velocity, and / or the like) and ends at a final goal state (e.g., a state that corresponds to a second spatiotemporal location that is different from the first spatiotemporal location) or goal region (e.g. a subspace of acceptable states (e.g., terminal states)). In some embodiments, the first state includes a location at which an individual or individuals are to be picked-up by the AV and the second state or region includes a location or locations at which Attorney Docket No. 46154-0567WO1 / 12023099 the individual or individuals picked-up by the AV are to be dropped-off. In some embodiments, routes 106 include a plurality of acceptable state sequences (e.g., a plurality of spatiotemporal location sequences), the plurality of state sequences associated with (e.g., defining) a plurality of trajectories. In an example, routes 106 include only high level actions or imprecise state locations, such as a series of connected roads dictating turning directions at roadway intersections. Additionally, or alternatively, routes 106 may include more precise actions or states such as, for example, specific target lanes or precise locations within the lane areas and targeted speed at those positions. In an example, routes 106 include a plurality of precise state sequences along the at least one high level action sequence with a limited lookahead horizon to reach intermediate goals, where the combination of successive iterations of limited horizon state sequences cumulatively correspond to a plurality of trajectories that collectively form the high level route to terminate at the final goal state or region.

[0042]

[0034] Area 108 includes a physical area (e.g., a geographic region) within which vehicles 102 can navigate. In an example, area 108 includes at least one state (e.g., a country, a province, an individual state of a plurality of states included in a country, etc.), at least one portion of a state, at least one city, at least one portion of a city, etc. In some embodiments, area 108 includes at least one named thoroughfare (referred to herein as a “road”) such as a highway, an interstate highway, a parkway, a city street, etc. Additionally, or alternatively, in some examples area 108 includes at least one unnamed road such as a driveway, a section of a parking lot, a section of a vacant and / or undeveloped lot, a dirt path, etc. In some embodiments, a road includes at least one lane (e.g., a portion of the road that can be traversed by vehicles 102). In an example, a road includes at least one lane associated with (e.g., identified based on) at least one lane marking.

[0043]

[0035] Vehicle-to-lnfrastructure (V2I) device 110 (sometimes referred to as a Vehicle-to-lnfrastructure or Vehicle-to-Everything (V2X) device) includes at least one device configured to be in communication with vehicles 102 and / or V2I infrastructure system 118. In some embodiments, V2I device 110 is configured to be in communication with vehicles 102, remote AV system 114, fleet Attorney Docket No. 46154-0567WO1 / 12023099 management system 116, and / or V21 system 118 via network 112. In some embodiments, V2I device 110 includes a radio frequency identification (RFID) device, signage, cameras (e.g., two-dimensional (2D) and / or three-dimensional (3D) cameras), lane markers, streetlights, parking meters, etc. In some embodiments, V2I device 110 is configured to communicate directly with vehicles 102. Additionally, or alternatively, in some embodiments V2I device 110 is configured to communicate with vehicles 102, remote AV system 114, and / or fleet management system 116 via V2I system 1 18. In some embodiments, V2I device 110 is configured to communicate with V2I system 118 via network 112.

[0044]

[0036] Network 112 includes one or more wired and / or wireless networks. In an example, network 112 includes a cellular network (e.g., a long term evolution (LTE) network, a third generation (3G) network, a fourth generation (4G) network, a fifth generation (5G) network, a code division multiple access (CDMA) network, etc.), a public land mobile network (PLMN), a local area network (LAN), a wide area network (WAN), a metropolitan area network (MAN), a telephone network (e.g., the public switched telephone network (PSTN), a private network, an ad hoc network, an intranet, the Internet, a fiber optic-based network, a cloud computing network, etc., a combination of some or all of these networks, and / or the like.

[0045]

[0037] Remote AV system 114 includes at least one device configured to be in communication with vehicles 102, V2I device 1 10, network 112, fleet management system 116, and / or V2I system 118 via network 112. In an example, remote AV system 114 includes a server, a group of servers, and / or other like devices. In some embodiments, remote AV system 114 is co-located with the fleet management system 116. In some embodiments, remote AV system 114 is involved in the installation of some or all of the components of a vehicle, including an autonomous system, an autonomous vehicle compute, software implemented by an autonomous vehicle compute, and / or the like. In some embodiments, remote AV system 114 maintains (e.g., updates and / or replaces) such components and / or software during the lifetime of the vehicle.

[0046]

[0038] Fleet management system 116 includes at least one device configured to be in communication with vehicles 102, V2I device 110, remote AV system 114, and / or V2I infrastructure system 118. In an example, fleet management Attorney Docket No. 46154-0567WO1 / 12023099 system 116 includes a server, a group of servers, and / or other like devices. In some embodiments, fleet management system 116 is associated with a ridesharing company (e.g., an organization that controls operation of multiple vehicles (e.g., vehicles that include autonomous systems and / or vehicles that do not include autonomous systems) and / or the like).

[0047]

[0039] In some embodiments, V2I system 118 includes at least one device configured to be in communication with vehicles 102, V2I device 110, remote AV system 114, and / or fleet management system 116 via network 112. In some examples, V2I system 118 is configured to be in communication with V2I device 110 via a connection different from network 112. In some embodiments, V2I system 118 includes a server, a group of servers, and / or other like devices. In some embodiments, V2I system 118 is associated with a municipality or a private institution (e.g., a private institution that maintains V2I device 110 and / or the like).

[0048]

[0040] The number and arrangement of elements illustrated in FIG. 1 are provided as an example. There can be additional elements, fewer elements, different elements, and / or differently arranged elements, than those illustrated in FIG. 1 . Additionally, or alternatively, at least one element of environment 100 can perform one or more functions described as being performed by at least one different element of FIG. 1. Additionally, or alternatively, at least one set of elements of environment 100 can perform one or more functions described as being performed by at least one different set of elements of environment 100.

[0049]

[0041] Referring now to FIG. 2, vehicle 200 (which may be the same as, or similar to vehicles 102 of FIG. 1 ) includes or is associated with autonomous system 202, powertrain control system 204, steering control system 206, and brake system 208. In some embodiments, vehicle 200 is the same as or similar to vehicle 102 (see FIG. 1). In some embodiments, autonomous system 202 is configured to confer vehicle 200 autonomous driving capability (e.g., implement at least one driving automation or maneuver-based function, feature, device, and / or the like that enable vehicle 200 to be partially or fully operated without human intervention including, without limitation, fully autonomous vehicles (e.g., vehicles that forego reliance on human intervention such as Level 5 ADS- operated vehicles), highly autonomous vehicles (e.g., vehicles that forego Attorney Docket No. 46154-0567WO1 / 12023099 reliance on human intervention in certain situations such as Level 4 ADS- operated vehicles), conditional autonomous vehicles (e.g., vehicles that forego reliance on human intervention in limited situations such as Level 3 ADS- operated vehicles) and / or the like. In one embodiment, autonomous system 202 includes operational or tactical functionality required to operate vehicle 200 in on-road traffic and perform part or all of Dynamic Driving Task (DDT) on a sustained basis. In another embodiment, autonomous system 202 includes an Advanced Driver Assistance System (ADAS) that includes driver support features. Autonomous system 202 supports various levels of driving automation, ranging from no driving automation (e.g., Level 0) to full driving automation (e.g., Level 5). For a detailed description of fully autonomous vehicles and highly autonomous vehicles, reference may be made to SAE International's standard J3016: Taxonomy and Definitions for Terms Related to On-Road Motor Vehicle Automated Driving Systems, which is incorporated by reference in its entirety. In some embodiments, vehicle 200 is associated with an autonomous fleet manager and / or a ridesharing company.

[0050]

[0042] Autonomous system 202 includes a sensor suite that includes one or more devices such as cameras 202a, LiDAR sensors 202b, radar sensors 202c, and microphones 202d. In some embodiments, autonomous system 202 can include more or fewer devices and / or different devices (e.g., ultrasonic sensors, inertial sensors, GPS receivers (discussed below), odometry sensors that generate data associated with an indication of a distance that vehicle 200 has traveled, and / or the like). In some embodiments, autonomous system 202 uses the one or more devices included in autonomous system 202 to generate data associated with environment 100, described herein. The data generated by the one or more devices of autonomous system 202 can be used by one or more systems described herein to observe the environment (e.g., environment 100) in which vehicle 200 is located. In some embodiments, autonomous system 202 includes communication device 202e, autonomous vehicle compute 202f, drive- by-wire (DBW) system 202h, and safety controller 202g.

[0051]

[0043] Cameras 202a include at least one device configured to be in communication with communication device 202e, autonomous vehicle compute 202f, and / or safety controller 202g via a bus (e.g., a bus that is the same as or Attorney Docket No. 46154-0567WO1 / 12023099 similar to bus 302 of FIG. 3). Cameras 202a include at least one camera (e.g., a digital camera using a light sensor such as a Charge-Coupled Device (CCD), a thermal camera, an infrared (IR) camera, an event camera, and / or the like) to capture images including physical objects (e.g., cars, buses, curbs, people, and / or the like). In some embodiments, camera 202a generates camera data as output. In some examples, camera 202a generates camera data that includes image data associated with an image. In this example, the image data may specify at least one parameter (e.g., image characteristics such as exposure, brightness, etc., an image timestamp, and / or the like) corresponding to the image. In such an example, the image may be in a format (e.g., RAW, JPEG, PNG, and / or the like). In some embodiments, camera 202a includes a plurality of independent cameras configured on (e.g., positioned on) a vehicle to capture images for the purpose of stereopsis (stereo vision). In some examples, camera 202a includes a plurality of cameras that generate image data and transmit the image data to autonomous vehicle compute 202f and / or a fleet management system (e.g., a fleet management system that is the same as or similar to fleet management system 116 of FIG. 1 ). In such an example, autonomous vehicle compute 202f determines depth to one or more objects in a field of view of at least two cameras of the plurality of cameras based on the image data from the at least two cameras. In some embodiments, cameras 202a is configured to capture images of objects within a distance from cameras 202a (e.g., up to 100 meters, up to a kilometer, and / or the like). Accordingly, cameras 202a include features such as sensors and lenses that are optimized for perceiving objects that are at one or more distances from cameras 202a.

[0052]

[0044] In an embodiment, camera 202a includes at least one camera configured to capture one or more images associated with one or more traffic lights, street signs and / or other physical objects that provide visual navigation information. In some embodiments, camera 202a generates traffic light data associated with one or more images. In some examples, camera 202a generates TLD (Traffic Light Detection) data associated with one or more images that include a format (e.g., RAW, JPEG, PNG, and / or the like). In some embodiments, camera 202a that generates TLD data differs from other systems described herein incorporating cameras in that camera 202a can include one or more cameras Attorney Docket No. 46154-0567WO1 / 12023099 with a wide field of view (e.g., a wide-angle lens, a fish-eye lens, a lens having a viewing angle of approximately 120 degrees or more, and / or the like) to generate images about as many physical objects as possible.

[0053]

[0045] Light Detection and Ranging (LiDAR) sensors 202b include at least one device configured to be in communication with communication device 202e, autonomous vehicle compute 202f, and / or safety controller 202g via a bus (e.g., a bus that is the same as or similar to bus 302 of FIG. 3). LiDAR sensors 202b include a system configured to transmit light from a light emitter (e.g., a laser transmitter). Light emitted by LiDAR sensors 202b include light (e.g., infrared light and / or the like) that is outside of the visible spectrum. In some embodiments, during operation, light emitted by LiDAR sensors 202b encounters a physical object (e.g., a vehicle) and is reflected back to LiDAR sensors 202b. In some embodiments, the light emitted by LiDAR sensors 202b does not penetrate the physical objects that the light encounters. LiDAR sensors 202b also include at least one light detector which detects the light that was emitted from the light emitter after the light encounters a physical object. In some embodiments, at least one data processing system associated with LiDAR sensors 202b generates an image (e.g., a point cloud, a combined point cloud, and / or the like) representing the objects included in a field of view of LiDAR sensors 202b. In some examples, the at least one data processing system associated with LiDAR sensor 202b generates an image that represents the boundaries of a physical object, the surfaces (e.g., the topology of the surfaces) of the physical object, and / or the like. In such an example, the image is used to determine the boundaries of physical objects in the field of view of LiDAR sensors 202b.

[0054]

[0046] Radio Detection and Ranging (radar) sensors 202c include at least one device configured to be in communication with communication device 202e, autonomous vehicle compute 202f, and / or safety controller 202g via a bus (e.g., a bus that is the same as or similar to bus 302 of FIG. 3). Radar sensors 202c include a system configured to transmit radio waves (either pulsed or continuously). The radio waves transmitted by radar sensors 202c include radio waves that are within a predetermined spectrum In some embodiments, during operation, radio waves transmitted by radar sensors 202c encounter a physical Attorney Docket No. 46154-0567WO1 / 12023099 object and are reflected back to radar sensors 202c. In some embodiments, the radio waves transmitted by radar sensors 202c are not reflected by some objects. In some embodiments, at least one data processing system associated with radar sensors 202c generates signals representing the objects included in a field of view of radar sensors 202c. For example, the at least one data processing system associated with radar sensor 202c generates an image that represents the boundaries of a physical object, the surfaces (e.g., the topology of the surfaces) of the physical object, and / or the like. In some examples, the image is used to determine the boundaries of physical objects in the field of view of radar sensors 202c.

[0055]

[0047] Microphones 202d includes at least one device configured to be in communication with communication device 202e, autonomous vehicle compute 202f, and / or safety controller 202g via a bus (e.g., a bus that is the same as or similar to bus 302 of FIG. 3). Microphones 202d include one or more microphones (e.g., array microphones, external microphones, and / or the like) that capture audio signals and generate data associated with (e.g., representing) the audio signals. In some examples, microphones 202d include transducer devices and / or like devices. In some embodiments, one or more systems described herein can receive the data generated by microphones 202d and determine a position of an object relative to vehicle 200 (e.g., a distance and / or the like) based on the audio signals associated with the data.

[0056]

[0048] Communication device 202e includes at least one device configured to be in communication with cameras 202a, LiDAR sensors 202b, radar sensors 202c, microphones 202d, autonomous vehicle compute 202f, safety controller 202g, and / or DBW (Drive- By- Wire) system 202h. For example, communication device 202e may include a device that is the same as or similar to communication interface 314 of FIG. 3. In some embodiments, communication device 202e includes a vehicle-to-vehicle (V2V) communication device (e.g., a device that enables wireless communication of data between vehicles).

[0057]

[0049] Autonomous vehicle compute 202f include at least one device configured to be in communication with cameras 202a, LiDAR sensors 202b, radar sensors 202c, microphones 202d, communication device 202e, safety controller 202g, and / or DBW system 202h. In some examples, autonomous vehicle compute Attorney Docket No. 46154-0567WO1 / 12023099

[0058] 202f includes a device such as a client device, a mobile device (e.g., a cellular telephone, a tablet, and / or the like), a server (e.g., a computing device including one or more central processing units, graphical processing units, and / or the like), and / or the like. In some embodiments, autonomous vehicle compute 202f is the same as or similar to autonomous vehicle compute 400, described herein. Additionally, or alternatively, in some embodiments autonomous vehicle compute 202f is configured to be in communication with an autonomous vehicle system (e.g., an autonomous vehicle system that is the same as or similar to remote AV system 114 of FIG. 1 ), a fleet management system (e.g., a fleet management system that is the same as or similar to fleet management system 116 of FIG. 1 ), a V2I device (e.g., a V2I device that is the same as or similar to V2I device 110 of FIG. 1 ), and / or a V2I system (e.g., a V2I system that is the same as or similar to V2I system 118 of FIG. 1 ).

[0059]

[0050] Safety controller 202g includes at least one device configured to be in communication with cameras 202a, LiDAR sensors 202b, radar sensors 202c, microphones 202d, communication device 202e, autonomous vehicle computer 202f, and / or DBW system 202h. In some examples, safety controller 202g includes one or more controllers (electrical controllers, electromechanical controllers, and / or the like) that are configured to generate and / or transmit control signals to operate one or more devices of vehicle 200 (e.g., powertrain control system 204, steering control system 206, brake system 208, and / or the like). In some embodiments, safety controller 202g is configured to generate control signals that take precedence over (e.g., overrides) control signals generated and / or transmitted by autonomous vehicle compute 202f.

[0060]

[0051] DBW system 202h includes at least one device configured to be in communication with communication device 202e and / or autonomous vehicle compute 202f. In some examples, DBW system 202h includes one or more controllers (e.g., electrical controllers, electromechanical controllers, and / or the like) that are configured to generate and / or transmit control signals to operate one or more devices of vehicle 200 (e.g., powertrain control system 204, steering control system 206, brake system 208, and / or the like). Additionally, or alternatively, the one or more controllers of DBW system 202h are configured to generate and / or transmit control signals to operate at least one different device Attorney Docket No. 46154-0567WO1 / 12023099

[0061] (e.g., a turn signal, headlights, door locks, windshield wipers, and / or the like) of vehicle 200.

[0062]

[0052] Powertrain control system 204 includes at least one device configured to be in communication with DBW system 202h. In some examples, powertrain control system 204 includes at least one controller, actuator, and / or the like. In some embodiments, powertrain control system 204 receives control signals from DBW system 202h and powertrain control system 204 causes vehicle 200 to make longitudinal vehicle motion, such as start moving forward, stop moving forward, start moving backward, stop moving backward, accelerate in a direction, decelerate in a direction or to make lateral vehicle motion such as performing a left turn, performing a right turn, and / or the like. In an example, powertrain control system 204 causes the energy (e.g., fuel, electricity, and / or the like) provided to a motor of the vehicle to increase, remain the same, or decrease, thereby causing at least one wheel of vehicle 200 to rotate or not rotate.

[0063]

[0053] Steering control system 206 includes at least one device configured to rotate one or more wheels of vehicle 200. In some examples, steering control system 206 includes at least one controller, actuator, and / or the like. In some embodiments, steering control system 206 causes the front two wheels and / or the rear two wheels of vehicle 200 to rotate to the left or right to cause vehicle 200 to turn to the left or right. In other words, steering control system 206 causes activities necessary for the regulation of the y-axis component of vehicle motion.

[0064]

[0054] Brake system 208 includes at least one device configured to actuate one or more brakes to cause vehicle 200 to reduce speed and / or remain stationary. In some examples, brake system 208 includes at least one controller and / or actuator that is configured to cause one or more calipers associated with one or more wheels of vehicle 200 to close on a corresponding rotor of vehicle 200. Additionally, or alternatively, in some examples brake system 208 includes an automatic emergency braking (AEB) system, a regenerative braking system, and / or the like.

[0065]

[0055] In some embodiments, vehicle 200 includes at least one platform sensor (not explicitly illustrated) that measures or infers properties of a state or a condition of vehicle 200. In some examples, vehicle 200 includes platform Attorney Docket No. 46154-0567WO1 / 12023099 sensors such as a global positioning system (GPS) receiver, an inertial measurement unit (IMU), a wheel speed sensor, a wheel brake pressure sensor, a wheel torque sensor, an engine torque sensor, a steering angle sensor, and / or the like. Although brake system 208 is illustrated to be located in the near side of vehicle 200 in FIG. 2, brake system 208 may be located anywhere in vehicle 200.

[0066]

[0056] Referring now to FIG. 3, illustrated is a schematic diagram of a device 300. As illustrated, device 300 includes processor 304, memory 306, storage component 308, input interface 310, output interface 312, communication interface 314, and bus 302. In some embodiments, device 300 corresponds to at least one device of vehicles 102 (e.g., at least one device of a system of vehicles 102), one or more devices of network 112 (e.g., one or more devices of a system of network 112), and / or any other device of the environment 100. In some embodiments, one or more devices of vehicles 102 (e.g., one or more devices of a system of vehicles 102), one or more devices of network 112 (e.g., one or more devices of a system of network 112), and / or any other device of the environment 100 include at least one device 300 and / or at least one component of device 300. As shown in FIG. 3, device 300 includes bus 302, processor 304, memory 306, storage component 308, input interface 310, output interface 312, and communication interface 314.

[0067]

[0057] Bus 302 includes a component that permits communication among the components of device 300. In some cases, processor 304 includes a processor (e.g., a central processing unit (CPU), a graphics processing unit (GPU), an accelerated processing unit (APU), and / or the like), a microphone, a digital signal processor (DSP), and / or any processing component (e.g., a field- programmable gate array (FPGA), an application specific integrated circuit (ASIC), and / or the like) that can be programmed to perform at least one function. Memory 306 includes random access memory (RAM), read-only memory (ROM), and / or another type of dynamic and / or static storage device (e.g., flash memory, magnetic memory, optical memory, and / or the like) that stores data and / or instructions for use by processor 304.

[0068]

[0058] Storage component 308 stores data and / or software related to the operation and use of device 300. In some examples, storage component 308 Attorney Docket No. 46154-0567WO1 / 12023099 includes a hard disk (e.g., a magnetic disk, an optical disk, a magneto-optic disk, a solid state disk, and / or the like), a compact disc (CD), a digital versatile disc (DVD), a floppy disk, a cartridge, a magnetic tape, a CD-ROM, RAM, PROM, EPROM, FLASH-EPROM, NV-RAM, and / or another type of computer readable medium, along with a corresponding drive.

[0069]

[0059] Input interface 310 includes a component that permits device 300 to receive information, such as via user input (e.g., a touchscreen display, a keyboard, a keypad, a mouse, a button, a switch, a microphone, a camera, and / or the like). Additionally or alternatively, in some embodiments input interface 310 includes a sensor that senses information (e.g., a global positioning system (GPS) receiver, an accelerometer, a gyroscope, an actuator, and / or the like). Output interface 312 includes a component that provides output information from device 300 (e.g., a display, a speaker, one or more lightemitting diodes (LEDs), and / or the like).

[0070]

[0060] In some embodiments, communication interface 314 includes a transceiver-like component (e.g., a transceiver, a separate receiver and transmitter, and / or the like) that permits device 300 to communicate with other devices via a wired connection, a wireless connection, or a combination of wired and wireless connections. In some examples, communication interface 314 permits device 300 to receive information from another device and / or provide information to another device. In some examples, communication interface 314 includes an Ethernet interface, an optical interface, a coaxial interface, an infrared interface, a radio frequency (RF) interface, a universal serial bus (USB) interface, a Wi-Fi® interface, a cellular network interface, and / or the like.

[0071]

[0061] In some embodiments, device 300 performs one or more processes described herein. Device 300 performs these processes based on processor 304 executing software instructions stored by a computer-readable medium, such as memory 305 and / or storage component 308. A computer-readable medium (e.g., a non-transitory computer readable medium) is defined herein as a non-transitory memory device. A non-transitory memory device includes memory space located inside a single physical storage device or memory space spread across multiple physical storage devices. Attorney Docket No. 46154-0567WO1 / 12023099

[0072]

[0062] In some embodiments, software instructions are read into memory 306 and / or storage component 308 from another computer-readable medium or from another device via communication interface 314. When executed, software instructions stored in memory 306 and / or storage component 308 cause processor 304 to perform one or more processes described herein. Additionally or alternatively, hardwired circuitry is used in place of or in combination with software instructions to perform one or more processes described herein. Thus, embodiments described herein are not limited to any specific combination of hardware circuitry and software unless explicitly stated otherwise.

[0073]

[0063] Memory 306 and / or storage component 308 includes data storage or at least one data structure (e.g., a database and / or the like). Device 300 is capable of receiving information from, storing information in, communicating information to, or searching information stored in the data storage or the at least one data structure in memory 306 or storage component 308. In some examples, the information includes network data, input data, output data, or any combination thereof.

[0074]

[0064] In some embodiments, device 300 is configured to execute software instructions that are either stored in memory 306 and / or in the memory of another device (e.g., another device that is the same as or similarto device 300). As used herein, the term “module” refers to at least one instruction stored in memory 306 and / or in the memory of another device that, when executed by processor 304 and / or by a processor of another device (e.g. , another device that is the same as or similar to device 300) cause device 300 (e.g., at least one component of device 300) to perform one or more processes described herein. In some embodiments, a module is implemented in software, firmware, hardware, and / or the like.

[0075]

[0065] The number and arrangement of components illustrated in FIG. 3 are provided as an example. In some embodiments, device 300 can include additional components, fewer components, different components, or differently arranged components than those illustrated in FIG. 3. Additionally or alternatively, a set of components (e.g., one or more components) of device 300 can perform one or more functions described as being performed by another component or another set of components of device 300. Attorney Docket No. 46154-0567WO1 / 12023099

[0076]

[0066] Referring now to FIG. 4A, illustrated is an example block diagram of an autonomous vehicle compute 400 (sometimes referred to as an “AV stack”). As illustrated, autonomous vehicle compute 400 includes perception system 402 (sometimes referred to as a perception module), planning system 404

[0077] (sometimes referred to as a planning module), localization system 406

[0078] (sometimes referred to as a localization module), control system 408

[0079] (sometimes referred to as a control module), and database 410. In some embodiments, perception system 402, planning system 404, localization system 406, control system 408, and database 410 are included and / or implemented in an autonomous navigation system of a vehicle (e.g., autonomous vehicle compute 202f of vehicle 200). Additionally, or alternatively, in some embodiments perception system 402, planning system 404, localization system 406, control system 408, and database 410 are included in one or more standalone systems (e.g., one or more systems that are the same as or similar to autonomous vehicle compute 400 and / or the like). In some examples, perception system 402, planning system 404, localization system 406, control system 408, and database 410 are included in one or more standalone systems that are located in a vehicle and / or at least one remote system as described herein. In some embodiments, any and / or all of the systems included in autonomous vehicle compute 400 are implemented in software (e.g., in software instructions stored in memory), computer hardware (e.g., by microprocessors, microcontrollers, application-specific integrated circuits (ASICs), Field Programmable Gate Arrays (FPGAs), and / or the like), or combinations of computer software and computer hardware. It will also be understood that, in some embodiments, autonomous vehicle compute 400 is configured to be in communication with a remote system (e.g., an autonomous vehicle system that is the same as or similar to remote AV system 114, a fleet management system 116 that is the same as or similar to fleet management system 116, a V2I system that is the same as or similar to V2I system 118, and / or the like).

[0080]

[0067] In some embodiments, perception system 402 receives data associated with at least one physical object (e.g., data that is used by perception system 402 to detect the at least one physical object) in an environment and classifies the at least one physical object. In some examples, perception system 402 Attorney Docket No. 46154-0567WO1 / 12023099 receives image data captured by at least one camera (e.g., cameras 202a), the image associated with (e.g., representing) one or more physical objects within a field of view of the at least one camera. In such an example, perception system 402 classifies at least one physical object based on one or more groupings of physical objects (e.g., bicycles, vehicles, traffic signs, pedestrians, and / or the like). In some embodiments, perception system 402 transmits data associated with the classification of the physical objects to planning system 404 based on perception system 402 classifying the physical objects.

[0081]

[0068] In some embodiments, planning system 404 receives data associated with a destination and generates data associated with at least one route (e.g., routes 106) along which a vehicle (e.g., vehicles 102) can travel along toward a destination. In some embodiments, planning system 404 periodically or continuously receives data from perception system 402 (e.g., data associated with the classification of physical objects, described above) and planning system 404 updates the at least one trajectory or generates at least one different trajectory based on the data generated by perception system 402. In other words, planning system 404 may perform tactical function-related tasks that are required to operate vehicle 102 in on-road traffic. Tactical efforts involve maneuvering the vehicle in traffic during a trip, including but not limited to deciding whether and when to overtake another vehicle, change lanes, or selecting an appropriate speed, acceleration, deacceleration, etc. In some embodiments, planning system 404 receives data associated with an updated position of a vehicle (e.g., vehicles 102) from localization system 406 and planning system 404 updates the at least one trajectory or generates at least one different trajectory based on the data generated by localization system 406.

[0082]

[0069] In some embodiments, localization system 406 receives data associated with (e.g., representing) a location of a vehicle (e.g., vehicles 102) in an area. In some examples, localization system 406 receives LiDAR data associated with at least one point cloud generated by at least one LiDAR sensor (e.g., LiDAR sensors 202b). In certain examples, localization system 406 receives data associated with at least one point cloud from multiple LiDAR sensors and localization system 406 generates a combined point cloud based on each of the point clouds. In these examples, localization system 406 compares the at least Attorney Docket No. 46154-0567WO1 / 12023099 one point cloud or the combined point cloud to two-dimensional (2D) and / or a three-dimensional (3D) map of the area stored in database 410. Localization system 406 then determines the position of the vehicle in the area based on localization system 406 comparing the at least one point cloud or the combined point cloud to the map. In some embodiments, the map includes a combined point cloud of the area generated prior to navigation of the vehicle. In some embodiments, maps include, without limitation, high-precision maps of the roadway geometric properties, maps describing road network connectivity properties, maps describing roadway physical properties (such as traffic speed, traffic volume, the number of vehicular and cyclist traffic lanes, lane width, lane traffic directions, or lane marker types and locations, or combinations thereof), and maps describing the spatial locations of road features such as crosswalks, traffic signs or other travel signals of various types. In some embodiments, the map is generated in real-time based on the data received by the perception system.

[0083]

[0070] In another example, localization system 406 receives Global Navigation Satellite System (GNSS) data generated by a global positioning system (GPS) receiver. In some examples, localization system 406 receives GNSS data associated with the location of the vehicle in the area and localization system 406 determines a latitude and longitude of the vehicle in the area. In such an example, localization system 406 determines the position of the vehicle in the area based on the latitude and longitude of the vehicle. In some embodiments, localization system 406 generates data associated with the position of the vehicle. In some examples, localization system 406 generates data associated with the position of the vehicle based on localization system 406 determining the position of the vehicle. In such an example, the data associated with the position of the vehicle includes data associated with one or more semantic properties corresponding to the position of the vehicle.

[0084]

[0071] In some embodiments, control system 408 receives data associated with at least one trajectory from planning system 404 and control system 408 controls operation of the vehicle. In some examples, control system 408 receives data associated with at least one trajectory from planning system 404 and control system 408 controls operation of the vehicle by generating and transmitting Attorney Docket No. 46154-0567WO1 / 12023099 control signals to cause a powertrain control system (e.g., DBW system 202h, powertrain control system 204, and / or the like), a steering control system (e.g., steering control system 206), and / or a brake system (e.g., brake system 208) to operate. For example, control system 408 is configured to perform operational functions such as a lateral vehicle motion control or a longitudinal vehicle motion control. The lateral vehicle motion control causes activities necessary for the regulation of the y-axis component of vehicle motion. The longitudinal vehicle motion control causes activities necessary for the regulation of the x-axis component of vehicle motion. In an example, where a trajectory includes a left turn, control system 408 transmits a control signal to cause steering control system 206 to adjust a steering angle of vehicle 200, thereby causing vehicle 200 to turn left. Additionally, or alternatively, control system 408 generates and transmits control signals to cause other devices (e.g., headlights, turn signal, door locks, windshield wipers, and / or the like) of vehicle 200 to change states.

[0085]

[0072] In some embodiments, perception system 402, planning system 404, localization system 406, and / or control system 408 implement at least one machine learning model (e.g., at least one multilayer perceptron (MLP), at least one convolutional neural network (CNN), at least one recurrent neural network (RNN), at least one autoencoder, at least one transformer, and / or the like). In some examples, perception system 402, planning system 404, localization system 406, and / or control system 408 implement at least one machine learning model alone or in combination with one or more of the above-noted systems. In some examples, perception system 402, planning system 404, localization system 406, and / or control system 408 implement at least one machine learning model as part of a pipeline (e.g., a pipeline for identifying one or more objects located in an environment and / or the like). An example of an implementation of a machine learning model is included below with respect to FIGS. 4B-4D.

[0086]

[0073] Database 410 stores data that is transmitted to, received from, and / or updated by perception system 402, planning system 404, localization system 406 and / or control system 408. In some examples, database 410 includes a storage component (e.g., a storage component that is the same as or similar to storage component 308 of FIG. 3) that stores data and / or software related to the operation and uses at least one system of autonomous vehicle compute 400. In Attorney Docket No. 46154-0567WO1 / 12023099 some embodiments, database 410 stores data associated with 2D and / or 3D maps of at least one area. In some examples, database 410 stores data associated with 2D and / or 3D maps of a portion of a city, multiple portions of multiple cities, multiple cities, a county, a state, a State (e.g., a country), and / or the like). In such an example, a vehicle (e.g., a vehicle that is the same as or similar to vehicles 102 and / or vehicle 200) can drive along one or more drivable regions (e.g., single-lane roads, multi-lane roads, highways, back roads, off road trails, and / or the like) and cause at least one LiDAR sensor (e.g, a LiDAR sensor that is the same as or similar to LiDAR sensors 202b) to generate data associated with an image representing the objects included in a field of view of the at least one LiDAR sensor.

[0087]

[0074] In some embodiments, database 410 can be implemented across a plurality of devices. In some examples, database 410 is included in a vehicle (e.g., a vehicle that is the same as or similar to vehicles 102 and / or vehicle 200), an autonomous vehicle system (e.g., an autonomous vehicle system that is the same as or similar to remote AV system 114, a fleet management system (e.g., a fleet management system that is the same as or similar to fleet management system 116 of FIG. 1 , a V2I system (e.g., a V2I system that is the same as or similar to V2I system 118 of FIG. 1 ) and / or the like.

[0088]

[0075] Referring now to FIG. 4B, illustrated is a diagram of an implementation of a machine learning model. More specifically, illustrated is a diagram of an implementation of a convolutional neural network (CNN) 420. For purposes of illustration, the following description of CNN 420 will be with respect to an implementation of CNN 420 by perception system 402. However, it will be understood that in some examples CNN 420 (e.g., one or more components of CNN 420) is implemented by other systems different from, or in addition to, perception system 402 such as planning system 404, localization system 406, and / or control system 408. While CNN 420 includes certain features as described herein, these features are provided for the purpose of illustration and are not intended to limit the present disclosure.

[0089]

[0076] CNN 420 includes a plurality of convolution layers including first convolution layer 422, second convolution layer 424, and convolution layer 426. In some embodiments, CNN 420 includes sub-sampling layer 428 (sometimes Attorney Docket No. 46154-0567WO1 / 12023099 referred to as a pooling layer). In some embodiments, sub-sampling layer 428 and / or other subsampling layers have a dimension (i.e. , an amount of nodes) that is less than a dimension of an upstream system. By virtue of sub-sampling layer 428 having a dimension that is less than a dimension of an upstream layer, CNN 420 consolidates the amount of data associated with the initial input and / or the output of an upstream layer to thereby decrease the amount of computations necessary for CNN 420 to perform downstream convolution operations. Additionally, or alternatively, by virtue of sub-sampling layer 428 being associated with (e.g., configured to perform) at least one subsampling function (as described below with respect to FIGS. 4C and 4D), CNN 420 consolidates the amount of data associated with the initial input.

[0090]

[0077] Perception system 402 performs convolution operations based on perception system 402 providing respective inputs and / or outputs associated with each of first convolution layer 422, second convolution layer 424, and convolution layer 426 to generate respective outputs. In some examples, perception system 402 implements CNN 420 based on perception system 402 providing data as input to first convolution layer 422, second convolution layer 424, and convolution layer 426. In such an example, perception system 402 provides the data as input to first convolution layer 422, second convolution layer 424, and convolution layer 426 based on perception system 402 receiving data from one or more different systems (e.g., one or more systems of a vehicle that is the same as or similar to vehicle 102), a remote AV system that is the same as or similar to remote AV system 114, a fleet management system that is the same as or similar to fleet management system 116, a V2I system that is the same as or similar to V2I system 118, and / or the like). A detailed description of convolution operations is included below with respect to FIG. 4C.

[0091]

[0078] In some embodiments, perception system 402 provides data associated with an input (referred to as an initial input) to first convolution layer 422 and perception system 402 generates data associated with an output using first convolution layer 422. In some embodiments, perception system 402 provides an output generated by a convolution layer as input to a different convolution layer. For example, perception system 402 provides the output of first convolution layer 422 as input to sub-sampling layer 428, second convolution Attorney Docket No. 46154-0567WO1 / 12023099 layer 424, and / or convolution layer 426. In such an example, first convolution layer 422 is referred to as an upstream layer and sub-sampling layer 428, second convolution layer 424, and / or convolution layer 426 are referred to as downstream layers. Similarly, in some embodiments perception system 402 provides the output of sub-sampling layer 428 to second convolution layer 424 and / or convolution Iayer426 and, in this example, sub-sampling layer 428 would be referred to as an upstream layer and second convolution layer 424 and / or convolution layer 426 would be referred to as downstream layers.

[0092]

[0079] In some embodiments, perception system 402 processes the data associated with the input provided to CNN 420 before perception system 402 provides the input to CNN 420. For example, perception system 402 processes the data associated with the input provided to CNN 420 based on perception system 402 normalizing sensor data (e.g., image data, LiDAR data, radar data, and / or the like).

[0093]

[0080] In some embodiments, CNN 420 generates an output based on perception system 402 performing convolution operations associated with each convolution layer. In some examples, CNN 420 generates an output based on perception system 402 performing convolution operations associated with each convolution layer and an initial input. In some embodiments, perception system 402 generates the output and provides the output as fully connected layer 430. In some examples, perception system 402 provides the output of convolution layer 426 as fully connected layer 430, where fully connected layer 430 includes data associated with a plurality of feature values referred to as F1 , F2 . . . FN. In this example, the output of convolution layer 426 includes data associated with a plurality of output feature values that represent a prediction.

[0094]

[0081] In some embodiments, perception system 402 identifies a prediction from among a plurality of predictions based on perception system 402 identifying a feature value that is associated with the highest likelihood of being the correct prediction from among the plurality of predictions. For example, where fully connected layer 430 includes feature values F1 , F2, . . . FN, and F1 is the greatest feature value, perception system 402 identifies the prediction associated with F1 as being the correct prediction from among the plurality of predictions. In some embodiments, perception system 402 trains CNN 420 to Attorney Docket No. 46154-0567WO1 / 12023099 generate the prediction. In some examples, perception system 402 trains CNN 420 to generate the prediction based on perception system 402 providing training data associated with the prediction to CNN 420.

[0095]

[0082] Referring now to FIGS. 4C and 4D, illustrated is a diagram of example operation of CNN 440 by perception system 402. In some embodiments, CNN 440 (e.g., one or more components of CNN 440) is the same as, or similar to, CNN 420 (e.g., one or more components of CNN 420) (see FIG. 4B).

[0096]

[0083] At step 450, perception system 402 provides data associated with an image as input to CNN 440 (step 450). For example, as illustrated, perception system 402 provides the data associated with the image to CNN 440, where the image is a greyscale image represented as values stored in a two-dimensional (2D) array. In some embodiments, the data associated with the image may include data associated with a color image, the color image represented as values stored in a three-dimensional (3D) array. Additionally, or alternatively, the data associated with the image may include data associated with an infrared image, a radar image, and / or the like.

[0097]

[0084] At step 455, CNN 440 performs a first convolution function. For example, CNN 440 performs the first convolution function based on CNN 440 providing the values representing the image as input to one or more neurons (not explicitly illustrated) included in first convolution layer 442. In this example, the values representing the image can correspond to values representing a region of the image (sometimes referred to as a receptive field). In some embodiments, each neuron is associated with a filter (not explicitly illustrated). A filter (sometimes referred to as a kernel) is representable as an array of values that corresponds in size to the values provided as input to the neuron. In one example, a filter may be configured to identify edges (e.g., horizontal lines, vertical lines, straight lines, and / or the like). In successive convolution layers, the filters associated with neurons may be configured to identify successively more complex patterns (e.g., arcs, objects, and / or the like).

[0098]

[0085] In some embodiments, CNN 440 performs the first convolution function based on CNN 440 multiplying the values provided as input to each of the one or more neurons included in first convolution layer 442 with the values of the filter that corresponds to each of the one or more neurons. For example, CNN Attorney Docket No. 46154-0567WO1 / 12023099

[0099] 440 can multiply the values provided as input to each of the one or more neurons included in first convolution layer 442 with the values of the filter that corresponds to each of the one or more neurons to generate a single value or an array of values as an output. In some embodiments, the collective output of the neurons of first convolution layer 442 is referred to as a convolved output. In some embodiments, where each neuron has the same filter, the convolved output is referred to as a feature map.

[0100]

[0086] In some embodiments, CNN 440 provides the outputs of each neuron of first convolutional layer 442 to neurons of a downstream layer. For purposes of clarity, an upstream layer can be a layer that transmits data to a different layer (referred to as a downstream layer). For example, CNN 440 can provide the outputs of each neuron of first convolutional layer 442 to corresponding neurons of a subsampling layer. In an example, CNN 440 provides the outputs of each neuron of first convolutional layer 442 to corresponding neurons of first subsampling layer 444. In some embodiments, CNN 440 adds a bias value to the aggregates of all the values provided to each neuron of the downstream layer. For example, CNN 440 adds a bias value to the aggregates of all the values provided to each neuron of first subsampling layer 444. In such an example, CNN 440 determines a final value to provide to each neuron of first subsampling layer 444 based on the aggregates of all the values provided to each neuron and an activation function associated with each neuron of first subsampling layer 444.

[0101]

[0087] At step 460, CNN 440 performs a first subsampling function. For example, CNN 440 can perform a first subsampling function based on CNN 440 providing the values output by first convolution layer 442 to corresponding neurons of first subsampling layer 444. In some embodiments, CNN 440 performs the first subsampling function based on an aggregation function. In an example, CNN 440 performs the first subsampling function based on CNN 440 determining the maximum input among the values provided to a given neuron (referred to as a max pooling function). In another example, CNN 440 performs the first subsampling function based on CNN 440 determining the average input among the values provided to a given neuron (referred to as an average pooling function). In some embodiments, CNN 440 generates an output based on CNN Attorney Docket No. 46154-0567WO1 / 12023099

[0102] 440 providing the values to each neuron of first subsampling layer 444, the output sometimes referred to as a subsampled convolved output.

[0103]

[0088] At step 465, CNN 440 performs a second convolution function. In some embodiments, CNN 440 performs the second convolution function in a manner similar to how CNN 440 performed the first convolution function, described above. In some embodiments, CNN 440 performs the second convolution function based on CNN 440 providing the values output by first subsampling layer 444 as input to one or more neurons (not explicitly illustrated) included in second convolution layer 446. In some embodiments, each neuron of second convolution layer 446 is associated with a filter, as described above. The filter(s) associated with second convolution layer 446 may be configured to identify more complex patterns than the filter associated with first convolution layer 442, as described above.

[0104]

[0089] In some embodiments, CNN 440 performs the second convolution function based on CNN 440 multiplying the values provided as input to each of the one or more neurons included in second convolution layer 446 with the values of the filter that corresponds to each of the one or more neurons. For example, CNN 440 can multiply the values provided as input to each of the one or more neurons included in second convolution layer 446 with the values of the filter that corresponds to each of the one or more neurons to generate a single value or an array of values as an output.

[0105]

[0090] In some embodiments, CNN 440 provides the outputs of each neuron of second convolutional layer 446 to neurons of a downstream layer. For example, CNN 440 can provide the outputs of each neuron of first convolutional layer 442 to corresponding neurons of a subsampling layer. In an example, CNN 440 provides the outputs of each neuron of first convolutional layer 442 to corresponding neurons of second subsampling layer 448. In some embodiments, CNN 440 adds a bias value to the aggregates of all the values provided to each neuron of the downstream layer. For example, CNN 440 adds a bias value to the aggregates of all the values provided to each neuron of second subsampling layer 448. In such an example, CNN 440 determines a final value to provide to each neuron of second subsampling layer 448 based on the Attorney Docket No. 46154-0567WO1 / 12023099 aggregates of all the values provided to each neuron and an activation function associated with each neuron of second subsampling layer 448.

[0106]

[0091] At step 470, CNN 440 performs a second subsampling function. For example, CNN 440 can perform a second subsampling function based on CNN 440 providing the values output by second convolution layer 446 to corresponding neurons of second subsampling layer 448. In some embodiments, CNN 440 performs the second subsampling function based on CNN 440 using an aggregation function. In an example, CNN 440 performs the first subsampling function based on CNN 440 determining the maximum input or an average input among the values provided to a given neuron, as described above. In some embodiments, CNN 440 generates an output based on CNN 440 providing the values to each neuron of second subsampling layer 448.

[0107]

[0092] At step 475, CNN 440 provides the output of each neuron of second subsampling layer 448 to fully connected layers 449. For example, CNN 440 provides the output of each neuron of second subsampling layer 448 to fully connected layers 449 to cause fully connected layers 449 to generate an output. In some embodiments, fully connected layers 449 are configured to generate an output associated with a prediction (sometimes referred to as a classification). The prediction may include an indication that an object included in the image provided as input to CNN 440 includes an object, a set of objects, and / or the like. In some embodiments, perception system 402 performs one or more operations and / or provides the data associated with the prediction to a different system, described herein.

[0108] Example Machine Learning-Based Vehicle Dynamic Simulator:

[0109]

[0093] In general, machine learning can be used to simulate vehicle dynamics, such as simulating the movement and position of a vehicle in response to operating commands.

[0110]

[0094] For instance, a neural network can be trained to predict the behavior of a vehicle, such as the acceleration and a yaw rate of a vehicle, based on an initial state of the vehicle and a command trajectory. The initial state of the vehicle can include information such as the vehicle’s initial position, speed, acceleration, heading, and / or any other information regarding a vehicle’s position, orientation, and / or movement. Attorney Docket No. 46154-0567WO1 / 12023099

[0111]

[0095] Further, the predicted behavior of the vehicle can be modified based on the physical constraints of the vehicle. As an example, the predicted acceleration of the vehicle can be modified based on constraints such as the maximum allowable acceleration and / or minimum allowable acceleration. This can be beneficial, for example, in increasing the accuracy of the predicted behavior based on the real-world characteristics and limitations of the vehicle.

[0112]

[0096] In addition, successive changes to the vehicle’s state can be used to simulate the vehicle’s trajectory over time. For example, based on the predicted changes to the vehicle’s state over time, the vehicle dynamic simulator can predict the locations of the vehicle over a period of time, including the final location of the vehicle at the end of the period of time and the specific time at which the vehicle arrived at that location. Further, the vehicle dynamic simulator can predict one or more intermediate locations leading up to the final location and respective times at which the vehicle arrived at those intermediate locations.

[0113]

[0097] In at least some implementations, the machine learning-based vehicle dynamic simulator described herein does not use complex physics-based models to represent the kinematics of various aspects of a vehicle over time (e.g., models that may be used by existing vehicle dynamic simulators). For instance, existing vehicle dynamic simulators may predict the behavior of vehicle by predicting the behavior of individual component sub-systems of the vehicle, and integrating the output of each of the models into a prediction. As an example, existing vehicle dynamic simulators models may include models that predict the behavior of a vehicle’s suspension and wheels / tires over time and in response to specific operating commands, the effects of environment conditions on the vehicle (e.g., wind, temperature, rain, etc.), and any other kinematic aspects of a vehicle’s operation. Further, the output of these models can be used to predict changes to the vehicle’s position, speed, acceleration, and heading over time. However, these models may be vehicle-specific due to the specific components and configurations of each vehicle. Further, these models may be difficult to implement in practice due to the variation and complexity of accurately modeling each of the aspects of a vehicle’s operation and integrating the models together. Further still, although the models may be Attorney Docket No. 46154-0567WO1 / 12023099 simplified in order to ease implementation, the output of the models may be biased and / or less accurate due to the simplification.

[0114]

[0098] In contrast, aspects of the vehicle dynamic simulator described herein can include one or more neural networks that are trained using data sets collected from test vehicles in a wide range of contexts. Example contexts include the locations of the vehicles, the operating conditions of the vehicles (e.g., weather, lighting, wind, road conditions, and / or other characteristics of the vehicle’ environment), and the maneuvers performed by the vehicles (e.g., accelerating, braking, turning, etc.), among others. Further, the trained neural networks can be used to predict a vehicle’s dynamics without explicit parameter modeling. Accordingly, the neural network can be trained based on the data sets to account for a wide array of characteristics of the vehicle’s operation, even if such characteristics are not expressly identified a human and expressly represented in a specialized kinematic model.

[0115]

[0099] Further, aspects of the vehicle dynamic simulator can include one or more LSTM networks that selectively regulate the flow of information through the network, such that information is maintained and updated in memory over time. This allows the network to retain information over successive states that may be beneficial in modeling a vehicle’s dynamics more accurately over time. For example, the LSTM networks can account for latent variables or other information that may represent changes of the characteristics of the vehicle and the context of its operation over time, even if those variations or information are not specifically identified by a human administrator during development and training of the network.

[0116]

[0100] In some implementations, the vehicle dynamic simulator can be used to facilitate the development of autonomous vehicle operations by allowing users to simulate the behavior of an autonomous vehicle in response to particular commands, without requiring that that test be performed on a physically deployed vehicle. Thus, the autonomous vehicle can be iteratively improved in a safer and more resource efficient manner. For example, a user can modify computer code that controls the operation of an autonomous vehicle, and test the effects of the modifications virtually using the vehicle dynamic simulator, rather than conducting the tests on a physically deployed vehicle. This reduces Attorney Docket No. 46154-0567WO1 / 12023099 the risk of damage to the vehicle and the environment, as well as the risk of injury to any passengers of vehicle or bystanders. Further, this reduces the consumption of physical resources needed to run a test vehicle, such as power and fuel, and reduces physical wear on the vehicle. Further, this allows a larger number of tests to be performed than might otherwise be possible using a physically deployed vehicle.

[0117]

[0101] FIG. 5 shows a simplified example system 500 for simulating the behavior of a vehicle (e.g., an autonomous vehicle). The system 500 includes an AV stack 510, a vehicle dynamic simulator module 520, and an environment model 530.

[0118]

[0102] During an example operation of the system 500, the AV stack 510 receives information regarding the state of the vehicle’s environment from the environment model 530, and information regarding the predicted current state of the vehicle (also referred to as the “ego vehicle state”) from the vehicle dynamic simulator module 520.

[0119]

[0103] In some implementations, information regarding the state of the vehicle’s environment can include information regarding weather conditions (e.g., wind, rain, snow, etc.), lighting conditions (e.g., day, evening, dusk, etc.), and / or road conditions (e.g., wet, icy, dry, snowy, etc.). In some implementations, information regarding the state of the vehicle’s environment can include the location of objects and other features in the vehicle’s environment (e.g., other vehicles, pedestrians, signs, roads, lane, buildings, obstacles, etc.), including their absolute geographical positions and / or their positions relative to the vehicle. The information can also include the dimensions and / or shapes of each of the objects or features, as well as each of the object’s or feature’s identities. If the objects are moving, the information can also include information regarding each of the object’s heading and / or speed (e.g., absolute heading / speed, and / or heading / speed relative to the vehicle).

[0120]

[0104] In some implementations, information regarding the predicted current state of the vehicle can include information regarding the vehicle’s current location, heading, speed, acceleration, yaw rate, and / or any other information regarding the position, orientation, and / or movement of the vehicle. Attorney Docket No. 46154-0567WO1 / 12023099

[0121]

[0105] Based on this information, the AV stack 510 determines the location of the vehicle (e.g., geographical location), as well as the spatial relationship between the vehicle and objects and / or other features of the vehicle’s environment. Further, the AV stack 510 generates one or more trajectories for traversing the environment (e.g., to a specific destination), evaluates the trajectories (e.g., by scoring the trajectories based on their safety, efficiency, comfort, legality, etc.), selects one of the trajectories for execution by the vehicle, and provides the selected trajectory to the vehicle dynamic simulator module 520. Example systems and techniques for performing these operations are described above, such as with reference to FIGS. 4A-4D.

[0122]

[0106] The vehicle dynamic simulator module 520 receives the selected trajectory from the AV stack 510, and predicts a new state of the vehicle in accordance with the selected trajectory. For instance, the vehicle dynamic simulator module 520 can predict, based on the selected trajectory, changes to the vehicle’s position, orientation, and / or movement when executing the trajectory, and determine the vehicle’s new position, orientation, and / or movement at some future point in time. As an example, the vehicle dynamic simulator module 520 can predict the vehicle’s new location, heading, speed, acceleration, and / or yaw rate at the future point in time (e.g., as a result of executing at least a portion of the selected trajectory).

[0123]

[0107] In some implementations, the vehicle dynamic simulator module 520 can generate a set of vehicle commands based on the selected trajectory (e.g., using a trajectory tracking module 522), and provide the vehicle commands to a vehicle dynamic simulator module 524. As an example, vehicle commands can include steering inputs, throttle inputs, and / or braking inputs that are provided to the vehicle’s steering, throttle, and / or breaking sub-systems in order to direct the vehicle along the selected trajectory. These commands can be generated using techniques similar to those described above with reference to the DBW system 202h. Based on the vehicle commands, the vehicle dynamic simulator module 524 can predict the new state of the vehicle.

[0124]

[0108] However, in some implementations, the vehicle dynamic simulator module 520 need not explicitly generate a set of vehicle commands based on the selected trajectory. Instead, the selected trajectory can be provided to Attorney Docket No. 46154-0567WO1 / 12023099 directly to the vehicle dynamic simulator module 524. Based on the selected trajectory, the vehicle dynamic simulator module 524 can directly predict the new state of the vehicle.

[0125]

[0109] An example implementation of the vehicle dynamic simulator module 524 is shown in greater detail in FIG. 6. In this example, the vehicle dynamic simulator module 524 is configured to directly predict the new state of the vehicle based on the selected trajectory provided by the AV stack 510, without requiring that the selected trajectory be first converted into a set of vehicle commands.

[0126]

[0110] As shown in FIG. 6, the vehicle dynamic simulator module 524 includes a neural network 602. The neural network 602 is configured to receive (i) information regarding an initial state of the vehicle, and (ii) information regarding the trajectory selected by the AV stack 510 (also referred to as the “optimized trajectory,” “commanded trajectory,” or “control trajectory”). Based on this information, the neural network 602 generates and outputs an initial prediction of the new state of the vehicle.

[0127]

[0111] As described above, information regarding the initial state of the vehicle can include information regarding the vehicle’s current position, orientation, and / or movement (e.g., location, speed, acceleration, yaw rate, etc.). Further, the initial prediction of the new state of the vehicle can include the vehicle’s predicted position, orientation, and / or movement at some point in the future (e.g., 0.05 seconds, 0.10 seconds. 0.15 seconds, or any other time interval). The vehicle’s predicted position, orientation, and / or movement can include the vehicle’s predicted location, speed, acceleration, yaw rate, etc. at the future time.

[0128]

[0112] In general, the vehicle’s initial state and / or predicted state can be expressed according to a single spatial direction or according to multiple spatial directions. For example, the vehicle’s current acceleration and predicted future acceleration can be expressed according to a single spatial direction (e.g., parallel of the vehicle’s direction of movement), or according to multiple spatial directions (e.g., x-direction and y-direction). For instance, as shown in FIG. 6, the vehicle’s predicted future acceleration can be represented by Axand Ay, indicating the vehicle’s acceleration in the x-direction and y-direction, respectively. In some implementations, the x-direction can refer to the direction Attorney Docket No. 46154-0567WO1 / 12023099 extending between the front and the back of the vehicle (e.g., the vehicle’s longitudinal axis), and the y-direction can refer to the direction extending between the left and right of the vehicle (e.g., the vehicle’s lateral axis). Further, as shown in FIG. 6, the vehicle’s current and initial predicted state can include the vehicle’s yaw rate a (e.g., the rate of change of the vehicle’s yaw or rotation in the x-y plane).

[0129]

[0113] As described above, the neural network 602 can be trained using training data sets collected from test vehicles in a wide range of contexts. Example contexts include the locations of the vehicles, the operating conditions of the vehicles (e.g., weather, lighting, wind, road conditions, and / or other characteristics of the vehicle’ environment), and the maneuvers performed by the vehicles (e.g., accelerating, braking, turning, etc.), among others. Further, the training data sets can include information regarding (i) the initial states of each of the test vehicles, (ii) the trajectory that was executed by each of the test vehicles, and (iii) one or more subsequent states of the test vehicle upon executing at least a portion of the trajectory (e.g., one or more future states of the test vehicle while following its respective trajectory). Based on this training data, the neural network 602 can determine one or more correlations and / or trends between the initial state of a vehicle, the trajectory provided for execution by the vehicle, and one or more subsequent states of the vehicle in accordance with the trajectory. Further, based on these correlations and / or trends, the neural network 602 can predict a future state of a vehicle, given the vehicle’s current state and a commanded trajectory.

[0130]

[0114] Example implementations of a neural network 602 are described above (e.g., with reference to FIGS. 4B-4D). Additional details regarding the neural network 602 are also provided below (e.g., with reference to FIGS. 7 and 8.

[0131]

[0115] Referring back to FIG. 6, the vehicle dynamic simulator module 524 also includes an acceleration limit thresholding module 604 configured to receive the initial predicted state generated by the neural network 602, and generate a modified predicted state based on one or more constraints associated with the vehicle.

[0132]

[0116] In general, the one or more constraints refer to limits regarding the vehicle, such performance or physical limitations of the vehicle, safety and / or Attorney Docket No. 46154-0567WO1 / 12023099 comfort limits regarding the operation of the vehicle, and / or legal limits regarding the operation of the vehicle. As an example, a vehicle may have a particular maximum speed, acceleration, or yaw rate beyond which it cannot exceed due to the performance or physical limitations of the vehicle itself (e.g., limits to the vehicle’s acceleration and / or braking capabilities). As another example, in order to maintain passenger comfort and safety, the vehicle may be limited to traversing at or below a particular threshold speed, accelerating at or below a particular threshold acceleration, braking at or below a particular braking threshold, and / or turning according to a rate at or below a particular yaw rate. As another example, in order to adhere to traffic rules, the vehicle may be limited to traversing at or below a particular threshold speed (e.g., a speed limit of the road along which the vehicle to traversing).

[0133]

[0117] In some implementations, the acceleration limit threshold module 604 determines whether the initial predicted state generated by the neural network 602 violates the constraints, and if so, generates a modified predicted state that conforms with the constraints. For example, if the initial predicted state includes an acceleration (e.g., in a single or multiple directions) and / or a yaw rate that exceeds their respective thresholds, the acceleration limit threshold module 604 can generate a modified predicted state in which the acceleration and / or yaw rate is set to the maximum allowable acceleration and / or a maximum allowable yaw rate, respectively.

[0134]

[0118] The vehicle dynamic simulator module 524 also includes a rigid body dynamics and kinematics model module 606. The module 606 is configured to receive the modified predicted state from the acceleration limit thresholding module 604 (or the original predicted state generated by the neural network 602 if no modifications were made by the acceleration limit thresholding module 604), and generate a predicted location, heading, and speed of the vehicle based on the predicted state. As an example, the module 606 can be configured to predict the vehicle’s new position in the x-direction and y-direction, the vehicle’s new heading, and vehicle’s new speed at a future time, upon executing at least a portion of the selected trajectory.

[0135]

[0119] As described above, in an at least some implementations, the module 606 does not use physics-based models to simulate the behavior of individual Attorney Docket No. 46154-0567WO1 / 12023099 components or sub-systems of the vehicle (e.g., the vehicle’s tires, wheels, suspension, etc.). Instead, in at least some implementations, the module 606 can use physics-based models to represent the simulate the behavior of the vehicle as a whole, based on the acceleration and / or yaw rate of the vehicle. For example, the module 606 can predict the vehicle’s new position, heading, and / or speed by using the equation F = ma, where m is the mass of the vehicle, a is the predicted acceleration of the vehicle (e.g., in a particular direction, and F is the force applied to the vehicle (e.g., along the particular direction) as a result of the acceleration. This information can be used to determine changes to the vehicle’s current speed and / or heading, which in turn can be used to determine the vehicle’s new position (e.g., relative to its current position).

[0136]

[0120] Further, the vehicle dynamic simulator module 524 also includes a simulated trajectory module 608 that generates a simulated trajectory of the vehicle based on the predicted new position, heading, and / or speed of the vehicle. For example, the module 524 can access the current position of the vehicle, and determine a new position of the vehicle based on the new speed of the vehicle and the new heading of the vehicle (e.g., by incrementing the position of the vehicle over the simulated time interval, in accordance with the predicted vehicle’s heading and speed during that time interval). As another example, the module 524 can predict, based on the information provided by the module 606, the new position of the vehicle relative to its previous location (e.g., in the x- direction and y-direction). Further, the module 608 can record the vehicle’s predicted locations for each of several simulated time intervals, and determine the vehicle’s trajectory over that period of time by interconnecting the predicted locations in order of time (e.g., to generate a line or trace representing the simulated trajectory of the vehicle over time).

[0137]

[0121] In general, the simulated trajectory can be compared to trajectory selected by the AV stack 510 in order to determine deviations between the vehicle’s commanded trajectory, the vehicle’s actual trajectory, and / or the vehicle’s predicted / simulated trajectory (e.g., as simulated by the vehicle dynamic simulator module 520). This information can be used to improve the operation of the vehicle (e.g., by improving the systems and techniques used to autonomously operate the vehicle given a particular commanded trajectory), Attorney Docket No. 46154-0567WO1 / 12023099 such that the vehicle traverses a trajectory in a more accurate and / or consistent manner. As an example, one or more of the perception system 402, the planning system 404, and localization system 406, and / or the control system 408 can be modified to improve the performance of the vehicle in executing a commanded trajectory.

[0138]

[0122] Further, the simulated trajectory can be used to simulate the behavior of a vehicle in response to a commanded trajectory, without requiring that the command trajectory be performed by a physically deployed vehicle. This allows the autonomous vehicle to be iteratively improved in a safer and more resource efficient manner. For example, one can modify computer code that controls the operation of an autonomous vehicle (e.g., the code implementing the planning system 404, and localization system 406, and / or the control system 408), and test the effects of the modifications virtually using the vehicle dynamic simulator module 520, rather than conducting the tests on a physically deployed vehicle. This reduces the risk of damage to the vehicle and the environment, as well as the risk of injury to any passengers of the vehicle or bystanders. Further, this reduces the consumption of physical resources needed to run a test vehicle, such as power and fuel, and reduces physical wear on the vehicle. Further, this allows a larger number of tests to be performed than might otherwise be possible using a physically deployed vehicle.

[0139]

[0123] As described above, aspects of the vehicle dynamic simulator module 520 can include one or more LSTM networks that selectively regulate the flow of information through the networks, such that information is maintained and updated in memory overtime. For example, the neural network 602 can include one or more LSTM networks.

[0140]

[0124] An example model architecture 700 is shown in FIG. 7. In this example shown in FIG. 7, the model architecture 700 includes several interconnected LSTM networks 702a-702n, each configured to receive state information from a previous LSTM network and / or initial state information (in the case of the first LSTM network 702a), and provide hidden state information to a following LSTM network (except in the case of the final LSTM network 702n). This configuration allows a neural network to retain information over successive states that may be beneficial in modeling a vehicle’s dynamics more accurately over time. For Attorney Docket No. 46154-0567WO1 / 12023099 example, by passing information between LSTM networks, the neural network can account for latent variables or other information that may represent changes of the characteristics of the vehicle and the context of its operation over time, even if those variations or information are not specifically identified by a human administrator during development and training of the network.

[0141]

[0125] Although FIG. 7 shows several LSTM networks that are interconnected with another, in practice, a neural network can also include a single LSTM network that is used iteratively over time to (i) receive information regarding the current state of the vehicle, a commanded trajectory, and previously generated hidden state information, and (ii) generate a predicted new state the vehicle and new hidden state information for use in future predictions. As an example, the hidden state information generated by the LSTM network according to one time point can be re-input into the LSTM network to generate predictions for a subsequent time point.

[0142]

[0126] Referring to FIG.7, in at least some implementations, an LSTM network 702a receives information regarding a measured state of the vehicle and a commanded trajectory for the vehicle at a first time point (e.g., t = 0 seconds). As described above, the measured state of the vehicle can include information regarding the acceleration, yaw rate, position, heading, and / or speed of the vehicle.

[0143]

[0127] Based on this information, the LSTM network 702a generates a predicted new state of the vehicle at a second time point (e.g., t = 0.05 seconds), such as the predicted acceleration of the vehicle, yaw rate, etc. at the second time point. The predicted new state of the vehicle is transformed according to a rigid body kinematic module 704 (e.g., in a similar manner as described with reference to module 606 in FIG. 6) in order to determine additional information regarding the predicted state of vehicle and the second time point (e.g., the position, heading, and speed of the vehicle at the second time point). The LSTM network 702a also generates hidden state information (e.g., representing latent variables or other information that may represent changes in the characteristics of the vehicle and the context of its operation over time).

[0144]

[0128] The information regarding the predicted state of the vehicle at the second time (e.g., acceleration, yaw rate, position, heading, and / or speed of the vehicle) Attorney Docket No. 46154-0567WO1 / 12023099 and the hidden state information at the second time point is provided to the LSTM network 702b. In turn, the LSTM network 702b and / or the rigid body kinematic module 704 generate a predicted new state of the vehicle at a third time (e.g., t = 0.1 seconds) and new hidden state information. The process is repeated over time in order to predict changes to the vehicle’s behavior over the time period.

[0145]

[0129] In some implementations, the predicted state for one time point can be transformed (at least in part) before being provided to an LSTM network for predicting a subsequent state. For example, shown in FIG. 7, the predicted state of the vehicle at the second time point (e.g., t = 0.05 seconds) can be transformed to obtain coordinates (e.g., spatial coordinates) representing the position of the vehicle at the second time point. This information can be provided to an LSTM network and rigid body kinematic module 704 to generate a new predicted state of the vehicle at the third time point (e.g., t = 0.1 seconds). This transformation process can be repeated until predictions have been generated for the entire time interval.

[0146]

[0130] Further, the LSTM networks 702a-702n can be trained in order to improve their predictive performance. For example, as shown in FIG. 7, the predicted states generated by the LSTM networks 702a-702n can be compared to the actual known states of the vehicle over time (e.g., accelerations, yaw rates, positions, headings, speeds, etc. over time). The difference between the predicted states and the actual states may be referred to as a “loss.” The LSTM networks 702a-702n can be trained (e.g., by modifying the relationships and transformations between inputs and outputs within each cell) in order to reduce the overall loss, such that the LSTM networks 702a-702n predict the states of a vehicle more accurately.

[0147]

[0131] FIG. 8 shows an example LSTM network 702 and a corresponding rigid body kinematic module 704.

[0148]

[0132] As shown in FIG. 8, the LSTM network 702 includes a LSTM network cell 802 and a fully connected (FC) layer 804. The LSTM network cell 802 is configured to receive information regarding the commanded trajectory and the measured state of the vehicle. This information can represent, for example, the position, heading, speed, acceleration, and / or yaw rate of the vehicle along one Attorney Docket No. 46154-0567WO1 / 12023099 or more directions). The output to the LSTM network cell 802 is transformed by a hyperbolic tangent function (tanh), and is input into the FC layer 804.

[0149]

[0133] In turn, the FC layer 804 transforms the input into a new predicted state of the vehicle (e.g., including the acceleration and yaw rate of the vehicle along one or more directions). As shown in FIG. 8, the FC layer 804 can perform various operations to generate the new predicted state, including performing the following sequential: (i) performing a first linear matrix transformation, (ii) performing a layer normalization operation, (iii) transforming the normalized data by a hyperbolic tangent function, (iv) applying a second linear matrix transformation, (v) transforming the normalized data by the hyperbolic tangent function, and (iv) performing element-wise matrix multiplication.

[0150]

[0134] Further, as described above, the rigid body kinematic module 704 receives the output of the LSTM network 702, and generates additional information regarding the predict state of the vehicle, including the vehicle’s predicted new position, heading, and speed (e.g., along one or more directions). As an example, the module 704 can determine relative changes (or “deltas”) between (i) the vehicle’s previous position, yaw, yaw rate, and speed, and (ii) the vehicle’s new predicted position, yaw, yaw rate, and speed (e.g., along one or more directions) in order to determine a vehicle’s relative change in state. Further, the module 704 can obtain absolute information regarding the vehicle’s previous state (e.g., absolute position, speed, yaw, yaw rate, etc.) and this information and the determined relative changes in the vehicle’s state in order to determine absolute information regarding the vehicle’s new predicted state (e.g., new position, heading, speed, etc.).

[0151]

[0135] Referring now to FIG. 9, illustrated is a flowchart of a process 900 for simulating an operation of an autonomous vehicle. In some embodiments, one or more of the steps described with respect to process 900 are performed (e.g., completely, partially, and / or the like) by one or more components of an autonomous vehicle, such as one or more processors that implement the autonomous vehicle compute 400 shown in FIG. 4A. Additionally, or alternatively, in some embodiments one or more steps described with respect to process 900 are performed (e.g., completely, partially, and / or the like) by another device or group of devices separate from an autonomous vehicle, such Attorney Docket No. 46154-0567WO1 / 12023099 as by one or more processors of a remote computer system (e.g., a remote AV system 114, fleet management system 116, V2I system 118, etc.).

[0152]

[0136] According to the process 900, one or more processors access (i) first input data representing a trajectory for traversal by an autonomous vehicle, and (ii) second input data representing a first state of the autonomous vehicle, including a position, a heading and a speed of the autonomous vehicle at a first time (902).

[0153]

[0137] The one or more processors estimate, using a machine learning system, a second state of the autonomous vehicle based on the first input data and the second input data (904). The second state of the autonomous vehicle includes an estimated position, heading, and speed of the autonomous vehicle at a second time subsequent to the first time.

[0154]

[0138] The machine learning system includes a long short-term memory (LSTM) network trained based on a training data set representing a plurality of operating scenarios for one or more additional autonomous vehicles. The training data set includes, for each of the operating scenarios: (i) first training data representing a respective trajectory for traversal by a respective additional autonomous vehicle, (ii) second training data indicating a beginning state and an ending state of the respective autonomous vehicle in traversing the trajectory.

[0155]

[0139] The one or more processors generate first output data representing the second state of the autonomous vehicle (906).

[0156]

[0140] In some implementations, the LSTM network can be configured to (i) receive the first input data and the second input data, and (ii) output, based on the first input data and the second input data, an estimated acceleration of the autonomous vehicle at the second time.

[0157]

[0141] In some implementations, the LSTM network can be configured to output, based on the first input data and the second input data, an estimated yaw rate of the autonomous vehicle at the second time.

[0158]

[0142] In some implementations, the machine learning system can also include an acceleration limit thresholding module configured to: (i) receive, from the LSTM network, the estimated acceleration of the autonomous vehicle at the second time, (ii) modify, based one or more first constraints, the estimated Attorney Docket No. 46154-0567WO1 / 12023099 acceleration of the autonomous vehicle at the second time, and (iii) output the modified estimated acceleration of the autonomous vehicle at the second time.

[0159]

[0143] In some implementations, the one or more first constraints can include a maximum acceleration of the autonomous vehicle.

[0160]

[0144] In some implementations, the acceleration limit thresholding module can be configured to: (i) receive, from the LSTM network, the estimated yaw rate of the autonomous vehicle at the second time, (ii) modify, based one or more second constraints, the estimated yaw rate of the autonomous vehicle at the second time, and (iii) output the modified estimated yaw rate of the autonomous vehicle at the second time.

[0161]

[0145] In some implementations, the one or more second constraints can include a maximum yaw rate of the autonomous vehicle.

[0162]

[0146] In some implementations, the machine learning system can also include a rigid body dynamics modeling module configured to: (i) receive, from the acceleration limit thresholding module, the modified estimated acceleration and the modified estimated yaw rate of the autonomous vehicle at the second time, and (ii) output, based on a rigid body dynamic model, the second state of the autonomous vehicle.

[0163]

[0147] In some implementations, the rigid body dynamic model can be configured to determine the second state of the autonomous vehicle based on one or more functions representing a relationship between (i) at least one of the modified estimated acceleration or the modified estimated yaw rate of the autonomous vehicle at the second time, and (ii) at least one of the estimated position, heading, or speed of the autonomous vehicle at the second time.

[0164]

[0148] In some implementations, the LSTM network can include a plurality of LSTM cells, where at least some of the LSTM cells is configured to output, to another one of the LSTM cells, an output representing a hidden state of the autonomous vehicle. Further, the hidden state of the autonomous vehicle can be determined based on at least one of an input to the LSTM cell and a hidden state received from among one of the LSTM cells.

[0165]

[0149] In some implementations, the process 900 can also include simulating a trajectory of the autonomous vehicle based on the output data, and generating Attorney Docket No. 46154-0567WO1 / 12023099 second output data representing the simulated trajectory of the autonomous vehicle.

[0166]

[0150] In some implementations, the process 900 also include (i) causing the autonomous vehicle to operating in accordance with the trajectory at the first time, (ii) measuring the position, heading, and speed of the autonomous vehicle at the second time, (iii) comparing (a) the measured position, heading, and speed of the autonomous vehicle at the second time and (b) the estimated position, heading, and speed of the autonomous vehicle at a second time, and (iv) modifying the machine learning system based on the comparison.

[0167]

[0151] In the foregoing description, aspects and embodiments of the present disclosure have been described with reference to numerous specific details that can vary from implementation to implementation. Accordingly, the description and drawings are to be regarded in an illustrative rather than a restrictive sense. The sole and exclusive indicator of the scope of the invention, and what is intended by the applicants to be the scope of the invention, is the literal and equivalent scope of the set of claims that issue from this application, in the specific form in which such claims issue, including any subsequent correction. Any definitions expressly set forth herein for terms contained in such claims shall govern the meaning of such terms as used in the claims. In addition, when we use the term “further comprising,” in the foregoing description or following claims, what follows this phrase can be an additional step or entity, or a sub- step / sub-entity of a previously-recited step or entity.

Claims

Attorney Docket No. 46154-0567WO1 / 12023099WHAT IS CLAIMED IS:1 . A method comprising: accessing, by one or more processors: first input data representing a trajectory for traversal by an autonomous vehicle, and second input data representing a first state of the autonomous vehicle, including a position, a heading and a speed of the autonomous vehicle at a first time; estimating, by the one or more processors using a machine learning system, a second state of the autonomous vehicle based on the first input data and the second input data, the second state of the autonomous vehicle including an estimated position, heading, and speed of the autonomous vehicle at a second time subsequent to the first time, wherein the machine learning system comprises a long short-term memory (LSTM) network trained based on a training data set representing a plurality of operating scenarios for one or more additional autonomous vehicles, wherein the training data set comprises, for each of the operating scenarios: first training data representing a respective trajectory for traversal by a respective additional autonomous vehicle, and second training data indicating a beginning state and an ending state of the respective autonomous vehicle in traversing the trajectory; and generating, by the one or more processors, first output data representing the second state of the autonomous vehicle.

2. The method of claim 1 , wherein the LSTM network is configured to: receive the first input data and the second input data, and output, based on the first input data and the second input data, an estimated acceleration of the autonomous vehicle at the second time.Attorney Docket No. 46154-0567WO1 / 120230993. The method of claim 2, wherein the LSTM network is configured to: output, based on the first input data and the second input data, an estimated yaw rate of the autonomous vehicle at the second time.

4. The method of claim 3, wherein the machine learning system further comprises an acceleration limit thresholding module configured to: receive, from the LSTM network, the estimated acceleration of the autonomous vehicle at the second time, modify, based on one or more first constraints, the estimated acceleration of the autonomous vehicle at the second time, and output the modified estimated acceleration of the autonomous vehicle at the second time.

5. The method of claim 4, wherein the one or more first constraints comprise a maximum acceleration of the autonomous vehicle.

6. The method of claim 3, wherein the acceleration limit thresholding module is configured to: receive, from the LSTM network, the estimated yaw rate of the autonomous vehicle at the second time, modify, based one or more second constraints, the estimated yaw rate of the autonomous vehicle at the second time, and output the modified estimated yaw rate of the autonomous vehicle at the second time.

7. The method of claim 6, wherein the one or more second constraints comprise a maximum yaw rate of the autonomous vehicle.

8. The method of claim 7, wherein the machine learning system further comprises a rigid body dynamics modeling module configured to: receive, from the acceleration limit thresholding module, the modified estimated acceleration and the modified estimated yaw rate of the autonomous vehicle at the second time, andAttorney Docket No. 46154-0567WO1 / 12023099 output, based on a rigid body dynamic model, the second state of the autonomous vehicle.

9. The method of claim 8, wherein the rigid body dynamic model is configured to determine the second state of the autonomous vehicle based on one or more functions representing a relationship between (i) at least one of the modified estimated acceleration or the modified estimated yaw rate of the autonomous vehicle at the second time, and (ii) at least one of the estimated position, heading, or speed of the autonomous vehicle at the second time.

10. The method of claim 1 , wherein the LSTM network comprises a plurality of LSTM cells, wherein at least some of the LSTM cells is configured to output, to another one of the LSTM cells, an output representing a hidden state of the autonomous vehicle, wherein the hidden state of the autonomous vehicle is determined based on at least one of an input to the LSTM cell and a hidden state received from among one of the LSTM cells.11 . The method of claim 1 , further comprising simulating a trajectory of the autonomous vehicle based on the output data, and generating second output data representing the simulated trajectory of the autonomous vehicle.

12. The method of claim 1 , further comprising: causing the autonomous vehicle to operating in accordance with the trajectory at the first time, measuring the position, heading, and speed of the autonomous vehicle at the second time, comparing (i) the measured position, heading, and speed of the autonomous vehicle at the second time and (ii) the estimated position, heading, and speed of the autonomous vehicle at a second time, and modifying the machine learning system based on the comparison.Attorney Docket No. 46154-0567WO1 / 1202309913. A system comprising: at least one processor; and at least one non-transitory storage media storing instructions that, when executed by the at least one processor, cause the at least one processor perform the method of any of claims 1 to 12.

14. At least one non-transitory storage media storing instructions that, when executed by at least one processor, cause the at least one processor perform the method of any of claims 1 to 12.

Images