Lane-shifting maneuvers for autonomous vehicles to negotiate curved roads
Lane deviation maneuvers enable autonomous vehicles to safely navigate around intruding vehicles and negotiate curved roads by driving off-center within their lane, addressing safety and efficiency challenges in autonomous driving.
Patent Information
- Application Number
- JP2024523167
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2022-10-24
- Filing Date
- 2022-10-25
- Publication Date
- 2024-11-08
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Autonomous vehicles face challenges in safely navigating around intruding, encroaching, or oversized vehicles in adjacent lanes, and in negotiating curved roads without changing lanes, which can lead to dangerous driving conditions.
Implementing lane deviation maneuvers that allow autonomous vehicles to drive off-center within their current lane to maintain a safe distance from intruding vehicles and compensate for trailer sway or road curvature, using sensors and control systems to determine the feasibility and safety of such maneuvers.
Enhances the safety and efficiency of autonomous vehicle navigation by allowing them to avoid collisions and maintain lane position, even in challenging scenarios, thereby improving overall driving performance and reducing risks.
Smart Images

Figure 2024541202000001_ABST
Abstract
Description
[Technical field]
[0001] (Priority) This application claims priority to U.S. Provisional Application No. 63 / 263,289, filed October 29, 2021, and entitled "Lane Bias Maneuver for Autonomous Vehicles to Avoid an Intruding Vehicle," and U.S. Provisional Application No. 63 / 263,303, filed October 29, 2021, and entitled "Lane Bias Maneuver for Autonomous Vehicles to Negotiate a Curved Road," both of which are incorporated herein by reference.
[0002] (Technical field) The present disclosure relates generally to autonomous vehicles and, more particularly, to lane-shifting maneuvers for autonomous vehicles to negotiate curved roads. [Background technology]
[0003] One aim of autonomous vehicle technology is to provide a vehicle that can safely navigate towards a destination. It is inevitable that an autonomous vehicle will encounter other vehicles while traveling on a road. Sometimes an autonomous vehicle may drive on a straight road and sometimes it may drive on a curved road. Summary of the Invention [Means for solving the problem]
[0004] The present disclosure recognizes various problems and heretofore unmet needs associated with implementing safe navigation for autonomous vehicles when the autonomous vehicle encounters an encroaching, encroaching, or oversized vehicle, or when the autonomous vehicle approaches a curved road. Certain embodiments of the present disclosure provide specific technical solutions to the technical problems of current autonomous vehicle technology, including those problems described above, for 1) implementing lane-shifting maneuvers to avoid an encroaching, encroaching, or oversized vehicle, and 2) implementing lane-shifting maneuvers to negotiate a curved road.
[0005] (Implementation of lane shifting maneuvers to avoid vehicles) The present disclosure contemplates a system and method for implementing lane-offset maneuvers to avoid vehicles. While traveling on a road, an autonomous vehicle may encounter an encroaching, encroaching, or oversized vehicle in an adjacent lane. The lane-offset maneuvers may enable the autonomous vehicle to drive off-center in its current lane to avoid the encountered vehicle and keep a safe distance (e.g., above a predetermined threshold distance) from the encountered vehicle. In other words, the lane-offset maneuvers may enable the autonomous vehicle to drive off-center in its current lane without changing into another lane.
[0006] In some cases, changing into another lane may not be a safe maneuver due to traffic on the road. In other cases, the autonomous vehicle may need to remain in its current lane to perform the next navigation maneuver in its given route plan, such as taking a particular exit, turning in a particular direction, etc. In such cases, changing into another lane may not be a practical solution for the autonomous vehicle.
[0007] Thus, the disclosed system of the present disclosure is integrated into a practical application to improve navigation of an autonomous vehicle by implementing lane-shifting maneuvers when the autonomous vehicle encounters an encroaching, encroaching, or oversized vehicle in an adjacent lane.
[0008] An aggressor vehicle in an adjacent lane may be classified as a vehicle that is always driving in close proximity to the autonomous vehicle, e.g., closer than a threshold distance defined by the control device of the autonomous vehicle, such as 6 feet, 7 feet, etc.
[0009] An intruding vehicle in an adjacent lane may be classified as a vehicle that temporarily or abruptly drives too close to the autonomous vehicle, e.g., closer than a threshold distance from the autonomous vehicle. An oversized vehicle in an adjacent lane may be classified as a vehicle that occupies more space in the adjacent lane due to its size, resulting in a distance between the oversized vehicle and the autonomous vehicle that is less than a threshold distance. Each of an intruding, encroaching, or oversized vehicle may generally be referred to herein as a vehicle or an encounter vehicle.
[0010] In any case where such a vehicle in an adjacent lane from the autonomous vehicle is encountered, the control device of the autonomous vehicle may determine that the lateral distance between the autonomous vehicle and each vehicle is less than a threshold distance.
[0011] In some cases, the autonomous vehicle's control device may determine that a vehicle in an adjacent lane from the autonomous vehicle has crossed a lane marker between the autonomous vehicle and the vehicle. In some cases, the autonomous vehicle's control device may determine that a vehicle in an adjacent lane has not crossed a lane marker but is driving within a threshold distance from the autonomous vehicle. In other cases, the autonomous vehicle's control device may determine that the vehicle has already crossed one or more lane markers on the road based on the vehicle's historical driving behavior or patterns. In other cases, the autonomous vehicle's control device may determine that the vehicle is in an emergency lane, either parked or in motion.
[0012] A vehicle may be detected on either side of the autonomous vehicle, or on either side and in front of the autonomous vehicle. The vehicle may be moving or stationary. For example, the control device of the autonomous vehicle may determine that a vehicle is driving on either side of the autonomous vehicle in an adjacent lane. In another example, the control device of the autonomous vehicle may determine that a vehicle is driving on either side of the autonomous vehicle in an adjacent lane and in front of the autonomous vehicle. In another example, the control device of the autonomous vehicle may determine that a vehicle is stopped (or stalled) on the side of the road in front of the autonomous vehicle, for example, if the vehicle is stopped on the side of the road to change its tires.
[0013] In any of these cases, the autonomous vehicle's control device may determine whether to command the autonomous vehicle to perform a lane-offset maneuver. A lane-offset maneuver may allow the autonomous vehicle to drive off-center in its current lane in the opposite direction to the vehicle until the lateral distance between the vehicle and the autonomous vehicle is at least equal to the threshold distance. In other words, the autonomous vehicle offsets toward the other side of the current lane. In some embodiments, the autonomous vehicle may perform a lane-offset maneuver until the autonomous vehicle and the vehicle are no longer adjacent to one another, for example, until either the autonomous vehicle passes the vehicle or the vehicle passes the autonomous vehicle. In one example, the autonomous vehicle may perform a lane-offset maneuver until it is determined that no portion of the autonomous vehicle overlaps with any portion of a vehicle traveling in an adjacent lane. In another example, the autonomous vehicle may perform a lane-offset maneuver until it is determined that a less-than-threshold portion of a vehicle traveling in an adjacent lane overlaps with any portion of the autonomous vehicle. The threshold portion of the vehicle may be one-third, one-half, two-thirds, or any other suitable portion of the vehicle's length.
[0014] If the control device of the autonomous vehicle determines that the lateral distance between the autonomous vehicle and the vehicle is less than a threshold distance, the control device may determine whether it is feasible and safe to perform a lane-shifting maneuver.
[0015] In this process, the control device of the autonomous vehicle may determine that it is feasible to perform a lane-shifting maneuver if it is determined that, depending on the traffic on the road, the lane-shifting maneuver can be performed within a threshold period of time, e.g., two minutes, five minutes, or any other suitable period of time. For example, if there is traffic congestion on the road, the control device may determine that the lane-shifting maneuver cannot be performed within the threshold period of time.
[0016] The control device may determine that a lane-shifting maneuver is feasible if it is determined that there is sufficient room or distance to the other side of the autonomous vehicle on its current lane to perform a lane-shifting maneuver.
[0017] In some embodiments, if the control device may determine that there is not enough distance on the other side of the autonomous vehicle to perform a lane-leaning maneuver, the control device may temporarily cross into the adjacent lane on the other side of the autonomous vehicle (relative to where the vehicle was detected) and occupy a similar amount of space in the adjacent lane (if traffic in the adjacent lane allows) until the lateral distance between the autonomous vehicle and the encountered vehicle equals the threshold distance.
[0018] The control device of the autonomous vehicle may determine that it is feasible to perform a lane-shifting maneuver based on the road structure. For example, if the road has a high curvature (e.g., greater than 50 degrees, 60 degrees, etc.), the control device may determine that it is not feasible to perform a lane-shifting maneuver.
[0019] The control device may determine that a lane-shifting maneuver is safe based on the historical driving behavior of the encountered vehicle. For example, if the historical driving behavior of the encountered vehicle indicates that the vehicle's driving pattern is highly unpredictable (e.g., the vehicle's driving pattern or trajectory prediction is below a threshold percentage, such as 70%, 65%, etc.), the control device may determine that a lane-shifting maneuver is unsafe.
[0020] If the control device determines that a lane-shifting maneuver is not feasible and / or safe, the control device may command the autonomous vehicle to perform a minimum-risk maneuver. The minimum-risk maneuver may include slowing down the autonomous vehicle or increasing the autonomous vehicle's speed until the autonomous vehicle and the vehicle are no longer adjacent to each other.
[0021] In some embodiments, the system may include an autonomous vehicle and a control device. The autonomous vehicle is configured to travel along a road. The autonomous vehicle includes at least one sensor configured to capture sensor data associated with one or more objects on the road. The control device is associated with the autonomous vehicle. The control device includes a processor. The processor may detect a presence of the vehicle from the sensor data. The processor may determine a lateral distance between the autonomous vehicle and the vehicle. The processor may compare the lateral distance between the autonomous vehicle and the vehicle to a threshold distance from the autonomous vehicle. The processor may determine whether to instruct the autonomous vehicle to perform a lane-offset maneuver based at least in part on the comparison between the lateral distance and the threshold distance. The lane-offset maneuver includes driving the autonomous vehicle off-center in a current lane traveled by the autonomous vehicle in an opposite direction relative to the vehicle until the lateral distance between the autonomous vehicle and the vehicle is at least equal to the threshold distance.
[0022] Thus, the disclosed system provides several practical applications and technical advantages, including: 1) improving navigation of an autonomous vehicle by allowing the autonomous vehicle to drive off-center in its current lane to avoid an encroaching, encroaching, or oversized vehicle in an adjacent lane, and 2) determining a lane offset distance at which the autonomous vehicle will drive off-center in its current lane such that the distance between the autonomous vehicle and the encountered vehicle is equal to a predetermined threshold distance.
[0023] (Implementation of lane deviation maneuver based on road curvature and trailer angle) The present disclosure further contemplates a system and method for implementing lane-leaning maneuvers for an autonomous vehicle based on road curvature and trailer angle. For example, the autonomous vehicle may be a semi-truck tractor unit attached with a trailer. While traveling on a curved road, the trailer of the autonomous vehicle may deviate from a straight line due to the road curvature. In other words, the trailer of the autonomous vehicle may sway to the left or right depending on the direction of the road curvature. Similarly, while traveling on a straight road, wind blowing across the autonomous vehicle may cause the trailer of the autonomous vehicle to sway or warp from a straight line. This may cause a trailer angle between the trailer of the autonomous vehicle and the semi-truck tractor unit of the autonomous vehicle. In such a case, the disclosed system may instruct the autonomous vehicle to perform a lane-leaning maneuver to compensate for the deviation of the trailer of the autonomous vehicle from a straight line.
[0024] The disclosed system calculates a first lane offset adjustment distance amount associated with the road curvature and a second lane offset adjustment distance amount associated with the trailer angle. The disclosed system calculates a total lane offset adjustment distance amount by combining the first and second lane offset adjustment distance amounts. The total lane offset adjustment distance amount is the distance the autonomous vehicle drives off-center in the current lane to compensate for the autonomous vehicle's trailer moving to the side from a straight line. The disclosed system may command the autonomous vehicle to offset toward the right or left (while in the current lane) based on the total lane offset adjustment distance amount.
[0025] According to an embodiment, a system includes an autonomous vehicle and a control device. The autonomous vehicle is configured to travel along a road. The autonomous vehicle is a semi-truck tractor unit attached with a trailer. The control device is associated with the autonomous vehicle. The control device includes a memory and a processor. The memory is configured to store map data comprising one or more roads ahead of the autonomous vehicle. The processor is operatively coupled to the memory. The processor may determine that the autonomous vehicle is approaching a curved road based at least in part on the map data. The processor may determine a road radius of the curved road from the map data. The processor may calculate a first lane offset adjustment amount associated with a road curvature of the curved road based at least in part on the road radius. The processor may determine a trailer angle between the trailer and the semi-truck tractor unit. The processor may calculate a second lane offset adjustment amount associated with the trailer angle based at least in part on the trailer angle. The processor may calculate a total lane offset adjustment amount by combining the first lane offset adjustment amount and the second lane offset adjustment amount. The processor may command the autonomous vehicle to perform a lane-offset maneuver, the lane-offset maneuver including driving the autonomous vehicle off-center within a curved lane currently being traveled by the autonomous vehicle based at least in part on the total lane-offset adjustment amount.
[0026] Thus, the disclosed system provides several practical applications and technical advantages, including: 1) improving the safety of operating an autonomous vehicle relative to surrounding vehicles and the like; 2) improving navigation of an autonomous vehicle on curved roads by allowing the autonomous vehicle to drive off-center within its current lane to compensate for the deviation of the autonomous vehicle's trailer from a straight line; and 3) determining the total lane offset adjustment distance that the autonomous vehicle will drive off-center within its current lane such that neither the autonomous vehicle's semi-truck tractor unit or trailer encroach on a side lane.
[0027] Thus, the systems described in this disclosure can be integrated into practical applications to determine more efficient, safe, and reliable navigation solutions for autonomous vehicles and other vehicles on the same road as the autonomous vehicle.
[0028] Certain embodiments of the present disclosure may include some, all, or none of these advantages. These advantages and other features will be more clearly understood from the following detailed description considered in conjunction with the accompanying drawings and claims. [Brief description of the drawings]
[0029] For a more complete understanding of the present disclosure, reference is now made to the following brief description taken in conjunction with the accompanying drawings and detailed description, in which like reference numerals represent like parts.
[0030] [Figure 1] FIG. 1 illustrates an embodiment of a system configured to implement lane-shifting maneuvers to avoid vehicles.
[0031] [Diagram 2] FIG. 2 illustrates an example flow chart of a method for implementing a lane-shifting maneuver to avoid a vehicle.
[0032] [Diagram 3] FIG. 3 illustrates an embodiment of a system configured to implement lane-shifting maneuvers for negotiating curved roads.
[0033] [Figure 4] FIG. 4 illustrates an example flow chart of a method for implementing lane-shifting maneuvers for negotiating curved roads.
[0034] [Diagram 5] FIG. 5 illustrates a block diagram of an example autonomous vehicle configured to implement autonomous driving operations.
[0035] [Figure 6] FIG. 6 illustrates an example system for providing autonomous driving operations for use by the autonomous vehicle of FIG. 5 .
[0036] [Figure 7] FIG. 7 illustrates a block diagram of an on-board control computer included within the autonomous vehicle of FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0037] In some cases, while the autonomous vehicle is traveling along a road, a vehicle may drive too close to the autonomous vehicle and encroach into the lane the autonomous vehicle is in. In some cases, the trailer of the autonomous vehicle may swerve left or right depending on the direction of the road. Such instances may lead to unsafe driving conditions for the autonomous vehicle and other vehicles.
[0038] As described above, previous techniques fail to provide an efficient, reliable, and safe navigation solution for an autonomous vehicle when the autonomous vehicle encounters an encroaching, encroaching, or oversized vehicle, or when the autonomous vehicle approaches a curved road. The present disclosure provides various systems, methods, and devices for 1) implementing lane-shifting maneuvers to avoid an encroaching, encroaching, or oversized vehicle, 2) implementing lane-shifting maneuvers to negotiate a curved road, and 3) providing a safe driving experience for autonomous vehicles, other vehicles, and pedestrians. FIG. 1 illustrates an embodiment of a system 100 for implementing lane-shifting maneuvers to avoid vehicles. FIG. 2 illustrates an embodiment of a method 200 for implementing lane-shifting maneuvers to avoid vehicles. FIG. 3 illustrates an embodiment of a system 300 for implementing lane-shifting maneuvers to negotiate a curved road. FIG. 4 illustrates an embodiment of a method 400 for implementing lane-shifting maneuvers to negotiate a curved road. 5-7 illustrate an example autonomous vehicle and its various systems and devices for implementing autonomous driving operations by the autonomous vehicle.
[0039] Exemplary Systems for Implementing Lane-Leaning Steering for Autonomous Vehicles 1 illustrates an embodiment of a system 100 for implementing a lane-shifting maneuver 130 for an autonomous vehicle 502. FIGURE 1 further illustrates a simplified schematic diagram of a road 102 navigated by the autonomous vehicle 502. In some embodiments, the system 100 comprises the autonomous vehicle 502 and its components, including a control device 550 and a sensor 546.
[0040] The control device 550 includes a processor 122 in signal communication with a memory 126. The memory 126 may store software instructions 128 that, when executed by the processor 122, cause the control device 550 to perform one or more functions described herein. For example, the software instructions 128, when executed, may cause the processor 122 to instruct the autonomous vehicle 502 to implement a lane-shifting maneuver 130 such that the autonomous vehicle 502 may drive off-center within its current lane and maintain a safe distance from one or more surrounding vehicles 108. The system 100 may be configured as shown or in any other suitable configuration.
[0041] In general, the system 100 may be configured to implement the lane shifting maneuver 130 in response to detecting: 1) the vehicle 108 is encroaching into the current lane 104a being traveled by the autonomous vehicle 502; 2) the distance between the vehicle 108 and a lane marker 106 between the autonomous vehicle 502 and the vehicle 108 is less than a threshold distance 132; and / or 3) historical driving behavior 162 associated with the vehicle 108 indicates that (in one or more instances) the vehicle 108 is encroaching or has encroached on one or more lanes 104.
[0042] While driving along the road 102, the autonomous vehicle 502 may face an encroaching, encroaching, or oversized vehicle 108 in an adjacent lane 104. There may be situations in which the distance between the autonomous vehicle 502 and such a vehicle 108 may be less than the threshold distance 132. For example, the encroaching or encroaching vehicle 108 may drive too close to (e.g., pass the threshold distance 132) or even cross the lane marker 106a between the autonomous vehicle 502 and the vehicle 108. In another example, the distance between the oversized vehicle 108 and the autonomous vehicle 502 may be less than the threshold distance 132 due to the larger space that the oversized vehicle 108 occupies.
[0043] In such cases, it may not be safe or feasible to move around to another lane 104. For example, the autonomous vehicle 502 may be in a side lane 104a on the road 102 and there may not be another lane 104 to move around. In another example, there may be traffic in an adjacent lane 104. In another example, the autonomous vehicle 502 may need to stay on the current lane 104a to follow its navigation or route plan 116 to reach its destination.
[0044] In such cases, a safer driving maneuver may be to perform a lane-offset maneuver 130. The lane-offset maneuver 130 may allow the autonomous vehicle 502 to drive off-center within the current lane 104a traveled by the autonomous vehicle 502 in the opposite direction relative to the vehicle 108 until the lateral distance 138 between the vehicle 108 and the autonomous vehicle 502 is at least equal to the threshold distance 132. In other words, the autonomous vehicle 502 offsets toward the other side of the current lane 104a (away from the vehicle 108) until the lateral distance 138 between the vehicle 108 and the autonomous vehicle 502 is at least equal to the threshold distance 132. In some embodiments, the autonomous vehicle 502 may perform the lane-offset maneuver 130 until the autonomous vehicle 502 and the vehicle 108 are no longer adjacent to each other, e.g., until the autonomous vehicle 502 passes the vehicle 108 or the vehicle 108 passes the autonomous vehicle 502. In one example, the autonomous vehicle 502 may perform a lane-shifting maneuver 130 until it is determined that no portion of the autonomous vehicle 502 overlaps with any portion of the vehicle 108 traveling in the adjacent lane 104. In another example, the autonomous vehicle 502 may perform a lane-shifting maneuver 130 until it is determined that a less-than-threshold portion of the vehicle 108 traveling in the adjacent lane 104 overlaps with any portion of the autonomous vehicle 502. The threshold portion of the vehicle 108 may be one-third, one-half, two-thirds, or any other suitable portion of the length of the vehicle 108.
[0045] System 100 may further be configured to perform a minimum risk maneuver 140 if it is determined that lane-shifting maneuver 130 is not feasible or that it is not safe to perform lane-shifting maneuver 130. Minimum risk maneuver 140 may include, among other things, slowing down autonomous vehicle 502, increasing the speed of autonomous vehicle 502 until autonomous vehicle 502 and vehicle 108 are no longer adjacent to one another.
[0046] Various use cases in which the autonomous vehicle 502 encounters a situation that may lead to performing a lane-shifting maneuver 130 are further described below in conjunction with the operational flow of the system 100. In the example use case described in FIG. 1, the autonomous vehicle 502 encounters an intruding, encroaching, or oversized vehicle 108. However, the example use case described in FIG. 1 is not meant to limit the scope of the present disclosure. Those skilled in the art will recognize other use cases and embodiments in light of the present disclosure. In some examples, the autonomous vehicle 502 may encounter an obstacle or object 142 obstructing at least a portion of the roadway 102, including a construction zone, a pedestrian on the side of the roadway 102, an emergency vehicle 108 parked or moving on an emergency lane, and a person standing on the side of the lane 104 paying attention to their vehicle 108. In any of these examples and use cases where an intruding, encroaching, or oversized vehicle 108 is encountered, the system 100 may treat each of these as a non-player character 144 that may result in performing a lane-shifting maneuver 130 and avoid each of these non-player characters 144. A non-player character 144 may be any object 142 with which the autonomous vehicle 502 interacts.
[0047] (System Components) In some embodiments, the autonomous vehicle 502 may include a semi-truck tractor unit (see FIG. 5) mounted on a trailer for transporting freight or cargo from one location to another. The autonomous vehicle 502 may be configured to travel generally along roads 102 in an autonomous mode. The autonomous vehicle 502 may be navigated using multiple components, which are described in detail in FIGS. 5-7. The operation of the autonomous vehicle 502 is described in more detail in FIGS. 5-7. The corresponding description below includes a brief description of certain components of the autonomous vehicle 502.
[0048] The control device 550 may generally be configured to control the operation of the autonomous vehicle 502 and its components and facilitate autonomous driving of the autonomous vehicle 502. The control device 550 may be further configured to determine a path ahead of the autonomous vehicle 502 that is safe to travel and free of objects / obstacles, and navigate the autonomous vehicle 502 to proceed within that path. This process is described in more detail in Figures 5-7. The control device 550 may generally include one or more data processors in signal communication with subsystem components of the autonomous vehicle 502 (see Figure 5).
[0049] The control device 550 may be configured to detect objects on and around the road 102 by analyzing the sensor data 134 and / or the map data 114. For example, the control device 550 may detect objects on and around the road 102 by implementing the object detection machine learning module 112. The object detection machine learning module 112 may be implemented using neural networks and / or machine learning algorithms to detect objects from images, videos, infrared images, point clouds, radar data, etc. The object detection machine learning module 112 is described in further greater detail below. The control device 550 may receive sensor data 134 from sensors 546 positioned on the autonomous vehicle 502 and determine a safe path to proceed. The sensor data 134 may include data captured by the sensors 546.
[0050] The sensors 546 may be configured to capture any objects within their detection zone or field of view, including, among others, landmarks, lane markers, lane boundaries, road boundaries, vehicles 108, pedestrians, road / traffic signs, etc. The sensors 546 may include cameras, LiDAR sensors, motion sensors, infrared sensors, etc. In some embodiments, the sensors 546 may be positioned around the autonomous vehicle 502 and capture the environment surrounding the autonomous vehicle 502. For further description of the sensors 546, please refer to the corresponding description of FIG.
[0051] (Control Device) The control device 550 is described in detail in FIG. 5. Briefly, the control device 550 may include a processor 122 in signal communication with a memory 126 and a network interface 124. The processor 122 may include one or more processing units to perform various functions as described herein. The memory 126 may store any data and / or instructions used by the processor 122 to perform its functions. For example, the memory 126 may store software instructions 128 that, when executed by the processor 122, cause the control device 550 to perform one or more functions described herein.
[0052] The processor 122 may be one of the data processors 570 illustrated in FIG. 5. The processor 122 comprises one or more processors operably coupled to the memory 126. The processor 122 is any electronic circuitry including, but not limited to, a state machine, one or more central processing unit (CPU) chips, a logic unit, a core (e.g., a multi-core processor), a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), or a digital signal processor (DSP). The processor 122 may be a programmable logic device, a microcontroller, a microprocessor, or any suitable combination of the foregoing. The processor 122 is communicatively coupled to and in signal communication with the network interface 124 and the memory 126. The one or more processors may be configured to process data and may be implemented in hardware or software. For example, the processor 122 may be 8-bit, 16-bit, 32-bit, 64-bit, or any other suitable architecture. The processor 122 may include an arithmetic logic unit (ALU) for performing arithmetic and logical operations, processor registers for supplying operands to the ALU and storing results of ALU operations, and a control unit for fetching instructions from memory and executing them by directing the coordinated operation of the ALU, registers, and other components. One or more processors may be configured to implement various instructions. For example, one or more processors may be configured to execute software instructions 128 for implementing functions disclosed herein, such as some or all of those described with respect to Figures 1-7. In some embodiments, the functions described herein are implemented using a logic unit, an FPGA, an ASIC, a DSP, or any other suitable hardware or electronic circuitry.
[0053] The network interface 124 may be a component of the network communications subsystem 592 illustrated in FIG. 5. The network interface 124 may be configured to enable wired and / or wireless communications. The network interface 124 may be configured to communicate data between the control device 550 and other network devices, systems, or domains. For example, the network interface 124 may comprise a WIFI interface, a local area network (LAN) interface, a wide area network (WAN) interface, a modem, a switch, or a router. The processor 122 may be configured to send and receive data using the network interface 124. The network interface 124 may be configured to use any suitable type of communications protocol.
[0054] The memory 126 may be one of the data storage devices 590 illustrated in Figure 5. The memory 126 may store any of the information illustrated in Figures 1-7 along with any other data, instructions, logic, rules, or code that, when executed by the processor 122, is operable to implement the functions described herein. For example, the memory 126 may store software instructions 128, lane deviation maneuvers 130, minimum risk maneuvers 140, locations of lane markers 106, threshold distances 132, sensor data 134, lateral distances 138a-c, distance 136, encroached distance 146, threshold duration 148, trajectories 150, 152, 158, and 168, road condition data 154, longitudinal distances 156 and 166, distance 160, historical driving behavior 162, obstacles / objects 142, non-player characters 144, driving pattern predictions 164, lane deviation amounts 110, object detection machine learning module 112, map data 114, route plans 116, driving instructions 118, and / or any other data / instructions. The software instructions 128 include code that, when executed by the processor 122, causes the control device 550 to perform the functions described herein, such as some or all of those described in FIGS. 1-7. The memory 126 may comprise one or more disks, tape drives, or solid state drives, and may be used as an overflow data storage device to store programs when such programs are selected for execution, and to store instructions and data read during program execution. The memory 126 may be volatile or non-volatile, and may comprise read only memory (ROM), random access memory (RAM), ternary content addressable memory (TCAM), dynamic random access memory (DRAM), and static random access memory (SRAM). The memory 126 may include one or more of a local database, a cloud database, a network attached storage device (NAS), and the like.
[0055] The object detection machine learning module 112 may be implemented by the processor 122 executing software instructions 128 and may generally be configured to detect objects and obstacles 142 from the sensor data 134. The object detection machine learning module 112 may be implemented using neural networks and / or machine learning algorithms to detect objects from any type of data, such as images, videos, infrared images, point clouds, radar data, etc.
[0056] In some embodiments, the object detection machine learning module 112 may be implemented using machine learning algorithms such as support vector machines (SVM), naive Bayes, logistic regression, k-nearest neighbors, decision trees, etc. In some embodiments, the object detection machine learning module 112 may utilize multiple neural network layers, convolutional neural network layers, etc., where the weights and biases of these layers are optimized in the training process of the object detection machine learning module 112. The object detection machine learning module 112 may be trained by a training dataset that may include samples of one or more object-labeled data types within each sample. For example, the training dataset may include sample images of objects (e.g., vehicles, lane markings, pedestrians, road signs, obstacles, etc.) labeled with objects within each sample image. Similarly, the training dataset may include samples of other data types, such as object-labeled videos, infrared images, point clouds, radar data, etc., within each sample data. The object detection machine learning module 112 may be trained, tested, and refined by the training dataset and the sensor data 134. The object detection machine learning module 112 uses the sensor data 134 (not labeled with objects) to increase the accuracy of its predictions in detecting objects. For example, supervised and / or unsupervised machine learning algorithms may be used to validate the predictions of the object detection machine learning module 112 in detecting objects in the sensor data 134.
[0057] Map data 114 may include a virtual map of a city or area that includes roads 102. In some examples, map data 114 may include map 658 and map database 636 (see FIG. 6 for a description of map 658 and map database 636). Map data 114 may include drivable areas, such as roads 102, routes, highways, and non-drivable areas, such as terrain (as determined by occupancy grid module 660 (see FIG. 6 for a description of occupancy grid module 660)). Map data 114 may define location coordinates of road signs, lanes, lane markings, lane boundaries, road boundaries, traffic lights, obstacles, and the like.
[0058] The route plan 116 is a plan for traveling from a starting location (e.g., a first autonomous vehicle departure / arrival point) to a destination (e.g., a second autonomous vehicle departure / arrival point). For example, the route plan 116 may specify a combination of one or more streets, roads, and highways in a particular order from the starting location to the destination. The route plan 116 may specify stages including a first stage (e.g., moving out from the starting location / origination point), multiple intermediate stages (e.g., proceeding along particular lanes of one or more particular streets / roads / highways), and a final stage (e.g., entering the destination / arrival point). The route plan 116 may include other information about the route from the starting location to the destination, such as road / traffic signs within the route plan 116.
[0059] The driving instructions 118 may be implemented by the planning module 662 (see description of the planning module 662 in FIG. 6). The driving instructions 118 may include instructions and rules for adapting the autonomous driving of the autonomous vehicle 502 according to the driving rules of each stage of the route plan 116. For example, the driving instructions 118 may include instructions for staying within a speed range of the road 102 traveled by the autonomous vehicle 502, for adapting the speed of the autonomous vehicle 502 to changes observed by the sensor 546, such as surrounding vehicles, object speeds, etc. within the detection area of the sensor 546.
[0060] The control device 550 may receive the object detection machine learning module 112, the map data 114, the route plan 116, the driving instructions 118, and / or any other data / instructions from a supervising server (not shown), which may be configured to supervise the operation of the autonomous vehicle 502, construct the map data 114, determine the route plan 116, determine the driving instructions 118, among other operations.
[0061] The threshold distance 132 may generally represent a safe distance that the control device 550 maintains (or attempts to maintain) between the autonomous vehicle 502 and its surrounding objects / obstacles 142. In one example, the control device 550 may define the threshold distance 132 based on road conditions 154, such as traffic and weather, on the road 102 traveled by the autonomous vehicle 502. For example, in heavy traffic, the threshold distance 132 from the autonomous vehicle 502 may be greater (e.g., 8 feet or any suitable distance) compared to a road without traffic. In another example, during bad weather conditions, the threshold distance 132 from the autonomous vehicle 502 may be greater (e.g., 6 feet, 7 feet, or any suitable distance) compared to normal weather conditions. In another example, the control device 550 may define different threshold distances 132 between the autonomous vehicle 502 and each object 142 based on one or more of the size of the object 142 and the type of the object 142 (e.g., vehicle, pedestrian, road sign, etc.). For example, if the first object 142 is a small vehicle 108, the first threshold distance 132 between the small vehicle 108 and the autonomous vehicle 502 may be determined to be smaller compared to the threshold distance 132 between the autonomous vehicle 502 and the oversized vehicle 108. In another example, if the second object 142 is a pedestrian or person on the side of the road, the second threshold distance 132 between the pedestrian and the autonomous vehicle 502 may be larger compared to the threshold distance 132 between the autonomous vehicle 502 and a road sign.
[0062] In some embodiments, the control device 550 may define a safety boundary or bounding box around each object 142 on and around the road 102 based on the size and type of object 142, such as a vehicle, pedestrian, road sign, etc. The bounding box around each object 142 may represent a safe distance that the control device 550 will keep (or attempt to keep) between the autonomous vehicle 502 and each object 142. The control device 550 may use the bounding box around each object 142 to determine various threshold distances 132 between the autonomous vehicle 502 and each object 142 on and around the road 102.
[0063] (Operation flow for implementing lane-shifting maneuver) The operational flow of the system 100 begins when the control device 550 receives sensor data 134 from the sensor 546 .
[0064] In an example operation, assume that an autonomous vehicle 502 is traveling along a road 102. While traveling, the sensors 546 capture sensor data 134 that describes the environment around the autonomous vehicle 502. The sensor data 134 is associated with one or more objects on the road 102. From the sensor data 134, the control device 550, by implementing the object detection machine learning module 112, may detect the position of the lane markers 106 and the distances 136a and 136b between the autonomous vehicle 502 and the lane markers 106a and 106b, respectively.
[0065] Assuming the vehicle 108 is on the road 102, the control device 550 may detect the presence of the vehicle 108 from the sensor data 134. The control device 550 may determine the lateral distance 138 between the autonomous vehicle 502 and the vehicle 108.
[0066] Control device 550 may compare lateral distance 138 to threshold distance 132. Control device 550 may determine whether to command autonomous vehicle 502 to perform lane-shifting maneuver 130 based on the comparison between lateral distance 138 and threshold distance 132. If control device 550 determines that it is safe and feasible to perform lane-shifting maneuver 130, control device 550 may command autonomous vehicle 502 to perform lane-shifting maneuver 130. Otherwise, control device 550 may command autonomous vehicle 502 to perform minimum-risk maneuver 140.
[0067] In some example scenarios, the autonomous vehicle 502 may encounter vehicles 108: 1) on either side adjacent to the autonomous vehicle 502, 2) on either side and in front of the autonomous vehicle 502, and 3) in front of the autonomous vehicle 502 if the vehicle 108 is stopped on the side of the road 102. The corresponding discussion below describes various example use cases of encountering a vehicle 108 (or, generally, a non-player character 144) that may lead to performing a lane-shifting maneuver 130.
[0068] (Encounter with adjacent vehicle to autonomous vehicle) In an example use case, assume that the control device 550, by implementing the object detection machine learning module 112, detects the presence of a vehicle 108a from the sensor data 134, and the vehicle 108a is detected on either side of the autonomous vehicle 502.
[0069] The control device 550 may determine whether to instruct the autonomous vehicle 502 to perform a lane-shifting maneuver 130 to avoid the vehicle 108a, whether the lane-shifting maneuver 130 is feasible and safe, as described below, and may instruct the autonomous vehicle 502 to perform the lane-shifting maneuver 130 if it is determined that the lane-shifting maneuver 130 is feasible and safe.
[0070] In this process, the control device 550 may determine a lateral distance 138a between the autonomous vehicle 502 and the vehicle 108a from the sensor data 134. The control device 550 may compare the lateral distance 138a to the threshold distance 132.
[0071] In the example of FIG. 1 , the control device 550 may determine that the lateral distance 138a between the autonomous vehicle 502 and the vehicle 108a is less than the threshold distance 132. In response, the control device 550 may determine whether to command the autonomous vehicle 502 to perform a lane-shifting maneuver 130. If the control device 550 determines that it is feasible and safe to perform the lane-shifting maneuver 130, the control device 550 may command the autonomous vehicle 502 to perform the lane-shifting maneuver 130. Otherwise, the control device 550 may command the autonomous vehicle 502 to perform a minimum-risk maneuver 140. The minimum-risk maneuver 140 may include slowing or increasing the speed of the autonomous vehicle 502 until the autonomous vehicle 502 and the vehicle 108a are no longer adjacent to one another.
[0072] 1, the vehicle 108a may be entering (or crossing) the lane marker 106a. To determine whether the vehicle 108a is crossing the lane marker 106a, the control device 550 may perform one or more operations described below.
[0073] The control device 550 may determine a distance 136a between the autonomous vehicle 502 and the lane marker 106a. The control device 550 may compare the distance 136a to the lateral distance 138a between the autonomous vehicle 502 and the vehicle 108a. If the control device 550 determines that the distance 136a is less than the lateral distance 138a between the autonomous vehicle 502 and the vehicle 108a, the control device 550 may determine that the vehicle 108a is encroaching into the lane 104a. Otherwise, the control device 550 may determine that the vehicle 108a is not encroaching into the lane 104a.
[0074] In this example, in response to determining that the lateral distance 138a between the autonomous vehicle 502 and the vehicle 108a is less than the distance 136a between the autonomous vehicle 502 and the lane marker 106a, the control device 550 may determine that the vehicle 108a is encroaching into the current lane 104a being traveled by the autonomous vehicle 502. In response, the control device 550 may determine whether to instruct the autonomous vehicle 502 to perform a lane-shifting maneuver 130 and whether it is safe and feasible to perform the lane-shifting maneuver.
[0075] (Determining whether it is feasible to perform a lane-shifting maneuver) To determine whether it is feasible to perform the lane-shifting maneuver 130, the control device 550 may perform one or more of the actions described below.
[0076] In some embodiments, the control device 550 may determine whether it is feasible to perform the lane-shifting maneuver 130 based on determining whether there is sufficient margin or distance 136 available on the other side of the autonomous vehicle 502 to perform the lane-shifting maneuver 130. To this end, the control device 550 may determine the amount of the current lane 104a encroached upon by the vehicle 108a. For example, the control device 550 may determine the amount of encroached distance 146 into the lane 104a encroached upon by the vehicle 108a.
[0077] The control device 550 may determine the available distance 136b (or available margin 136b) on the other side of the autonomous vehicle 502 on the current lane 104a. The control device 550 may determine whether there is sufficient margin 136b or available distance 136b on the other side of the autonomous vehicle 502 to perform a lane-shifting maneuver, i.e., to drive off-center and shift toward the lane marker 106b. To this end, the control device 550 may compare the invaded distance 146 to the available distance 136b.
[0078] The control device 550 may determine whether the lane-leaning maneuver 130 may be performed based on a comparison between the intruded distance 146 and the available distance 136b. If the control device 550 determines that the available distance 136b is greater than the intruded distance 146 (and / or that there is sufficient available distance 136b to perform the lane-leaning maneuver 130), the control device 550 may command the autonomous vehicle 502 to perform the lane-leaning maneuver 130.
[0079] The control device 550 may instruct the autonomous vehicle 502 to perform the lane-leaning maneuver 130 if the control device 550 determines that the lane-leaning maneuver 130 can be performed or is feasible and safe. The process of determining whether it is safe to perform the lane-leaning maneuver 130 is further described below. Briefly, the control device 550 may determine that it is safe to perform the lane-leaning maneuver 130 by determining whether there is traffic (e.g., another vehicle 108) on the road 102, i.e., whether the traffic would allow the autonomous vehicle 502 to perform the lane-leaning maneuver 130. If it is determined that there is no or little traffic on the road 102, the control device 550 may determine that it is safe to perform the lane-leaning maneuver 130 and instruct the autonomous vehicle 502 to perform the lane-leaning maneuver 130.
[0080] If the control device 550 determines that the available distance 136b is less than the intruded distance 146 (and / or that there is not enough available distance 136b to perform the lane-shifting maneuver 130), the control device 550 may determine that the lane-shifting maneuver 130 cannot be performed or is not feasible. In response, the control device 550 may instruct the autonomous vehicle 502 to perform the minimum-risk maneuver 140. In these operations, the control device 550 takes into account the size, width, and length of the autonomous vehicle 502 when determining whether there is enough available distance 136b to perform the lane-shifting maneuver 130.
[0081] In some embodiments, the control device 550 may determine whether it is feasible to perform a lane-shifting maneuver 130 based on determining that if the lane-shifting maneuver 130 is performed, the future lateral distance 138a between the autonomous vehicle 502 and the vehicle 108a will be at least equal to the threshold distance 132 within the threshold period 148.
[0082] If control device 550 determines that the future lateral distance 138a between autonomous vehicle 502 and vehicle 108a will be at least equal to threshold distance 132 within threshold period 148 if lane-shifting maneuver 130 is performed, control device 550 may determine that lane-shifting maneuver 130 is feasible. In response, control device 550 may instruct autonomous vehicle 502 to perform lane-shifting maneuver 130 if it is determined to be safe. Otherwise, control device 550 may instruct autonomous vehicle 502 to perform minimum-risk maneuver 140.
[0083] The threshold period 148 may be two minutes, five minutes, or any other suitable duration. The control device 550 may define the threshold period 148 based on one or more of the road condition data 154 (such as traffic and weather data associated with the road 102), the size of the autonomous vehicle 502, the speed of the autonomous vehicle 502, the trajectory 150 of the autonomous vehicle 502, and the size of the vehicle 108a, the speed of the vehicle 108a, and the trajectory 152 of the vehicle 108a.
[0084] To determine whether the lane-shifting maneuver 130 can be implemented within the threshold period 148, the control device 550 may perform one or more of the following actions.
[0085] The control device 550 may, for example, determine the speed (or estimated speed) and position of the vehicle 108a from the sensor data 134. Based on the speed and position of the vehicle 108a, the control device 550 may determine the trajectory 152 of the vehicle 108a. Similarly, based on the speed and position of the autonomous vehicle 502, the control device 550 may determine the trajectory 150 of the autonomous vehicle 502 when the lane-shifting maneuver 130 is performed.
[0086] The control device 550 predicts the future lateral distance 138a between the autonomous vehicle 502 and vehicle 108a if the autonomous vehicle 502 follows trajectory 150 and vehicle 108a follows trajectory 152.
[0087] Control device 550 may compare future lateral distance 138a to threshold distance 132. Control device 550 may determine that lane-shifting maneuver 130 can be performed within threshold period 148 if future lateral distance 138a will be at least equal to threshold distance 132 within threshold period 148. In response, control device 550 may command autonomous vehicle 502 to perform lane-shifting maneuver 130 (if determined to be safe).
[0088] In some embodiments, if control device 550 determines that predicted future lateral distance 138a will be less than threshold distance 132 within threshold time period 148, control device 550 may determine that lane-shifting maneuver 130 cannot be performed and / or is unsafe to be performed, for example, due to traffic on road 102. In response, control device 550 may instruct autonomous vehicle 502 to perform minimum-risk maneuver 140.
[0089] In another embodiment, if the control device 550 determines that the predicted future lateral distance 138a will be less than the threshold distance 132 within the threshold period 148, the control device 550 may determine whether it is safe to cross over into the adjacent lane 104c based on the traffic on the lane 104c. If the control device 550 determines that there is no or little traffic (e.g., another vehicle 108) in the adjacent lane 104c, the control device 550 may determine that it is safe to cross over temporarily into lane 104c. In response, the control device 550 may command the autonomous vehicle 502 to perform a lane-shifting maneuver 130 and command the autonomous vehicle 502 to temporarily cross over into lane 104c and occupy a similar amount of space from lane 104c until the lateral distance 138a between the autonomous vehicle 502 and the vehicle 108a is equal to the threshold distance 132. The control device 550 may command the autonomous vehicle 502 to drive back into its original lane 104a once it determines that the autonomous vehicle 502 and vehicle 108a are no longer adjacent to one another.
[0090] In some embodiments, the control device 550 may determine whether it is feasible to perform the lane-shifting maneuver 130 based on the road structure. For example, when the autonomous vehicle 502 is traveling on a curved road, the trailer attached to the semi-truck tractor unit of the autonomous vehicle 502 takes up more space in the lane. Thus, the control device 550 may take the road structure into account when determining whether it is feasible to perform the lane-shifting maneuver 130. This use case is described in detail in FIGS. 3 and 4.
[0091] (Implementation of lane deviation maneuver) In some embodiments, in response to determining that it is feasible and safe to perform lane-shifting maneuver 130, control device 550 may instruct autonomous vehicle 502 to perform lane-shifting maneuver 130.
[0092] To do this, the control device 550 causes the autonomous vehicle 502 to drive off-center from the current lane 104a and offset toward the lane marker 106b (i.e., the other side of the lane 104a compared to where the vehicle 108a was detected).
[0093] The control device 550 may determine the lane deviation amount 110. The lane deviation amount 110 is the distance between the centerline of the current lane 104a and the trajectory line 150. In other words, the lane deviation amount 110 is the distance that the autonomous vehicle 502 deviates from the centerline of the lane 104a until the lateral distance 138a between the autonomous vehicle 502 and the vehicle 108a is at least equal to the threshold distance 132. The autonomous vehicle 502 may perform the lane deviation maneuver 130 until the lateral distance 138a between the autonomous vehicle 502 and the vehicle 108a is at least equal to the threshold distance 132.
[0094] The control device 550 may maintain a consistent lane offset 110 until the autonomous vehicle 502 is no longer adjacent to the vehicle 108. In some embodiments, the control device 550 may maintain a consistent lane offset 110 even as the vehicle 108 changes lateral distance 138. In some embodiments, the control device 550 may adjust the lane offset 110 to keep at least a threshold distance 132 with the vehicle 108.
[0095] (Implementation of minimum risk maneuvers) In some embodiments, in response to determining that performing lane-shifting maneuver 130 is not feasible (and is not safe to be performed), control device 550 may command autonomous vehicle 502 to perform minimum-risk maneuver 140. Minimum-risk maneuver 140 may include, among other maneuvers, slowing autonomous vehicle 502, increasing autonomous vehicle 502's speed until autonomous vehicle 502 and vehicle 108a are no longer adjacent to one another.
[0096] In an example where the autonomous vehicle 502 encounters a vehicle 108a on either side of the autonomous vehicle 502, if the control device 550 determines that the speed of the vehicle 108a exceeds the speed of the autonomous vehicle 502, the minimum risk maneuver 140 may include slowing down the autonomous vehicle 502 and letting the vehicle 108a pass the autonomous vehicle 502. In another example, if the control device 550 determines that the speed of the vehicle 108a is less than the speed of the autonomous vehicle 502, the minimum risk maneuver 140 may include increasing the speed of the autonomous vehicle 502. In another example, if the vehicle 108a is behind the autonomous vehicle 502, the minimum risk maneuver 140 may include increasing the speed of the autonomous vehicle 502. In another example, if the vehicle 108a is in front of the autonomous vehicle 502, the minimum risk maneuver 140 may include slowing down the autonomous vehicle 502.
[0097] In some embodiments, the control device 550 may perform a minimum risk maneuver 140 in addition to the lane-shifting maneuver 130 if the lane-shifting maneuver is determined to be feasible and safe.
[0098] (Encountering a vehicle in front of and on either side of the autonomous vehicle) In another use case, assume that the control device 550, by implementing the object detection machine learning module 112, detects the presence of a vehicle 108b from the sensor data 134, and that the vehicle 108b is detected in front of the autonomous vehicle 502 and on either side of the lane 104a being traveled by the autonomous vehicle 502.
[0099] The control device 550 may determine whether to instruct the autonomous vehicle 502 to perform a lane-shifting maneuver 130 to avoid vehicle 108b, whether the lane-shifting maneuver 130 is feasible and safe, as described below, and may instruct the autonomous vehicle 502 to perform the lane-shifting maneuver 130 if it is determined that the lane-shifting maneuver 130 is feasible and safe.
[0100] In this process, the control device 550 may determine a lateral distance 138b between the autonomous vehicle 502 and the vehicle 108b by analyzing the sensor data 134. The control device 550 may compare the lateral distance 138b to the threshold distance 132.
[0101] 1, the control device 550 may determine that the lateral distance 138b between the autonomous vehicle 502 and the vehicle 108b is less than the threshold distance 132. In the illustrated example, the control device 550 may determine that the vehicle 108b is driving too close to the lane marker 106a between the autonomous vehicle 502 and the vehicle 108b. For example, the control device 550 may determine that the distance 160 between the vehicle 108b and the lane marker 106a is less than a threshold, e.g., less than 20 inches, 40 inches, or any other suitable distance.
[0102] In response, control device 550 may determine whether to command autonomous vehicle 502 to perform lane-shifting maneuver 130. Similar to what is described above with respect to the example of vehicle 108a, if control device 550 determines that it is feasible and safe to perform lane-shifting maneuver 130, control device 550 may command autonomous vehicle 502 to perform lane-shifting maneuver 130. Otherwise, control device 550 may command autonomous vehicle 502 to perform minimum-risk maneuver 140. These operations are described below.
[0103] (Determining whether it is feasible to perform a lane-shifting maneuver) In some embodiments, the control device 550 may determine whether it is feasible to perform the lane-shifting maneuver 130 within the threshold period 148 based on the lateral distance 138b, the longitudinal distance 156, the trajectory 150 of the autonomous vehicle 502 when the lane-shifting maneuver 130 is performed, and the trajectory 158 of the vehicle 108b, as described below.
[0104] The control device 550 may determine the longitudinal distance 156 between the autonomous vehicle 502 and the vehicle 108b from the sensor data 134.
[0105] Control device 550 may determine trajectory 150 of autonomous vehicle 502 when lane-shifting maneuver 130 is performed based on the speed and position of autonomous vehicle 502, similar to that described above. Similarly, control device 550 may determine trajectory 158 of vehicle 108b based on the position and speed (or estimated speed) of vehicle 108b.
[0106] The control device 550 predicts the future lateral distance 138b between the autonomous vehicle 502 and vehicle 108b if the autonomous vehicle 502 follows trajectory 150 and vehicle 108b follows trajectory 158.
[0107] Control device 550 may compare predicted future lateral distance 138b to threshold distance 132. If control device 550 determines that predicted future lateral distance 138b will be greater than or equal to threshold distance 132 within threshold period 148, control device 550 may determine that lane-shifting maneuver 130 can be performed. In response, control device 550 may command autonomous vehicle 502 to perform lane-shifting maneuver 130 (if determined to be safe).
[0108] In some embodiments, if control device 550 determines that predicted future lateral distance 138b will be less than threshold distance 132 within threshold period 148, control device 550 may determine that lane shifting maneuver 130 cannot be performed within threshold period 148. In response, control device 550 may instruct autonomous vehicle 502 to perform minimum risk maneuver 140.
[0109] In another embodiment, similar to that described above with respect to the exemplary use case of encountering vehicle 108a, if control device 550 determines that the predicted future lateral distance 138b will be less than threshold distance 132 within threshold period 148, control device 550 may determine whether it is safe to temporarily cross over into adjacent lane 104c. If it is determined that it is safe to temporarily cross over into lane 104c, control device 550 may drive autonomous vehicle 502 beyond lane marker 106b until the lateral distance 138b between autonomous vehicle 502 and vehicle 108b is at least equal to threshold distance 132. Once autonomous vehicle 502 and vehicle 108b are no longer adjacent to one another, control device 550 may drive autonomous vehicle 502 back into its original lane 104a.
[0110] (Determining whether it is safe to perform a lane-shifting maneuver) In some embodiments, the control device 550 may determine whether it is safe to perform the lane-shifting maneuver 130 based on road condition data 154 (e.g., traffic and weather data) and / or historical driving behaviors 162 associated with the surrounding vehicles 108. To this end, while traveling along the road 102, the control device 550 may record the driving behaviors 162 associated with the vehicles 108 on the road 102.
[0111] In one example, the control device 550 may determine that it is not safe to perform the lane-shifting maneuver 130 if it is determined that the driving pattern prediction 164 of the vehicle 108 is below a threshold percentage, e.g., below 70%, 60%, etc., i.e., the driving pattern of the vehicle 108 is highly unpredictable. The control device 550 may determine the driving pattern prediction 164 based on the historical driving behavior 162. In another example, the control device 550 may determine that it is not safe to perform the lane-shifting maneuver 130 if it is determined that the historical driving behavior 162 of the vehicle 108 indicates that the vehicle 108 is encroaching or encroaching on one or more lanes 104, or is driving too close to the lane markers 106 (e.g., driving less than a threshold distance from the lane markers 106).
[0112] In one example with respect to vehicle 108b, control device 550 may determine that historical driving behavior 162 of vehicle 108b indicates that vehicle 108b is encroaching or has encroached on one or more lanes 104. In another example with respect to vehicle 108b, control device 550 may determine that historical driving behavior 162 of vehicle 108b indicates that vehicle 108b is driving too close to lane markers 106 (e.g., less than a threshold distance from lane marker 106).
[0113] In such a case, the control device 550 may determine that it is not safe to perform the lane-shifting maneuver 130.
[0114] In some embodiments, the control device 550 may determine that it is unsafe to perform the lane-shifting maneuver 130 if it is determined that the lane-shifting maneuver 130 cannot be performed within the threshold period 148 due to road conditions 154, such as traffic congestion or undesirable weather conditions on the road 102.
[0115] (Autonomous vehicle encounters a stopped vehicle in front of it) In another use case, assume that the control device 550, by implementing the object detection machine learning module 112, detects the presence of a vehicle 108c from the sensor data 134, and the vehicle 108c is stopped in front of the autonomous vehicle 502 on the side of the road 102 being traveled by the autonomous vehicle. Similarly, the control device 550 may detect the presence of a person on the side of the road 102 from the sensor data 134.
[0116] In a similar use case, the sensor 546 may detect any stationary object 142 on the side of the road 102 that: 1) occupies at least a portion of the road; 2) encroaches on a lane 104 of the road 102; 3) is on an emergency lane; or 4) is on the other side of the lane marker 106b but is too close to the lane marker 106b (e.g., the distance between the object 142 and the lane marker 106b is less than a threshold value of 20 inches, 40 inches, etc.). The stationary object 142 may include a construction cone, a construction barrier, a construction worker, construction equipment, a pedestrian, a vehicle, and / or any other object 142.
[0117] The control device 550 may determine whether to instruct the autonomous vehicle 502 to perform a lane-shifting maneuver 130 to avoid the vehicle 108c, whether the lane-shifting maneuver 130 is feasible and safe, as described below, and may instruct the autonomous vehicle 502 to perform the lane-shifting maneuver 130 if it is determined that the lane-shifting maneuver 130 is feasible and safe.
[0118] In this process, the control device 550 may determine a lateral distance 138c between the autonomous vehicle 502 and the vehicle 108c by analyzing the sensor data 134. The control device 550 may compare the lateral distance 138c to the threshold distance 132.
[0119] 1, the control device 550 may determine that the lateral distance 138c between the autonomous vehicle 502 and the vehicle 108c is less than the threshold distance 132. In the illustrated example, the control device 550 determines that the vehicle 108c is crossing the lane boundary 106b based on determining that the lateral distance 138c is less than the distance 136b between the autonomous vehicle 502 and the lane boundary 106b.
[0120] In response, control device 550 may determine whether to command autonomous vehicle 502 to perform lane-shifting maneuver 130. Similar to what is described above with respect to the example of vehicle 108a, if control device 550 determines that it is feasible and safe to perform lane-shifting maneuver 130, control device 550 may command autonomous vehicle 502 to perform lane-shifting maneuver 130. Otherwise, control device 550 may command autonomous vehicle 502 to perform minimum-risk maneuver 140. These operations are described below.
[0121] (Determining whether it is feasible to perform a lane-shifting maneuver) In some embodiments, the control device 550 may determine whether it is feasible to perform the lane-shifting maneuver 130 within the threshold period 148 based on determining the lateral distance 138c, the longitudinal distance 166, the trajectory 150 of the autonomous vehicle 502 when the lane-shifting maneuver 130 is executed, and whether there is sufficient room or distance 136a on the other side of the autonomous vehicle 502 to perform the lane-shifting maneuver 130, as described below.
[0122] The control device 550 may determine the trajectory 166 of the autonomous vehicle 502 based on the speed and position of the autonomous vehicle 502 when the lane shifting maneuver 130 is performed, similar to that described above with respect to determining the trajectory 150.
[0123] The control device 550 predicts a future lateral distance 138c between the autonomous vehicle 502 and the lane 108c if the autonomous vehicle 502 follows the trajectory 166. The control device 550 may compare the predicted future lateral distance 138c to the threshold distance 132.
[0124] If control device 550 determines that the predicted future lateral distance 138c will be greater than or equal to threshold distance 132 within threshold period 148, control device 550 may determine that lane-shifting maneuver 130 can be implemented. In response, control device 550 may implement lane-shifting maneuver 130 (if determined to be safe).
[0125] In some embodiments, the control device 550 may determine a classification of the vehicles 108 based on their size. For example, the control device 550 may classify oversized vehicles 108, such as buses, into a first class, regular-sized vehicles 108, such as family cars, into a second class, etc. Thus, determining whether to perform a lane-shifting maneuver 130 may be further based on the particular class to which the encountered vehicle 108 belongs.
[0126] In some embodiments, if the control device 550 determines that the predicted future lateral distance 138c will be less than the threshold distance 132 within the threshold period 148, the control device 550 may determine that the lane shifting maneuver 130 cannot be performed within the threshold period 148. In response, the control device 550 may instruct the autonomous vehicle 502 to perform the minimum risk maneuver 140. In another embodiment, similar to that described above with respect to the example use case of encountering the vehicle 108a, if the control device 550 determines that the predicted future lateral distance 138c will be less than the threshold distance 132 within the threshold period 148, the control device 550 may determine whether it is safe to cross over into the adjacent lane 104d.
[0127] (Example Method for Implementing Lane Offset Steering) 2 illustrates an example flowchart of a method 200 for implementing a lane-shifting maneuver 130 for an autonomous vehicle 502. Modifications, additions, or omissions may be made to method 200. Method 200 may include more, fewer, or other operations. For example, operations may be performed in parallel or in any suitable order. At times, autonomous vehicle 502, control device 550, or any component thereof are discussed as performing operations, although any suitable system or component of a system may perform one or more operations of method 200. For example, one or more operations of method 200 may be implemented, at least in part, in the form of software instructions 128 and processing instructions 580 from Figures 1 and 5, respectively, stored on a non-transitory, tangible, machine-readable medium (e.g., memory 126 and data storage device 590 from Figures 1 and 5, respectively) that, when invoked by one or more processors (e.g., processors 122 and 570 from Figures 1 and 5, respectively), can cause the one or more processors to perform operations 202-218.
[0128] The method 200 begins at operation 202, where a control device 550 receives sensor data 134 from a sensor 546 associated with the autonomous vehicle 502. The control device 550 may receive the sensor data 134 from the sensor 546 continuously, periodically (e.g., every second, every five seconds, or any suitable duration), and / or on demand.
[0129] In operation 204, the control device 550 may detect the presence of a vehicle 108 from the sensor data 134. For example, the control device 550 may implement the object detection machine learning module 112 to detect the vehicle 108 from the sensor data 134. The control device 550 may detect the presence of any of the vehicles 108a, 108b, or 108c illustrated in FIG. 1 . Similarly, the control device 550 may detect the presence of any obstacles / objects 142 and / or non-player characters 144 from the sensor data 134.
[0130] In operation 206, the control device 550 may determine the lateral distance 138 between the autonomous vehicle 502 and the vehicles 108. For example, the control device 550 may determine the lateral distances 138a, 138b, or 138c between the autonomous vehicle 502 and each of the vehicles 108a, 108b, and 108c, respectively, similar to that illustrated in FIG.
[0131] In operation 208, the control device 550 may compare the lateral distance 138 to the threshold distance 132 from the autonomous vehicle 502. The control device 550 may define the threshold distance 132 similar to that described in FIG.
[0132] In operation 210, the control device 550 may determine whether the lateral distance 138 between the autonomous vehicle 502 and the vehicle 108 is less than the threshold distance 132. In this process, the control device 550 may determine whether to command the autonomous vehicle 502 to perform a lane-shifting maneuver 130. If the control device 550 determines that the lateral distance 138 between the autonomous vehicle 502 and the vehicle 108 is less than the threshold distance 132, the method 200 may proceed to operation 212. Otherwise, the method 200 may return to operation 202. In other words, if the control device 550 determines that the lateral distance 138 between the autonomous vehicle 502 and the vehicle 108 is equal to or greater than the threshold distance 132, the control device 550 may determine that the autonomous vehicle 502 is keeping a safe distance from the vehicle 108. Thus, the control device 550 may continue to monitor and evaluate the distance between the autonomous vehicle 502 and the other vehicle 108 (or object 142).
[0133] In operation 212, the control device 550 may determine whether a lane-shifting maneuver 130 is feasible.
[0134] In some embodiments, control device 550 may determine that lane-offset maneuver 130 is feasible if it is determined that lane-offset maneuver 130 can be performed within threshold period 148. In other words, control device 550 may determine that lane-offset maneuver 130 is feasible if, by performing lane-offset maneuver 130, similar to that illustrated in FIG. 1, future lateral distance 138 between autonomous vehicle 502 and vehicle 108 will be at least equal to threshold distance 132 within threshold period 148.
[0135] In some embodiments, the control device 550 may determine whether the lane-leaning maneuver 130 is feasible based on the road structure. For example, if the road 102 has a high curvature (e.g., greater than 60 degrees, 70 degrees, etc.), the control device 550 may determine that the lane-leaning maneuver 130 is not feasible. If the control device 550 determines that the lane-leaning maneuver 130 is feasible, the method 200 may proceed to operation 216. Otherwise, the method 200 may proceed to operation 214.
[0136] In operation 214, the control device 550 may command the autonomous vehicle 502 to perform a minimum risk maneuver 140. For example, the minimum risk maneuver 140 may include slowing or increasing the speed of the autonomous vehicle 502 until the autonomous vehicle 502 and the vehicle 108 are no longer adjacent to each other.
[0137] In operation 216, the control device 550 may determine whether it is safe to perform the lane-shifting maneuver 130.
[0138] In some embodiments, control device 550 may determine whether it is safe to perform lane-shifting maneuver 130 based on historical driving behavior 162 associated with vehicle 108. For example, similar to that illustrated in FIG. 1, if control device 550 determines that historical driving behavior 162 associated with vehicle 108 indicates that the driving behavior of vehicle 108 is highly unpredictable, e.g., if control device 550 determines that driving pattern prediction 164 of vehicle 108 is below a threshold percentage, such as 60%, 55%, etc., control device 550 may determine that it is not safe to perform lane-shifting maneuver 130.
[0139] In another example, the control device 550 may determine whether it is safe to perform the lane-shifting maneuver 130 based on road conditions 154, such as traffic and weather on the road 102. For example, if the control device 550 determines that there is heavy traffic on the road 102 and / or bad weather conditions to the extent that performing the lane-shifting maneuver 130 may cause the risk of a collision with another vehicle 108 to be greater than a threshold percentage (e.g., greater than 10%, 15%, etc.), the control device 550 may determine that it is not safe to perform the lane-shifting maneuver 130. If it is determined that it is not safe to perform the lane-shifting maneuver 130, the method 200 may proceed to operation 214. Otherwise, the method 200 may proceed to operation 218.
[0140] In operation 218, control device 550 may command autonomous vehicle 502 to perform lane-offset maneuver 130. In this process, control device 550 drives autonomous vehicle 502 off-center (with a distance of lane-offset 110) in the opposite direction relative to vehicle 108 in current lane 104a until lateral distance 138 between vehicle 108 and autonomous vehicle 502 is at least equal to threshold distance 132, similar to that described above in FIG.
[0141] (An exemplary system for implementing lane-shifting maneuvers to negotiate curved roads) 3 illustrates an embodiment of a system 300 for implementing a lane-shifting maneuver 130 for an autonomous vehicle 502 to negotiate a curved road 302. FIG. 3 further illustrates a simplified schematic diagram of the curved road 302 being navigated by the autonomous vehicle 502. In some embodiments, the system 300 comprises the autonomous vehicle 502 and its components, including a control device 550 and a sensor 546.
[0142] The control device 550 includes a processor 122 in signal communication with a memory 126. The memory 126 may store software instructions 312 that, when executed by the processor 122, cause the control device 550 to perform one or more functions described herein. For example, the software instructions 312, when executed, cause the processor 122 to implement a lane-shifting maneuver 130 for negotiating a curved road 302. The system 300 may be configured as shown or in any other suitable configuration.
[0143] In general, the system 300 may be configured to implement the lane-leaning maneuver 130 in response to detecting that the autonomous vehicle 502 is approaching a curved road, such as the exemplary illustrated curved road 302, and / or when the trailer 318 of the autonomous vehicle 502 veers out of a straight line or causes a trailer angle 342 between the semi-truck tractor unit 316 and the trailer 318 of the autonomous vehicle 502, for example, due to wind blowing across the autonomous vehicle 502 even on a straight road. The semi-truck tractor unit 316 is synonymously referred to herein as the cab 316 of the autonomous vehicle 502.
[0144] As briefly illustrated in FIG. 1 , in some cases, when an autonomous vehicle 502 is driving on a curved lane 304 on a curved road 302, the trailer 318 of the autonomous vehicle 502 may inadvertently swerve or swing away from the centerline of the curved lane 304.
[0145] In some cases, the distance between the autonomous vehicle 502 and the vehicle on a curved road 302 may be less than the threshold distance 132 due to the trailer 318 of the autonomous vehicle 502 deviating from the centerline of the curved lane 304. Similarly, when wind is traveling across the autonomous vehicle 502 (when the autonomous vehicle 502 is on either a curved or straight road), the wind may cause the trailer 318 of the autonomous vehicle 502 to deviate or swing away from the centerline of the current lane, causing a trailer angle 342 between the trailer 318 of the autonomous vehicle 502 and the cab 316. For example, the distance between the autonomous vehicle 502 and the vehicle on a curved or straight road may be less than the threshold distance 132 due to the trailer 318 of the autonomous vehicle 502 deviating from the centerline of the curved or straight lane.
[0146] These situations can lead to dangerous driving conditions for the autonomous vehicle 502 on curved or straight roads and vehicles on either side of the autonomous vehicle 502. In such cases, the control device 550 may implement lane-offset maneuvers 130 to drive the autonomous vehicle 502 off-center in the curved lane 304 so that the autonomous vehicle 502 does not encroach on the side lane. In other words, the control device 550 may implement lane-offset maneuvers 130 so that neither the cab 316 nor the trailer 318 veering into the side lane.
[0147] To this end, the control device 550 calculates a total lane offset adjustment 320, which is a distance the autonomous vehicle 502 should be driven off-center in the current lane to perform the lane offset maneuver 130. The process of calculating the distance the autonomous vehicle 502 should be driven off-center in the current lane 304 (i.e., the total lane offset adjustment 320) is further described below in conjunction with the operational flow of the system 300.
[0148] Briefly, to calculate the total lane offset adjustment 320, the control device 550 calculates a first lane offset adjustment 330 associated with the road curvature 328, calculates a second lane offset adjustment 340 associated with the trailer angle 342, and combines them together. In calculating the first lane offset adjustment 330, the trailer angle 342 is represented (or assumed) to be zero, and the control device 550 calculates the first lane offset adjustment 330 alone. In calculating the second lane offset adjustment 340, the road 302 is represented (or assumed) to be straight, and the control device 550 calculates the second lane offset adjustment 340 alone. In the present disclosure, the first lane offset adjustment amount 330 may be synonymously referred to herein as a first lane offset adjustment distance amount 330, the second lane offset adjustment amount 340 may be synonymously referred to herein as a second lane offset adjustment distance amount 340, and the total lane offset adjustment amount 320 may be synonymously referred to herein as a total lane offset adjustment distance amount 320.
[0149] In an embodiment, system 100 of FIG. 1 or system 300 of FIG. 3 may perform one or more operations of the operational flow illustrated in FIG. 1, one or more operations of method 200 illustrated in FIG. 2, one or more operations of the operational flow illustrated in FIG. 3, and one or more operations of method 400 illustrated in FIG. 4.
[0150] In an embodiment, the system may include one or more components of system 100 of FIG. 1 and system 300 of FIG. 3 and may be configured to perform one or more operations of the operational flow described in FIG. 1, one or more operations of method 200 described in FIG. 2, one or more operations of operational flow 300 described in FIG. 3, and one or more operations of method 400 described in FIG. 4.
[0151] (System Components) 1 and 2, and additional aspects are described below. The memory 126 may be further configured to store software instructions 312, road radius 322, trailer angle 342, trailer length 326, lane offset maneuver 130, map data 114, sensor data 314, first lane offset adjustment amount 330, second lane offset adjustment amount 340, total lane offset adjustment amount 320, lane offset amount 110, final lane offset amount 350, trailer offset amount 352, and threshold distance 132.
[0152] The corresponding description below describes the operational flow of the system 300 for implementing the lane-shifting maneuver 130 when the autonomous vehicle 502 is approaching a curved road 302.
[0153] (Operation flow for implementing lane-shifting maneuvers to negotiate curved roads) The operational flow of the system 300 begins when the control device 550 may determine that the autonomous vehicle 502 is approaching a curved road 302.
[0154] In some embodiments, the control device 550 may determine that the autonomous vehicle 502 is approaching a curved road 302 based on analyzing the map data 114. The map data 114 is illustrated in FIG. 1 . For example, the map data 114 may include a virtual map of a city in which the autonomous vehicle 502 is driving. The map data 114 may include one or more roads ahead of the autonomous vehicle 502. The control device 550 may implement data processing algorithms, such as image processing algorithms, to analyze the map data 114 to determine the shape of the one or more roads ahead of the autonomous vehicle 502.
[0155] In some embodiments, the control device 550 may determine that the autonomous vehicle 502 is approaching a curved road 302 based on the sensor data 314. In this process, the control device 550 may receive sensor data 314 from the sensors 546 of the autonomous vehicle 502, where the sensor data 314 describes the environment surrounding the autonomous vehicle 502. The sensor data 314 may include data indicative of a set of locations of lane markers 306 on the curved road 302. The control device 550 may implement the object detection machine learning module 166 to process the sensor data 314 and determine the set of locations of the lane markers 306 from the sensor data 314. The control device 550 may determine that the autonomous vehicle 502 is approaching a curved road 302 based on determining that the set of locations of the lane markers 306 follow a curve.
[0156] Because the control device 550 has determined that the autonomous vehicle 502 is approaching the curved road 302, the control device 550 may determine a distance to drive the autonomous vehicle 502 off the centerline of the curved lane 304 to perform the lane offset maneuver 130. To do this, the control device 550 may determine the total lane offset adjustment amount 320.
[0157] To determine the total lane offset adjustment amount 320, the control device 550 combines the first lane offset adjustment amount 330 and the second lane offset adjustment amount 340. The first lane offset adjustment amount 330 may be associated with the road curvature 328. The second lane offset adjustment amount 340 may be associated with the trailer angle 342. The control device 550 may command the autonomous vehicle 502 to perform a lane offset maneuver 130, which includes driving the autonomous vehicle 502 off-center within the curved lane currently being traveled by the autonomous vehicle 502 based on the total lane offset adjustment amount 320.
[0158] The corresponding description below describes calculating the first lane offset adjustment amount 330 and the second lane offset adjustment amount 340.
[0159] (Calculation of the first lane offset adjustment amount associated with the road curvature) In some embodiments, in calculating the first lane offset adjustment 330, the trailer angle 342 is represented to be zero. As can be seen in Figure 3, in calculating the first lane offset adjustment 330, the cab 316 and trailer 318 of the autonomous vehicle 502 are substantially aligned with one another such that the trailer angle 342 between the cab 316 and trailer 318 of the autonomous vehicle 502 is zero.
[0160] To determine the first lane offset adjustment amount 330, the control device 550 may determine a road radius 322 of the curved road 302. In some embodiments, the control device 550 may determine the road radius 322 from the map data 114. In this process, the control device 550 may determine the virtual circle 324 on the map data 114 such that the curved road 302 is part of the circumference of the virtual circle 324. In other words, the centerline of the road 302 may be part of the circumference of the virtual circle 324.
[0161] The control device 550 may determine the road radius 322 of the curved road 302 by calculating the distance between the center of the virtual circle 324 and a point 332 where the cab 316 of the autonomous vehicle 502 meets the trailer 318 of the autonomous vehicle 502. The control device 550 uses the road radius 322 and the trailer length 326 as described below to calculate a first lane offset adjustment amount 330 associated with the road curvature 328. The trailer length 326 may be provided to the control device 550 in software instructions 312.
[0162] In some embodiments, the control device 550 may calculate the first lane offset adjustment amount 330 according to the following equation (1): First lane offset adjustment value = (Road radius 2 -Trailer length 2 ) 1 / 2 -Road radius Equation (1)
[0163] The trailer length in equation (1) is trailer length 326 , which is the length of trailer 318 .
[0164] In some embodiments, the control device 550 may adjust the sign of the first lane offset adjustment amount 330 based on the direction of the road curvature 328 and / or the direction of the first lane offset adjustment amount 330. As can be inferred from equation (1), the first lane offset adjustment amount 330 calculated using equation (1) will always be positive. However, the sign of the first lane offset adjustment amount 330 may need to be adjusted based on the direction of the road curvature 328 and / or the direction of the first lane offset adjustment amount 330.
[0165] 3, if the road curvature 328 is to the left, the sign associated with the road curvature 328 is assumed to be positive, and if the road curvature 328 is to the right, the sign associated with the road curvature 328 is assumed to be negative. Furthermore, if the first lane offset adjustment amount 330 is to the left, the sign of the first lane offset adjustment amount 330 is assumed to be negative, and if the first lane offset adjustment amount 330 is to the right, the sign of the first lane offset adjustment amount 330 is assumed to be positive.
[0166] In this example, the first lane offset adjustment 330 is to the left because the midpoint of the trailer 318 is to the right of the centerline of the lane 304. Thus, the sign of the first lane offset adjustment 330 is negative. Thus, in this example situation where the direction of the road curvature 328 is to the left and the direction of the first lane offset adjustment 330 is to the left, the first lane offset adjustment 330 with the adjusted sign may be calculated according to the following equation (2): First lane offset adjustment value with adjusted sign = ((road radius 2 -Trailer length 2 ) 1 / 2 -road radius)) x sign (-road curvature) Equation (2)
[0167] The sign (road curvature) in equation (2) indicates the sign associated with the road curvature 328.
[0168] (Calculation of second lane offset adjustment amount associated with trailer angle) In some embodiments, in calculating the second lane offset adjustment 340, the road 302 is represented as being a straight line. As can be seen in Figure 3, in calculating the second lane offset adjustment 340, the autonomous vehicle 502 is assumed to be on a straight line 348. The straight line 348 may be the centerline of the road, which is assumed to be a straight line.
[0169] To calculate the second lane offset adjustment amount 340, the control device 550 may determine a trailer angle 342 between the trailer 318 and the cab 316. The control device 550 may determine the trailer angle 342 from sensor data 314 received from a sensor 546, which may be configured to measure the trailer angle 342 between the trailer 318 and the cab 316 by measuring mechanical rotations and converting them to scaled electrical signals. The control device 550 uses the trailer angle 342 and the trailer length 326 to calculate the second lane offset adjustment amount 340 associated with the trailer angle 342, as described below.
[0170] In some embodiments, the control device 550 may calculate the second lane offset adjustment amount 340 according to the following equation (3): Second lane offset adjustment value = trailer length × sin(trailer angle) Equation (3)
[0171] In some embodiments, the control device 550 may adjust the sign of the second lane offset adjustment amount 340 based on the direction of the trailer angle 342 and / or the direction of the second lane offset adjustment amount 340 .
[0172] The sign of the trailer angle 342 may depend on the direction that the trailer 318 deviates from the straight line 348. For example, if the trailer 318 is deflected to the left, the sign associated with the trailer angle 342 is negative, and if the trailer 318 is deflected to the right, the sign associated with the trailer angle 342 is positive. Further, if the second lane offset adjustment amount 340 is to the left, the sign of the second lane offset adjustment amount 340 is negative, and if the second lane offset adjustment amount 340 is to the right, the sign of the second lane offset adjustment amount 340 is positive.
[0173] 3, the trailer 318 is veered to the left, so the direction of the second lane offset adjustment amount 340 is to the right. Thus, a rightward direction of the second lane offset adjustment amount 340 means that the sign of the second lane offset adjustment amount 340 is positive.
[0174] Thus, in this example where the direction of the trailer angle 342 is to the left and the direction of the second lane offset adjustment amount 340 is to the right, the control device 550 may calculate the second lane offset adjustment amount 340 with an adjusted sign according to the following equation (4): Second lane offset adjustment value with assigned code = - trailer length × sin(trailer angle) Equation (4)
[0175] As explained above, each of the first lane offset adjustment amount 330 and the second lane offset adjustment amount 340 may be toward the left or right and is associated with a sign.
[0176] If the lane offset direction is to the left, the sign for the lane offset adjustment amount is negative, and if the lane offset direction is to the right, the sign for the lane offset adjustment amount is positive. The direction and amount of the total lane offset adjustment amount 320 may depend on the sign and amount of each of the first lane offset adjustment amount 330 and the second lane offset adjustment amount 340.
[0177] In other words, the direction and amount of the total lane offset adjustment 320 may depend on whether the first lane offset adjustment 330 or the second lane offset adjustment 340 has more effect on the autonomous vehicle 502. For example, if the first lane offset adjustment 330 has a negative sign (i.e., heading left) and the second lane offset adjustment 340 has a positive sign (i.e., heading right), and the first lane offset adjustment 330 is greater than the second lane offset adjustment 340, then the total lane offset adjustment 320 will be negative (i.e., heading left).
[0178] (Implementation of lane deviation maneuver) The control device 550 calculates the total lane offset adjustment amount 320 by combining the first lane offset adjustment amount 330 and the second lane offset adjustment amount 340 .
[0179] Once the total lane offset adjustment 320 is calculated, the control device 550 may use the total lane offset adjustment 320 to command the autonomous vehicle 502 to perform a lane offset maneuver 130. In this maneuver, the control device 550 drives the autonomous vehicle 502 off the centerline of the turned lane 304 based on the calculated total lane offset adjustment 320 so that the autonomous vehicle 502 does not encroach on a side lane.
[0180] If the autonomous vehicle 502 encounters a vehicle 310 along a curved road 302, the control device 550 may consider keeping a safe distance from the vehicle 310 in addition to considering the lane offset adjustment amount 330 and / or 340 associated with the road curvature 328 and / or trailer angle 342.
[0181] To this end, the control device 550 may determine a lane offset amount 110 to maintain a safe distance from the vehicle 310 (i.e., until the distance between the autonomous vehicle 502 and the vehicle 310 is at least equal to the threshold distance 132), similar to that described in Figures 1 and 2. The control device 550 may also determine a total lane offset adjustment amount 320, similar to that described above in Figure 3. The control device 550 may combine or add the total lane offset adjustment amount 320 with the lane offset amount 110 described in Figures 1 and 2 to determine a final lane offset amount 350 and drive the autonomous vehicle 502 off the centerline of the lane 304.
[0182] If there are no vehicles 310 on the curved road 302 , the total lane offset adjustment 320 may be equal to the final lane offset 350 .
[0183] An example scenario in which an autonomous vehicle 502 encounters a vehicle 310 on a curved road 302 is described below.
[0184] (Encountering a vehicle on a curved road) In an example scenario, assume that an autonomous vehicle 502 encounters a vehicle 310 while traveling on a curved road 302. In such a case, the control device 550 may calculate a lane deviation 110 to keep at least a threshold distance 132 from the vehicle 310, similar to that described in FIGS.
[0185] The control device 550 may also calculate the total lane offset adjustment amount 320 by calculating and combining the first lane offset adjustment amount 330 and the second lane offset adjustment amount 340, similar to that described above. The control device 550 may determine the final lane offset amount 350 by combining the lane offset amount 110 and the total lane offset adjustment amount 320.
[0186] In the example of FIG. 3, if the autonomous vehicle 502 encounters the outside vehicle 310a on the curved road 302, the control device 550 may combine the lane offset amount 110 with the total lane offset adjustment amount 320 to calculate a final lane offset amount 350.
[0187] If the autonomous vehicle 502 encounters vehicle 310b on the inside of the curved road 302, the control device 550 may decide not to combine the lane offset amount 110 into the total lane offset adjustment amount 320 and use only the total lane offset adjustment amount 320 to drive the autonomous vehicle 502 off-center from the lane 304. One reason for deciding not to combine the lane offset amount 110 into the total lane offset adjustment amount 320 is to not reduce the distance between the autonomous vehicle 502 and vehicle 310b and therefore not reduce the final lane offset amount 350.
[0188] (Performing lane-shifting maneuvers on a straight road) In some embodiments, the control device 550 may determine whether to command the autonomous vehicle 502 to perform a lane-offset maneuver 130 while traveling on a straight road. The control device 550 may determine that the autonomous vehicle 502 is traveling along a straight road by analyzing the map data 114 and / or the sensor data 314, similar to that described above with respect to determining that the autonomous vehicle 502 is approaching a curved road 302. Thus, the control device 550 may determine that the first lane-offset adjustment amount 330 is zero because the road radius 322 is sufficiently large, such as greater than a threshold distance, e.g., 1,000 meters.
[0189] In some cases, the wind may push the trailer 318 of the autonomous vehicle 502 away from the centerline of a straight road. In such a case, the control device 550 may detect that the wind is causing the trailer 318 of the autonomous vehicle 502 to deviate from a straight line. The control device 550 may determine that the wind is causing the trailer 318 of the autonomous vehicle 502 to deviate from a straight line based on the sensor data 314 received from the sensor 546 indicating that the trailer angle 342 is greater than zero. In response, the control device 550 may determine the trailer angle 342 associated with the wind from the sensor data 314 received from the sensor 546, which may be configured to measure the trailer angle 342.
[0190] The control device 550 calculates the wind induced second lane offset adjustment amount 340 in a manner similar to that described above using equations (3) and (4). Because the first lane offset adjustment amount 330 is determined to be zero, the control device 550 may determine that the total lane offset adjustment amount 320 is equal to the second lane offset adjustment amount 340.
[0191] (Calculating Total Lane Offset Adjustment Using Trailer Offset) In another embodiment, the control device 550 may calculate the final lane deviation 350 as described below.
[0192] In this embodiment, the control device 550 may calculate the trailer offset 352 according to the following equation (5): Trailer offset = (Road radius 2 + Trailer length 2 -2 x road radius x trailer length x cos(π / 2-trailer angle) 1 / 2 -Road radius Equation (5)
[0193] The calculated trailer offset 352 can be positive or negative. When the trailer offset 352 is positive, it means the trailer 318 is offset or biased toward the outside of the curved road 302. When the trailer offset 352 is negative, it means the trailer 318 is offset or biased toward the inside of the curved road 302. For each of the trailer angle 342, road curvature 328, and final lane offset 350, a positive sign means leftward and a negative sign means rightward.
[0194] Therefore, the direction or sign of trailer offset 352 may need to be adjusted to match the correct lane offset direction. Thus, the control device 550 may calculate the final lane offset 350 by combining lane offset 110 with the trailer offset 352 with the adjusted sign.
[0195] The process of adjusting the sign of trailer offset 352 is described below. In some cases, the autonomous vehicle 502 may encounter a vehicle 310 on either side and on a curved road 302, where the curve of the curved road 302 may be toward the left or right. For example, the autonomous vehicle 502 may encounter: 1) a road with a left curve, the vehicle 310 may be on the right side of the autonomous vehicle 502; 2) a road with a left curve, the vehicle 310 may be on the left side of the autonomous vehicle 502; 3) a road with a right curve, the vehicle 310 may be on the left side of the autonomous vehicle 502; 4) a road with a right curve, the vehicle 310 may be on the right side of the autonomous vehicle 502.
[0196] The control device 550 may determine the direction and sign of each of the trailer angle 342, the curve, the trailer offset 352, and the final lane offset 350, as illustrated in Table 1 below. [Table 1]
[0197] As can be seen from Table 1, for a left curve, the direction of trailer offset 352 has an opposite sign compared to the direction of lane offset 110, while for a right curve, the direction of trailer offset 352 has the same sign as the direction of lane offset 110. Thus, the control device 550 may reverse the sign of trailer offset 352 when it is proceeding through a left curve. This can be done by multiplying the sign of trailer angle 342 by trailer offset 352.
[0198] In some cases, trailer angle 342 may be too small, e.g., less than a threshold degree, e.g., less than 5 degrees, 4 degrees, etc. In such cases, the sign of road curvature 328 (opposite to trailer angle 342) may be used to adjust the sign of trailer offset 352.
[0199] In another use case, assume that the autonomous vehicle 502 is traveling along a straight road and the trailer 318 is or is forced to be offset from a straight line due to road bank (roll angle) or wind. In such a case, the road radius 322 is very large (e.g., above a threshold amount), so the trailer offset amount 352 will always be negative (inside compared to the large circle associated with the large road radius 322).
[0200] In such a case, the control device 550 may determine the direction and sign of each of the trailer angle 342, the curve, the trailer offset 352, and the final lane offset 350, as illustrated in Table 2 below. [Table 2]
[0201] As can be seen from Table 2, road curvature 328 is not usable since the road is straight. As can be further seen from Table 2, if trailer 318 is deflected to the left, trailer offset 352 has an opposite sign compared to final lane offset 350, while if trailer 318 is deflected to the right, trailer offset 352 has the same sign as final lane offset 350. In this case, control device 550 may multiply the sign of trailer angle 342 by trailer offset 352 to adjust the sign of trailer offset 352.
[0202] In an example scenario, assume that the road radius 322 (or turning radius) is 620 meters (m) and the average trailer angle 342 is 0.01 radians (rad) with a mean deviation or spike of up to 0.02 rad. In this example, the average trailer offset 352 may be 0.0m with a mean deviation or spike of up to -0.14m. In this example, negative values mean on the inside. Thus, the trailer offset 352 is halfway between the lanes with some variation or deviation to the inside.
[0203] In another example scenario, assume the road radius 322 (or turning radius) is 502 m and the average trailer angle 342 is 0.018 rad with an average deviation or spike of up to -0.3 rad. The deviation or spike may be caused by wind, control adjustments to the autonomous vehicle 502, and / or road bumps, but the trailer offset 352 is toward the inside.
[0204] The difference between offsetting towards the inside and outside of the curved road 302 may be explained by different roll angles: the inside of the road has a larger roll angle, and a larger roll angle may lead to the trailer 318 offsetting towards the inside.
[0205] In some cases, trailer angle 342 may be too small to be measured due to the shape of road curvature 328, such as a sharp turn. In such a case, road radius 322 may be at least 1,000 m. Thus, trailer offset 352 may be less than 0.15 m. Here, trailer offset 352 is outward (according to ignoring very small trailer angles 342). One possible reason is that such a curved road may be flat, and the roll angle is much smaller than the angle of the turn.
[0206] In some embodiments, trailer offset 352 may be affected by the speed of the autonomous vehicle 502, road curvature 328 (or turning radius), roll angle, and / or wind moving across the autonomous vehicle 502. To reduce the complexity of calculating trailer offset 352, trailer angle 342 may be used. If the turning radius is greater than 1,000 m, trailer offset 352 may be calculated to be less than 0.15 m. This amount of offset is less than the lateral error range preconfigured in the control device 550. Therefore, it may not be noted by the control device 550.
[0207] In some embodiments, the control device 550 may be configured to reduce signal noise and inherent control adjustments and driving behaviors in calculating the total lane offset adjustment amount 320 and / or the final lane offset amount 350. In this operation, the control device 550 may implement a smoothing filter to smooth the command signal indicating to perform the lane offset maneuver 130, thereby reducing signal noise and inherent control adjustments and driving behaviors. In some embodiments, the control device 550 may use a digital filter, such as a moving average filter, a finite impulse response filter, an infinite impulse response filter, etc., to reduce noise or deviations in calculating the trailer angle 342.
[0208] In one embodiment, control device 550 may perform similar operations with respect to determining the lateral distance (138a-c in FIG. 1) between autonomous vehicle 502 and vehicle 310a and / or vehicle 310b, compare the lateral distance to a threshold distance (132 in FIG. 1), and determine whether the lateral distance is less than the threshold distance. If the lateral distance is determined to be less than the threshold distance, control device 550 may command the autonomous vehicle to perform a lane-shifting maneuver 130.
[0209] 1 and 2 when encountering a straight road 102, and vice versa. For example, control device 550 may adjust one or more actions for navigating autonomous vehicle 502 described herein according to the shape of the road to achieve a more optimal and safer navigation path for autonomous vehicle 502, surrounding vehicles, and pedestrians.
[0210] In one embodiment, control device 550 may perform similar operations with respect to determining the lateral distance (138a-c in FIG. 1) between autonomous vehicle 502 and vehicle 310a and / or vehicle 310b, compare the lateral distance to a threshold distance (132 in FIG. 1), and determine whether the lateral distance is less than the threshold distance. If the lateral distance is determined to be less than the threshold distance, control device 550 may command the autonomous vehicle to perform a lane-shifting maneuver 130.
[0211] 1 and 2 when encountering a straight road 102, and vice versa. For example, control device 550 may adjust one or more actions for navigating autonomous vehicle 502 described herein according to the shape of the road to achieve a more optimal and safer navigation path for autonomous vehicle 502, surrounding vehicles, and pedestrians.
[0212] The embodiments, examples, and operations described in this disclosure are not mutually exclusive: in an embodiment, any and all combinations of the embodiments, examples, and operations may be implemented in conjunction with one another as the circumstances encountered by the autonomous vehicle 502 require.
[0213] (An exemplary method for implementing lane-shifting maneuvers to negotiate curved roads) FIG. 4 illustrates an example flow chart of a method 400 implementing a lane-shifting maneuver 130 for negotiating a curved road. Modifications, additions, or omissions may be made to the method 400. The method 400 may include more, fewer, or other operations. For example, the operations may be performed in parallel or in any suitable order. At times, the autonomous vehicle 502, the control device 550, or any component thereof is discussed as performing operations, but any suitable system or component of a system may perform one or more operations of the method 400. For example, one or more operations of the method 400 may be implemented in the form of software instructions 312 and processing instructions 580, respectively, from FIGS. 3 and 5, stored on a non-transitory tangible machine-readable medium (e.g., memory 126 and data storage 590, respectively, from FIGS. 3 and 5) that, when initiated at least in part by one or more processors (e.g., processors 122 and 570, respectively, from FIGS. 3 and 5), may cause the one or more processors to perform operations 402-418.
[0214] Method 400 begins at operation 402, where control device 550 may determine whether autonomous vehicle 502 is approaching a curved road 302. In this process, control device 550 may determine whether autonomous vehicle 502 is approaching a curved road 302 based on analyzing map data 114 and / or sensor data 314, similar to that described in FIG. 3. If control device 550 determines that autonomous vehicle 502 is approaching a curved road 302, method 400 may proceed to operation 404. Otherwise, method 400 may proceed to operation 416.
[0215] In operation 404, the control device 550 may determine a road radius 322 of the curved road 302, similar to that illustrated in FIG.
[0216] In operation 406, the control device 550 may calculate a first lane offset adjustment amount 330 associated with the road curvature 328 of the curved road 302 based on the road radius 322. In this process, the control device 550 may calculate the first lane offset adjustment amount 330 according to equations (1) and (2) illustrated in FIG.
[0217] In operation 408, the control device 550 may determine the trailer angle 342 between the trailer 318 and the cab 316 of the autonomous vehicle 502, similar to that illustrated in FIG. 3.
[0218] In operation 410, the control device 550 may calculate, based on the trailer angle 342, a second lane offset adjustment amount 340 associated with the trailer angle 342. In this process, the control device 550 may calculate the second lane offset adjustment amount 340 according to equations (3) and (4), similar to those described in FIG.
[0219] In operation 412 , the control device 550 may calculate the total lane offset adjustment amount 320 by combining the first and second lane offset adjustment amounts 330 and 340 .
[0220] In operation 414, the control device 550 may combine the total lane offset adjustment value 320 with the original lane offset amount 110. The control device 550 may command the autonomous vehicle 502 to perform a lane offset maneuver 130 based on the total lane offset adjustment amount 320 and the original lane offset amount 110. The lane offset maneuver 130 includes driving the autonomous vehicle 502 off-center within the curved lane 304 currently being traveled by the autonomous vehicle 502 based on the total lane offset adjustment amount 320 and the original lane offset amount 110.
[0221] If the autonomous vehicle 502 is on a straight road, the control device 550 may perform operations 416-420, described below.
[0222] In operation 416, the control device 550 may determine whether the trailer angle 342 between the cab 316 and the trailer 318 of the autonomous vehicle 502 is greater than zero. For example, the control device 550 may determine whether the trailer angle 342 is greater than zero if the sensor data 314 comprises data indicating that the trailer angle 342 is greater than zero. In this process, the control device 550 may determine whether a wind moving across the autonomous vehicle 502 is pushing the trailer 318 of the autonomous vehicle left or right, similar to that described in FIG. 3, causing the trailer angle 342 between the cab 316 and the trailer 318 of the autonomous vehicle 502 to be greater than zero. If the control device 550 determines that the trailer angle 342 is greater than zero, the method 400 may proceed to operation 418. Otherwise, the method 400 may proceed to operation 420.
[0223] In operation 418, the control device 550 may determine that the first lane offset adjustment value 330 is zero. In response, the method 200 may proceed to 414.
[0224] In operation 420, control device 550 may determine that total lane offset adjustment value 320 is zero. In one embodiment, control device 750 may determine not to command autonomous vehicle 502 to perform lane offset maneuver 130 if lane offset amount 110 and total lane offset adjustment value 320 are zero.
[0225] Exemplary Autonomous Vehicles and Their Operation FIG. 5 illustrates a block diagram of an example system 500 in which autonomous driving operations may be implemented. As illustrated in FIG. 5, the autonomous vehicle 502 may be a semi-trailer truck. The system 500 may include several subsystems and components that may generate and / or deliver one or more sources of information / data and related services to an on-board control computer 550 that may be located within the autonomous vehicle 502. The on-board control computer 550 may be in data communication with multiple vehicle subsystems 540, all of which may reside within the autonomous vehicle 502. A vehicle subsystem interface 560 may be provided to facilitate data communication between the on-board control computer 550 and the multiple vehicle subsystems 540. In some embodiments, the vehicle subsystem interface 560 may include a controller area network (CAN) controller for communicating with devices within the vehicle subsystem 540.
[0226] Autonomous vehicle 502 may include various vehicle subsystems that support operation of autonomous vehicle 502. Vehicle subsystems 540 may include a vehicle driving subsystem 542, a vehicle sensor subsystem 544, a vehicle control subsystem 548, and / or a network communication subsystem 592. The components or devices of vehicle driving subsystem 542, vehicle sensor subsystem 544, and vehicle control subsystem 548 shown in FIG. 5 are examples. Autonomous vehicle 502 may be configured as shown or in any other configuration.
[0227] Vehicle driving subsystem 542 may include components operable to provide powered motion for autonomous vehicle 502. In an exemplary embodiment, vehicle driving subsystem 542 may include engine / motor 542a, wheels / tires 542b, transmission 542c, electrical subsystem 542d, and power source 542e.
[0228] The vehicle sensor subsystem 544 may include a number of sensors 546 configured to sense information about the environment or conditions of the autonomous vehicle 502. The vehicle sensor subsystem 544 may include one or more cameras 546a or image capture devices, a radar unit 546b, one or more temperature sensors 546c, a wireless communication unit 546d (e.g., a cellular communication transceiver), an inertial measurement unit (IMU) 546e, a laser range finder / LiDAR unit 546f, a global positioning system (GPS) transceiver 546g, and / or a wiper control system 546h. The vehicle sensor subsystem 544 may also include sensors configured to monitor internal systems of the autonomous vehicle 502 (e.g., an O2 monitor, a fuel gauge, engine oil temperature, etc.).
[0229] The IMU 546e may include any combination of sensors (e.g., accelerometers and gyroscopes) configured to sense the position and orientation changes of the autonomous vehicle 502 based on inertial acceleration. The GPS transceiver 546g may be any sensor configured to estimate the geographic location of the autonomous vehicle 502. To this end, the GPS transceiver 546g may include a receiver / transmitter operable to provide information regarding the position of the autonomous vehicle 502 relative to the Earth. The radar unit 546b may represent a system that utilizes radio signals and senses objects within the local environment of the autonomous vehicle 502. In some embodiments, in addition to sensing objects, the radar unit 546b may be further configured to sense the speed and heading of objects proximate to the autonomous vehicle 502. The laser range finder or LiDAR unit 546f may be any sensor configured to use lasers and sense objects within the environment in which the autonomous vehicle 502 is located. The camera 546a may include one or more devices configured to capture multiple images of the environment of the autonomous vehicle 502. The camera 546a may be a still image camera or a motion video camera.
[0230] The vehicle control subsystem 548 may be configured to control the operation of the autonomous vehicle 502 and its components. Thus, the vehicle control subsystem 548 may include various elements such as a throttle and gear selector 548a, a brake unit 548b, a navigation unit 548c, a steering system 548d, and / or an autonomous control unit 548e. The throttle and gear selector 548a may be configured to control, for example, the operating speed of the engine, which in turn controls the speed of the autonomous vehicle 502. The throttle and gear selector 548a may be configured to control gear selection of the transmission. The brake unit 548b may include any combination of mechanisms configured to slow the autonomous vehicle 502. The brake unit 548b may slow the autonomous vehicle 502 in a standard manner, including using friction or engine braking to slow the wheels. The brake unit 548b may include an anti-lock braking system (ABS) that may prevent the brakes from locking when the brakes are being applied. The navigation unit 548c may be any system configured to determine a driving path or route for the autonomous vehicle 502. The navigation unit 548c may be further configured to dynamically update the driving path while the autonomous vehicle 502 is operating. In some embodiments, the navigation unit 548c may be configured to incorporate data from the GPS transceiver 546g and one or more predefined maps to determine a driving path for the autonomous vehicle 502. The steering system 548d may represent any combination of mechanisms that may be operable to adjust the heading of the autonomous vehicle 502 in an autonomous mode or a driver-controlled mode.
[0231] The autonomous control unit 548e may represent a control system configured to identify, evaluate, and avoid or otherwise negotiate potential obstacles or obstructions in the environment of the autonomous vehicle 502. In general, the autonomous control unit 548e may be configured to control the autonomous vehicle 502 for driverless operation or to provide driver assistance in controlling the autonomous vehicle 502. In some embodiments, the autonomous control unit 548e may be configured to incorporate data from the GPS transceiver 546g, the radar unit 546b, the LiDAR unit 546f, the camera 546a, and / or other vehicle subsystems to determine a driving path or trajectory for the autonomous vehicle 502.
[0232] Network communications subsystem 592 may include network interfaces such as routers, switches, modems, etc. Network communications subsystem 592 may be configured to establish communications between autonomous vehicle 502 and other systems, including a supervision server that may be configured to oversee operation of autonomous vehicle 502. Network communications subsystem 592 may be further configured to transmit data to and receive data from the other systems.
[0233] Many or all of the functions of the autonomous vehicle 502 can be controlled by an on-board control computer 550. The on-board control computer 550 may include at least one data processor 570 (which may include at least one microprocessor) that executes processing instructions 580 stored in a data storage device 590 or a non-transitory computer-readable medium such as a memory. The on-board control computer 550 may also represent multiple computing devices that may be responsible for controlling individual components or subsystems of the autonomous vehicle 502 in a distributed manner. In some embodiments, the data storage device 590 may include processing instructions 580 (e.g., program logic) executable by the data processor 570 to implement various methods and / or functions of the autonomous vehicle 502, including those described with respect to FIGS. 1-7.
[0234] Data storage device 590 may similarly include additional instructions including instructions for transmitting data to, receiving data from, interacting with, or controlling one or more of vehicle driving subsystem 542, vehicle sensor subsystem 544, and vehicle control subsystem 548. On-board control computer 550 may be configured to include a data processor 570 and data storage device 590. On-board control computer 550 may control functions of autonomous vehicle 502 based on inputs received from various vehicle subsystems (e.g., vehicle driving subsystem 542, vehicle sensor subsystem 544, and vehicle control subsystem 548).
[0235] FIG. 6 illustrates a system 600 for providing precise autonomous driving operations. The system 600 may include several modules that may operate within an on-board control computer 550, as described in FIG. 5. The on-board control computer 550 may include a sensor fusion module 602, shown in the upper left corner of FIG. 6, which may perform at least four image or signal processing operations. The sensor fusion module 602 may acquire images from a camera disposed on the autonomous vehicle, perform image segmentation 604, and detect the presence of moving objects (e.g., other vehicles, pedestrians, etc.) and / or static obstacles (e.g., stop signs, speed bumps, terrain, etc.) located around the autonomous vehicle. The sensor fusion module 602 may acquire LiDAR point cloud data items from a LiDAR sensor disposed on the autonomous vehicle, perform LiDAR segmentation 606, and detect the presence of objects and / or obstacles located around the autonomous vehicle.
[0236] The sensor fusion module 602 may perform instance segmentation 608 on the images and / or point cloud data items to identify contours (e.g., boxes) around objects and / or obstacles located around the autonomous vehicle. The sensor fusion module 602 may perform temporal fusion 610, where objects and / or obstacles from one image and / or frame of a point cloud data item are correlated or associated with objects and / or obstacles from one or more images or frames received subsequently in time.
[0237] The sensor fusion module 602 may fuse objects and / or obstacles from images acquired from the cameras and / or point cloud data items acquired from the LiDAR sensor. For example, the sensor fusion module 602 may determine, based on the locations of the two cameras, that an image from one of the cameras with a half of the vehicle located in front of the autonomous vehicle is the same as the vehicle captured by another camera. The sensor fusion module 602 may send the fused object information to the interference module 646 and send the fused obstacle information to the occupancy grid module 660. The on-board control computer may include an occupancy grid module 660, which may retrieve landmarks from a map database 658 stored in the on-board control computer. The occupancy grid module 660 may determine a drivable area and / or obstacles from the fused obstacles acquired from the sensor fusion module 602 and the landmarks stored in the map database 658. For example, the occupancy grid module 660 may determine that the drivable area may include a speed bump obstacle.
[0238] Under the sensor fusion module 602, the on-board control computer 550 may include a LiDAR-based object detection module 612, which may perform object detection 616 based on point cloud data items acquired from a LiDAR sensor 614 disposed on the autonomous vehicle. The object detection 616 technique may provide a location (e.g., in 3D world coordinates) of an object from the point cloud data items. Under the LiDAR-based object detection module 612, the on-board control computer may include an image-based object detection module 618, which may perform object detection 624 based on images acquired from a camera 620 disposed on the autonomous vehicle. The object detection 618 technique may employ deep machine learning techniques 624 to provide a location (e.g., in 3D world coordinates) of an object from images provided by the camera 620.
[0239] A radar 656 on the autonomous vehicle may scan the area in front of the autonomous vehicle or the area towards which the autonomous vehicle is driven. The radar data is transmitted to the sensor fusion module 602, which may use the radar data to correlate objects and / or obstacles detected by the radar 656 with objects and / or obstacles detected from both the LiDAR point cloud data items and the camera images. The radar data may also be transmitted to the interference module 646, which may perform data processing on the radar data and track objects via the object tracking module 648, as described further below.
[0240] The on-board control computer may include an interference module 646 that receives locations of objects from the point cloud and from the images and fused objects from the sensor fusion module 602. The interference module 646 also receives radar data that can be used by an object tracking module 648 to track objects from one point cloud data item and one image acquired at one time instance to another (or next) point cloud data item and another image acquired at another subsequent time instance.
[0241] The interference module 646 may perform object attribute estimation 650 to estimate one or more attributes of an object detected in an image or point cloud data item. The one or more attributes of the object may include a type of object (e.g., a pedestrian, a car, or a truck, etc.). The interference module 646 may perform behavior prediction 652 to estimate or predict a movement pattern of an object detected in the image and / or point cloud. The behavior prediction 652 may be performed to detect a location of an object in a set of images received at different times (e.g., sequential images) or a set of point cloud data items received at different times (e.g., sequential point cloud data items). In some embodiments, the behavior prediction 652 may be performed for each image received from a camera and / or each point cloud data item received from a LiDAR sensor. In some embodiments, the interference module 646 can be implemented (e.g., activated or executed) to reduce computational load by performing behavior prediction 652 on every other or a predetermined number of images received from a camera or point cloud data items received from a LiDAR sensor (e.g., every second image or every third point cloud data item).
[0242] The behavior prediction 652 feature may determine from the radar data the speed and direction of the object surrounding the autonomous vehicle, and the speed and direction information can be used to predict or determine the object's motion pattern. The motion pattern may comprise the object's predicted trajectory information for a predetermined length of time into the future after the image is received from the camera. Based on the predicted motion pattern, the interference module 646 may assign a motion pattern situation tag to the object (e.g., "located at coordinate (x,y)", "stopped", "driving at 50 mph", "accelerating", or "decelerating"). The situation tag may describe the object's motion pattern. The interference module 646 may send one or more object attributes (e.g., type of object) and the motion pattern situation tag to the planning module 662. The interference module 646 may perform the environment analysis 654 using any information obtained by the system 600 and any number or combination of its components and components.
[0243] The on-board control computer may include a planning module 662 that receives object attributes and motion pattern situation tags, drivable areas and / or obstacles from the interference module 646, and vehicle location and attitude information from the fused localization module 626 (described further below).
[0244] The planning module 662 may perform a navigation plan 664 to determine a set of trajectories on which the autonomous vehicle may be driven. The set of trajectories may be determined based on the drivable area information, one or more object attributes of the object, the object's motion pattern situation tag, the location of the obstacle, and the drivable area information. In some embodiments, the navigation plan 664 may include determining an area adjacent to a road where the autonomous vehicle may be safely parked in case of an emergency. The planning module 662 may include a behavior decision-making 666 that determines a driving action (e.g., steering, braking, throttling) in response to determining a changing condition on the road (e.g., a traffic light has turned yellow or the autonomous vehicle is in an unsafe driving condition because another vehicle has encroached in front of the autonomous vehicle and entered an area within a predetermined safe distance of the location of the autonomous vehicle). The planning module 662 may perform a trajectory generation 668 to select a trajectory from the set of trajectories determined by the navigation planning operation 664. The selected trajectory information may be transmitted by the planning module 662 to the control module 670.
[0245] The on-board control computer may include a control module 670, which receives the proposed trajectory from the planning module 662 and the autonomous vehicle location and attitude from the fusion localization module 662. The control module 670 may include a system identifier 672. The control module 670 may perform model-based trajectory refinement 674 to refine the proposed trajectory. For example, the control module 670 may apply filtering (e.g., a Kalman filter) to smooth the proposed trajectory data and / or minimize noise. The control module 670 may perform robust control 676 by determining the amount of brake pressure to apply, the steering angle, the amount of throttle to control the vehicle's speed, and / or transmission gears based on the refined proposed trajectory information and the current location and / or attitude of the autonomous vehicle. The control module 670 can transmit the determined brake pressure, steering angle, throttle amount, and / or transmission gears to one or more devices within the autonomous vehicle to control and facilitate precise driving behavior of the autonomous vehicle.
[0246] Deep image-based object detection 624 performed by image-based object detection module 618 can also be used to detect landmarks on the road (e.g., stop signs, speed bumps, etc.). The on-board control computer can include a fusion localization module 626 that obtains landmarks detected from the images, landmarks obtained from a map database 636 stored on the on-board control computer, landmarks detected from point cloud data items by the LiDAR-based object detection module 612, speed and displacement from an odometer sensor 644, and an estimated location of the autonomous vehicle from GPS / IMU sensors 638 (i.e., GPS sensor 640 and IMU sensor 642) located on or in the autonomous vehicle. Based on this information, the fusion localization module 626 can perform a localization operation 628 and determine the location of the autonomous vehicle, which can be transmitted to the planning module 662 and the control module 670.
[0247] The fusion localization module 626 can estimate the attitude 630 of the autonomous vehicle based on the GPS and / or IMU sensors 638. The attitude of the autonomous vehicle can be transmitted to the planning module 662 and the control module 670. The fusion localization module 626 can also estimate the status (e.g., location, possible movement angles) of the trailer unit based on, for example, information provided by the IMU sensors 642 (e.g., angular velocity and / or linear velocity) (e.g., trailer status estimation 634). The fusion localization module 626 may further check the map content 632.
[0248] FIG. 7 illustrates an example block diagram of an on-board control computer 550 included within an autonomous vehicle 502. The on-board control computer 550 may include at least one processor 704 and a memory 702 having stored instructions (e.g., software instructions 128, 312 and processing instructions 580 of FIGS. 1, 3, and 5, respectively). The instructions, when executed by the processor 704, configure the on-board control computer 550 and / or various modules of the on-board control computer 550 to perform the operations described in FIGS. 1-7. The transmitter 706 may transmit or send information or data to one or more devices within the autonomous vehicle. For example, the transmitter 706 may transmit instructions for one or more motors of a steering wheel to steer the autonomous vehicle. The receiver 708 may receive information or data transmitted or sent by one or more devices. For example, the receiver 708 may receive a status of the current speed from an odometer sensor or the current transmission gear from a transmission. The transmitter 706 and receiver 708 may also be configured to communicate with multiple vehicle subsystems 540 and on-board control computer 550 described above in FIGS.
[0249] Although several embodiments are provided in this disclosure, it should be understood that the disclosed systems and methods may be embodied in many other specific forms without departing from the spirit or scope of the disclosure. The examples are to be considered illustrative rather than restrictive, and the intention is not to be limited to the details given herein. For example, various elements or components may be combined or integrated into another system, or certain features may be omitted or not implemented.
[0250] In addition, the techniques, systems, subsystems, and methods described and illustrated in various embodiments, either separately or separately, may be combined or integrated with other systems, modules, techniques, or methods without departing from the scope of the present disclosure. Other items shown or discussed as coupled or directly coupled or in communication with one another may be indirectly coupled or in communication through some interface, device, or intermediate component, whether electrically, mechanically, or otherwise. Other examples of changes, substitutions, and alterations are ascertainable by those skilled in the art and may be made without departing from the spirit and scope disclosed herein.
[0251] In order to assist the Patent Office and any reader of any patent issued on this application in interpreting the claims appended hereto, Applicant emphasizes that Applicant does not intend that any of the appended claims invoke 35 U.S.C. § 112(f), as they exist as of the filing date of this application, unless the words "means for" or "step for" are expressly used in a particular claim.
[0252] Implementations of the present disclosure can be described in terms of the following notes, the features of which can be combined in any reasonable manner.
[0253] Clause 1. A system comprising a control device associated with an autonomous vehicle, the control device comprising: a memory configured to store sensor data associated with one or more objects on a roadway, the sensor data being captured by at least one sensor associated with the autonomous vehicle; and at least one processor operatively coupled to the memory; and wherein at least one processor includes at least Detecting the presence of a vehicle from sensor data; Determining a lateral distance between the autonomous vehicle and the vehicle; Comparing the lateral distance between the autonomous vehicle and the vehicle to a threshold distance from the autonomous vehicle; determining whether to command the autonomous vehicle to perform a lane-shifting maneuver based at least in part on a comparison between the lateral distance and a threshold distance; It is configured as follows: The lane-offset maneuver includes driving the autonomous vehicle off-center within a current lane traveled by the autonomous vehicle in an opposite direction relative to the vehicle until a lateral distance between the autonomous vehicle and the vehicle is at least equal to a threshold distance, the system.
[0254] Clause 2. The system of Clause 1, wherein the at least one processor is further configured to determine at least a lane offset amount, the lane offset amount being a distance that the autonomous vehicle moves off-center within a current lane until a lateral distance between the autonomous vehicle and the vehicle is at least equal to a threshold distance.
[0255] Addendum 3. At least one processor includes at least: Determining locations of lane markers between the autonomous vehicle and the vehicle; determining that the vehicle is encroaching into a current lane in response to determining that a lateral distance between the autonomous vehicle and the vehicle is less than a distance between the autonomous vehicle and a lane marker; determining an amount of the current lane encroached by the vehicle; determining an available distance in the current lane on the other side of the autonomous vehicle compared to where the vehicle was detected; Comparing the intrusion distance in the current lane by the vehicle to the available distance in the current lane on the other side of the autonomous vehicle; determining whether there is sufficient available distance within the current lane on the other side of the autonomous vehicle to perform a lane-shifting maneuver based at least in part on a comparison between the encroached distance of the current lane by the vehicle and the available distance on the other side of the autonomous vehicle; instructing the autonomous vehicle to perform a lane-shifting maneuver in response to determining that there is sufficient available distance in the current lane on the other side of the autonomous vehicle to perform a lane-shifting maneuver; In response to determining that there is not sufficient available distance in the current lane on the other side of the autonomous vehicle to perform a lane-shifting maneuver, command the autonomous vehicle to perform a minimum-risk maneuver. 2. The system of claim 1, further configured as follows:
[0256] Clause 4. The system of Clause 3, wherein the minimum risk maneuver includes slowing down the autonomous vehicle so that the autonomous vehicle is not driving adjacent to the vehicle, or increasing the autonomous vehicle's speed so that the autonomous vehicle is not driving adjacent to the vehicle.
[0257] Appendix 5. The system of Appendix 3, wherein the at least one processor is further configured to determine that there is sufficient available distance on the current lane on the other side of the autonomous vehicle to perform a lane shifting maneuver, at least if the available distance on the current lane on the other side of the autonomous vehicle is greater than or equal to the encroached distance on the current lane by the vehicle.
[0258] Appendix 6. The system of Appendix 3, wherein the at least one processor is further configured to determine that there is not sufficient available distance within the current lane on the other side of the autonomous vehicle to perform a lane-shifting maneuver, at least if the available distance on the current lane on the other side of the autonomous vehicle is less than a penetrated distance on the current lane by the vehicle.
[0259] Addendum 7. At least one processor includes at least: determining that a lateral distance between the autonomous vehicle and the vehicle is less than a threshold distance; determining that there is not sufficient available distance within the current lane on the other side of the autonomous vehicle to perform a lane-shifting maneuver; In response to determining that there is not sufficient available distance within the current lane on the other side of the autonomous vehicle to perform a lane-shifting maneuver. determining whether another vehicle is in an adjacent lane of the autonomous vehicle; in response to determining that there are no other vehicles in the adjacent lane, Commanding the autonomous vehicle to perform a lane-shifting maneuver; commanding the autonomous vehicle to temporarily drive into the adjacent lane until a distance between the autonomous vehicle and the vehicle is equal to a threshold distance; When the autonomous vehicle is no longer adjacent to the vehicle, command the autonomous vehicle to drive back into its current lane. 2. The system of claim 1, further configured as follows:
[0260] Appendix 8. A method, comprising: Detecting a presence of the vehicle from sensor data captured by at least one sensor associated with the autonomous vehicle; Determining a lateral distance between the autonomous vehicle and the vehicle; comparing a lateral distance between the autonomous vehicle and a vehicle to a threshold distance from the autonomous vehicle, the autonomous vehicle being configured to travel along a road; determining whether to command the autonomous vehicle to perform a lane-shifting maneuver based at least in part on a comparison between the lateral distance and a threshold distance, the lane-shifting maneuver including driving the autonomous vehicle off-center in a current lane traveled by the autonomous vehicle in an opposite direction relative to the vehicle until a lateral distance between the autonomous vehicle and the vehicle is at least equal to the threshold distance; A method comprising:
[0261] Clause 9. The method of clause 8, further comprising instructing the autonomous vehicle to perform a lane-shifting maneuver in response to determining that a lateral distance between the autonomous vehicle and a vehicle is less than a threshold distance, the vehicle being detected in an adjacent lane on either side of the autonomous vehicle.
[0262] Addendum 10. Determining whether a lane-shifting maneuver may be performed within a threshold period; and instructing the autonomous vehicle to perform a lane-shifting maneuver in response to determining that a lane-shifting maneuver can be performed within a threshold time period, where the vehicle is detected on an adjacent lane in front of and on either side of the autonomous vehicle; 9. The method of claim 8, further comprising:
[0263] Addendum 11. Determining whether a lane-shifting maneuver may be performed within a threshold period includes: Determining a longitudinal distance between the autonomous vehicle and the vehicle; Determining a first velocity and a first position of the vehicle; determining a first trajectory of the vehicle based at least in part on the first velocity and the first position of the vehicle; Determining a second velocity and a second position of the autonomous vehicle; determining a second trajectory for the autonomous vehicle based at least in part on a second speed and a second position of the autonomous vehicle if a lane-shifting maneuver is performed; predicting a future lateral distance between the autonomous vehicle and the vehicle based at least in part on the first trajectory of the vehicle, the second trajectory of the vehicle, and the longitudinal distance between the autonomous vehicle and the vehicle; comparing a predicted lateral distance between the autonomous vehicle and the vehicle to a threshold distance; performing a lane-shifting maneuver in response to determining that a predicted lateral distance between the autonomous vehicle and the vehicle will be at least equal to a threshold distance within a threshold time period, the threshold time period being according to at least one of traffic on the roadway, a speed of the autonomous vehicle, and a size of the vehicle; 11. The method of claim 10, comprising:
[0264] Clause 12. The method of clause 10, further comprising instructing the autonomous vehicle to perform a minimum risk maneuver in response to determining that a lane deviation maneuver cannot be performed within the threshold period, the minimum risk maneuver comprising slowing down the autonomous vehicle so that the autonomous vehicle is not driving adjacent to the vehicle, or increasing the autonomous vehicle's speed so that the autonomous vehicle is not driving adjacent to the vehicle.
[0265] Addendum 13. Determining a longitudinal distance between an autonomous vehicle and a vehicle, the vehicle being stopped on the side of a road ahead of the autonomous vehicle; determining an amount of the current lane encroached by the vehicle; determining an available distance on the current lane on the other side of the autonomous vehicle compared to where the vehicle was detected; determining that there is sufficient available distance on the current lane on the other side of the current lane to perform a lane-shifting maneuver; Commanding the autonomous vehicle to perform a lane-shifting maneuver; 9. The method of claim 8, further comprising:
[0266] Clause 14. The method of clause 8, further comprising determining not to instruct the autonomous vehicle to perform a lane-shifting maneuver in response to determining that the vehicle's driving pattern indicates that a vehicle driving pattern prediction is less than a threshold percentage and the vehicle's driving pattern is highly unpredictable.
[0267] Appendix 15. The method of Appendix 14, wherein the driving pattern of the vehicle is determined based at least in part on historical driving behavior associated with the vehicle, and the historical driving behavior indicates that the vehicle has encroached into another lane.
[0268] Clause 16. A non-transitory computer-readable medium storing instructions that, when executed by one or more processors, Detecting a presence of the vehicle from sensor data captured by at least one sensor associated with the autonomous vehicle; determining a lateral distance between an autonomous vehicle and a vehicle, the autonomous vehicle being configured to travel along a road; comparing a lateral distance between the autonomous vehicle and the vehicle to a threshold distance from the autonomous vehicle; determining whether to command the autonomous vehicle to perform a lane-shifting maneuver based at least in part on a comparison between the lateral distance and a threshold distance; on one or more processors, The lane-shifting maneuver includes driving the autonomous vehicle off-center within a current lane traveled by the autonomous vehicle in an opposite direction relative to the vehicle until a lateral distance between the autonomous vehicle and the vehicle is at least equal to a threshold distance.
[0269] Appendix 17. The non-transitory computer-readable medium of Appendix 16, wherein the instructions, when executed by the one or more processors, in response to implementing a lane-offset maneuver, further cause the one or more processors to maintain consistent lane-offset until the autonomous vehicle is no longer adjacent to the vehicle.
[0270] Addendum 18. The autonomous vehicle includes a semi-truck tractor unit mounted on a trailer, the road is a curved road, and the instructions, when executed by the one or more processors, determining a road curvature associated with a road; determining a trailer angle between the semi-truck tractor unit and the trailer when the autonomous vehicle will reach a road curve; calculating a total lane offset adjustment based at least in part on road curvature and trailer angle; commanding the autonomous vehicle to perform a lane-shifting maneuver based at least in part on the total lane-shifting adjustment amount; 17. The non-transitory computer-readable medium of claim 16, further causing one or more processors to:
[0271] Appendix 19. The non-transitory computer-readable medium of Appendix 16, wherein the instructions, when executed by the one or more processors, further cause the one or more processors to determine a classification of the vehicles based, at least in part, on a size of each vehicle, and determining whether to perform a lane-shifting maneuver is further based, at least in part, on a particular class to which the vehicle belongs.
[0272] Clause 20. The non-transitory computer-readable medium of Clause 16, wherein the threshold distance is according to at least one of traffic on the road, a speed of the autonomous vehicle, and a size of the vehicle.
[0273] Addendum 21. A system comprising a control device associated with an autonomous vehicle, the control device comprising: a memory configured to store map data comprising one or more roads ahead of an autonomous vehicle, the autonomous vehicle comprising a semi-truck tractor unit mounted on a trailer; and at least one processor operatively coupled to the memory; Equipped with At least one processor includes at least determining, based at least in part on the map data, that the autonomous vehicle is approaching a curved road; determining a road radius for the curved road based at least in part on the map data; Calculating a first lane offset adjustment amount associated with a road curvature of the curved road based at least in part on a road radius; determining a trailer angle between the trailer and the semi-truck tractor unit; calculating a second lane offset adjustment amount associated with the trailer angle based at least in part on the trailer angle; calculating a total lane offset adjustment amount by combining the first lane offset adjustment amount and the second lane offset adjustment amount; Commanding the autonomous vehicle to perform a lane-shifting maneuver;
[0023] The system, wherein the lane offset maneuver includes driving the autonomous vehicle off-center within a curved lane currently being traveled by the autonomous vehicle based at least in part on the total lane offset adjustment amount.
[0274] Clause 22. The autonomous vehicle comprises at least one sensor configured to capture sensor data describing an environment surrounding the autonomous vehicle; At least one processor includes at least receiving sensor data from at least one sensor; determining a set of locations of lane markers on a road traveled by the autonomous vehicle from the sensor data;
[0023] 20. The method according to claim 1, further comprising: 22. The system of claim 21, wherein to determine that the autonomous vehicle is approaching a curved road, the at least one processor is further configured to determine that at least the set of lane marker locations follow a curve.
[0275] Addendum 23. To determine a road radius for a curved road, the at least one processor includes at least: determining a virtual circle on the map data such that the curved road is part of a circumference of the virtual circle; Calculating the distance between the center of the imaginary circle and the point where the semi-truck tractor unit is tangent to the trailer; 22. The system of claim 21, further configured to:
[0276] Note 24. The first lane offset adjustment amount caused by the road curvature is calculated by the following first formula: First lane offset adjustment value = (Road radius 2 -Trailer length 2 ) 1 / 2 -road radius where trailer length is the length of the trailer.
[0277] Appendix 25. The at least one processor is further configured to at least adjust a sign of the first lane offset adjustment amount based at least in part on a direction of the road curvature and a direction of the first lane offset adjustment amount, where if the direction of the road curvature is to the left and the direction of the first lane offset adjustment amount is to the left, the first lane offset adjustment amount with the adjusted sign is calculated based on a second equation below: First lane offset adjustment value with adjusted sign = ((road radius 2 -Trailer length 2 ) 1 / 2 -road radius)) x sign (-road curvature) is calculated according to the formula: If the direction of the road curvature is to the left, the sign associated with the road curvature is a positive sign; if the direction of the first lane offset adjustment amount is to the left, a sign associated with the first lane offset adjustment amount is a negative sign; The system of claim 24, wherein the sign (road curvature) indicates a sign associated with a direction of the road curvature.
[0278] Clause 26. The system of Clause 21, wherein the trailer angle is determined from sensor data received from a sensor associated with the autonomous vehicle.
[0279] Clause 27. The system of clause 21, wherein in calculating the first lane offset adjustment amount, the trailer angle is represented as being zero.
[0280] Addendum 28. A method, comprising: determining that an autonomous vehicle is approaching a curved road based at least in part on map data comprising one or more roads ahead of the autonomous vehicle, the autonomous vehicle comprising a semi-truck tractor unit mounted on a trailer; determining a road radius for the curved road based at least in part on the map data; Calculating a first lane offset adjustment amount associated with a road curvature of the curved road based at least in part on a road radius; determining a trailer angle between the trailer and the semi-truck tractor unit; calculating a second lane offset adjustment amount associated with the trailer angle based at least in part on the trailer angle; calculating a total lane offset adjustment amount by combining the first lane offset adjustment amount and the second lane offset adjustment amount; When you command an autonomous vehicle to perform a lane-shifting maneuver, Including, The method, wherein the lane offset maneuver includes driving the autonomous vehicle off-center within a curved lane currently being traveled by the autonomous vehicle based at least in part on a total lane offset adjustment amount.
[0281] Clause 29. Receiving sensor data from at least one sensor associated with an autonomous vehicle; Detecting the presence of a vehicle on a road from sensor data; Determining a lateral distance between the autonomous vehicle and the vehicle; comparing a lateral distance between the autonomous vehicle and the vehicle to a threshold distance; determining that the lateral distance is less than a threshold distance; instructing the autonomous vehicle to perform a lane-shifting maneuver in response to determining that the lateral distance is less than a threshold distance; Further comprising: 29. The method of claim 28, wherein the lane-shifting maneuver includes driving the autonomous vehicle off-center within a current lane traveled by the autonomous vehicle in an opposite direction relative to the vehicle until a lateral distance between the autonomous vehicle and the vehicle is at least equal to a threshold distance.
[0282] Addendum 30. Determining a lane offset, the lane offset being a distance that the autonomous vehicle moves off-center until a lateral distance between the autonomous vehicle and the vehicle is at least equal to a threshold distance; Combining the total lane offset adjustment amount with the lane offset amount; 30. The method of claim 29, further comprising:
[0283] Clause 31. The method of clause 30, wherein the amount of lane deviation is determined based at least in part on the amount of the current lane that has been encroached upon by the vehicle and the available distance on the current lane on the other side of the autonomous vehicle compared to where the vehicle was detected.
[0284] Appendix 32. The method of Appendix 30, wherein the total lane offset adjustment amount is combined with the lane offset amount when the vehicle is on the right side of the autonomous vehicle and the road curvature is to the left.
[0285] Clause 33. The method of clause 30, wherein the total lane offset adjustment amount is not combined with the lane offset amount when the vehicle is to the left of the autonomous vehicle and the road curvature is to the left.
[0286] Addendum 34. The method of Addendum 28, wherein in calculating the second lane offset adjustment amount, the road traveled by the autonomous vehicle is represented as being straight.
[0287] Addendum 35. The method of Addendum 29, wherein the at least one sensor comprises at least one of a camera, a light detection and ranging (LiDAR) sensor, and an infrared sensor.
[0288] Clause 36. A non-transitory computer-readable medium storing instructions that, when executed by one or more processors, determining that an autonomous vehicle is approaching a curved road based at least in part on map data comprising one or more roads ahead of the autonomous vehicle, the autonomous vehicle comprising a semi-truck tractor unit mounted on a trailer; determining a road radius for the curved road based at least in part on the map data; Calculating a first lane offset adjustment amount associated with a road curvature of the curved road based at least in part on a road radius; determining a trailer angle between the trailer and the semi-truck tractor unit; calculating a second lane offset adjustment amount associated with the trailer angle based at least in part on the trailer angle; calculating a total lane offset adjustment amount by combining the first lane offset adjustment amount and the second lane offset adjustment amount; Commanding the autonomous vehicle to perform a lane-shifting maneuver; on one or more processors, The non-transitory computer-readable medium, wherein the lane offset maneuver includes driving the autonomous vehicle off-center within a curved lane currently being traveled by the autonomous vehicle based at least in part on a total lane offset adjustment amount.
[0289] Note 37. The second lane offset adjustment amount caused by the trailer angle is calculated using the following third formula: Second lane offset adjustment value = trailer length x sin(trailer angle) where trailer length is the length of the trailer.
[0290] Clause 38. The instructions, when executed by the one or more processors, further cause the one or more processors to adjust a sign of the second lane offset adjustment amount based at least in part on a direction of the trailer angle and a direction of the second lane offset adjustment amount, such that if the direction of the trailer angle is to the left and the direction of the second lane offset adjustment amount is to the right, then the second lane offset adjustment amount with the adjusted sign is determined to be in accordance with a fourth equation below: Second lane offset adjustment value with adjusted sign = - trailer length x sin(trailer angle) where trailer length is the length of the trailer.
[0291] Addendum 39. The instructions, when executed by one or more processors, determining, based at least in part on the map data, that the autonomous vehicle is traveling along a straight road; determining that the first lane offset adjustment amount is zero because the road radius is substantially large; and detecting that wind is causing a trailer of the autonomous vehicle to deviate from a straight line; determining a trailer angle caused by wind; calculating a second lane offset adjustment caused by the trailer angle; determining that the total lane offset adjustment amount is equal to the second lane offset adjustment amount; 37. The non-transitory computer-readable medium of claim 36, further causing one or more processors to:
[0292] Clause 40. The non-transitory computer-readable medium of Clause 36, wherein when the trailer sways to the left, a sign associated with the trailer angle is negative, and when the trailer sways to the right, a sign associated with the trailer angle is positive.
[0293] Addendum 41. The system of any of Addendums 1-7, wherein at least one processor is further configured to perform one or more operations of the methods described in any of Addendums 8-15.
[0294] Addendum 42. The system of any of Addendums 1-7, wherein the processor is further configured to perform one or more operations of any of Addendums 16-20.
[0295] Addendum 43. An apparatus comprising means for carrying out the method according to any one of Addendums 8-15.
[0296] Clause 44. An apparatus comprising means for carrying out one or more instructions according to any of clauses 16-20.
[0297] Addendum 45. A non-transitory computer-readable medium as described in any of Addendums 16-20 storing instructions that, when executed by one or more processors when launched on a system, further cause the one or more processors to perform one or more operations of the methods described in any of Addendums 8-15.
[0298] Addendum 46. The system of any of Addendums 21-27, wherein at least one processor is further configured to perform one or more operations of any of Addendums 28-35.
[0299] Addendum 47. The system of any of Addendums 21-27, wherein the processor is further configured to perform one or more operations of any of Addendums 36-40.
[0300] Addendum 48. An apparatus comprising means for carrying out the method according to any one of Addendums 28-25.
[0301] Clause 49. An apparatus comprising means for carrying out one or more instructions according to any of clauses 36-40.
[0302] Addendum 50. A non-transitory computer-readable medium as described in any of Addendums 36-40 storing instructions that, when executed by one or more processors when launched on a system, further cause the one or more processors to perform one or more operations of the methods described in any of Addendums 28-35.
[0303] Addendum 51. The system of any of Addendums 1-7 and / or 21-27.
[0304] Addendum 52. A method including the operations described in any of Addendums 8-15 and / or 28-35.
[0305] Addendum 53. An apparatus comprising means for carrying out the method according to any of Addendums 8-15 and / or 28-35.
[0306] Appendix 54. An apparatus comprising means for carrying out one or more instructions according to any of Appendixes 16-20 and / or 36-40.
[0307] Addendum 55. A non-transitory computer-readable medium as described in any of Addendums 16-20 and / or 36-40 storing instructions that, when executed by one or more processors when activated on a system, further cause the one or more processors to perform one or more operations of the methods described in any of Addendums 8-15 and / or 28-35.
Claims
1. A system (300) comprising a control device (550) associated with an autonomous vehicle (502), the control device comprising: a memory (126, 590) configured to store map data (114), the map data comprising one or more roads ahead of the autonomous vehicle (502), the autonomous vehicle (502) comprising a trailer-mounted semi-truck tractor unit; and At least one processor (122, 570) operatively coupled to said memory (126, 590); Equipped with The at least one processor includes at least determining, based at least in part on the map data (114), that the autonomous vehicle (502) is approaching a curved road (302); determining a road radius (322) of the curved road based at least in part on the map data (114); calculating a first lane offset adjustment amount (330) associated with a road curvature of the curved road (302) based at least in part on the road radius (322); determining a trailer angle (342) between the trailer and the semi-truck tractor unit; calculating a second lane offset adjustment amount (340) associated with the trailer angle (342) based at least in part on the trailer angle (342); calculating a total lane offset adjustment amount (320) by combining the first lane offset adjustment amount and the second lane offset adjustment amount; commanding the autonomous vehicle (502) to perform a lane-shifting maneuver (130); configured to: The lane offset maneuver (130) includes driving the autonomous vehicle (502) off-center within a curved lane currently being traveled by the autonomous vehicle (520) based at least in part on the total lane offset adjustment amount (320), in the system (300).
2. the autonomous vehicle (520) comprises at least one sensor (546) configured to capture sensor data (314) describing an environment surrounding the autonomous vehicle (502); The at least one processor (122, 570) includes at least receiving said sensor data (314) from said at least one sensor (546); determining a set of locations of lane markers (306) on the curved road from the sensor data (314); [0023] 20. The method according to claim 1, further comprising:
2. The system of claim 1, wherein to determine that the autonomous vehicle is approaching the curved road, the at least one processor is further configured to determine that at least the set of lane marker locations follows a curve.
3. To determine the road radius (322) of the curved road, the at least one processor (122, 570) performs at least determining the virtual circle (324) on the map data (114) such that the curved road is part of a circumference of the virtual circle; Calculating the distance between the center of the imaginary circle and the point where the semi-truck tractor unit is tangent to the trailer (332); The system of claim 1 , further configured to:
4. The first lane offset adjustment amount (330) caused by the road curvature is calculated based on a first equation: First lane offset adjustment value=(road radius 2 - Trailer length 2 ) 1/2 -Road radius 2. The system of claim 1, wherein the trailer length is calculated according to:
5. The at least one processor (122, 570) is further configured to at least adjust a sign of the first lane offset adjustment amount (330) based at least in part on a direction of the road curvature and a direction of the first lane offset adjustment amount (330), where if the direction of the road curvature is to the left and the direction of the first lane offset adjustment amount (330) is to the left, the first lane offset adjustment amount (330) with the adjusted sign is calculated based on a second equation: First lane offset adjustment value with adjusted sign= ((road radius 2 - Trailer length 2 ) 1/2 - road radius)) x sign (- road curvature) is calculated according to the formula: if the direction of the road curvature is to the left, then the sign associated with the road curvature is a positive sign; if the direction of the first lane offset adjustment amount (330) is to the left, a sign associated with the first lane offset adjustment amount (330) is a negative sign; The system of claim 4 , wherein the sign (road curvature) indicates a sign associated with the direction of the road curvature.
6. The system of claim 1 , wherein the trailer angle (342) is determined from sensor data (314) received from a sensor (546) associated with the autonomous vehicle (502).
7. 2. The system of claim 1, wherein in calculating the first lane offset adjustment (330), the trailer angle (342) is represented as being zero.
8. 1. A method (400), comprising: determining (402) that an autonomous vehicle (502) is approaching a curved road (302) based at least in part on map data (114) comprising one or more roads ahead of the autonomous vehicle (502), the autonomous vehicle (502) comprising a semi-truck tractor unit mounted on a trailer; determining (404) a road radius (322) of the curved road (302) based at least in part on the map data (114); calculating (406) a first lane offset adjustment amount (330) associated with a road curvature of the curved road (302) based at least in part on the road radius (322); Determining (408) a trailer angle (342) between the trailer and the semi-truck tractor unit; Calculating (410) a second lane offset adjustment amount (340) associated with the trailer angle (342) based at least in part on the trailer angle (342); Calculating (412) a total lane offset adjustment amount (320) by combining the first lane offset adjustment amount (33) and the second lane offset adjustment amount (340); commanding the autonomous vehicle (502) to perform a lane-shifting maneuver (130); Including, The method (400), wherein the lane offset maneuver (130) includes driving the autonomous vehicle (502) off-center within a curved lane currently being traveled by the autonomous vehicle (502) based at least in part on the total lane offset adjustment amount (320).
9. Receiving (202) sensor data (314) from at least one sensor (546) associated with the autonomous vehicle (502); Detecting (204) the presence of vehicles (108a-c, 310a-b) on a roadway from said sensor data (314); Determining (206) a lateral distance (138a-c) between the autonomous vehicle (502) and the vehicles (108a-c, 310a-b); comparing (208) the lateral distance (138a-c) between the autonomous vehicle (502) and the vehicles (108a-c, 310a-b) to a threshold distance (132); determining (210) that the lateral distance (138a-c) is less than the threshold distance (132); instructing (218) the autonomous vehicle (502) to perform the lane-shifting maneuver (130) in response to determining that the lateral distance (138a-c) is less than the threshold distance (132); Further comprising:
10. The method of claim 8, wherein the lane-shifting maneuver includes driving the autonomous vehicle off-center within a current lane being traveled by the autonomous vehicle in an opposite direction relative to the vehicle until the lateral distance between the autonomous vehicle and the vehicle is at least equal to the threshold distance.
10. determining a lane deviation (110), the lane deviation (110) being a distance that the autonomous vehicle (502) moves off-center until the lateral distance (138a-c) between the autonomous vehicle (502) and the vehicles (108a-c, 310a-b) is at least equal to the threshold distance (132); combining (414) said total lane offset adjustment amount (320) with said lane offset amount (110); The method of claim 9 further comprising:
11. 11. The method of claim 10, wherein the lane deviation amount (110) is determined based at least in part on an amount of the current lane encroached by the vehicle (108a-c, 310a-b) and an available distance on the current lane on the other side of the autonomous vehicle (502) compared to where the vehicle (108a-c, 310a-b) was detected.
12. 11. The method of claim 10, wherein the total lane offset adjustment amount is combined with the lane offset amount when the vehicle is on the right side of the autonomous vehicle and the road curvature is to the left.
13. 11. The method of claim 10, wherein the total lane offset adjustment amount is not combined with the lane offset amount when the vehicle is on the left side of the autonomous vehicle and the road curvature is to the left.
14. 10. The method of claim 8, wherein in calculating the second lane offset adjustment (340), a road traveled by the autonomous vehicle is represented to be a straight line.
15. The method of claim 9, wherein the at least one sensor (546) comprises at least one of a camera, a light detection and ranging (LiDAR) sensor, and an infrared sensor.
16. A non-transitory computer readable medium (126, 590) storing instructions (312, 580) that, when executed by one or more processors (122, 570), perform the following: determining that an autonomous vehicle (502) is approaching a curved road (302) based at least in part on map data (114) comprising one or more roads ahead of the autonomous vehicle (502), the autonomous vehicle (502) comprising a trailer-mounted semi-truck tractor unit; determining a road radius (322) of the curved road based at least in part on the map data (114); calculating a first lane offset adjustment amount (330) associated with a road curvature of the curved road based at least in part on the road radius; determining a trailer angle (342) between the trailer and the semi-truck tractor unit; calculating a second lane offset adjustment amount (340) associated with the trailer angle based at least in part on the trailer angle; calculating a total lane offset adjustment amount (320) by combining the first lane offset adjustment amount (330) and the second lane offset adjustment amount (340); commanding the autonomous vehicle (502) to perform a lane-shifting maneuver (130); causing the one or more processors to The lane offset maneuver (130) includes driving the autonomous vehicle (502) off-center within a curved lane currently being traveled by the autonomous vehicle (502) based at least in part on the total lane offset adjustment amount (320).
17. The second lane offset adjustment amount (340) caused by the trailer angle (342) is calculated according to a third equation: Second lane offset adjustment value = trailer length x sin (trailer angle) 17. The non-transitory computer readable medium of claim 16, wherein the trailer length (342) is a length of the trailer.
18. The instructions (312, 580), when executed by the one or more processors (122, 570), cause the one or more processors (122, 570) to further adjust a sign of the second lane offset adjustment amount (340) based at least in part on a direction of the trailer angle (342) and a direction of the second lane offset adjustment amount (340), such that when the direction of the trailer angle (342) is to the left and the direction of the second lane offset adjustment amount (340) is to the right, the second lane offset adjustment amount (340) with the adjusted sign is satisfied by a fourth equation: Second Lane Offset Adjustment Value with Adjusted Sign=-Trailer Length×sin(Trailer Angle) 17. The non-transitory computer readable medium of claim 16, wherein the trailer length is calculated according to:
19. The instructions (312, 580), when executed by the one or more processors (122, 570), determining, based at least in part on the map data (114), that the autonomous vehicle (502) is traveling along a straight road; determining that the road radius (322) is substantially large such that the first lane offset adjustment amount (330) is zero; detecting that wind is causing a trailer of the autonomous vehicle (502) to deviate from a straight line; determining the trailer angle (342) caused by the wind; Calculating the second lane offset adjustment (340) caused by the trailer angle (342); determining that the total lane offset adjustment amount (320) is equal to the second lane offset adjustment amount (340); 20. The non-transitory computer readable medium of claim 16, further causing the one or more processors to:
20. When the trailer swings to the left, the sign associated with the trailer angle (342) is negative; 17. The non-transitory computer readable medium (126, 590) of claim 16, wherein the sign associated with the trailer angle (342) is positive when the trailer yaws to the right.
21. The system (300) of any of claims 1-7, wherein the at least one processor (122, 570) is further configured to perform one or more operations of the method (400) of any of claims 8-15.
22. The system (300) of any of claims 1-7, wherein the processor (122, 570) is further configured to perform one or more operations of any of claims 16-20.
23. Apparatus comprising means for performing the method (400) according to any of claims 8-15.
24. Apparatus comprising means for implementing one or more instructions (312, 580) according to any of claims 16-20.
25. A non-transitory computer readable medium according to any of claims 16-20, storing instructions (312, 580) which, when run on a system (300), when executed by the one or more processors, further cause the one or more processors to perform one or more operations of a method (400) according to any of claims 8-15.
