Autonomous driving system and autonomous driving method

Abstract

To increase the possibility of enabling a vehicle to recover from a travel impossible state by maximizing the performance of on-vehicle sensors.SOLUTION: An autonomous traveling system 300 includes: a general route planning unit 305 that plans a general route from the a travel start position to a target position of an own vehicle by referring to a normal resolution map of the travel environment that has been created in advance; a travel impossibility determination unit 307 that determines whether the own vehicle is in a travel impossible state when the own vehicle autonomously travels along the general route; a high-resolution surrounding map generation unit 308 that generates a high-resolution surrounding map having a higher resolution than that of the normal resolution map when the travel impossibility determination unit determines that the own vehicle is in the travel impossible state.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present invention relates to an autonomous driving system Mu and and an autonomous driving method.

Background Art

[0002] With the labor shortage due to the declining birthrate and aging population and the expansion of the e-commerce market, there is a demand for labor-saving and improved work efficiency in logistics warehouses and factories. Therefore, the introduction of an autonomous driving system that operates an unmanned autonomous driving vehicle is being promoted. In conventional autonomous driving systems, from the perspective of safety, the range within which the vehicle can move has been limited, and it has been common for the working range of on-site workers, etc., to be different from the movement range of the vehicle. On the other hand, with the development of autonomous driving technology in the field of robotics, even when the traffic lines of vehicles and people overlap, the vehicle itself can judge the necessity of avoidance or detour, etc., and can select flexible actions according to the surrounding situation. For this reason, an autonomous driving system in which the movement range of the vehicle is not limited in the future and a plurality of vehicles and people can share the entire area of one site is desirable.

[0003] By the way, in order to realize wide-area autonomous driving, autonomous driving of a vehicle in an environment that changes every moment is required. For example, in a warehouse or factory related to transportation work, depending on the destination, vehicle driving in between shelves or in a complicated passageway, etc., is required. Generally, an autonomous driving vehicle is equipped with an in-vehicle sensor, and the in-vehicle sensor compares the current measurement information with the map information created in advance, and detects the position of the vehicle in the created map coordinate system and obstacles that impede the vehicle's driving.

[0004] ​However, in environments such as warehouses and factories, the arrangement of shelves, pallets, and carts changes constantly, so there may be discrepancies between the information currently detected by the vehicle's onboard sensors and the pre-created map information. In this case, the route planned based on the pre-created map information may be blocked by obstacles, making it impossible to reach the destination. Therefore, when a vehicle becomes unable to move, a technology has been proposed, such as Patent Document 1, that recovers from the immobile state by regenerating the map using information detected by the onboard sensors and resetting an appropriate driving trajectory according to the changes in the surrounding environment.

[0005] Patent Document 1 states that, "When the track simulator determines that the track to be reset by the resetting determination unit, it re-executes the track simulation using the surrounding environment generated by the surrounding environment generation unit, the track evaluation unit newly determines the vehicle's track based on the results of the track simulation re-executed by the track simulator, and the vehicle coordination unit transmits the information of the newly determined track to the onboard device by the track evaluation unit." [Prior art documents] [Patent Documents]

[0006] [Patent Document 1] Japanese Patent Publication No. 2020-140534 [Overview of the project] [Problems that the invention aims to solve]

[0007] As mentioned above, when a vehicle becomes inoperable, a technology has been proposed to recover from this state by regenerating the map and resetting an appropriate driving trajectory in accordance with changes in the surrounding environment. Incidentally, the map information is generally a grid map generated using on-board sensors such as LiDAR (Light Detection and Ranging). A grid map divides real space into a grid and classifies it into drivable areas where the on-board sensors did not detect obstacles, areas where observation values ​​from the on-board sensors were obtained, and unknown areas before the obstacle areas that have not yet been observed, and expresses the probability of the presence of obstacles in each grid numerically.

[0008] The granularity of a map divided by a grid is expressed as the resolution of the grid map, specifically as the size of each grid cell in real space (cm / pixel). By increasing the resolution of the map, obstacle shapes can be captured in more detail, allowing for a finer representation of the drivable area and increasing the likelihood of driving on narrow roads. However, for example, if a low-resolution map is 5 cm / pixel and a high-resolution map is 1 cm / pixel, the map information has a 25-fold difference in data capacity. Therefore, due to the limitations of the in-vehicle controller's processing power (memory, CPU performance, etc.), generally, a map with a resolution lower than the maximum resolution achievable from the performance of the in-vehicle sensor is generated. The maximum resolution can be estimated from the beam width of the LiDAR (in-vehicle sensor), the horizontal resolution (distance between adjacent beams), and the measurement error provided by each LiDAR manufacturer. Furthermore, since the map resolution may depend on the algorithm used for map generation, it is also possible to set multiple resolutions and select a map resolution that can represent the shapes of surrounding obstacles without distortion.

[0009] Furthermore, in the technology described in Patent Document 1, when the vehicle becomes inoperable, the map regeneration process does not take steps to change the resolution of the grid map to maximize the performance of the on-board sensors. As a result, in narrow roads and complex environments, the shape of obstacles on the map cannot be captured with greater precision, making it highly likely that the vehicle will become inoperable again.

[0010] This invention has been made in view of the above circumstances, and its objective is to maximize the performance of on-board sensors and improve the possibility of restoring a vehicle from a state where it is unable to drive. [Means for solving the problem]

[0011] The present invention includes a global route planning unit that plans a global route from the vehicle's starting position to a target position by referring to a pre-created normal-resolution map of the driving environment; a vehicle immobility determination unit that determines whether or not the vehicle has become immobile while autonomously driving along the global route; and a high-resolution surrounding map generation unit that generates a high-resolution surrounding map with a higher resolution than the normal-resolution map if the vehicle immobility determination unit determines that the vehicle has become immobile. A normal-resolution map update unit manages a normal-resolution map, receives a high-resolution surrounding map from the vehicle, converts the received high-resolution surrounding map to the resolution of a normal-resolution map, and updates the normal-resolution map with the converted high-resolution surrounding map; a map distribution unit transmits the normal-resolution map updated by the normal-resolution map update unit to vehicles in the driving environment; an on-board sensor information management unit manages the performance information of the vehicle's on-board sensors; and a target position management unit manages the vehicle's target position. If the vehicle is determined to be unable to reach its target position by the driving impossibility determination unit, the unit receives from the vehicle status information including the status of being unable to reach the target position, the vehicle's position and attitude on the normal-resolution map at the time of the determination, and the vehicle's target position. Based on the driving status information and the information from the on-board sensor information management unit, the unit searches for other vehicles that can travel to the vehicle's target position in its place and transmits the vehicle's target position to the other vehicles found. It is equipped with. [Effects of the Invention]

[0012] According to the present invention with the above configuration, it is possible to maximize the performance of the on-board sensor and improve the possibility of restoring the vehicle from a state where it is unable to drive. Other issues, configurations, and effects will be clarified by the following description of the embodiments. [Brief explanation of the drawing]

[0013] [Figure 1] This figure shows an example of a driving environment according to the first embodiment of the present invention. [Figure 2] This figure shows an example configuration of an autonomous vehicle according to the first embodiment of the present invention. [Figure 3] This is a block diagram showing an example of the functional configuration of an autonomous driving system according to the first embodiment of the present invention. [Figure 4] It is a block diagram showing the hardware configuration of an in-vehicle controller of an autonomous driving system according to the first embodiment of the present invention. [Figure 5] It is a conceptual diagram showing a part of the driving environment according to the first embodiment of the present invention. [Figure 6] It is a conceptual diagram showing an example of a state where an obstacle is newly installed in the driving environment according to the first embodiment of the present invention. [Figure 7] It is a conceptual diagram showing a normal resolution map and a global route to a target position according to the first embodiment of the present invention. [Figure 8] It is a diagram showing a state where an obstacle and a global route are expanded in a normal resolution map according to the first embodiment of the present invention. [Figure 9] It is a conceptual diagram showing a state where an obstacle and a global route are expanded when a new obstacle is installed in a normal resolution map according to the first embodiment of the present invention. [Figure 10] It is a conceptual diagram showing an example of a high-resolution surrounding map around the host vehicle according to the first embodiment of the present invention. [Figure 11] It is a flowchart showing the procedure of autonomous driving processing in an autonomous driving system according to the first embodiment of the present invention. [Figure 12] It is a block diagram showing an example of the functional configuration of an autonomous driving system according to the second embodiment of the present invention. [Figure 13] It is a block diagram showing an example of the functional configuration of an autonomous driving system according to the third embodiment of the present invention. [Figure 14] It is a conceptual diagram showing vehicles of different vehicle types according to the third embodiment of the present invention.

Embodiments for Carrying Out the Invention

[0014] Hereinafter, embodiments for carrying out the present invention will be described with reference to the accompanying drawings. In this specification and the drawings, components having substantially the same function or configuration are denoted by the same reference numerals, and redundant descriptions are omitted.

[0015] <First Embodiment> First, an example of a driving environment according to the first embodiment of the present invention will be described. Figure 1 is a diagram showing an example of a driving environment according to this embodiment. The driving environment 100 is a driving environment in an example of a warehouse related to transport operations. In Figure 1, the horizontal direction on the drawing, "X", indicates the width direction of the warehouse, and the "Y" direction intersecting the "X" direction indicates the vertical direction of the warehouse. Also, in Figure 1, the same figure represents the same thing, and only one of them is labeled with a reference numeral for explanation.

[0016] As shown in Figure 1, the driving environment 100 is an environment where shelves 111, pallets 113, charging units 114 for each vehicle, and conveyors 115 for transporting goods from other areas are located. The goods 112, indicated by the gray squares, are placed on shelves 111, pallets 113, and conveyors 115, etc. In the driving environment 100, there are narrow paths enclosed by shelves 111, pallets 113, conveyors 115, etc. The transport operation is carried out by autonomous driving of vehicles such as vehicle 101, a trolley-type transport vehicle, vehicle 102, a towing vehicle, and vehicle 103, a forklift, as shown in Figure 1.

[0017] [Example configuration of an autonomous vehicle] Next, an example of the configuration of an autonomous vehicle will be described. The autonomous vehicle is, for example, vehicle 101, vehicle 102, and vehicle 103, as shown in Figure 1. Since the main components related to autonomous driving and their operation are basically the same for vehicles 101, 102, and 103, in the following description, vehicle 101 will be described as an example of an autonomous vehicle, and the descriptions of vehicles 102 and 103 will be omitted. Figure 2 is a diagram showing an example of the configuration of an autonomous vehicle according to this embodiment. Vehicle 101 includes a vehicle frame 201 equipped with a bumper 202, a transfer device 203 that can move up and down and left and right on the vehicle frame 201, and a cargo handling member 204 that supports a load 112. Vehicle 101 also includes drive wheels 206, driven wheels 207, and on-board sensors 211 arranged on the vehicle frame 201. Furthermore, the vehicle frame 201 is equipped with a transfer motor 205 that drives the transfer device 203 and a travel motor 208 that drives the drive wheels 206. Also, the vehicle frame 201 is equipped with an encoder 209 that measures movement information including the distance traveled and current speed of the vehicle 101 based on the amount of rotation of the drive wheels 206, and an on-board controller 220.

[0018] The on-board sensor 211 is, for example, composed of a LiDAR. A LiDAR is a sensor that measures objects present in the irradiation range and the distance to those objects while changing the irradiation angle of the laser beam. The LiDAR acquires point cloud information indicating the position of an object from information on the optical axis angle at the time of laser irradiation and the distance information to the object. In this embodiment, a 2D (Dimensions)-LiDAR that acquires only one layer of 2D information in the direction perpendicular to the laser beam is used as an example of the on-board sensor 211, but the present invention is not limited to this. The on-board sensor 211 may also be a 3D-LiDAR that irradiates in multiple layers and acquires 3D information. Furthermore, although only one on-board sensor 211 is installed in the vehicle 101 shown in Figure 2, the present invention is not limited to this, and multiple on-board sensors 211 may be installed in the vehicle 101.

[0019] The in-vehicle controller 220 is an example of an information processing device that controls the operation of each component of the vehicle 101 so that the vehicle 101 can drive autonomously. The functional configuration of the in-vehicle controller 220 will be explained in Figure 3 below, and the hardware configuration of the in-vehicle controller 220 will be explained in Figure 4 below.

[0020] [Example of an autonomous driving system's functional configuration] Next, an example of the functional configuration of the autonomous driving system according to this embodiment will be described. Figure 3 is a block diagram showing an example of the functional configuration of the autonomous driving system 300 according to this embodiment. As shown in Figure 3, the autonomous driving system 300 includes an on-board controller 220, drive wheels 206 of the vehicle 101, a driving motor 208, an encoder 209, and an on-board sensor 211. The on-board controller 220 is connected to the driving motor 208, the encoder 209, and the on-board sensor 211, respectively. The driving motor 208 and the encoder 209 are connected to the drive wheels 206.

[0021] The in-vehicle controller 220 includes a drive unit 301, a normal resolution map generation unit 302, a self-position estimation unit 303, a target position input unit 304, a global route planning unit 305, and a local route planning unit 306. The autonomous driving system 300 includes a driving impossibility determination unit 307, a high-resolution surrounding map generation unit 308, a local target position and route calculation unit 309, and a control command generation unit 310.

[0022] The drive unit 301 is connected to the travel motor 208 and outputs drive commands for the drive wheels 206 to the travel motor 208. The normal resolution map generation unit 302, self-position estimation unit 303, high-resolution surrounding map generation unit 308, and control command generation unit 310 are connected to the output side of the encoder 209 and receive vehicle movement information from the encoder 209. In addition, the normal resolution map generation unit 302, self-position estimation unit 303, local path planning unit 306, and high-resolution surrounding map generation unit 308 are connected to the output side of the on-board sensor 211 and receive measurement information including point cloud information of obstacles and distance to obstacles measured by the on-board sensor 211.

[0023] The normal-resolution map generation unit 302, target position input unit 304, global route planning unit 305, local route planning unit 306, drivability determination unit 307, control command generation unit 310, and drive unit 301 are connected in this order. The high-resolution surrounding map generation unit 308 and local target position and route calculation unit 309 are connected in this order. The local target position and route calculation unit 309 is further connected to the input side of the local route planning unit 306. The normal-resolution map generation unit 302 and the high-resolution surrounding map generation unit 308 are each connected to the input side of the self-position estimation unit 303, and the self-position estimation unit 303 is further connected to the input side of the global route planning unit 305, local route planning unit 306, and control command generation unit 310. The normal-resolution map generation unit 302 is also connected to the input side of the self-position estimation unit 303 and the global route planning unit 305. Furthermore, the overall route planning unit 305 and the vehicle impassability determination unit 307 are connected to the input side of the high-resolution surrounding map generation unit 308, respectively.

[0024] The drive unit 301 converts the drive command for the vehicle 101 input from the control command generation unit 310 into a predetermined current value and outputs it to the drive motor 208. The encoder 209 acquires the amount of rotation of the drive motor 208 output at each time point, and, taking into account the wheel diameter of the drive wheels 206, measures movement information including the vehicle's travel distance and current speed. The movement information measured by the encoder 209 is transmitted to the normal resolution map generation unit 302, the self-position estimation unit 303, the high-resolution surrounding map generation unit 308, and the control command generation unit 310, respectively.

[0025] The normal-resolution map generation unit 302 receives measurement information from the on-board sensor 211 and vehicle movement information from the encoder 209. Based on the measurement information from the on-board sensor 211 and the vehicle movement information from the encoder 209, the normal-resolution map generation unit 302 creates a grid map with normal resolution using, for example, a SLAM (Simultaneous Localization and Mapping) method. Here, the normal-resolution grid map is assumed to be a grid map with a resolution of 5 cm / pixel (see Figures 7 to 9 below). In the following description, the normal-resolution grid map will be referred to as the "normal-resolution map". Before the vehicle 101 starts autonomous driving, the normal-resolution map generation unit 302 creates a normal-resolution map of the driving environment, for example, the driving environment 100 shown in Figure 1, i.e., a map of the entire driving environment 100, and registers it in the storage device 22d (see Figure 4) below. Furthermore, the normal-resolution map generation unit 302 outputs the generated normal-resolution map (whole-area map) to the self-position estimation unit 303, the target position input unit 304, and the global path planning unit 305.

[0026] The self-position estimation unit 303 receives measurement information from the on-board sensor 211, a normal-resolution map from the normal-resolution map generation unit 302, and vehicle 101 movement information from the encoder 209. The self-position estimation unit 303 also performs map matching processing between the measurement information from the on-board sensor 211 and obstacle information on the normal-resolution map using the measurement information from the on-board sensor 211 and the normal-resolution map. Based on the matching processing results and the vehicle 101 movement information, the self-position estimation unit 303 estimates the self-position and attitude of the vehicle 101 on the normal-resolution map. Any method used for position estimation, such as the AMCL (Adaptive Monte Carlo Localization) method, may be used as the self-position estimation technique. The self-position estimation unit 303 also outputs the estimated current self-position of the vehicle 101 to the global route planning unit 305, the local route planning unit 306, and the control command generation unit 310.

[0027] The target position input unit 304 receives a normal resolution map from the normal resolution map generation unit 302 and sets the coordinates of the destination of the task currently being performed by the vehicle 101 on the normal resolution map as the target position for autonomous driving. Here, the target position for autonomous driving may be directly input by an operator, for example, via an external device such as a mobile display. Alternatively, the target position for autonomous driving may be automatically calculated by inputting the destination of the task pre-set on the control server (see Figures 12 and 13 below) that manages multiple vehicles, for example, on the control server side. The target position input unit 304 also outputs the target position for autonomous driving to the global route planning unit 305.

[0028] The global route planning unit 305 receives a pre-created normal-resolution map from the normal-resolution map generation unit 302, the current position of the vehicle from the self-position estimation unit 303, and the target position from the target position input unit 304. The global route planning unit 305 also refers to a pre-created normal-resolution map of the driving environment and plans a global route (see Figure 7 below) from the vehicle's starting position to the target position. The global route is represented by a set of grid coordinates (matrix numbers, etc.) on the normal-resolution map, and the vehicle reaches the target position by traveling in the order of the coordinates in the set. Any route planning method may be used, such as Dijkstra's algorithm or the A* (A-star) method. The global route planning unit 305 also outputs the planned global route to the local route planning unit 306 and the high-resolution surrounding map generation unit 308.

[0029] The local route planning unit 306 is an example of the first local route planning unit, and receives measurement information from the on-board sensor 211, the current position of the vehicle from the self-position estimation unit 303, and the global route from the global route planning unit 305. The local route planning unit 306 also sets a first local target position located between the current position of the vehicle and the target position by referring to a normal resolution map along the global route, and plans a first local route from the current position to the local target position (see Figure 8 described later).

[0030] Specifically, the local path planning unit 306 uses the measurement information from the on-board sensor 211, the current position, and the global path to set a grid coordinate on the global path located a predetermined distance (e.g., 2m) from the current position as the first local target position. The local path planning unit 306 also plans and generates a first local path to the set first local target position. Here, the first local path is not a set of grid coordinates, but rather a set of finer coordinates or a function representing a curve, and is a smooth path that takes into account the motion characteristics of the vehicle 101. Specifically, the local path planning unit 306 plans the local path using methods such as spline function interpolation or DWA (Dynamic Window Approach) that takes into account the vehicle's motion model. Note that the local path planning method is not limited to the above methods; any method that can plan a more detailed path than the global path can be applied.

[0031] Furthermore, the first local target position may not be a grid coordinate on a global path located at a predetermined distance (e.g., 2m) from the current position, but may be dynamically set according to the current vehicle speed, the conditions around the vehicle, etc. Alternatively, for example, the predetermined distance from the current position may be set as a predetermined coefficient × vehicle speed. In other words, the predetermined distance from the current position may be dynamically changed according to the vehicle's speed. Alternatively, for example, information on surrounding obstacles detected by the on-board sensor 211 may be superimposed on a normal-resolution map, and the current surrounding obstacle shapes may be taken into account when setting the first local target position.

[0032] Furthermore, the local route planning unit 306 outputs the first local target position and the first local route to the control command generation unit 310 via the vehicle immobility determination unit 307. Also, if the local route planning unit 306 cannot generate the first local route to the first local target position within a certain time, or if the current position input from the self-position estimation unit 303 does not change for a certain period of time or longer, that is, if the vehicle 101 is unable to move due to an obstacle or the like, it notifies the vehicle immobility determination unit 307 of this fact. After notifying the vehicle immobility determination unit 307, if the local route planning unit 306 receives input from the local target position and the second local target position and second local route described later from the route calculation unit 309, it outputs the second local target position and second local route to the control command generation unit 310 via the vehicle immobility determination unit 307.

[0033] The vehicle immobility determination unit 307 receives the first local target position and the first local route from the local route planning unit 306 and outputs them to the control command generation unit 310. Furthermore, if the vehicle immobility determination unit 307 receives the second local target position and the second local route (described later) from the local target position and route calculation unit 309, it outputs the second local target position and the second local route to the control command generation unit 310. The vehicle immobility determination unit 307 also determines whether the vehicle has entered an immobility state while autonomously traveling along the global route. The processing in the vehicle immobility determination unit 307 will be described in detail in Figure 11 below.

[0034] The high-resolution surrounding map generation unit 308 receives vehicle movement information from the encoder 209, measurement information from the on-board sensor 211, and the global route from the global route planning unit 305. Furthermore, if the vehicle immobility determination unit 307 determines that the vehicle is immobile, the high-resolution surrounding map generation unit 308 generates a high-resolution surrounding map (see Figure 10 below) with a higher resolution than the normal-resolution map. In this case, to maximize the performance of the on-board sensor 211 (horizontal resolution of the laser light, measurement range, measurement error, etc.), the high-resolution surrounding map generation unit 308 sets a higher resolution than the normal-resolution map. The vehicle 101 is then driven to perform translation and turning movements, and the high-resolution surrounding map generation unit 308 generates a high-resolution surrounding map of the surroundings in accordance with the vehicle 101's translation and turning movements. The high-resolution surrounding map includes a portion of the global route generated by the global route planning unit 305. Furthermore, the high-resolution surrounding map generation unit 308 transmits the generated high-resolution surrounding map to the local target position and path calculation unit 309 and the self-position estimation unit 303.

[0035] The local target position and route calculation unit 309 is an example of a second local route planning unit and receives a high-resolution surrounding map from the high-resolution surrounding map generation unit 308. Furthermore, if the driving impossibility determination unit 307 determines that the vehicle is in a driving impossibility state, the local target position and route calculation unit 309 sets a second local target position on the high-resolution surrounding map along the global route and plans a second local route from the vehicle's current position to the second local target position. For example, the local target position and route calculation unit 309 sets the endpoint of the high-resolution surrounding map on the vehicle 101's target position side as the second local target position and plans a second local route (see Figure 10 below). The local target position and route calculation unit 309 also outputs the generated second local target position and second local route to the local route planning unit 306. Furthermore, the second local target location is not limited to the endpoint of the high-resolution surrounding map on the target location side, but may be set to a position coordinate on the global route at a predetermined distance from the current location in the high-resolution surrounding map, or it may be dynamically changed according to the vehicle speed.

[0036] The control command generation unit 310 receives the first local target position and first local route, as well as the second local target position and second local route, from the local route planning unit 306 via the driveability determination unit 307. The control command generation unit 310 also receives movement information of the vehicle 101 from the encoder 209 and the current position from the self-position estimation unit 303. When the first local target position and first local route are received from the local route planning unit 306, the control command generation unit 310 generates a drive command for the vehicle 101 using a route-following algorithm so that it travels from the current position to the first local target position and along the first local route. When the second local target position and second local route are received from the local route planning unit 306, the control command generation unit 310 generates a drive command for the vehicle 101 using a route-following algorithm so that it travels from the current position to the second local target position and along the second local route. Furthermore, the control command generation unit 310 outputs the generated drive command for the vehicle 101 to the drive unit 301. The path-following algorithm may include, for example, the Pure Pursuit method or the DWA (Dynamic Window Approach) method.

[0037] [Example hardware configuration for an in-vehicle controller] Next, the hardware configuration of the on-board controller 220 of the autonomous driving system 300 will be described. Figure 4 is a block diagram showing an example of the hardware configuration of the on-board controller 220 of the autonomous driving system 300 according to this embodiment. The on-board controller 220 performs various calculations to realize autonomous driving of the vehicle 101.

[0038] As shown in Figure 4, the in-vehicle controller 220 includes a CPU (Central Processing Unit) 22a, a ROM (Read Only Memory) 22b, a RAM (Random Access Memory) 22c, a storage device 22d, and an input / output interface 22e. The CPU 22a, ROM 22b, RAM 22c, storage device 22d, and input / output interface 22e are connected by a bus 22f to enable them to send and receive information data from one another.

[0039] The CPU 22a controls the operation of each part within the in-vehicle controller 220. For example, the CPU 22a controls the operation of each functional component of the in-vehicle controller 220 shown in Figure 3. Specifically, it controls processes such as driving route planning, self-position estimation, generation of high-resolution surrounding maps, and generation of drive commands for the vehicle 101. Note that a GPU (Graphics Processing Unit) may be used instead of the CPU 22a, or the CPU 22a and GPU (Graphics Processing Unit) may be used in combination.

[0040] ROM22b is composed of a storage medium such as non-volatile memory and stores programs and data that the CPU22a executes and references.

[0041] RAM22c is composed of a storage medium such as volatile memory, and temporarily stores information (data) necessary for each process performed by CPU22a.

[0042] The storage device 22d is a computer-readable, non-transient recording medium that stores programs executed by the CPU 22a, and is composed of a storage device such as an HDD (Hard Disk Drive). The storage device 22d stores programs for the CPU 22a to control various parts, an OS (Operating System), controllers, and other programs and data. The storage device 22d also stores normal-resolution maps, high-resolution surrounding maps, etc. Some of the programs and data stored in the storage device 22d may be stored in ROM 22b. Furthermore, the computer-readable, non-transient recording medium that stores programs executed by the CPU 22a is not limited to an HDD, but may be a recording medium such as an SSD (Solid State Drive), CD (Compact Disc)-ROM, or DVD (Digital Versatile Disc)-ROM.

[0043] The input / output interface 22e transmits and receives information data to and from the external environment of the in-vehicle controller 220 under the control of the CPU 22a.

[0044] Next, we will explain the driving environment assumed in the autonomous driving process described later in the autonomous driving system 300. Figure 5 is a conceptual diagram representing a part of the driving environment 100 shown in Figure 1. In Figure 5, the "X" and "Y" directions are the same as those shown in Figure 1, so a redundant explanation is omitted. The "Z" direction shown in Figure 5 is perpendicular to the [X,Y] plane and indicates the height direction perpendicular to the ground in the driving environment 100. The target position G501 shown in Figure 5 indicates the target position of the vehicle 101 in the driving environment 100. Figure 6 is a conceptual diagram showing an example of how a pallet 801 (obstacle) is newly installed in the driving environment 100 shown in Figure 5. As shown in Figure 6, the pallet 801 blocks a part of the passage between the vehicle 101 and the target position G501.

[0045] Here, for the driving environment 100 shown in Figure 5, it is assumed that the normal resolution map created in advance by the normal resolution map generation unit 302 and the global route generated by the global route planning unit 305 are the normal resolution map M700 and the global route 701 shown in Figure 7, respectively. Figure 7 is a conceptual diagram showing the normal resolution map and the global route to the target position according to this embodiment. In Figure 7, the "X" and "Y" directions are the same as those shown in Figure 1, so a redundant explanation is omitted. In Figure 7, the black grid represents obstacles. The hatched grid represents the global route 701. As shown in Figure 7, by referring to the normal resolution map M700, the vehicle 101 can reach the target position G501 by driving according to the global route 701. Note that in Figure 7, the obstacles represented by the black grid correspond to the actual size of the obstacles.

[0046] Figure 8 is a conceptual diagram showing the appearance of an inflated obstacle and the global route 701 in the normal-resolution map M700 according to this embodiment. In Figure 8, the "X" and "Y" directions are the same as those shown in Figure 1, so a redundant explanation is omitted. In Figure 8, the portion of the obstacle that is inflated beyond its actual size is shown by a gray grid, which represents the outline of the obstacle (black portion) detected by the on-board sensor 211. LG502 shown in Figure 8 indicates the first local target position set by the local route planning unit 306. The first local route 702 planned by the local route planning unit 306 is shown by a solid black line connecting the vehicle 101 and the first local target position LG502.

[0047] Figure 9 shows the expansion of the obstacle, including the pallet 801 detected by the on-board sensor 211, and the global route 711 in the normal resolution map M700 when the pallet 801 is installed (see Figure 6). In Figure 9, the "X" and "Y" directions are the same as those shown in Figure 1, so a redundant explanation is omitted. Since the pallet 801 is a newly installed obstacle, it becomes an unknown obstacle. In this case, the global route planning unit 305 plans and generates the global route 711 shown in Figure 8. The vehicle 101 can travel according to the global route 711 to the grid 902 in front of the contour of the pallet 801 detected by the on-board sensor 211. However, according to the global route 701, the distance between the contours of the obstacle, i.e., the width of the route, from the grid one grid beyond grid 902 (the grid one grid above grid 902 in the "Y" direction in the drawing), becomes smaller than the width of the vehicle 101. Therefore, the local route planning unit 306 is unable to plan a local route that proceeds beyond grid 902 to G501 according to the global route 701. In other words, vehicle 101 becomes unable to move at grid 902.

[0048] Figure 10 is a conceptual diagram showing an example of a high-resolution surrounding map of the vehicle according to the first embodiment of the present invention. Here, the high-resolution surrounding map generation unit 308 is assumed to generate a high-resolution grid map of 1 cm / pixel, taking into account the performance of the on-board sensor 211. The normal resolution map M700 shown in Figure 10 is a scaled-down version of the normal resolution map M700 shown in Figure 9, so a redundant explanation is omitted. The area enclosed by the dashed line on the normal resolution map M700 is the maximum range that the on-board sensor 211 can detect, centered on the position where the vehicle 101 became immobile (grid 902 shown in Figure 9).

[0049] The high-resolution surrounding map M710 shown in Figure 10 is a 1cm / pixel high-resolution surrounding map generated by the high-resolution surrounding map generation unit 308 for the area enclosed by the dashed line on the normal-resolution map M700. The hatched grid on the high-resolution surrounding map M710 includes a portion of the global route 711 shown on the normal-resolution map M700. LG1002 is a second local target location set by the local target location and route calculation unit 309 based on the high-resolution surrounding map M710. The local route 1003 generated by the local target location and route calculation unit 309 based on the high-resolution surrounding map M710 is shown by a solid black line connecting the vehicle 101 and the second local target location LG1002. As shown in Figure 10, since the grid constituting the high-resolution surrounding map M710 is finer than that of the normal-resolution map M700, the shapes of obstacles such as the pallet 801 are represented in more detail. Therefore, the local target position and route calculation unit 309 can refer to the high-resolution surrounding map M710 to plan a local route 1003 for the vehicle 101 to the second local target position LG1002, and restore the vehicle 101, which has become immobile, to a drivable state.

[0050] [Procedure for autonomous driving in an autonomous driving system] Next, the procedure for autonomous driving in the autonomous driving system 300 will be described. Figure 11 is a flowchart showing the procedure for autonomous driving in the autonomous driving system 300 according to this embodiment. The process described below starts when the vehicle 101 begins autonomous driving.

[0051] First, the target position input unit 304 inputs the destination of the task currently being performed by the vehicle 101, and sets the coordinates of that destination on a normal resolution map as the target position for autonomous driving (step S11).

[0052] Next, the global route planning unit 305 plans and generates a global route (see Figure 7) from the starting position to the target position on a normal resolution map, based on the current position of the vehicle 101 (starting position) and the set target position (step S12).

[0053] Next, the local route planning unit 306 sets a first local target position using the measurement information from the on-board sensor 211, the current position, and the global route, and plans and generates a first local route to the first local target position (see Figure 8) (step S13). The local route planning unit 306 also outputs the first local target position and the first local route to the control command generation unit 310 via the vehicle impassability determination unit 307.

[0054] Next, the control command generation unit 310 generates a drive command for the vehicle 101 using a path-following algorithm so that the vehicle 101 travels along a first local path from its current position toward a first local target position (step S14). The control command generation unit 310 also outputs the generated drive command for the vehicle 101 to the drive unit 301. The drive unit 301 drives the travel motor 108 according to the drive command, causing the vehicle 101 to autonomously travel along the first local path.

[0055] Next, the vehicle inoperability determination unit 307 determines whether or not the vehicle 101 has reached the target position (step S15). In this process, the vehicle inoperability determination unit 307 compares the current position with the target position, and if the distance between the two points is less than or equal to a predetermined threshold, for example, 0.1m, it determines that the target position has been reached, and step S15 becomes a YES determination. Also, the vehicle inoperability determination unit 307 compares the current position with the target position, and if the distance between the two points exceeds the predetermined threshold, it determines that the target position has not been reached, and step S15 becomes a NO determination. Note that the predetermined threshold set for the distance between the current position and the target position is not limited to 0.1m and may be changed to an appropriate value depending on the actual situation.

[0056] In step S15, if the vehicle non-movement determination unit 307 determines that the vehicle 101 has reached the target position (YES determination in step S15), the autonomous driving process ends in a completed driving state.

[0057] On the other hand, in the process of step S15, if the vehicle 101 has not reached the target position, the vehicle 101 determination unit 307 determines whether the vehicle 101 is in a state where it cannot be driven (step S16). In this process, if the local route planning unit 306 cannot generate the first local route within a certain time, or if the position of the vehicle 101 has not changed for a certain time or longer, the vehicle 101 determination unit 307 determines that the vehicle 101 is in a state where it cannot be driven, and step S16 becomes a YES determination. If the vehicle 101 is not in a state where it cannot be driven, the vehicle 101 determination unit 307 becomes a NO determination.

[0058] In step S16, if the vehicle non-operational determination unit 307 determines that the vehicle 101 is not in an inoperable state (NO determination in step S16), the process returns to step S12, and steps S12 to S16 are repeatedly executed.

[0059] On the other hand, in the process of step S16, if the vehicle inoperability determination unit 307 determines that the vehicle 101 is in an inoperable state (YES determination in step S16), the high-resolution surrounding map generation unit 308 clears the reference relationship with the normal-resolution map (step S17). Here, clearing means temporarily removing the reference relationship with the normal-resolution map, but the normal-resolution map itself is not deleted. In addition, in this process, the high-resolution surrounding map generation unit 308 saves the current position and orientation of the vehicle 101 relative to the coordinate system of the normal-resolution map. That is, it inherits the coordinate system of the normal-resolution map. This is to match the coordinate system of the high-resolution surrounding map created by the high-resolution surrounding map generation unit 308 with the coordinate system of the normal-resolution map as closely as possible.

[0060] Next, the high-resolution surrounding map generation unit 308 changes the map resolution from the normal resolution (5 cm / pixel) to the high resolution (1 cm / pixel) (step S18).

[0061] Next, the high-resolution surrounding map generation unit 308 sets up a map creation operation to create a high-resolution surrounding map of the vehicle's surroundings (step S19). In this process, the high-resolution surrounding map generation unit 308 transmits a command to the control command generation unit 310 to set up the map creation operation. The control command generation unit 310 generates a drive command as the map creation operation, which causes the vehicle 101 to turn twice (720 degrees) in place while maintaining an angular velocity of, for example, 0.3 rad / s, and outputs it to the drive unit 301. The drive unit 301 then drives the travel motor 108 to move the vehicle 101 according to the drive command for the map creation operation from the control command generation unit 310.

[0062] Next, the high-resolution surrounding map generation unit 308 generates a high-resolution surrounding map of the surroundings based on information acquired by the on-board sensor 211 when the vehicle 101 is moving, turning, or performing other actions (step S20).

[0063] Next, the high-resolution surrounding map generation unit 308 determines whether the map generation operation of the vehicle 101 is complete, that is, whether the creation of the high-resolution surrounding map is complete (step S21). In this process, the high-resolution surrounding map generation unit 308 determines whether the turning operation of the vehicle 101 is complete, for example, by using the value obtained by integrating the angular velocity obtained from the encoder 209 with respect to the travel time. If the high-resolution surrounding map generation unit 308 determines that the turning operation is complete, it determines that the map generation operation is complete, and step S21 is a YES determination. If the high-resolution surrounding map generation unit 308 determines that the turning operation is not complete, it determines that the map generation operation is not complete, and step S21 is a NO determination. Normally, high-resolution maps are created by manually operating the vehicle 101 and driving it over the entire driving environment. Therefore, the determination of the completion of map creation is left to the operator. However, in this embodiment, since the high-resolution surrounding map is generated while the vehicle 101 is operating itself, it is necessary to determine and register the conditions for completion of the map creation operation in advance. Furthermore, if it is desired to determine the map generation operation based on more accurate values, the high-resolution surrounding map generation unit 308 may also perform self-position estimation based on a normal-resolution map in parallel and determine whether the map generation operation is complete while monitoring changes in the position and attitude of the vehicle 101.

[0064] In step S21, if the high-resolution surrounding map generation unit 308 determines that the map generation operation is not complete (NO determination in step S21), it returns to step S20 and repeats the process from step S20 to step S21.

[0065] On the other hand, in the process of step S21, if the high-resolution surrounding map generation unit 308 determines that the map generation operation is complete (YES determination in step S21), it sets the generated high-resolution surrounding map (see Figure 10) as the reference map (step S22).

[0066] Next, the local target position and route calculation unit 309 sets a second local target position, which is a local target position in the high-resolution surrounding map (step S23). In this process, the local target position and route calculation unit 309 sets, for example, the endpoint of the high-resolution surrounding map on the autonomous driving target position side as the second local target position.

[0067] Next, the local target position and route calculation unit 309 generates a second local route, which is the local route from the vehicle's current position to the second local target position (step S24). In this process, the local target position and route calculation unit 309 outputs the second local target position and the generated second local route to the local route planning unit 306. The local route planning unit 306 then outputs the second local target position and the second local route to the control command generation unit 310 via the vehicle impassability determination unit 307.

[0068] Next, the control command generation unit 310 generates a drive command for the vehicle 101 using a path-following algorithm so that the vehicle 101 travels along a second local path from its current position toward a second local target position (step S25). The control command generation unit 310 also outputs the generated drive command for the vehicle 101 to the drive unit 301. The drive unit 301 drives the travel motor 108 according to the drive command to make the vehicle 101 travel along the second local path.

[0069] Next, the vehicle inoperability determination unit 307 determines whether the vehicle 101 has reached the second local target location (step S26). In this process, the vehicle inoperability determination unit 307 compares the current location with the second local target location, and if the distance between the two points is less than or equal to a predetermined threshold, it determines that the vehicle has reached the second local target location, and step S26 becomes a YES determination. Conversely, the vehicle inoperability determination unit 307 compares the current location with the second local target location, and if the distance between the two points exceeds a predetermined threshold, it determines that the vehicle has not reached the second local target location, and step S26 becomes a NO determination. Here, it is assumed that the endpoint of the high-resolution surrounding map closest to the target location is the second local target location.

[0070] In step S26, if the vehicle inoperability determination unit 307 determines that the vehicle 101 has reached the second local target position (YES determination in step S26), the high-resolution surrounding map generation unit 308 resets the normal resolution map as the reference map (step S27). In this process, the high-resolution surrounding map generation unit 308 clears the reference relationship with the high-resolution surrounding map, but the high-resolution surrounding map itself is not deleted. After the processing in step S27, the autonomous driving process returns to step S12, and the processes from step S12 to step S26 are repeated.

[0071] On the other hand, in the process of step S26, if the vehicle 101 has not reached the second local target position (NO determination in step S26), the vehicle 101

[0072] In step S28, if the drivability determination unit 307 determines that the drivability condition has not continued for a specified time or longer (NO determination in step S28), it returns to step S24, and steps S24 to S28 are executed repeatedly.

[0073] On the other hand, in the process of step S28, if the vehicle inoperability determination unit 307 determines that the vehicle inoperability state has continued for a specified time or longer (YES determination in step S28), the autonomous driving process ends in a state where the vehicle cannot reach the target position.

[0074] [effect] As described above, when the vehicle becomes immobile, the autonomous driving system 300 according to this embodiment creates a high-resolution surrounding map by making maximum use of the performance of the on-board sensor 211 (horizontal resolution of the laser light, distance measurement range, distance measurement error, etc.). Since the grid that makes up the high-resolution surrounding map is finer than that of a normal-resolution map, the shape of obstacles in front of the vehicle can be represented with more detail. For this reason, the autonomous driving system 300 can make maximum use of the performance of the on-board sensor and refer to the high-resolution surrounding map to plan a more precise route and improve the possibility of recovering the vehicle from the immobile state.

[0075] <Second Embodiment> Here, an autonomous driving system 2000 according to a second embodiment of the present invention will be described. Figure 12 is a block diagram showing an example of the functional configuration of the autonomous driving system 2000 according to this embodiment. In addition to the components of the autonomous driving system 300 shown in Figure 3, the autonomous driving system 2000 further includes a control server 2001 and a map data transmission unit 2002 installed in the on-board controller 220, as shown in Figure 12. The control server 2001 is installed outside the on-board controller 220 and connected to the on-board controller 220. Note that in Figure 12, the components and their functions other than the normal resolution map generation unit 302, the high resolution surrounding map generation unit 308, and the map data transmission unit 2002 of the on-board controller 220 are the same as the components and their functions shown in Figure 3, so their illustration and description are omitted.

[0076] The map data transmission unit 2002 is connected to the output side of the high-resolution surrounding map generation unit 308. The map data transmission unit 2002 receives the high-resolution surrounding map created by the high-resolution surrounding map generation unit 308 and transmits it to the control server 2001's later-described area map update unit 2003. The timing for transmitting the high-resolution surrounding map to the area map update unit 2003 is set, for example, to the timing after the vehicle 101 has become immobile and recovered from that state. Specifically, for example, the high-resolution surrounding map is transmitted to the control server 2001's later-described area map update unit 2003 after the processing of step S27 of the autonomous driving process shown in Figure 11. Alternatively, the timing for transmitting the high-resolution surrounding map to the area map update unit 2003 may be after the high-resolution surrounding map generation unit 308 has created the high-resolution surrounding map, for example, after the processing of step S22 of the autonomous driving process shown in Figure 11.

[0077] As shown in Figure 12, the control server 2001 has a full-area map update unit 2003 and a map distribution unit 2004. The full-area map update unit 2003 is connected to the map distribution unit 2004 and is also connected to the map data transmission unit 2002 of the in-vehicle controller 220 so that it can receive information data. In addition, the map distribution unit 2004 is also connected to the normal-resolution map generation unit 302 of the in-vehicle controller 220 so that it can transmit information data.

[0078] The area-wide map update unit 2003 is an example of a normal-resolution map update unit and manages normal-resolution maps. The area-wide map update unit 2003 also receives a high-resolution surrounding map from the map data transmission unit 2002 of the vehicle-mounted controller 220 of the vehicle 101, converts the received high-resolution surrounding map to the resolution of a normal-resolution map, and updates the normal-resolution map with the converted high-resolution surrounding map.

[0079] Specifically, the full-area map update unit 2003 updates the area of ​​the normal-resolution map corresponding to the high-resolution surrounding map to a high-resolution surrounding map (grid information) in accordance with the position and orientation of the vehicle 101. That is, each time a high-resolution surrounding map is generated, the portion of the normal-resolution map corresponding to the generated high-resolution surrounding map is updated. As described in the first embodiment, when the high-resolution surrounding map generation unit 308 generates a high-resolution surrounding map, it saves the current position and orientation of the vehicle 101 relative to the coordinate system of the normal-resolution map. When the map data transmission unit 2002 transmits the high-resolution surrounding map to the full-area map update unit 2003, it simultaneously transmits the saved position and orientation of the vehicle 101. In this embodiment, when converting the high-resolution surrounding map to the resolution of the normal-resolution map, the 1cm / pixel high-resolution surrounding map is downsampled (compressed) to 1 / 5 of its size by image processing to match the resolution of the full-area map (a normal-resolution map with 5cm / pixel).

[0080] The map distribution unit 2004 transmits the normal resolution map, i.e., the full-area map, updated by the full-area map update unit 2003, to the normal resolution map generation unit 302 in the vehicle's onboard controller 220. The normal resolution map generation unit 302 replaces the current normal resolution map with the latest normal resolution map received at a predetermined timing, for example, when the vehicle is stopped in a task waiting state. If multiple vehicles are present in the driving environment, the map distribution unit 2004 transmits the normal resolution map updated by the full-area map update unit 2003 to all vehicles in the driving environment, rather than transmitting it only to vehicle 101.

[0081] [effect] As described above, the autonomous driving system 2000 according to this embodiment updates the entire map of the same driving environment that all vehicles refer to, using a high-resolution surrounding map created by the high-resolution surrounding map generation unit 308 of the onboard controller 220 of the vehicle that has become inoperable. Therefore, since the autonomous driving system 2000 according to this embodiment updates the entire map of the area around the location where a vehicle has become inoperable with a high-resolution surrounding map, it is possible to improve the likelihood of preventing other vehicles from becoming inoperable at that location when passing through it.

[0082] <Third Embodiment> Here, an autonomous driving system 3000 according to a third embodiment of the present invention will be described. Figure 13 is a block diagram showing an example of the functional configuration of the autonomous driving system 3000 according to a third embodiment of the present invention. As shown in Figure 13, the autonomous driving system 3000 comprises a control server 2010, an on-board controller 220, and an on-board controller 220A. The on-board controller 220 and the on-board controller 220A are each connected to the control server 2010. The on-board controller 220A has the same configuration as the on-board controller 220 and is installed in a vehicle different from vehicle 101 (it may be a vehicle of a different type). Although Figure 13 shows an example in which the on-board controllers of two vehicles are connected to the control server 2010, the present invention is not limited to this. The number of vehicles connected to the control server 2010 is the number of autonomous driving vehicles in the same driving environment and can be set to any number.

[0083] As can be seen by comparing Figure 13 and Figure 12, the control server 2010 includes, in addition to the components of the control server 2001 shown in Figure 12, an on-board sensor information management unit 3002 and a target position management unit 3003.

[0084] The on-board sensor information management unit 3002 stores performance information and installation location information for each vehicle's on-board sensor 211. The on-board sensor performance information includes at least one of the following: the maximum irradiation distance of the on-board sensor 211, the horizontal resolution of the laser beam, the distance measurement range, and the distance measurement error. The installation location information is information regarding the installation location of the on-board sensor 211 in each vehicle, for example, the installation position and orientation information of the on-board sensor 211 relative to a predetermined part position on each vehicle (for example, the center of the rear wheel).

[0085] The target position management unit 3003 receives and manages the target position of each vehicle from the target position input unit 304A of the in-vehicle controller 220A and the target position input unit 304B of the in-vehicle controller 220B. Furthermore, if vehicle 101 cannot recover from the inoperable state even after referring to the high-resolution surrounding map, that is, if the inoperable determination unit 307 determines that the target position cannot be reached (YES determination in step S28 of Figure 11), the target position management unit 3003 searches for other vehicles that can travel to the target position of vehicle 101. The target position management unit 3003 receives the driving status information described later from vehicle 101 and, based on the driving status information and the information from the in-vehicle sensor information management unit 3002, searches for other vehicles that are more advanced than vehicle 101, specifically those with finer horizontal resolution of the beam light of the in-vehicle sensor 211, and smaller in size (vehicle width and vehicle length) than vehicle 101. The target position management unit 3003 transmits the target position of vehicle 101 to the other vehicle it has found, and instructs the other vehicle to drive to the target position of vehicle 101 in place of vehicle 101. After the other vehicle receives the target position of vehicle 101, it autonomously drives to that target position as described in the first embodiment (see Figure 11).

[0086] Since the in-vehicle controller 220 and the in-vehicle controller 220A have the same configuration, the following description will use the in-vehicle controller 220 as an example and omit the description of the in-vehicle controller 220A.

[0087] The on-board controller 220 includes, in addition to the components shown in Figure 12, a driving status transmission unit 3001. When the vehicle becomes unable to reach its target position, the driving status transmission unit 3001 transmits the vehicle's driving status information to the target position management unit 3003 of the control server 2010. Here, the driving status information includes information indicating that the target position is unreachable, the vehicle's position and attitude on a normal-resolution map at the time it is determined that the target position is unreachable, and the vehicle's target position.

[0088] Furthermore, the high-resolution surrounding map generated by the high-resolution surrounding map generation unit 308 before the vehicle becomes unable to reach the target position, and when it becomes unable to move, has already been transmitted to the control server 2010's full-area map update unit 2003 (see Figure 12). Therefore, when the target position management unit 3003 searches for another vehicle that can travel to the target position in place of vehicle 101, it refers to the full-area map updated with the high-resolution surrounding map to find another vehicle of a size that can pass through the area where the vehicle became unable to move.

[0089] Next, the vehicle types of the autonomous vehicles according to this embodiment will be described. Figure 14 is a conceptual diagram showing an example of the vehicle types of the autonomous vehicles according to this embodiment. Vehicle 101 shown in Figure 14 is the same as vehicle 101 shown in Figure 2, so a redundant explanation will be omitted. Vehicle 102 shown in Figure 14 is a towing vehicle, and vehicle 103 is a forklift. Vehicles 101 to 103 have different vehicle classes. Furthermore, as shown in Figure 14, the on-board sensors 211, 211A, and 211B of vehicles 101 to 103 are installed in different positions. In this embodiment, it is assumed that the performance of on-board sensors 211, 211A, and 211B are different.

[0090] [effect] As described above, in the autonomous driving system 3000 according to this embodiment, when the control server 2010 receives driving status information from a vehicle that has become unable to reach its target location, it searches for another vehicle that can drive in its place. Specifically, based on the driving status information received from the vehicle that has become unable to reach its target location and the information from the on-board sensor information management unit 3002, the control server 2010 searches for another vehicle that has higher performance on-board sensors than the vehicle in question and is of a vehicle size that can pass through the area where the vehicle became unable to drive. Therefore, the autonomous driving system 3000 according to this embodiment can improve the operational efficiency of the vehicles on site by taking into account the performance of the on-board sensors of other vehicles and grasping the cause of the vehicle becoming unable to drive (changes in surrounding obstacles, vehicle size, etc.), and by utilizing other vehicles that can reach the target location.

[0091] It should be noted that the present invention is not limited to the embodiments described above, and various other applications and modifications can be taken as long as they do not deviate from the gist of the present invention as described in the claims. For example, the embodiments described above are detailed and specific explanations of the configuration of an autonomous driving system in order to clearly illustrate the present invention, and are not necessarily limited to having all the configurations described. Furthermore, it is possible to replace some of the configurations of the embodiments described here with the configurations of other embodiments, and it is also possible to add configurations from other embodiments to the configuration of one embodiment. In addition, it is possible to add, delete, or replace some of the configurations of each embodiment with other configurations. Furthermore, the control lines and information lines shown are those deemed necessary for explanatory purposes, and not all control lines and information lines are necessarily shown in the actual product. In reality, it is safe to assume that almost all components are interconnected. [Explanation of Symbols]

[0092] 101, 102, 103…Vehicle, 206…Drive wheels, 208…Traction motor, 209…Encoder, 211, 211B, 211C…On-board sensors, 220, 220A…On-board controller, 300, 2000, 3000…Autonomous driving system, 301…Drive unit, 302, 302A…Normal resolution map generation unit, 303…Self-position estimation unit, 304, 304A…Target position input unit, 305…Global route planning unit, 3 06...Local route planning unit, 307...Drivability determination unit, 308...High-resolution surrounding map generation unit, 309...Local target position and route calculation unit, 310...Control command generation unit, 2001,2010...Control server, 2002,2002A...Map data transmission unit, 2003...Whole-area map update unit, 2004...Map distribution unit, 3001,3001A...Driving status transmission unit, 3002...On-board sensor information management unit, 3003...Target position management unit

Claims

1. A global route planning unit plans the overall route from the vehicle's starting position to the target position by referring to a pre-created normal-resolution map of the driving environment. A vehicle determination unit that determines whether or not it has entered a state where it cannot move while autonomously traveling along the overall route, If the vehicle is determined to be in a state of being unable to move by the vehicle determination unit, the high-resolution surrounding map generation unit generates a high-resolution surrounding map with a higher resolution than the normal resolution map, A normal resolution map update unit manages the normal resolution map, receives the high-resolution surrounding map from the vehicle, converts the received high-resolution surrounding map to the resolution of the normal resolution map, and updates the normal resolution map with the converted high-resolution surrounding map. A map distribution unit transmits the normal resolution map updated by the normal resolution map update unit to a vehicle in the driving environment. An on-board sensor information management unit manages performance information of on-board sensors in the aforementioned vehicle, The system includes a target position management unit that manages the target position of the vehicle, and when the vehicle is determined to be unable to reach the target position by the vehicle inability determination unit, it receives from the vehicle status information including the inability to reach the target position, the vehicle's position and orientation on the normal resolution map at the time the inability to reach the target position was determined, and the vehicle's target position. Based on the vehicle status information and the information from the on-board sensor information management unit, the system searches for another vehicle that can travel to the vehicle's target position in place of the vehicle, and transmits the vehicle's target position to the other vehicle that was found. Autonomous driving system.

2. The system includes a first local route planning unit that, by referring to the aforementioned normal resolution map, sets a first local target position located between the current position of the vehicle and the target position along the global route, and plans a first local route from the current position of the vehicle to the local target position. The autonomous driving system according to claim 1.

3. If the vehicle is determined to be in a state where it cannot be driven by the vehicle determination unit, the vehicle is provided with a second local route planning unit that sets a second local target position on the high-resolution surrounding map along the global route and plans a second local route from the vehicle's current position to the second local target position. The autonomous driving system according to claim 2.

4. The performance information of the in-vehicle sensor includes at least one of the following: the maximum irradiation distance of the in-vehicle sensor, the horizontal resolution of the laser beam, the distance measurement range, and the distance measurement error. An autonomous driving system according to any one of claims 1 to 3.

5. The process involves planning the overall route from the vehicle's starting position to the target position by referring to a pre-created normal-resolution map of the driving environment, and The steps include determining whether the vehicle has become unable to move while autonomously driving along the aforementioned general route, If it is determined that the vehicle itself has become inoperable, the steps include generating a high-resolution surrounding map with a higher resolution than the normal resolution map, The steps include: managing the standard resolution map, receiving the high-resolution surrounding map from the vehicle, converting the received high-resolution surrounding map to the resolution of the standard resolution map, and updating the standard resolution map with the converted high-resolution surrounding map; The steps include transmitting the updated normal resolution map to a vehicle in the driving environment, The steps include managing performance information of the on-board sensors of the vehicle, The process includes managing the target position of the vehicle, and if it is determined that the vehicle cannot reach the target position, receiving from the vehicle status information that the target position cannot be reached, the vehicle's position and orientation on the normal resolution map at the time the target position cannot be reached, and driving status information including the vehicle's target position; searching for another vehicle that can travel to the vehicle's target position in its place based on the driving status information and the performance information of the on-board sensors; and transmitting the vehicle's target position to the other vehicle that was found. Autonomous driving methods.

Images