Mobile vehicle driving control system

Abstract

The present invention aims to provide a mobile vehicle travel control system that can stably control the movement of a mobile vehicle regardless of the area in which the mobile vehicle is located. [Solution] The movement of the mobile body is controlled by classifying the area in which the mobile body is located and switching between an infrastructure-based mode in which the management device performs self-position estimation processing and formation generation processing according to the area, and a standalone mode in which the mobile body performs self-position estimation processing and formation generation processing. Alternatively, the movement of the mobile body is controlled by classifying the area in which the mobile body is located in the management device and switching between an infrastructure-based mode and a standalone mode according to the area.

Description

Technical Field

[0001] The present invention relates to a mobile body travel control system that controls the travel of a mobile body moving in physical space.

Background Art

[0002] In recent years, rapid urbanization has increased the demand for public transportation. The global micromobility market size was valued at $40.11 billion in 2022, and the global micromobility market size is expected to grow. Initiatives for micromobility are expected to increase for three reasons. First, there is a labor shortage. With the expansion of the e-commerce industry, the shortage of drivers and delivery personnel has become increasingly severe. Second, we are facing an aging society. Due to physical limitations and driving difficulties, the elderly have limited mobility options. Finally, there is a need for carbon neutrality. To mitigate climate change, reducing greenhouse gas emissions and achieving carbon neutrality are indispensable. Given such a social context, there is an urgent need to develop a sustainable and efficient urban transportation system. In particular, advanced technologies are required to address challenges such as labor shortages, aging populations, and environmental issues. Among these technologies, digital twin technology has attracted attention, and it has the potential to greatly improve the safety and efficiency of micromobility solutions.

[0003] A digital twin is a model that is a counterpart to the real world, created in digital space. The benefits that can be expected from a digital twin are threefold: estimation of object behavior, accident prediction, and analysis for system optimization. Furthermore, one initiative to improve mobility safety using digital twins is the LiDAR (Light Detection and Ranging) sensor network. LiDAR sensors are gaining popularity, and discussions are ongoing regarding the construction of LiDAR sensor networks in the environment (see, for example, Non-Patent Document 1). Since in-vehicle LiDAR is often obscured by obstacles, there is ongoing discussion about outsourcing the vehicle's sensing functions to an in-environment LiDAR sensor network (see Non-Patent Documents 1 and 2). In addition, there is edge computing. Because in-vehicle computers have limited processing power, there is currently discussion about outsourcing the functions of in-vehicle computers to edge computers (see Non-Patent Document 3). [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Kuon Akiyama, Ryoichi Shinkuma, Chotaro Yamamoto, Mai Saito, Toshio Ito, Koichi Nihei, and Takanori Iwai. Edge computing system with multi-lidar sensor network for robustness of autonomous personal mobility. In 2022 IEEE 42nd International Conference on Distributed Computing Systems Workshops (ICDCSW), pp. 290-295. IEEE, 2022. [Non-Patent Document 2] Masaki Wago, Kuon Akiyama, Ryoichi Shinkuma, Gabriele Trovato, Koichi Nihei, and Takanori Iwai. Prototype of edge sensing and computing system with multi-lidar network for autonomous micro-mobility. In 2023 IEEE 20th Consumer Communications & Networking Conference (CCNC), pp. 927-928. IEEE, 2023. [Non-Patent Document 3] Ziran Wang, Xishun Liao, Xuanpeng Zhao, Kyungtae Han, Prashant Tiwari, Matthew J Barth, and Guoyuan Wu. A digital twin paradigm:Vehicle-to-cloud based advanced driver assistance systems. In 2020 IEEE 91st Vehicular Technology Conference (VTC2020-Spring), pp. 1-6. IEEE, 2020. [Overview of the project] [Problems that the invention aims to solve]

[0005] However, in conventional systems, autonomous driving of vehicles was only considered within the area where the vehicle and the edge computer could communicate, or within the area where the LiDAR sensor could detect the vehicle.

[0006] This invention has been made in view of the above-mentioned technical background, and aims to provide a mobile vehicle travel control system that can stably control the movement of a mobile vehicle regardless of the area in which the mobile vehicle is located. [Means for solving the problem]

[0007] To achieve the above objective, the present invention provides a mobile body travel control system comprising: one or more sensor devices arranged on the environmental side of a predetermined real space divided into multiple areas; a management device connected to the sensor devices in a communicative manner; and one or more mobile bodies connected to the management device in a communicative manner and traveling in the predetermined real space, wherein the sensor device comprises a first sensor unit for acquiring first sensor data in the predetermined real space, and a sensor-side transmission unit for transmitting the first sensor data in the predetermined real space acquired by the first sensor unit to the management device, the management device comprises an aggregation unit for aggregating the first sensor data transmitted from the sensor devices, a map generation unit for generating first map data of the real space based on the first sensor data aggregated by the aggregation unit, a first self-position estimation unit for estimating the self-position of the mobile body based on the first map data generated by the map generation unit and the first sensor data acquired by the sensor devices or the second sensor data acquired by the mobile body, and information regarding the self-position of the mobile body estimated by the first self-position estimation unit The mobile body comprises a first path generation unit that generates a path for the mobile body to travel, the mobile body comprises a second sensor unit that acquires second sensor data around the mobile body, a map storage unit that stores second map data of real space, a second self-position estimation unit that estimates the mobile body's own position based on the second map data stored by the map storage unit and the second sensor data acquired by the second sensor unit, a second path generation unit that generates a path for the mobile body to travel based on information regarding the mobile body's own position estimated by the second self-position estimation unit, and a mode switching unit that switches between an infrastructure-based mode in which the management device performs a process to estimate the mobile body's own position and a process to generate a path, and a standalone mode in which the mobile body performs a process to estimate the mobile body's own position and a process to generate a path, the mobile body or the management device further comprises an area classification unit that classifies the area where the mobile body is located based on information regarding each area in a predetermined real space and information regarding the mobile body's own position, and the mode switching unit isThe system is characterized by switching between the infrastructure-based mode and the standalone mode according to the area classified by the area classification unit.

[0008] Furthermore, the system may include a travel control switching unit that switches the control of the mobile body's movement so that, when the system switches to the infrastructure-based mode by the switching unit according to the area classified by the area classification unit, the mobile body travels along the route generated by the first route generation unit of the management device, while when the system switches to the standalone mode by the switching unit, the mobile body travels along the route generated by the second route generation unit of the mobile body.

[0009] Furthermore, the area classification unit may be provided on the mobile body and classify the area in which the mobile body is located based on information about the area in real space and information about the mobile body's own position.

[0010] Furthermore, the management device may include an area information transmission unit that transmits information about a predetermined area in real space to the mobile body, the mobile body may include an area information receiving unit that receives information about a predetermined area in real space transmitted from the management device, and the area classification unit may classify the area in which the mobile body is located based on the information about the area in real space received by the area information receiving unit and information about the mobile body's own position.

[0011] Furthermore, the area classification unit may be provided in the management device and classify the area in which the mobile body is located based on information regarding the area in real space and information regarding the self-position of the mobile body.

[0012] Furthermore, the mobile body may include a mobile body information transmission unit that transmits information about the mobile body's own position, and the management device may include a mobile body information receiving unit that receives information about the mobile body's position transmitted from the mobile body, and the area classification unit may classify the area in which the mobile body is located based on information about an area in a predetermined physical space and information about the mobile body's own position received by the mobile body information receiving unit.

[0013] Furthermore, the information relating to an area in a predetermined physical space includes at least an area (Rc) where the management device and the mobile body can communicate but the management device does not detect the mobile body; an area (Rs) where the management device detects the mobile body but the management device and the mobile body cannot communicate; and an area (Rc and Rs) where the management device and the mobile body can communicate and the management device detects the mobile body. The mode switching unit may switch to the standalone mode when the area classification unit classifies the area where the mobile body is located as either area (Rc) or area (Rs), while switching to the infrastructure-based mode when the area classification unit classifies the area where the mobile body is located as either area (Rc and Rs).

[0014] Furthermore, information regarding areas in a predetermined real space includes the area (Rc), the area (Rs), and areas that do not fall under any of the areas (Rc and Rs) (other). The mode estimation unit may switch to the standalone mode if the area classification unit classifies the area where the moving object is located as the area (other).

[0015] Furthermore, the first sensor unit may acquire two-dimensional or three-dimensional first sensor data consisting of a point cloud in a predetermined real space, and the second sensor unit may acquire two-dimensional or three-dimensional first sensor data consisting of a point cloud in a predetermined real space. [Effects of the Invention]

[0016] According to the present invention, since the location for performing self-localization estimation processing and route generation processing of a moving body is switched between the moving body or a management device in accordance with the area where the moving body is located, the traveling of the moving body can be stably controlled regardless of the area where the moving body is located. Moreover, it is possible to simultaneously propose a method of classifying areas of a moving body in the moving body and a method of classifying areas of a moving body in the management device. [BRIEF DESCRIPTION OF THE DRAWINGS]

[0017] [Figure 1] FIG. 1 is an overall configuration diagram showing a moving body travel control system according to an embodiment of the present invention. [Figure 2] FIG. 2 is a schematic diagram showing the situation of sensor devices and a moving body in real space. [Figure 3] FIG. 3 is a diagram showing the configuration of a sensor device. [Figure 4] FIG. 4 is a diagram showing the configuration of a management device and a moving body according to the first embodiment. [Figure 5] FIG. 5 is a schematic diagram showing areas in real space. [Figure 6] FIG. 6 is a flowchart showing self-localization estimation processing and route generation processing in the management device. [Figure 7] FIG. 7 is a flowchart showing self-localization estimation processing and route generation processing in the moving body. [Figure 8] FIG. 8 is a flowchart showing area classification, mode switching, and travel control according to the first embodiment. [Figure 9] FIG. 9 is a diagram showing the configuration of a management device and a moving body according to a second embodiment. [Figure 10] FIG. 10 is a flowchart showing area classification, mode switching, and travel control according to the second embodiment. [Figure 11] FIG. 11 is a diagram showing an experimental site of the present example. [Figure 12] FIG. 12 is a diagram showing a sequence in an edge trigger method of the present example. [Figure 13] FIG. 13 is a diagram showing a sequence in a vehicle trigger method of the present command. [Figure 14]This figure shows the results of the sequence processing time for this embodiment. [Modes for carrying out the invention]

[0018] <First Embodiment> Next, a first embodiment of the mobile vehicle driving control system according to the present invention (hereinafter referred to as "this system") will be described with reference to Figures 1 to 8.

[0019] [Overall structure] As shown in Figure 1, this system comprises a plurality of sensor devices 1 arranged on the environmental side of a predetermined real space, a management device 2 connected to the sensor devices 1 in a communication-enabled manner, and a plurality of mobile bodies 3 connected to the management device 2 in a communication-enabled manner, and controls the movement of the mobile bodies 3.

[0020] Real space refers to existing, primarily three-dimensional spaces related to social life or the environment, such as roads, streets, buildings, indoors, rivers, and mountains, and is divided into multiple areas as described later. Mobile objects 3 refer to mechanically controllable objects such as automobiles, micromobility vehicles, drones, and transport robots.

[0021] The configurations of the sensor device 1, the management device 2, and the mobile unit 3 will be described in detail below.

[0022] [Configuration of Sensor Device 1] As shown in Figure 2, the sensor device 1 is installed on a support column or the like on the environment side in real space, and as shown in Figure 3, it includes a first sensor unit 11 that acquires first sensor data consisting of a point cloud in real space at predetermined time intervals, and a sensor-side transmission unit 12 that sequentially transmits the first sensor data consisting of a point cloud in real space acquired by the first sensor unit 11 to the management device 2.

[0023] The first sensor unit 11 is a sensor known as LiDAR (light detection and ranging). This LiDAR is a type of sensor that uses laser light, and compared to radio waves, it has a higher density of radiant flux. By scanning and irradiating an object with short-wavelength laser light, it acquires image sensor data consisting of point clouds in real space, and accurately detects not only the distance to the object but also the position and shape of the object.

[0024] Examples of LiDAR systems include those that acquire 360-degree omnidirectional image sensor data by rotating multiple laser emitters, and those that acquire image sensor data by directly irradiating with laser light within a predetermined light irradiation angle range. Furthermore, while increasing the number of laser emitters improves the accuracy of the sensor data, it also increases the cost, so inexpensive LiDAR systems with fewer emitters may be used.

[0025] Furthermore, if sensor data is acquired by a single sensor device 1, there is a risk that a blind spot will occur on the opposite side of the wall or moving object 3 from the perspective of the sensor device 1. Therefore, in this embodiment, by aggregating the sensor data consisting of point clouds acquired by multiple sensor devices 1 as described later, the blind spot area can be reduced. Thus, as shown in Figure 2, it is preferable to configure a sensor network by providing multiple sensor devices 1 (LiDAR) that acquire sensor data consisting of point clouds from different directions in the same real space.

[0026] [Configuration of Control Device 2] As shown in Figure 4, the management device 2 is a so-called edge computer positioned to communicate with each sensor device 1 via wired or wireless means, and includes an aggregation unit 21 that aggregates first sensor data transmitted from the sensor devices 1, a map generation unit 22 that generates first map data of real space, a first self-position estimation unit 23 that estimates the self-position of the mobile body 3, and a first path generation unit 24 that generates a path for the mobile body 3 to travel.

[0027] The aggregation unit 21 aggregates the first two-dimensional or three-dimensional sensor data consisting of point clouds in real space transmitted from each sensor device 1. Specifically, the aggregation unit 21 synthesizes the first sensor data consisting of point clouds in real space transmitted from each sensor device 1 in a time series based on timestamps, aggregates it by aligning it in three-dimensional space, and then saves the first image sensor data consisting of the aggregated point clouds. The aggregation unit 21 may perform other processing, such as smoothing the point cloud of the first sensor data, when aggregating the first sensor data consisting of point clouds in real space.

[0028] The map generation unit 22 generates first map data of real space based on the first sensor data aggregated by the aggregation unit 21. This map data is data relating to the environment side of real space, and may be generated directly from the first sensor data consisting of a point cloud, or it may be generated by converting the first sensor data into other data formats. Furthermore, if the map data includes obstacles that hinder the movement of the moving body 3 in real space, it may be generated with the obstacles included, or it may be generated without the obstacles included.

[0029] The first self-position estimation unit 23 estimates the self-position of the mobile body 3 in real space based on the first map data generated by the map generation unit 22 and the second sensor data acquired by the mobile body 3, which will be described in detail later. In this embodiment, the first self-position estimation unit 23 uses the second sensor data acquired by the mobile body 3 when estimating the self-position of the mobile body, but it may also use the first sensor data acquired by the sensor device 1.

[0030] The first path generation unit 24 generates a path for the mobile body 3 to travel, based on the information regarding the mobile body 3's own position estimated by the first self-position estimation unit 23. This path is the route from the mobile body 3's current position to a predetermined destination or waypoint. Preferably, if the first path generation unit 24 includes any obstacles on the path, it generates a path that avoids such obstacles.

[0031] Furthermore, in this embodiment, the management device 2 includes an area information transmission unit 25 that transmits information regarding areas in a predetermined physical space to the mobile body 3. In this embodiment, this area information is information including location information consisting of the following four areas demarcated in a predetermined physical space: area (Rc), area (Rs), area (Rc and Rs), and area (other), and is stored in advance in the management device 2.

[0032] • Area (Rc): An area where the control device 2 and the mobile device 3 can communicate, but the control device 2 cannot detect the mobile device 3. • Area (Rs): An area where the mobile object 3 is detected by the management device 2, but the management device 2 and the mobile object 3 cannot communicate. • Area (Rc and Rs): The area where the management device 2 and the mobile body 3 can communicate, and where the management device 2 can sense the mobile body 3. • Area (Other): Areas that do not fall under any of the above areas (Rc), (Rs), or (Rc and Rs)

[0033] The area in which the moving object 3 is detected by the management device 2 refers to the area in which the moving object 3 can be detected by the sensor device 1 (LiDAR) connected to the management device 2.

[0034] [Configuration of Mobile Unit 3] As shown in Figure 4, the mobile body 3 includes a second sensor unit 31 for acquiring second sensor data around the mobile body 3, a map storage unit 32 for storing second map data of a predetermined real space, and an onboard computer 33.

[0035] The second sensor unit 31, like the first sensor unit 11, is a so-called LiDAR (light detection and ranging) sensor mounted on the mobile body 3, and acquires two-dimensional or three-dimensional sensor data consisting of a point cloud around the mobile body 3 as it travels through real space.

[0036] The map storage unit 32 stores a second map data of a predetermined real space. This second map data may consist of a point cloud, like the first map data, or it may be in another format. Furthermore, this second map data may be stored in the map storage unit 32 of the mobile unit 3 in advance, or it may be acquired periodically or as needed from an external system and stored in the map storage unit 32.

[0037] The onboard computer 3 includes a second self-position estimation unit 331 for estimating the self-position of the mobile body 3, a second path generation unit 332 for generating a path for the mobile body 3 to travel, an area information receiving unit 333 for receiving information about an area in a predetermined real space, an area classification unit 334 for classifying the area in which the mobile body 3 is located, a mode switching unit 335 for switching between infrastructure-based mode and standalone mode, a driving control switching unit 336 for switching the control of the mobile body 3's movement, a driving control unit 337 for controlling the movement of the mobile body 3, and a hardware control unit 338 for controlling hardware such as wheels.

[0038] The second self-position estimation unit 331 estimates the position of the moving object 3 in real space based on the second map data stored by the map storage unit 32 and the second sensor data acquired by the second sensor unit 31. The second sensor data is also transmitted from the moving object 3 to the management device 2 as appropriate and is used for the estimation of the moving object 3's position by the first self-position estimation unit 23.

[0039] The second path generation unit 332 generates a path for the mobile body 3 to travel on, based on the information regarding the mobile body 3's own position estimated by the second self-position estimation unit 331. This path is the route from the mobile body 3's current position to a predetermined destination or waypoint. Preferably, if the second path generation unit 332 includes any obstacles on the path, it generates a path that avoids such obstacles.

[0040] The area information receiving unit 333 receives information about areas in a predetermined physical space (area (Rc), area (Rs), area (Rc and Rs), area (other)) transmitted from the management device 2.

[0041] The area classification unit 334 classifies the area in which the mobile body 3 is located based on information about a predetermined real space received by the area information receiving unit 333 and information about the self-position of the mobile body 3 estimated by the second self-position estimation unit 331.

[0042] For example, as shown in Figure 5, if the mobile body 3 is located within the range of areas (Rc and Rs) (corresponding to 1 in Figure 5), the area classification unit 334 classifies the area where the mobile body 3 is located as area (Rc and Rs). Also, if the mobile body 3 is located within the range of area (Rc) (corresponding to 2 in Figure 5), the area classification unit 334 classifies the area where the mobile body 3 is located as area (Rc). Also, if the mobile body 3 is located within the range of area (Rs) (corresponding to 3 in Figure 5), the area classification unit 334 classifies the area where the mobile body 3 is located as area (Rs). Furthermore, as shown in Figure 5, if the mobile body 3 is located within the range of area (other) (corresponding to 4 in Figure 5), the area classification unit 334 classifies the area where the mobile body 3 is located as area (other).

[0043] The mode switching unit 335 switches between infrastructure-based mode and standalone mode according to the area classified by the area classification unit 334. Infrastructure-based mode is a mode in which the management device 2 executes a process to estimate the self-position of the mobile body 3 using the first self-position estimation unit 23 (self-position estimation process) and a process to generate a path for the mobile body 3 using the first path generation unit 24 (path generation process). On the other hand, standalone mode is a mode in which the mobile body 3 executes a process to estimate the self-position of the mobile body 3 using the second self-position estimation unit 331 (self-position estimation process) and a process to generate a path for the mobile body 3 using the second path generation unit 332 (path generation process).

[0044] For example, if the area classification unit 334 classifies the area where the mobile body 3 is located as Area (Rc and Rs), the mode switching unit 335 switches to infrastructure-based mode, in which the management device 2 performs self-position estimation and route generation processing for the mobile body 3. Alternatively, if the area classification unit 334 classifies the area where the mobile body 3 is located as Area (Rs), Area (Rc), or Area (Other), the mode switching unit 335 switches to standalone mode, in which the mobile body 3 performs self-position estimation and route generation processing for the mobile body 3.

[0045] The aforementioned travel control switching unit 336 switches the travel control of the mobile body 3 according to the area classified by the area classification unit 334.

[0046] For example, if the mode switching unit 335 switches the driving control switching unit 336 to infrastructure-based mode according to the area classified by the area classification unit 334, it obtains the route of the mobile body 3 from the first route generation unit 24 of the management device 2 and controls the mobile body 3 to travel along that route. On the other hand, if the mode switching unit 335 switches the driving control unit 337 to standalone mode according to the area classified by the area classification unit 334, it obtains the route of the mobile body 3 from the second route generation unit 332 of the mobile body 3 and controls the mobile body 3 to travel along that route.

[0047] The travel control unit 337 controls the movement of the mobile body 3 based on the route selected by the travel control switching unit 336, either in infrastructure-based mode or standalone mode.

[0048] The hardware control unit 338 controls hardware such as wheels based on commands from the travel control unit 337 so that the mobile body 3 travels along a predetermined path.

[0049] [How this system works] Next, we will explain each operation of this system, referring to the following:

[0050] <Self-position estimation and path generation processing in management device 2> The self-position estimation and path generation processes in the management device 2 will be explained below with reference to Figure 6.

[0051] First, in the sensor device 1, the first sensor unit 11 acquires first sensor data consisting of a point cloud in real space at predetermined time intervals (S11).

[0052] The sensor-side transmission unit 12 then sequentially transmits the first sensor data, which consists of a point cloud in real space acquired by the first sensor unit 11, to the management device 2 (S12).

[0053] Next, in the management device 2, the aggregation unit 21 aggregates the first sensor data consisting of point clouds in real space transmitted from each sensor device 1 (S13).

[0054] Then, the map generation unit 22 generates first map data of real space based on the first sensor data aggregated by the aggregation unit 21 (S14).

[0055] Then, the first self-position estimation unit 23 estimates the self-position of the mobile body 3 in real space based on the first map data generated by the map generation unit 22 and the second sensor data consisting of a point cloud acquired by the mobile body 3 (S15).

[0056] Then, the first path generation unit 24 generates a path for the mobile body 3 to travel on, based on the information regarding the mobile body 3's self-position estimated by the first self-position estimation unit 23 (S16).

[0057] <Self-position estimation and path generation processing in mobile object 3> The self-position estimation and path generation processes for the mobile object 3 will be explained below with reference to Figure 7.

[0058] First, in the mobile body 3, the second sensor unit 31 acquires two-dimensional sensor data consisting of a point cloud around the mobile body 3 as it travels through real space (S21).

[0059] Then, the second self-position estimation unit 331 estimates the self-position of the moving object 3 in real space based on the second map data stored by the map storage unit 32 and the second sensor data consisting of a point cloud acquired by the second sensor unit 31 (S22).

[0060] Then, the second path generation unit 332 generates a path for the mobile body 3 to travel, based on the information regarding the mobile body 3's self-position estimated by the second self-position estimation unit 331 (S23).

[0061] <Area classification, mode switching, and driving control> The following describes the area classification, mode switching, and driving control of the mobile unit 3, with reference to Figure 8.

[0062] First, in the management device 2, the area information transmission unit 25 transmits information about each area in a predetermined physical space ((Rc), area (Rs), area (Rc and Rs), area (other)) to the mobile device 3 (S31).

[0063] Next, in the mobile unit 3, the area information receiving unit 333 receives information transmitted from the management device 2 regarding areas in a predetermined physical space (area (Rc), area (Rs), area (Rc and Rs), area (other)) (S32).

[0064] Then, the area classification unit 334 classifies the area where the mobile body 3 is located based on information about areas in a predetermined real space (area (Rc), area (Rs), area (Rc and Rs), area (other)) received by the area receiving unit and information about the self-position of the mobile body 3 estimated by the second self-position estimation unit 331 (S33).

[0065] Then, the mode switching unit 335 switches between the infrastructure-based mode and the standalone mode according to the area classified by the area classification unit 334 (S34).

[0066] Then, the travel control switching unit 336 switches the control of the travel of the mobile body 3 so that the mobile body 3 travels along an infrastructure-based mode or standalone mode path generated in the management device 2 or the mobile body 3 according to the area classified by the area classification unit 334 (S35).

[0067] Then, the travel control unit 337 controls the movement of the mobile body 3 based on the path switched by the travel control switching unit 336 (S36).

[0068] Then, the hardware control unit 338 controls hardware such as the wheels based on the commands from the driving control unit 337 (S37).

[0069] <Second Embodiment> Next, a second embodiment of this system will be described, mainly with reference to Figures 9 and 10. In the following, only configurations that differ from the above embodiment will be described, and identical configurations will be omitted from description and given the same reference numerals.

[0070] In the first embodiment, the mobile body 3 classified the area in which it was located, but in this embodiment, the management device 2 classifies the area in which the mobile body 3 was located.

[0071] As shown in Figure 3, the sensor device 1 includes a first sensor unit 11 that acquires sensor data consisting of a point cloud in real space at predetermined time intervals, and a sensor-side transmission unit 12 that sequentially transmits the sensor data consisting of a point cloud in real space acquired by the first sensor unit 11 to the management device 2.

[0072] As shown in Figure 9, the management device 2 includes an aggregation unit 21 that aggregates first sensor data transmitted from the sensor device 1, a map generation unit 22 that generates first map data of real space, a first self-position estimation unit 23 that estimates the self-position of the mobile body 3, and a first path generation unit 24 that generates a path for the mobile body 3 to travel.

[0073] In this embodiment, the management device 2 includes a mobile information receiving unit 26 that receives mobile information including information about the location of the mobile body 3 transmitted from the mobile body 3, and an area classification unit 27 that classifies the area in which the mobile body 3 is located.

[0074] The area classification unit 27 classifies the area where the mobile body 3 is located based on information about an area in a predetermined real space and information about the self-position of the mobile body 3 included in the mobile body information received by the mobile body information receiving unit 26. The method of classifying the area is the same as that of the area classification unit 334 in the first embodiment.

[0075] As shown in Figure 9, the mobile body 3 includes a second sensor unit 31 for acquiring second sensor data around the mobile body 3, a map storage unit 32 for storing second map data of a predetermined real space, and an onboard computer 33.

[0076] The onboard computer 33 includes a second self-position estimation unit 331 for estimating the self-position of the mobile body 3, a second path generation unit 332 for generating a path for the mobile body 3 to travel, a mode switching unit 335 for switching between infrastructure-based mode and standalone mode, a travel control switching unit 336 for switching the control of the travel of the mobile body 3, a travel control unit 337 for controlling the travel of the mobile body 3, and a hardware control unit 338 for controlling hardware such as wheels.

[0077] In this embodiment, the onboard computer 3 includes a mobile information transmission unit 339 that transmits mobile information, including information regarding the mobile body's own position, to the management device 2. This information regarding the mobile body's own position is estimated by a second self-position estimation unit 331.

[0078] Furthermore, the mode switching unit 335 switches between the infrastructure-based mode and the standalone mode according to the area classified by the area classification unit 27 of the management device 2. Also, the travel control switching unit 336 switches the travel control of the mobile body 3 according to the area classified by the area classification unit 27. Information regarding this area classification is transmitted from the management device 2 to the mobile body 3 via a network or the like.

[0079] Next, we will explain each operation in this system with reference to Figure 10. Note that the operations of <Self-position estimation processing and route generation processing in the management device 2> and <Self-position estimation processing and route generation processing in the mobile body 3> are the same as in the first embodiment 1, so we will omit their explanation and only explain <Area classification, mode switching, and driving control>.

[0080] <Area classification, mode switching, and driving control> First, in the mobile body 3, the mobile body information transmission unit 339 transmits mobile body information, which includes at least information regarding the self-position of the mobile body 3, to the management device 2 (S41).

[0081] Next, in the management device 2, the mobile information receiving unit 26 receives mobile information, including information about the location of the mobile body 3, which has been transmitted from the mobile body 3 (S42).

[0082] The area classification unit 27 then classifies the area in which the mobile body 3 is located based on information regarding areas in a predetermined real space (area (Rc), area (Rs), area (Rc and Rs), area (other)) and information regarding the self-position of the mobile body 3 included in the mobile body information received by the mobile body information receiving unit 26 (S43).

[0083] Next, in the mobile unit 3, the mode switching unit 335 switches between the infrastructure-based mode and the standalone mode according to the area classified by the area classification unit 27 (S44).

[0084] Then, the travel control switching unit 336 switches the control of the mobile body 3's movement so that it travels along an infrastructure-based mode or standalone mode path generated in the management device 2 or the mobile body 3 according to the area classified by the area classification unit 27 (S45).

[0085] Then, the driving control unit 337 controls the movement of the mobile body 3 based on the route selected by the driving control switching unit 336, either in infrastructure-based mode or standalone mode (S46).

[0086] Then, the hardware control unit 338 controls hardware such as the wheels based on the commands from the driving control unit 337 (S47). [Examples]

[0087] Next, the experiment according to the present invention will be described with reference to Figures 11 to 14.

[0088] [Experimental scenario] This experiment was conducted in Room 14Q32 of the Shibaura Institute of Technology Research Building on the Toyosu Campus, using a vehicle (LIMO ROS2) as the mobile device 3 and an edge computer (Jetson AGX Orin) as the management device 2, in the experimental area shown in Figure 11. Figure 11(a) is an image of the experimental area, and Figure 11(b) shows the dimensions and location of the experimental area.

[0089] In this experimental area, two types of areas are defined: Area (Rc and Rs) and Area (Rc), and a vehicle is placed in one of these areas. Goal information is constantly provided to the vehicle and the edge computer. One vehicle was used in the experiment, and during the experiment, the vehicle was manually moved to various positions. The vehicle was placed in the positions shown in Figure 11A and B. The tires were lifted off the ground to prevent the vehicle from moving.

[0090] When switching to an infrastructure-based system, the vehicle was moved from point A to point B in Figure 11. When switching to a standalone system, the vehicle was moved from point B to point A in Figure 11.

[0091] In this experiment, the execution time of processing was measured for two patterns. The execution time of switching from standalone to infrastructure-based was measured. The vehicle was initially positioned at point A in Figure 11, confirmed its manually provided self-positioning information, and started standalone autonomous driving processing. Next, the vehicle was moved to point B in Figure 11, and its self-positioning information was manually provided. The switch was confirmed when wheel control information was provided.

[0092] We also measured the execution time for switching from infrastructure-based to standalone mode. The vehicle was initially positioned at point B in Figure 11, where it confirmed its manually entered self-position information and initiated the infrastructure-based automated driving process. Next, the vehicle was moved to point A in Figure 11, where its self-position information was manually provided. We confirmed the switch when wheel control information was provided.

[0093] In summary, each measurement pattern was repeated five times to collect data.

[0094] [Experimental metrics] In this experiment, we focused on two approaches—edge-triggered and vehicle-triggered—and compared and evaluated the processing time of the switching method.

[0095] The edge trigger method corresponds to the first embodiment, in which area information is transmitted from the edge computer (management device 2) to the vehicle (mobile body 3), and the vehicle (mobile body 3) performs area classification. On the other hand, the vehicle trigger method corresponds to the second embodiment, in which vehicle information (including vehicle location information) is transmitted from the vehicle (mobile body 3) to the edge computer (management device 2), and the edge computer performs area classification.

[0096] Figure 12(a) shows a sequence diagram of module interactions when switching to standalone mode in an edge-triggered experimental system. In this experiment, the processing time between modules when switching to standalone mode was measured in order to evaluate the processing time associated with switching to the automatic driving mode.

[0097] Figure 12(b) shows a sequence diagram of module interactions when switching to an infrastructure-based system in an edge-triggered experimental setup. In this experiment, the processing time between modules during the switch to the infrastructure-based system was also measured.

[0098] On the other hand, Figure 13(a) shows a sequence diagram of module interactions when the vehicle-triggered experimental system is switched to standalone mode. In this experiment, the processing time between modules during the switch to standalone mode was also evaluated.

[0099] Figure 13(b) also shows a sequence diagram of module interactions when switching to an infrastructure-based system in the vehicle-based experimental setup. In this experiment, the processing time between modules during the switch to the infrastructure-based system was also evaluated.

[0100] [Experimental Results] Figure 14(a) is a table showing the processing time for each sequence and the average total time when switching from infrastructure-based to standalone using two switching methods. The numbers (1) to (8) in the table in Figure 14(a) correspond to the sequences (1) to (8) assigned in Figures 12 and 13. In the table in Figure 14(a), sequences (3) and (4), and (5) and (6) are processed simultaneously.

[0101] Observing the table in Figure 14(a), in the edge-triggered method, the sequence with the longest processing time is sequence (5)(6), which represents the transition from infrastructure-based to standalone and the subsequent vehicle self-localization process. In the vehicle-triggered method as well, the sequence with the longest processing time is sequence (5)(6), which also represents the transition from infrastructure-based to standalone and the subsequent vehicle self-localization process. Thus, sequences (5)(6), which include the transition from infrastructure-based to standalone and the subsequent self-localization process, had long processing times in both the edge-triggered and vehicle-triggered methods. This is thought to be due to the time required to start up the application used in this sequence.

[0102] Furthermore, Sequence 2, which includes information transmission and area classification, showed the largest difference in processing time between the edge-triggered method and the vehicle-triggered method. In the edge-triggered method, area classification is performed on the vehicle using area information, while in the vehicle-triggered method, the edge computer performs area classification using vehicle information. This is thought to be because the edge computer, with its larger processing resources, can perform area classification faster.

[0103] Figure 14(b) is a table showing the processing time for each sequence and the average total time when switching from standalone to infrastructure-based using two switching methods. The numbers (1) to (8) in the table in Figure 14(b) correspond to the sequences (1) to (8) assigned in Figures 12 and 13. In the table in Figure 14(b), sequences (3) and (4), and (5) and (6) are processed simultaneously.

[0104] Observing the table in Figure 14(b), in the edge-triggered method, the sequence with the longest processing time is sequence (7), which represents the transition from self-localization processing to route generation processing. Similarly, in the vehicle-triggered method, the sequence with the longest processing time is also sequence (7), which represents the transition from self-localization processing to route generation processing. Thus, sequence (7), which includes the transition from self-localization processing to route generation processing, had the longest processing time in both the edge-triggered and vehicle-triggered methods. This is thought to be due to increased communication overhead caused by feedback from the edge computer to the vehicle, and its reliance on network interaction.

[0105] The sequence (2) including information transmission for area classification showed the largest difference in processing time between the edge-triggered method and the vehicle-triggered method. In the edge-triggered method, area classification is performed on the vehicle using area information, whereas in the vehicle-triggered method, area classification is performed on the edge computer using movement information. The faster classification in the vehicle-triggered method is thought to be due to the larger processing resources available on the edge computer.

[0106] This invention proposes an adaptive switching mechanism for a hybrid autonomous driving system that integrates both infrastructure-based mode processing and standalone mode processing. This system can dynamically switch between these modes for self-localization and path generation depending on the area in which the vehicle is traveling.

[0107] Furthermore, two methods can be introduced: an edge-triggered method that classifies the driving area using information from edge computers, and a vehicle-triggered method that relies on location information transmitted from vehicles. In both cases, the area classification framework allows for adaptive switching between infrastructure-based mode processing and standalone mode processing.

[0108] Although embodiments of the present invention have been described above with reference to the drawings, the present invention is not limited to the illustrated embodiments. Various modifications and variations can be made to the illustrated embodiments within the same scope as the present invention, or within the equivalent scope. [Explanation of Symbols]

[0109] 1...Sensor device 11...First sensor unit 12...Sensor-side transmitter 2…Management device 21... Consolidation Department 22...Map generation unit 23...First self-position estimation unit 24...First path generation unit 25... Area Information Transmission Unit 3… Mobile 31...Second sensor unit 32...Map storage section 33…Onboard computer 331...Second self-position estimation unit 332...Second path generation unit 333... Area Information Receiving Unit 334... Area Classification Department 335...Mode switching section 336... Driving control switching unit 337... Driving control unit 338…Hardware Control Unit

Claims

1. A mobile body travel control system comprising one or more sensor devices positioned on the environmental side of a predetermined physical space divided into multiple areas, a management device connected to the sensor devices in a communicative manner, and one or more mobile bodies connected to the management device in a communicative manner and traveling within the predetermined physical space, The aforementioned sensor device is A first sensor unit that acquires first sensor data in a predetermined real space, The system includes a sensor-side transmission unit that transmits first sensor data in a predetermined real space acquired by the first sensor unit to the management device, The aforementioned control device is An aggregation unit that aggregates the first sensor data transmitted from the sensor device, A map generation unit generates first map data of real space based on the first sensor data aggregated by the aggregation unit, A first self-position estimation unit estimates the self-position of the moving object based on the first map data generated by the map generation unit and the first sensor data acquired by the sensor device or the second sensor data acquired by the moving object. The system comprises a first path generation unit that generates a path for the moving object to travel, based on information regarding the self-position of the moving object estimated by the first self-position estimation unit, The aforementioned moving body is A second sensor unit that acquires second sensor data around the moving object, A map storage unit that stores second map data of real space, A second self-position estimation unit estimates the self-position of the moving object based on the second map data stored in the map storage unit and the second sensor data acquired by the second sensor unit. A second path generation unit generates a path for the moving object to travel, based on information regarding the self-position of the moving object estimated by the second self-position estimation unit, The management device includes a mode switching unit that switches between an infrastructure-based mode in which the mobile body performs processes to estimate its own position and generate a path, and a standalone mode in which the mobile body performs processes to estimate its own position and generate a path. The mobile body or the management device further comprises an area classification unit that classifies the area in which the mobile body is located based on information relating to each area in a predetermined real space and information relating to the self-position of the mobile body. The mode switching unit is characterized by switching between the infrastructure-based mode and the standalone mode according to the area classified by the area classification unit, thereby enabling mobile vehicle driving control.

2. A mobile vehicle travel control system according to claim 1, comprising a travel control switching unit that switches the control of the mobile vehicle's travel so that when the switching unit switches to the infrastructure-based mode according to the area classified by the area classification unit, the mobile vehicle travels along a route generated by the first route generation unit of the management device, and when the switching unit switches to the standalone mode, the mobile vehicle travels along a route generated by the second route generation unit of the mobile vehicle.

3. The mobile body travel control system according to claim 1, wherein the area classification unit is provided on the mobile body and classifies the area in which the mobile body is located based on information relating to an area in real space and information relating to the self-position of the mobile body.

4. The management device includes an area information transmission unit that transmits information regarding an area in a predetermined physical space to the mobile body. The mobile unit includes an area information receiving unit that receives information about a predetermined area in a real space transmitted from the management device, The mobile vehicle driving control system according to claim 3, wherein the area classification unit classifies the area in which the mobile vehicle is located based on information about an area in real space received by the area information receiving unit and information about the self-position of the mobile vehicle.

5. The mobile body travel control system according to claim 1, wherein the area classification unit is provided in the management device and classifies the area in which the mobile body is located based on information relating to an area in real space and information relating to the self-position of the mobile body.

6. The mobile body includes a mobile body information transmission unit that transmits information relating to the mobile body's own position, The aforementioned control device is A mobile information receiving unit that receives information about the position of the mobile object transmitted from the mobile object, The mobile vehicle driving control system according to claim 5, wherein the area classification unit classifies the area in which the mobile vehicle is located based on information relating to an area in a predetermined real space and information relating to the self-position of the mobile vehicle received by the mobile vehicle information receiving unit.

7. Information regarding an area in a given physical space is, The management device and the mobile object can communicate, but there is an area (Rc) where the mobile object is not detected by the management device, The moving object is detected by the management device, but there is an area (Rs) where the management device and the moving object cannot communicate, The management device and the mobile body can communicate with each other, and the management device can detect the mobile body in at least the area (Rc and Rs), The mobile vehicle travel control system according to claim 1, wherein the mode switching unit switches to the standalone mode when the area where the mobile vehicle is located is classified by the area classification unit as area (Rc) or area (Rs), and switches to the infrastructure-based mode when the area classification unit classifies the area where the mobile vehicle is located as areas (Rc and Rs).

8. Information regarding areas in a given physical space includes the area (Rc), the area (Rs), and areas that do not fall under any of the areas (Rc and Rs) (others). The mobile body travel control system according to claim 7, wherein the mode estimation unit switches to the standalone mode when the area in which the mobile body is located is classified as the area (other) by the area classification unit.

9. The first sensor unit acquires two-dimensional or three-dimensional first sensor data consisting of a point cloud in a predetermined real space. The mobile vehicle driving control system according to claim 1, wherein the second sensor unit acquires two-dimensional or three-dimensional first sensor data consisting of a point cloud in a predetermined real space.

Images