Autonomous travel vehicle control system and autonomous travel vehicle control method
The autonomous vehicle control system uses a camera to update environmental maps with no-entry zones, addressing the challenge of movable obstacles without additional sensors, ensuring safe navigation.
Patent Information
- Application Number
- JP2023083269
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-05-19
- Publication Date
- 2024-11-29
AI Technical Summary
Existing autonomous robots struggle to update prohibited areas effectively in environments where obstacles, especially movable ones, without requiring additional sensors or re-running the robot, leading to potential contact risks.
An autonomous vehicle control system that uses a camera to capture a bird's-eye view of the environment, creates a flattened image, identifies obstacles, and updates an environmental map with no-entry zones, allowing the robot to avoid obstacles without additional sensor configurations or re-running.
Enables the autonomous robot to update prohibited areas accurately and efficiently, reducing the risk of contact with obstacles by using a centralized server to process images and set no-entry zones based on real-time obstacle positions.
Smart Images

Figure 2024166875000001_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to an autonomous vehicle control system and an autonomous vehicle control method. [Background technology]
[0002] Conventionally, there are service robots (hereinafter referred to as autonomous robots) that operate in residential spaces or commercial facilities. The autonomous robot detects obstacles using various sensors and is controlled to run while avoiding contact with the obstacles. However, depending on the shape or material of the obstacle, the various sensors may not be able to detect it correctly, and the autonomous robot may come into contact with the obstacle while running. Examples of obstacles that are difficult to detect include glass or mesh objects, objects with complex vertical cross-sectional shapes, and tables or chairs with thin legs. In general, when such obstacles that are difficult to detect exist within the area in which the autonomous robot runs, a no-entry area may be set for the location where the obstacle exists on the environmental map for autonomous running. As a result, the autonomous robot is controlled not to run through the no-entry area, so that even if the autonomous robot cannot detect the obstacle, it can avoid contact with the obstacle. However, even if a no-entry area is set around an obstacle, if the obstacle moves or if the number of obstacles increases or decreases, the no-entry area needs to be reset. In particular, when the area in which an autonomous robot travels is a living space, there is a high possibility that the obstacle will move. [Prior art documents] [Patent documents]
[0003] [Patent Document 1] JP 2020-184148 A Summary of the Invention [Problem to be solved by the invention]
[0004] For example, Patent Document 1 shows an example in which a no-entry area is set around a shelf, which is an obstacle with a complex vertical cross-sectional shape, to reduce the risk of contact between an autonomous robot and the obstacle. In Patent Document 1, in order to correctly set a no-entry area around a shelf, an autonomous robot equipped with many sensors is made to travel near the shelf to obtain the outline of the shelf, and the no-entry area is set based on that outline. When the area in which the autonomous mobile robot travels is a residential space, the autonomous mobile robot needs to travel in the area each time the no-entry area is updated in Patent Document 1. In addition, the autonomous mobile robot needs to be equipped with many sensors.
[0005] The object of the present invention is to provide an autonomous vehicle control system and an autonomous vehicle control method that can update no-entry areas and reduce the risk of contact with obstacles in an environment where obstacles installed within the area in which the autonomous robot travels, particularly obstacles that are difficult for the autonomous robot to detect, are moving, without the need to drive the autonomous robot to update no-entry areas, and without limiting the sensor configuration or performance of the autonomous robot itself. [Means for solving the problem]
[0006] In order to solve the above problems, the present invention employs the following means. In other words, the autonomous vehicle control system of the present invention is characterized by comprising an imaging device that captures an overhead image of the driving environment including obstacles, a server that acquires a flattened image of the driving environment captured by the imaging device at a predetermined timing, identifies obstacles in the driving environment from the flattened image, and creates an environmental map that sets out no-entry areas that encompass the identified obstacles, and a driving vehicle that drives through the driving environment based on the environmental map in which the no-entry areas are set.
[0007] In addition, the autonomous vehicle control method of the present invention is characterized in that it includes capturing an overhead image of a driving environment including obstacles using an imaging device, acquiring a flattened image of the driving environment captured by the imaging device at a predetermined timing, identifying obstacles in the driving environment from the flattened image, creating an environmental map that sets a no-entry area that includes the identified obstacles, and driving a driving vehicle based on the environmental map in which the no-entry area is set. Effect of the Invention
[0008] According to the present invention, it is possible to provide an autonomous vehicle control system and an autonomous vehicle control method that can update no-entry areas without limiting the sensor configuration or performance of the autonomous robot itself in an environment where obstacles installed within the area in which the autonomous robot travels, particularly obstacles that are difficult for the autonomous robot to detect, are moving, thereby reducing the risk of contact with obstacles. [Brief description of the drawings]
[0009] [Figure 1] 1 is a schematic diagram showing a configuration of an autonomous vehicle control system 1 according to an embodiment of the present invention. [Diagram 2] 1 is a schematic diagram showing an example of the appearance of an autonomous mobile robot 2 according to the present embodiment. [Diagram 3] 1 is a block diagram showing a configuration of an autonomous vehicle control system 1 according to an embodiment of the present invention. [Figure 4] 4 is a flowchart for explaining the operation of the autonomous vehicle control system 1 according to the present embodiment. [Diagram 5] FIG. 2 is a conceptual diagram illustrating a real space and a flattened image according to the present embodiment. [Figure 6] 10 is a conceptual diagram for explaining a method for matching the orientation and scale of an environment map and a flattened image according to the present embodiment. FIG. [Figure 7] 1 is a conceptual diagram for explaining a method for identifying an obstacle and setting a no-entry area according to the present embodiment. FIG. [Figure 8] 13 is a conceptual diagram for explaining a method for setting a boundary area of chairs 41 and 42 that have been moved within an obstacle movable area according to this embodiment. FIG. [Figure 9] 11A and 11B are conceptual diagrams for explaining another method for setting a no-entry area according to an embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an embodiment of an autonomous mobile robot control system according to the present invention will be described with reference to the accompanying drawings.
[0011] FIG. 1 is a schematic diagram showing the configuration of an autonomous vehicle control system 1 according to an embodiment of the present invention. In FIG. 1, the autonomous vehicle control system 1 includes an autonomous robot 2 (traveling vehicle), a server 3, and a camera 4 (imaging device). The autonomous robot 2 travels while estimating its own position based on an environmental map and detection results of various sensors (described later) (in the following description, the area in which the autonomous robot 2 travels is called the robot operation area). At that time, if a no-entry area (logical wall) is set in the environmental map, the autonomous robot 2 is controlled not to travel in the no-entry area. That is, the no-entry area is an area that includes obstacles 5 and 6 in the robot operation area, and the risk of contacting the obstacles 5 and 6 is not high unless the robot enters the no-entry area. In this embodiment, the obstacles 5 and 6 are assumed to be movable furniture (tables, chairs, shelves, etc.) installed in the robot operation area, but the no-entry area may also include other obstacles such as walls, pillars, and partitions. The no-entry area can also be set as an area other than the area including the obstacles 5 and 6, for example, an area in which the autonomous mobile robot 2 should not be allowed to travel, such as near a structure that may be difficult to see, such as a window glass. In the following, only the movement of the obstacles 5 and 6 will be described, but the present invention may also include cases in which the obstacles 5 and 6 are not only moved, but are also added or removed.
[0012] The server 3 creates an environmental map that sets a no-entry area in which obstacles 5 and 6 exist within the obstacle moving area from a flattened image obtained by flattening the obstacle moving area photographed by the camera 4, and provides the map to the autonomous mobile robot 2. The camera 4 is installed on the ceiling of the robot operating area, and photographs the entire obstacle moving area (at least the area where the autonomous mobile robot 2 moves) from the ceiling toward the floor. Note that, although one camera 4 is installed in FIG. 1, multiple cameras may be installed, and images photographed by the multiple cameras may be combined to form a single image that covers the entire obstacle moving area (at least the area where the autonomous mobile robot 2 moves), and the resulting image may be flattened and used as the flattened image.
[0013] The autonomous mobile robot 2 and the server 3 are capable of data communication via wireless communication 7. In FIG. 1, before entering the obstacle movable area (in the obstacle immobile area), the autonomous mobile robot 2 requests the server 3 to update the environmental map via wireless communication 7. Here, the obstacle immobile area is an area (fixed area) in which the obstacle does not move. The obstacle movable area is an area in which the obstacle may move, and the obstacle is assumed to be, for example, a table or a chair. When the server 3 receives a request to update the environmental map from the autonomous mobile robot 2, the server 3 photographs the entire obstacle movable area with the camera 4, identifies the obstacles 5 and 6 in the obstacle movable area from the flattened image, creates an environmental map in which the identified obstacles 5 and 6 are included in the no-entry area, and transmits the created environmental map to the autonomous mobile robot 2 via wireless communication 7. The autonomous mobile robot 2 travels in the obstacle movable area according to the environmental map in which the no-entry area is set. In this embodiment, as is clear from the above description, the robot operating area is made up of an obstacle immobile area and an obstacle movable area. Here, the obstacle movable area constitutes the traveling environment of the autonomous traveling robot 2.
[0014] FIG. 2 is a schematic diagram showing an example of the appearance of the autonomous traveling robot 2 according to this embodiment. As shown in FIG. 2, a LiDAR (Light Detection and Ranging, Laser Imaging Detection and Ranging) 11, a depth camera 12, and a plurality of ultrasonic sensors 13, 13, ... are installed on the outside of a housing 10 of the autonomous traveling robot 2, and wheels 14, 14 for traveling are arranged on the lower part of the housing 10. The LiDAR 11 is arranged at a predetermined position of the housing 10 and is a sensor that emits pulsed laser light in a predetermined angle range, measures reflected light from the laser irradiation, and detects the distance, position, and shape of obstacles 5, 6 in the robot operation area. When traveling within the obstacle movable area, the autonomous traveling robot 2 according to this embodiment can travel while avoiding the obstacles 5, 6 according to the environmental map created by the server 3 using the results detected by the LiDAR 11. The depth camera 12 has a built-in depth sensor and acquires depth information (three-dimensional) of the obstacles 5, 6 within the obstacle movable area, but this can be omitted. Each of the ultrasonic sensors 13, 13, ... measures the presence or absence of and distance to the obstacles 5, 6 within the obstacle movable area by measuring the time from transmitting ultrasonic waves to receiving the ultrasonic waves, but this can be omitted. Note that the autonomous traveling robot 2 may be provided with sensors other than the depth camera 12 or the ultrasonic sensors 13, 13, ... to obtain information such as the shapes of the obstacles 5, 6 within the obstacle movable area.
[0015] FIG. 3 is a block diagram showing the configuration of an autonomous vehicle control system 1 according to this embodiment. Note that parts corresponding to those in FIG. 1 or FIG. 2 are given the same reference numerals and the description thereof is omitted. In FIG. 3, the autonomous robot 2 includes a control unit (CPU) 15, a storage unit 16, an operation unit 17, a wireless communication unit 18, and a wheel drive unit 19 in addition to the LiDAR 11 described above. The control unit 15 controls each unit so that the autonomous robot 2 runs while avoiding obstacles 5 and 6, using an environmental map created by the server 3 and the results of detection and measurement by the LiDAR 11, according to the autonomous running program stored in the storage unit 16. The storage unit 16 stores the autonomous running program and various data, in particular an environmental map (data) in which the position, size, and no-entry area of an obstacle in the obstacle movable area are set. Note that the environmental map is not limited to one image data, and may be composed of a layer of a base environmental map created by the LiDAR 11 and a layer of an environmental map in which the no-entry area is indicated. The environmental map may be updated by the server 3 before the autonomous robot 2 enters the obstacle movable area.
[0016] The operation unit 17 is composed of various switches and the like that allow the user to give instructions such as to start and stop traveling. The wireless communication unit 18 establishes wireless communication with the server 3 and performs data communication. In particular, in this embodiment, the wireless communication unit 18 is used when the autonomous mobile robot 2 requests the server 3 to update the environmental map, or when receiving the updated environmental map transmitted from the server 3 in response to the request to update the environmental map. The wheel drive unit 19 includes a drive system that drives the wheels 14, 14, and drives the wheels 14, 14 to run, stop, turn, etc. under the control of the control unit 15.
[0017] The server 3 is a general computer, and includes a control unit (CPU) 21, a storage unit 22, a wireless communication unit 23, a display unit 24, and an operation unit 25. When the autonomous mobile robot 2 requests to update the environmental map, the control unit 15 photographs the entire obstacle movable area with the camera 4 according to the environmental map creation program stored in the storage unit 16, flattens the photographed image by image processing, and aligns the orientation and scale of the environmental map and the flattened image, identifies an obstacle from the flattened image by image recognition processing, and creates an environmental map in which an entry-prohibited area is set based on the identified obstacle. At this time, if an obstacle in the obstacle movable area moves, a new entry-prohibited area is set according to the moved obstacle. The storage unit 22 stores the environmental map creation program, various data, an environmental map (data) in which an entry-prohibited area is set, and the like. The wireless communication unit 23 establishes wireless communication with the autonomous mobile robot 2 and performs data communication. In particular, in this embodiment, it is used when receiving an update request for the environmental map from the autonomous mobile robot 2, or when transmitting an updated environmental map created in response to the update request for the environmental map. The display unit 24 displays various information such as the operating status. The operation unit 25 is composed of a keyboard and the like for the user to give instructions for operations.
[0018] 4 is a flowchart for explaining the operation of the autonomous vehicle control system 1 according to this embodiment. In the autonomous robot 2, the control unit 15 recognizes its own position when traveling in an obstacle immobile area (step S10) and determines whether the traveling route from the current point onward is an obstacle movable area (step S12). If it is not an obstacle movable area (NO in step S12), the process returns to step S10 and continues traveling.
[0019] On the other hand, if the travel route beyond the current point is an obstacle movable area (YES in step S12), the control unit 15 temporarily stops travel via the wheel drive unit 19 (step S14). Next, the control unit 15 transmits an update request for the environmental map to the server 3 via the wireless communication unit 18 (step S16). At this time, the control unit 15 transmits the environmental map currently stored in the memory unit 16 of the autonomous traveling robot 2. Then, the control unit 15 determines whether or not a response has been received from the server 3 (step S18), and if a response has not been received (NO in step S18), the control unit 15 waits until a response is received.
[0020] Meanwhile, in the server 3, the control unit 21 determines whether or not an update request has been received from the autonomous mobile robot 2 (step S30), and if an update request has not been received (NO in step S30), the control unit 21 waits until an update request is received. On the other hand, if an update request (and an environmental map) has been received from the autonomous mobile robot 2 (YES in step S30), the control unit 21 acquires a captured image from the camera 4 (step S32) and flattens the acquired captured image by image processing (step S34).
[0021] Here, FIG. 5 is a conceptual diagram for explaining the real space and flattened image according to this embodiment. FIG. 5(a) shows a simulated real space 30 including an obstacle movable area, in which blocks simulating a plurality of obstacles are placed. FIG. 5(b) shows a flattened image 31 obtained by flattening an image of the real space 30 shown in FIG. 5(a) captured by a camera 4 installed on the ceiling. In this way, in the flattened image 31, the obstacles placed in the obstacle movable area are viewed from almost directly above. For example, as shown in FIG. 5(c), when the obstacles 5 and 6 have a step in the cross-sectional shape in the height direction, depending on the sensing positions SP1 and SP2 of the autonomous traveling robot 2, there is a possibility that the recessed portion is recognized as the size of the obstacles 5 and 6, rather than the actual size (maximum size) of the obstacles 5 and 6. In this way, if the obstacles are recognized as smaller than their actual size, there is a risk of contacting the obstacles 5 and 6 when actually traveling. In contrast, by using a flattened image 31 obtained by flattening an image taken by a camera 4 installed on the ceiling, it is possible to identify the areas R1 and R2 viewed from above the obstacles 5 and 6 as obstacles, as shown in Figure 5(c), making it possible to set no-entry areas to accurately avoid the obstacles 5 and 6 regardless of the sensing positions SP1 and SP2 of the autonomous mobile robot 2.
[0022] Next, the control unit 21 aligns the orientation and scale of the environmental map and the flattened image (step S36). Here, Figs. 6(a) and (b) are conceptual diagrams for explaining a method for aligning the orientation and scale of the environmental map and the flattened image according to this embodiment. As shown in Fig. 6(a), in real space 30, flattened image 31 is the range DL indicated by the dotted line, and environmental map 32 is the range BL indicated by the dashed line. By setting the following information (1) to (6) in advance, it is possible to superimpose (match) flattened image 31 and environmental map 32 as shown in Fig. 6(b) by basic geometric transformations (image processing) such as translation, rotation, and enlargement / reduction.
[0023] (1) A reference direction vector in the flattened image (e.g., N1 (north)) (2) Scale of flattened image (pixel / meter) (3) A reference direction vector in the environmental map (e.g., N2 (north)) (4) Scale of the environmental map (pixel / meter) (5) A reference point of the flattened image (e.g., point D1) (6) A reference point of the flattened image in the environmental map (e.g., point D2)
[0024] In addition, when applying this to different types of autonomous robots with different driving performance, housing sizes, etc., the flattened image is the same, so the above (3), (4), and (6) must be set in advance for each of the different types of autonomous robots.
[0025] Next, the control unit 21 identifies obstacles from the flattened image with the adjusted orientation and scale using image recognition processing (step S38), and sets a no-entry area in the flattened image based on the identified obstacle (step S40). The control unit 21 applies the no-entry area to the environmental map (step S42), and determines whether the environmental map has been updated (step S44).
[0026] Here, Fig. 7 is a conceptual diagram for explaining a method for identifying obstacles and setting a no-entry area according to this embodiment. An example of an autonomous mobile robot 2 for serving food that operates in cafes, restaurants, etc. is explained. As shown in Fig. 7(a), assume that the autonomous mobile robot 2 travels in a real space 40 in an environment (obstacle movable area) in which obstacles that may move, such as chairs and tables (e.g., chairs 41, 42), exist. If the no-entry area is set on an environmental map in advance, the autonomous mobile robot 2 will not travel within the no-entry area, and therefore can avoid contact with the obstacle regardless of whether it can be detected by a sensor.
[0027] However, as shown in FIG. 7(b), it is assumed that the obstacles, in the illustrated example, two chairs 41 and 42, move and go outside the boundaries L1 and L2 of the no-entry area in the environmental map 43. The obstacles, such as chairs and tables, have shapes that are difficult to recognize with the various sensors equipped in the autonomous mobile robot 2, and may not be recognized accurately depending on the mounting position of the sensor (especially LiDAR). Therefore, as shown by the dotted arrow in FIG. 7(b), even if the autonomous mobile robot 2 travels based on the no-entry area set in the environmental map 43 and uses various sensors, there is an increased risk that the autonomous mobile robot 2 will come into contact with the obstacles, the chairs 41 and 42, at the positions marked with x in FIG. 7(b).
[0028] In response to such a problem, in this embodiment, immediately before the autonomous mobile robot 2 travels through its environment (obstacle movable area), the position and size of the obstacle in the real space (chairs 41 and 42 in FIG. 7(b)) are recognized from an image captured by the camera 4 on the ceiling, and as shown in FIG. 7(c), a new no-entry area is set with boundaries L1' and L2' including the boundary area (dotted rectangle) of the moved chairs 41 and 42, and the new no-entry area is applied to the environmental map 43. In this way, by updating the environmental map 43, even if the chairs 41 and 42, which are obstacles, move in the real space 40, the movement can be accommodated. Therefore, the autonomous mobile robot 2 can travel through the real space 40 based on the updated environmental map 43, as shown by the solid arrow in FIG. 7(d), and the risk of contact with the moved obstacles (chairs 41 and 42) can be significantly reduced compared to the conventional method.
[0029] Here, FIG. 8 is a conceptual diagram for explaining a method for setting a boundary area of chairs 41, 42 moved in the obstacle movable area according to this embodiment. Obstacles 5, 6 as shown in FIG. 8(a) are photographed by the camera 4 on the ceiling, and the photographed image is flattened to obtain a flattened image as shown in FIG. 8(b). In the flattened image, flattened images 5', 6' correspond to the obstacles 5, 6, respectively. The server 3 specifies the positions of the obstacles 5, 6 (flattened images 5', 6') and their apparent sizes A1, A2 shown in FIG. 8(c) from the flattened image shown in FIG. 8(b). Then, the no-entry area is set based on the sizes A1, A2 of the obstacles 5, 6. At this time, the boundary area is set for the obstacles 5, 6 (flattened images 5', 6') taking into consideration individual differences of the autonomous traveling robot 2 (such as the size of the housing 10, the accuracy of self-position recognition (various sensors), and the accuracy of the driving system) so that the autonomous traveling robot 2 does not come into contact with the obstacles 5, 6 even if it approaches them.
[0030] For example, when the size of the housing 10 of the autonomous mobile robot 2 is relatively small, when the accuracy of self-position recognition (various sensors) is high, or when the accuracy of the drive system is high, relatively small (distance to the obstacle is small) boundary areas A1', A2' are set around the obstacles 5, 6 (flattened images 5', 6') as shown in FIG. 8(d). On the other hand, when the size of the housing 10 of the autonomous mobile robot 2 is relatively large, when the accuracy of self-position recognition (various sensors) is low, or when the accuracy of the drive system is low, relatively large (distance to the obstacle is large) boundary areas A1", A2" are set around the obstacles 5, 6 (flattened images 5', 6') as shown in FIG. 8(e). The no-entry area is set so as to include the boundary areas A1', A2' or the boundary areas A1", A2" set around the obstacles 5, 6 (flattened images 5', 6').
[0031] Returning to the explanation of Fig. 4, if the environmental map has been updated, that is, if there has been a change (movement, addition, etc.) in the obstacle movable area in the obstacle movable area (YES in step S44), the control unit 21 of the server 3 transmits the updated environmental map in which the new no-entry area has been applied to the autonomous mobile robot 2 via the wireless communication unit 23 (step S46). On the other hand, if the environmental map has not been updated, that is, if there has been no change in the obstacle in the obstacle movable area (NO in step S44), the control unit 21 transmits a notification that no update is required to the autonomous mobile robot 2 via the wireless communication unit 23 (step S48). Thereafter, the control unit 21 ends the process (or returns to step S30 and continues the process until an end instruction is received).
[0032] On the other hand, when the control unit 15 of the autonomous mobile robot 2, which has been waiting for a response from the server 3 in step S18, receives the response from the server 3 (YES in step S18), it determines whether or not it has received an environmental map to which an updated no-entry area has been applied from the server 3 (step S20). If it has received an environmental map (YES in step S20), the control unit 15 updates the environmental map stored in the storage unit 16 to the received new environmental map (step S22) and resumes traveling via the wheel drive unit 19 (step S24). That is, the autonomous mobile robot 2 travels within the obstacle movable area according to the environmental map to which a new no-entry area has been applied. In the updated environmental map, a new no-entry area has been applied in response to a change (movement, addition, etc.) in an obstacle, so that the autonomous mobile robot 2 can travel without coming into contact with the moved obstacle (chairs 41, 42). Thereafter, when the preset destination is reached or the preset job is completed, the control unit 15 ends the process (or returns to step S10 and continues the process until an end instruction is received).
[0033] On the other hand, when the control unit 15 receives a notification that updating is not required from the server 3 (NO in step S20), the control unit 15 resumes traveling via the wheel drive unit 19 without updating the environmental map stored in the storage unit 16 (step S24). That is, the autonomous mobile robot 2 travels within the obstacle movable area according to the no-entry areas of the environmental map already stored. In this case, since there is no change (movement, addition, etc.) in the obstacles, the autonomous mobile robot 2 can travel without coming into contact with obstacles (tables, chairs) by traveling according to the no-entry areas of the environmental map already stored. Thereafter, when the control unit 15 reaches a destination set in advance or the job set in advance is completed, the control unit 15 ends the process (or returns to step S10 and continues the process until an end instruction is received).
[0034] According to the above-described embodiment, when the autonomous traveling robot 2 requests an environmental map from the server 3 before traveling through the obstacle moving area, the server 3 photographs the obstacle moving area with the camera 4, flattens the photographed image, identifies the obstacles 5, 6 within the obstacle moving area from the flattened image, and creates an environmental map that sets a no-entry area that encompasses the identified obstacles 5, 6. Therefore, even if it is difficult to detect the obstacles 5, 6 or the obstacles 5, 6 have moved, the autonomous traveling robot 2 can travel through the obstacle moving area based on the environmental map in which the latest no-entry area is set, thereby reducing the risk of contacting the obstacles 5, 6.
[0035] According to the above-described embodiment, when the autonomous traveling robot 2 approaches the obstacle moving area, it requests an environmental map from the server 3. Therefore, the latest environmental map of the obstacle moving area can be obtained in advance. Even if it is difficult to detect the obstacles 5, 6 or the obstacles 5, 6 have moved, the autonomous traveling robot can travel through the obstacle moving area based on the latest environmental map in which more accurate no-entry areas are set, thereby reducing the risk of contacting the obstacles 5, 6.
[0036] In the above embodiment, the server 3 updates the environmental map to reflect a new no-entry area when an update of the environmental map is requested by the autonomous traveling robot 2, but the server 3 may update the environmental map to reflect a new no-entry area at a predetermined time period instead of at a request from the autonomous traveling robot 2, and transmit the most recently updated environmental map when a request is received from the autonomous traveling robot 2. According to this, the server 3 creates an environmental map in which the latest no-entry areas are set at a predetermined time period, so that the autonomous traveling robot 2 can always travel through the obstacle movable area according to the environmental map in which the latest no-entry areas are set, even when it is difficult to detect the obstacles 5, 6 or the obstacles 5, 6 have moved, and the risk of contacting the obstacles 5, 6 can be reduced.
[0037] Also, an environmental map in which the latest no-entry areas are set may be created by the server 3 both when the autonomous robot 2 approaches the obstacle movable area and at a predetermined time interval. In this way, even if it is difficult for the autonomous robot 2 to detect the obstacles 5, 6 or if the obstacles 5, 6 have moved, the autonomous robot 2 can travel through the obstacle movable area according to the environmental map in which the latest no-entry areas are set, thereby reducing the risk of contacting the obstacles 5, 6.
[0038] Furthermore, according to the above-described embodiment, the server 3 sets boundary areas A1', A2' or boundary areas A1", A2" of different sizes around the obstacles 5, 6 depending on the individual differences of the autonomous traveling robot 2 (such as the size of the housing 10, the accuracy of self-position recognition (various sensors), and the accuracy of the drive system), and sets no-entry areas so that the boundary areas A1', A2' encompass the boundary areas A1", A2". Therefore, it is possible to set no-entry areas depending on the individual differences of the autonomous traveling robot 2, and even when it is difficult to detect the obstacles 5, 6 or the obstacles 5, 6 have moved, the risk of contacting the obstacles 5, 6 can be reduced.
[0039] In the above-described embodiment, the server 3 updates the environmental map to reflect the new no-entry area when an update of the environmental map is requested by the autonomous traveling robot 2. However, the server 3 may periodically monitor the obstacle movement area, and create an environmental map that reflects the new no-entry area when it detects that an obstacle has moved (moving object detection).
[0040] In the above-described embodiment, the autonomous traveling robot 2 requests an update of the environmental map in the obstacle immovable area before entering the obstacle movable area as shown in Fig. 1, but is not limited to this. For example, it is possible to configure the robot to have a plurality of obstacle movable areas in the robot operating area, and to request an update of the environmental map of the obstacle movable area B in the obstacle movable area A when entering from one of the plurality of obstacle movable areas (referred to as "obstacle movable area A") to another of the plurality of obstacle movable areas (referred to as "obstacle movable area B").
[0041] In the above-mentioned embodiment, the environmental map does not need to be updated in an area where obstacles are fixed. As shown in Fig. 7(a) to (d), an environment where obstacles move is set as an obstacle movable area, and the autonomous traveling robot control system described above is applied only to the obstacle movable area, thereby streamlining the entire system by reducing the processing load, limiting the number of cameras, etc.
[0042] In addition, in the above-described embodiment, an environmental map with no-entry areas set is created on a server 3 outside the autonomous traveling robot 2 as shown in FIG. 1, but this is not limited to the above. The server may be provided inside the autonomous traveling robot 2, and an image captured by the camera 4 may be acquired just before the autonomous traveling robot 2 enters an obstacle movable area, and the server of the autonomous traveling robot 2 may flatten the captured image to create (update) an environmental map with new no-entry areas set from the flattened image.
[0043] In addition, in the above-described embodiment, a flattened image is generated from an image captured by at least one camera 4 installed on the ceiling, and the position, size, etc. of an obstacle are recognized from the flattened image. However, in order to recognize obstacles more accurately or to exclude things other than obstacles, such as people, a depth camera or a sensor such as LiDAR other than the camera 4 may be installed, and the position, size, etc. of the obstacle may be recognized from information acquired by these cameras or sensors.
[0044] In the above embodiment, a flattened image is generated from an image captured by at least one camera 4 installed on the ceiling, and the position, size, etc. of each obstacle is recognized from the flattened image, and each obstacle that moves to the no-entry area is included, but this is not limited to the above. Here, FIG. 9 is a conceptual diagram for explaining another setting method of the no-entry area according to the embodiment. Based on the position and size of each obstacle (dotted line surrounding the obstacle) recognized from the flattened image shown in FIG. 9(a), a concave envelope (or convex envelope) 50, 51, 52 that includes a plurality of adjacent obstacles may be derived each time as shown in FIG. 9(b), and the concave envelope (or convex envelope) 50, 51, 52 may be derived to create a no-entry area.
[0045] Therefore, according to this embodiment, obstacles 5 and 6 are detected by analyzing camera images, so there is no need to drive the autonomous robot 2 to update the no-entry areas, and an autonomous vehicle control system and an autonomous vehicle control method can be provided that can update the no-entry areas without restricting the sensor configuration or performance of the autonomous robot 2 itself, thereby reducing the risk of contacting obstacles.
[0046] In addition, it is possible to select and discard the configurations given in the above embodiment, or to change them to other configurations as appropriate, without departing from the spirit of the present invention. [Explanation of symbols]
[0047] 1 Autonomous vehicle control system 30, 40 Real space 2 Autonomous robot (moving vehicle) 31 Flattened image 3 Server 32, 43 Environment Map 4 Camera (imaging device) 41, 42 Chair 5, 6 Obstacles 50, 51, 52 Concave envelope 7 Wireless communication R1, R2 area 11 LiDAR SP1, SP2 sensing position 12 Depth Camera L1, L2, L1', L2' Boundaries 13 Ultrasonic sensor S1, S2 size 14 wheels A1', A2', A1”, A2” boundary area 15, 21 Control unit (CPU) 16, 22 Storage section 17, 25 Operation section 18, 23 Wireless Communication Department 19 Wheel drive unit 24 Display
Claims
1. An imaging device that captures an overhead image of a driving environment including obstacles; a server that acquires a flattened image of the driving environment captured by the imaging device at a predetermined timing, identifies obstacles in the driving environment from the flattened image, and creates an environmental map that sets no-entry areas that include the identified obstacles; a traveling vehicle that travels in the traveling environment based on the environmental map in which the no-entry area is set; An autonomous vehicle control system comprising:
2. The predetermined timing is The time when the traveling vehicle approaches the traveling environment.
2. The autonomous vehicle control system according to claim 1 .
3. The predetermined timing is a predetermined time period, 3. The autonomous vehicle control system according to claim 1 or 2.
4. The server, When setting the no-entry area, a boundary area is set around the obstacle, and the no-entry area is set so as to include the boundary area; The size of the boundary area is set according to individual differences of the traveling vehicle.
2. The autonomous vehicle control system according to claim 1 .
5. Taking an overhead photograph of the driving environment including obstacles using an imaging device; Acquiring a planarized image of the traveling environment captured by the imaging device at a predetermined timing; identifying obstacles in the driving environment from the flattened image; creating an environmental map that defines a no-entry area that includes the identified obstacle; causing a traveling vehicle to travel based on the environmental map in which the no-entry area is set; 2. An autonomous vehicle control method comprising:
Citation Information
Patent Citations
Information processing device and information processing method
JP2020184148A
