Mobile object management system, mobile object control system, mobile object management method and program

Abstract

A mobile body management system according to an embodiment of the present disclosure includes a prediction unit that, in response to acquisition of route information for a mobile body from the mobile body, predicts presence of an obstruction situated on a route indicated by the route information, and an output unit that outputs, to the mobile body, information relating to the presence of the obstruction predicted by the prediction unit. The mobile body management system can suppress communication load when deciding a movement route of the mobile body.

Description

MOBILE BODY MANAGEMENT SYSTEM, MOBILE BODY CONTROL SYSTEM, AND MOBILE BODY MANAGEMENT METHOD

[0001] The present disclosure relates to a mobile object management system, a mobile object control system, and a mobile object management method.

[0002] In recent years, advances have been made in digital twin technology, which acquires information about the real world using sensors and then constructs a virtual space that reflects the acquired information. By using digital twin technology, it is possible to perform highly accurate simulations without actually conducting experiments in the real world.

[0003] For example, Patent Literature 1 describes a system using a digital behavior twin for collision avoidance. The system using the digital behavior twin is mounted on a vehicle, and is capable of identifying the risk of collision between either the remote vehicle or the own vehicle based on the digital behavior twin of the remote vehicle, the digital behavior twin of the own vehicle, and digital data of driving behavior.

[0004] Japanese Patent Application Laid-Open No. 2020-009428

[0005] In Patent Document 1, the amount of information used for vehicle control, such as information related to digital behavior twins or digital data on driving behavior, can be enormous. Therefore, when a vehicle system acquires information used for vehicle control, the communication load between the system and the network can increase.

[0006] One of the objectives to be achieved by the embodiments of the present disclosure is to provide a mobile object management system, a mobile object control system, and a mobile object management method that are capable of reducing communication load when determining the movement route of a mobile object. It should be noted that this objective is only one of multiple objectives to be achieved by multiple embodiments disclosed herein. Other objectives or problems and novel features will become apparent from the description of this specification or the accompanying drawings.

[0007] A mobile object management system according to one embodiment includes a prediction means for predicting the presence of an obstacle located on a route indicated by the route information of a mobile object in response to obtaining the route information of the mobile object from the mobile object, and an output means for outputting information regarding the presence of the obstacle predicted by the prediction means to the mobile object.

[0008] A mobile body control system according to one embodiment includes a prediction means for predicting the presence of an obstacle located on a route indicated by route information of a mobile body in response to acquisition of the route information of the mobile body from the mobile body, an output means for outputting information regarding the presence of the obstacle predicted by the prediction means to the mobile body, and the mobile body for determining the route of the mobile body based on the information regarding the presence of the obstacle.

[0009] A mobile body management method according to one embodiment predicts the presence of an obstacle located on the route indicated by the route information of a mobile body in response to obtaining the route information of the mobile body from the mobile body, and outputs information regarding the predicted presence of the obstacle to the mobile body.

[0010] The present disclosure makes it possible to provide a mobile object management system, a mobile object control system, and a mobile object management method that are capable of reducing communication load when determining a movement route for a mobile object.

[0011] FIG. 1 is a block diagram showing an example of a mobile object management system according to the present disclosure. FIG. 2 is a flowchart showing an example of representative processing of a mobile object management system according to the present disclosure. FIG. 3 is a block diagram showing an example of a mobile object control system according to the present disclosure. FIG. 4 is a block diagram showing an example of a robot control system according to the present disclosure. FIG. 5 is a block diagram showing an example of a robot according to the present disclosure. FIG. 6 is a block diagram showing an example of a management server according to the present disclosure. FIG. 7 is a diagram showing an example of setting small areas for candidate travel routes. FIG. 8 is a diagram showing an example of a situation in which a target area determination unit makes a determination regarding a certain obstacle. FIG. 9 is a diagram showing an example of a situation in which the target area determination unit makes a determination regarding another obstacle. FIG. 10 is a sequence diagram showing an example of representative processing of a robot control system according to the present disclosure. FIG. 11 is a block diagram showing another example of a management server according to the present disclosure. FIG. 12 is a block diagram showing an example of a hardware configuration of a device according to the present disclosure.

[0012] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. Note that the following descriptions and drawings in the embodiments have been omitted or simplified as appropriate for clarity of explanation. Furthermore, in this disclosure, unless otherwise specified, when multiple items are defined as "at least one of multiple items," the definition may mean any one item, or any multiple items including all items.

[0013] Each drawing referenced in the embodiments is merely an example for describing one or more embodiments. Each drawing is not related to only one particular embodiment, but may also be related to one or more other embodiments. As will be understood by those skilled in the art, various features or steps described with reference to any one drawing can be combined with features or steps shown in one or more other drawings to create, for example, an embodiment not explicitly shown or described. Not all features or steps shown in any one drawing are necessarily required to describe an exemplary embodiment, and some features or steps may be omitted. The order of steps described in any drawing may be changed as appropriate.

[0014] Additionally, common definitions used throughout this disclosure are explained below.

[0015] [Explanation of Definitions] In this disclosure, a digital twin refers to the creation of a virtual space that reproduces a real space using information about the real space acquired using sensors, etc., or a virtual space that reproduces a real space. In particular, this disclosure further defines two types of digital twin: a normal digital twin and a probabilistic digital twin.

[0016] In a typical digital twin, a system uses data on an object in real space acquired from sensors or the like to recreate the object in a virtual space. Specifically, the system acquires information on the object's movement, such as at least one of the object's position, direction of travel, speed, or acceleration in real space, from sensors or the like. By reflecting the object's movement information in virtual space, the system can recreate the object's behavior in real space in virtual space. Furthermore, by using the movement information, the system can predict parameters such as the object's position, direction of travel, speed, or acceleration from the predicted time into the future. At this time, the system generates information such as the object's future position as a unique value. The system can also acquire information indicating the object's unique properties, such as its shape, size, color, and type, from sensors or the like, and recreate the object's properties in virtual space.

[0017] In a probabilistic digital twin, as in the case of a typical digital twin, the system acquires information about the object's movement in real space, such as its position, direction of travel, speed, or acceleration, from sensors or the like. By reflecting the information about the object's movement in virtual space, the system can reproduce the object's behavior in real space in virtual space. However, a difference between a probabilistic digital twin and a typical digital twin is that the system treats the object's information as a probability distribution. Examples of object information include information about the object's position, movement, shape, size, color, and type. For example, the system treats information about whether an object exists in a certain area as a probability of existence, such as a 90% probability that the object is currently in area 1 and a 10% probability that it exists in area 2, which is different from area 1.

[0018] The system then uses the movement information to generate information such as the object's position from the time of prediction into the future as a probabilistic value. For example, the system predicts whether the object will be present in a certain area as a presence probability, such as a 50% probability that the object will be present in area 3 and a 30% probability that the object will be present in area 4 different from area 3, regarding the object's position after a predetermined period of time has elapsed from the time of prediction.

[0019] When using a probabilistic digital twin, the system may define the position of an object in a virtual space, for example, as follows. The system sets the virtual space in the form of a space-time. The space-time is provided with a plurality of, for example, a huge number of, unit areas in which a predetermined spatial area and a predetermined time area are determined in the space-time. In other words, the space-time is defined as a mesh structure composed of a plurality of unit areas. The system calculates the existence probability of an object at the time of prediction or the existence probability of an object after a predetermined period has elapsed from the time of prediction for each unit area in the space-time, and stores the calculated probability in association with information identifying the unit area. The system calculates the existence probability of an object after a predetermined period has elapsed from the time of prediction using the existence probability of the object at the time of prediction and information on the movement of the object. The information identifying the unit area may be, for example, coordinate information of the unit area in the space-time. As described above, the system is capable of calculating the existence probability of an object for a unit area in which an object may exist at the time of prediction or may exist after a predetermined period has elapsed from the time of prediction.

[0020] By their very nature, information acquired by sensors and information on 3D (three-dimensional) models of objects created in a virtual space by a digital twin may contain some degree of error. When a system employs a probabilistic digital twin approach, it takes into account errors in the acquired information to generate information such as an object's future location as a probabilistic value rather than a uniquely determined value. Therefore, when a system employs a probabilistic digital twin approach, it can predict an object's future location in a manner that tolerates ambiguity compared to when a conventional digital twin is employed. However, when a system employs a probabilistic digital twin approach, the amount of data required to construct a virtual space may be larger compared to when a conventional digital twin is employed. As described above, with a probabilistic digital twin, the system not only retains information on the presence or absence of an object, but also needs to handle information on the probability of the object's presence.

[0021] The virtual space reproduced by the digital twin may be any space, such as two-dimensional or three-dimensional. The virtual space may be expressed in any format as long as the position of an object existing in the virtual space can be identified. For example, the position information of an object in the virtual space may be coordinate information assigned on a two-dimensional or three-dimensional map of the digital twin, or may be represented by geographic coordinates such as latitude and longitude.

[0022] Furthermore, in the present disclosure, the term "mobile body" refers to any machine capable of movement, including, but not limited to, various mobile robots, vehicles, ships, and aircraft. Furthermore, the term "mobile body" may also refer to a machine capable of autonomous movement using a computer without the need for human operation or instruction. Examples of autonomously moving machines include, but are not limited to, automated guided vehicles, autonomous mobile robots, autonomous transport robots, unmanned forklifts, unmanned cranes, and drones. However, the term "mobile body" may also refer to a machine equipped with a system that assists humans in moving the mobile body by informing them of information output by a system of the present disclosure, as described below, to the mobile body.

[0023] In this disclosure, an "obstacle" refers to any object that, when present on the path of a moving object, prevents the moving object from moving along the path or requires the moving object to take action to avoid it. Actions to avoid it include, for example, detouring or temporarily stopping. An "obstacle" may be any movable object, such as various mobile robots, vehicles, ships, aircraft, as well as humans and other animals. Furthermore, an "obstacle" may also include stationary objects, such as structures such as racks and pillars.

[0024] In this disclosure, a "point" refers to a specific location in space, and ideally a specific point with no extension. Furthermore, an "area" refers to a specific range with a predetermined extension in space. For example, an "area" in a two-dimensional map of a virtual space in a digital twin refers to a range with a certain area, and an "area" in a three-dimensional map refers to a range with a certain volume. Furthermore, a "position" may refer to either a "point" or an "area."

[0025] First Embodiment (1A) Hereinafter, a first embodiment of the present disclosure will be described. In this (1A), a mobile object management system that does not include a mobile object and is realized by one or more computers will be described.

[0026] An example of the configuration of a mobile object management system will be described using Fig. 1. The mobile object management system 10 includes a prediction unit 11 and an output unit 12, and each unit of the mobile object management system 10 is controlled by a hardware controller (not shown). The prediction unit 11 and the output unit 12 will be described below.

[0027] The prediction unit 11 predicts the presence of an obstacle located on the route indicated by the route information of the mobile object, in response to receiving the route information of the mobile object from the mobile object. An obstacle being located on the route means that the obstacle is present in a predetermined area that includes part of the route. The route information of the mobile object may simply include information indicating the destination position of the mobile object. Alternatively, the route information of the mobile object may include information indicating the destination position of the mobile object and information indicating the time when the mobile object will arrive at the destination position. Hereinafter, the destination position will also be simply referred to as the destination position. The information indicating the destination position may be, for example, any of the following information: - Information indicating a position in a virtual space managed by the mobile object management system 10 - Information indicating a position in a virtual space managed by the mobile object - Information indicating a position in real space, such as latitude and longitude. Furthermore, the mobile object may receive not only the route information of the mobile object, but also trigger information that determines the content of information to be returned from the mobile object management system 10 to the mobile object. Details of the trigger information will be described later in embodiment 2. The mobile object management system 10 obtains the route information of the mobile object from the mobile object via wireless or wired communication.

[0028] For example, the prediction unit 11 references the digital twin data held by the mobile object management system 10 in response to acquiring the route information of the mobile object from the mobile object. Hereinafter, the digital twin data will also be referred to as twin data. By referencing the twin data, the prediction unit 11 determines which location in the virtual space represented by the digital twin the mobile object will move to. Then, the prediction unit 11 considers one or more objects other than the mobile object included in the twin data as obstacles, and predicts whether the obstacle is present at a predetermined position on the route of the mobile object in the virtual space.

[0029] Here, when the mobile object management system 10 constructs a virtual space using a typical digital twin technique, the prediction unit 11 may predict whether or not an obstacle is present at a position on the path of the mobile object in the virtual space. The prediction unit 11 uses information on the movement of the obstacle to generate information on the position of the obstacle in the virtual space at a time point a predetermined period of time has elapsed from the time of prediction as a unique value at the predetermined time point. Here, the prediction information at the predetermined time point generated by the prediction unit 11 is expressed as binary information indicating whether or not an obstacle is present.

[0030] When the route information of the moving body simply includes information indicating the moving position of the moving body, the prediction unit 11 predicts whether or not an obstacle is present at the moving position at one or more arbitrary time points. When an obstacle is present at the moving position of the moving body at any time point, the prediction unit 11 generates prediction information indicating that "an obstacle is present." On the other hand, when no obstacle is present at the moving position of the moving body at any time point, the prediction unit 11 generates prediction information indicating that "an obstacle is not present."

[0031] Furthermore, when the route information of the moving object includes information indicating the movement position and information indicating the time when the moving object will reach the movement position, the prediction unit 11 performs prediction as follows: The prediction unit 11 predicts whether or not an obstacle will be present at the movement position indicated in the route information at the time when the moving object will reach the movement position.

[0032] The prediction unit 11 may predict whether or not an obstacle exists, not for the destination position indicated by the route information, but for an area including the destination position. Hereinafter, the area including the destination position indicated by the route information or the area indicating the destination position will also be referred to simply as a destination area. The destination area is an area on the route indicated by the route information.

[0033] The prediction unit 11 can perform the above-described prediction process for one or more obstacles. When the prediction unit 11 performs the prediction process for multiple obstacles included in the twin data, the output unit 12 may output a prediction result indicating whether or not an obstacle is present at the movement position for each obstacle.

[0034] Alternatively, the prediction unit 11 may generate prediction results for multiple obstacles, indicating whether or not an obstacle is present at the movement position, and then generate a prediction result by integrating the generated prediction results. Here, "integration" refers to a process of reducing the number of prediction results for N obstacles, indicating whether or not an obstacle is present at the movement position, to N-1 or less, where N is a natural number equal to or greater than 2. For example, when the twin data includes multiple obstacles and one or more obstacles are present at the movement position, the prediction unit 11 may generate a single prediction result that "an obstacle is present."

[0035] On the other hand, when the mobile object management system 10 constructs a virtual space using a probabilistic digital twin technique, the prediction unit 11 may predict the probability that an obstacle will be present in an area on the path of the mobile object in the virtual space. The prediction unit 11 uses information about the movement of the obstacle to generate information about the obstacle's position in the virtual space from the predicted time to the future as a probability value between 0 and 1. Hereinafter, the probability value will be expressed as between 0 and 1.

[0036] When the route information of the moving object simply includes information indicating the moving position of the moving object, the prediction unit 11 performs prediction as follows: The prediction unit 11 predicts the probability that an obstacle exists in the target movement area for one or more arbitrary time points.

[0037] Furthermore, when the route information of the moving object includes information indicating the moving position of the moving object and information indicating the time when the moving object will reach the moving position, the prediction unit 11 performs prediction as follows: The prediction unit 11 predicts the probability that an obstacle will be present in the moving target area at the time when the moving object will reach the moving position.

[0038] The prediction unit 11 can execute the prediction process described above for one or more obstacles. When the prediction unit 11 executes the prediction process for multiple obstacles, the numerical values ​​generated individually as the existence probabilities of each obstacle may be used. Alternatively, the prediction unit 11 may integrate the existence probabilities of multiple obstacles. Here, "integration" refers to a process in which, when N numerical values ​​are generated as the existence probabilities of N obstacles, the existence probability numerical values ​​are reduced to N-1 or less. Note that N is a natural number of 2 or greater.

[0039] For example, assume that two obstacles O1 and O2 exist, and the prediction unit 11 predicts that the presence probabilities of the obstacles O1 and O2 in the target movement area are E1 and E2, respectively, where 1 > E1 > E2 > 0. In this case, the prediction unit 11 may integrate the presence probabilities E1 and E2 to calculate E1, which is the higher presence probability, as a single presence probability.

[0040] As another example, the prediction unit 11 may calculate αE1 and βE2, which are values ​​obtained by weighting E1 and E2 by predetermined coefficients α and β, respectively, and calculate the sum of the two values, E3 = αE1 + βE2, as a single existence probability. Note that α and β are determined, for example, so that E3 satisfies E1 + E2 > E3 > E1 and 1 > E3. As yet another example, the prediction unit 11 may calculate 1 - ((1 - E1) * (1 - E2)) as a single existence probability. In this way, the prediction unit 11 calculates the numerical value of the single existence probability.

[0041] The prediction unit 11 may determine the coefficients α and β by determining whether the obstacles O1 and O2 are objects having predetermined characteristics. The predetermined characteristics may refer to, for example, the type of obstacle, the speed of the obstacle, or the object carried by the mobile robot that is the obstacle. More specifically, the predetermined characteristics may refer to whether the obstacle is a small robot, whether the obstacle is a large unmanned crane, whether the obstacle has a slow speed of 5 km / h, whether the obstacle has a fast speed of 50 km / h, whether the obstacle is not carrying any cargo, or whether the obstacle is carrying any expensive cargo. Whether an obstacle has the predetermined characteristics may be determined as a result of classifying the obstacle using at least one of the following perspectives, for example: (1) Whether the obstacle is a human; (2) Whether the obstacle is an object smaller than a predetermined size; (3) Whether the obstacle is a predetermined type of machine; and (4) The speed of the obstacle at the time when the obstacle reaches the target movement area, as predicted by the prediction unit 11. Regarding (1) to (3), it can also be defined that the coefficients α and β are determined according to the type of obstacle. Note that (3) may be determined, for example, from the perspective of whether the obstacle is an expensive or important machine. As another example, (3) may be determined from the perspective of whether the robot that is the obstacle is carrying luggage, or whether the luggage being carried is an expensive or important machine. Furthermore, when using the perspective of (4), the prediction unit 11 may group multiple obstacles by determining whether the speed of the obstacle at the time when the obstacle reaches the target movement area is equal to or less than a predetermined threshold.

[0042] Furthermore, the prediction unit 11 is capable of executing the following processing regardless of whether a virtual space is constructed using a normal digital twin method or a probabilistic digital twin method.

[0043] The prediction unit 11 may calculate a single existence probability value for three or more obstacles by performing the same calculation as for two obstacles. However, the prediction unit 11 may, for example, divide the obstacles into two or more groups and, for each group, perform a calculation that integrates the existence probabilities of multiple obstacles included in the group. In this way, the prediction unit 11 can calculate the existence probability value for each group. Note that the prediction unit 11 may group the obstacles based on at least one of the above-mentioned viewpoints (1) to (4).

[0044] The prediction unit 11 may also change the size of a target movement area on a route for which the presence of an obstacle is predicted, depending on the size of the obstacle. The size of the target movement area refers to the size in the virtual space of the digital twin. For example, if the virtual space is two-dimensional, the size of the target movement area indicates the size of the area. If the virtual space is three-dimensional, the size of the target movement area indicates the size of the volume. The size of the obstacle indicates, for example, the volume of the obstacle, the longest side of the obstacle, or the largest area of ​​each face of the obstacle. The prediction unit 11 may acquire information on the size of the obstacle from a database within the mobile object management system 10 or from outside the mobile object management system 10. For example, the prediction unit 11 may acquire information on the size of the obstacle from a sensor that detects the external environment.

[0045] For example, the prediction unit 11 may set two or more stages of the size of the target area to be moved and one or more thresholds corresponding to the sizes of the target areas to be moved. The prediction unit 11 can flexibly set the size of the target area to be moved according to the size of the obstacle by comparing the size of the obstacle with the threshold. As another example, the prediction unit 11 may continuously change the size of the target area to be moved so that it monotonically increases or becomes monotonically non-increasing with the size of the obstacle. In this way, the prediction unit 11 can arbitrarily change the size of the target area to be moved according to the size of the obstacle.

[0046] The prediction unit 11 may also set whether to predict the presence of an obstacle in a nearby area located near the target area of ​​movement, depending on at least one of, for example, the characteristics of the obstacle, the characteristics of the moving body, the width of the route traveled by the moving body, or the distance between the moving body and the obstacle. The characteristics of the obstacle refer to, for example, the type of obstacle, the speed of the obstacle, etc. The characteristics of the moving body refer to, for example, the type of moving body, the speed of the moving body, the type of cargo carried by the moving body, etc. The nearby area is, for example, an area adjacent to the target area of ​​movement, or an area within a predetermined distance from the target area of ​​movement. The width of the route traveled by the moving body may be, for example, the width of a route that the moving body can travel, which is preset in the digital twin. Alternatively, the width of the route traveled by the moving body may be defined using information received from the moving body. Furthermore, the width of the route traveled by the moving body may be defined using the size of the moving body itself, or may be defined based on the size of the cargo carried by the moving body.

[0047] For example, the prediction unit 11 may compare the width of the route traveled by the moving object with a predetermined threshold value. When the width of the route traveled by the moving object is greater than the predetermined threshold value, the prediction unit 11 may set a nearby area located near the target area of ​​movement as a target for predicting the presence of an obstacle. On the other hand, when the width of the route traveled by the moving object is equal to or smaller than the predetermined threshold value, the prediction unit 11 may set a nearby area located near the target area of ​​movement as a target for predicting the presence of an obstacle.

[0048] As another example, the prediction unit 11 may compare the distance between the moving body and the obstacle with a predetermined threshold. When the distance between the moving body and the obstacle is equal to or less than the predetermined threshold, the prediction unit 11 may set a nearby area located near the target area to be included in the prediction of the presence of an obstacle. On the other hand, when the distance between the moving body and the obstacle is greater than the predetermined threshold, the prediction unit 11 may set a nearby area located near the target area not to be included in the prediction of the presence of an obstacle.

[0049] As another example, the prediction unit 11 may compare the speed or acceleration of the obstacle at the time of prediction with a predetermined threshold value. If the speed or acceleration of the obstacle is equal to or greater than the predetermined threshold value, the prediction unit 11 may set a nearby area located near the target movement area as a target for predicting the presence of an obstacle. On the other hand, if the speed or acceleration of the obstacle is less than the predetermined threshold value, the prediction unit 11 may set a nearby area located near the target movement area as a target for predicting the presence of an obstacle.

[0050] As a more detailed example, the prediction unit 11 may change the size of the neighborhood area in which the presence of an obstacle is predicted, depending on at least one of the width of the route along which the moving object will pass, the distance between the moving object and the obstacle at the time of prediction, or the speed or acceleration of the obstacle at the time of prediction. For example, the prediction unit 11 may set two or more levels of neighborhood area size and one or more thresholds corresponding to the sizes of each neighborhood area. The prediction unit 11 can set two or more levels of neighborhood area size by comparing at least one of the width of the route along which the moving object will pass, the distance between the moving object and the obstacle at the time of prediction, or the speed or acceleration of the obstacle at the time of prediction with a threshold. However, the method for setting the size of the neighborhood area is not limited to this. In this way, the prediction unit 11 can arbitrarily change the size of the neighborhood area.

[0051] Furthermore, the prediction unit 11 may include not only moving objects but also stationary objects as obstacles to be predicted. That is, the prediction unit 11 may predict whether an obstacle, which is a stationary object, is located at the movement position or the target movement area of ​​the moving object in the virtual space. When the mobile object management system 10 constructs the virtual space using a conventional digital twin method and an obstacle is present at the movement position or the target movement area of ​​the moving object, the prediction unit 11 generates prediction information indicating that "an obstacle is present." On the other hand, when an obstacle is not present at the movement position or the target movement area of ​​the moving object, the prediction unit 11 generates prediction information indicating that "an obstacle is not present."

[0052] Furthermore, when the mobile object management system 10 constructs a virtual space using a probabilistic digital twin technique and an obstacle is present at the moving position or in the target area of ​​the moving object, the prediction unit 11 generates prediction information indicating that the probability of the obstacle's presence is 1 or a value close to 1. On the other hand, when an obstacle is not present at the moving position or in the target area of ​​the moving object, the prediction unit 11 generates prediction information indicating that the probability of the obstacle's presence is 0 or a value close to 0. For example, the prediction unit 11 determines whether an object included in the digital twin is stationary based on information about the object's movement. If the object is stationary, the prediction unit 11 executes the above process.

[0053] However, the prediction unit 11 does not need to include stationary objects as obstacles in the prediction of their presence. For example, the prediction unit 11 determines whether an object included in the digital twin is stationary, and if the object is a moving object, sets the object as an obstacle in the prediction of its presence. The prediction unit 11 then generates information indicating whether an obstacle exists or the probability of the obstacle's existence for the set object. If the object is stationary, it is easy for a moving body to detect the object's presence and take evasive action to prevent contact. Therefore, it is not necessary for the prediction unit 11 to include stationary objects as obstacles in the prediction of their presence, and for the output unit 12 to provide information about the stationary object to the moving body.

[0054] Furthermore, the prediction unit 11 may use, for example, a pre-trained AI (Artificial Intelligence) model to generate information indicating whether an obstacle exists or information indicating the probability of the existence of an obstacle. The AI ​​model used by the prediction unit 11 is trained by inputting training data to the AI ​​model, the training data including information on the position and movement of each sample obstacle at a predetermined time point and information on the position of each obstacle after a predetermined time has elapsed from the predetermined time point. The information on the position of each obstacle after a predetermined time has elapsed from the predetermined time point is the correct label in the training data. Any technique, such as logistic regression or a neural network, can be used as the training method.

[0055] When the prediction unit 11 acquires route information from the moving body, it inputs information on the position and movement of each obstacle currently included in the twin data to an AI model trained using teacher data. Based on the input information, the AI ​​model outputs information on the position of each obstacle after a predetermined time has elapsed since the time of prediction. The time after the predetermined time has elapsed since the time of prediction is, for example, the time when the moving body arrives at the movement position indicated in the route information. The AI ​​model outputs the position information of each obstacle as information indicating whether or not the obstacle exists, or information indicating the probability of the obstacle's existence. The prediction unit 11 generates information indicating whether or not the obstacle exists, or information indicating the probability of the obstacle's existence, by comparing the position information output by the AI ​​model with the movement position information indicated in the route information.

[0056] However, the prediction unit 11 may generate information indicating whether an obstacle is present or information indicating the probability of the obstacle's presence by using an algorithm based on a predefined rule base instead of an AI model. For example, assume that the mobile object management system 10 acquires information that a transport robot carrying luggage passes through a certain point A at time t. In this case, the algorithm may use information constituting the virtual space of the digital twin to generate information indicating whether the transport robot is present in the area surrounding point A before and after time t or information indicating the probability of the obstacle's presence. Here, the information constituting the virtual space includes, for example, information on one or more routes that the transport robot can travel.

[0057] The output unit 12 outputs to the mobile body information relating to the presence of an obstacle predicted by the prediction unit 11. The output unit 12 is configured as, for example, an interface used for wireless or wired communication with the mobile body.

[0058] As described above, the information regarding the presence of an obstacle output by the output unit 12 may be information indicating whether or not an obstacle exists, or may be information indicating the probability of the obstacle's existence. When multiple obstacles are included in the twin data, the output unit 12 may output a numerical value obtained by individually calculating the probability of the presence of each obstacle. Alternatively, the output unit 12 may output a numerical value obtained by integrating the probabilities of the presence of multiple obstacles.

[0059] When the prediction unit 11 generates information indicating the probability of an obstacle's existence, the output unit 12 may compare the generated value of the probability of the obstacle's existence with a predetermined threshold value to determine whether or not to output the information generated by the prediction unit 11. A detailed description will be given later in the second embodiment.

[0060] [Explanation of Processing Flow] Fig. 2 is a flowchart showing an example of a typical process of the mobile object management system 10, and the flowchart in Fig. 2 explains an overview of the process of the mobile object management system 10. Note that the details of each process are as described above, and therefore will not be explained as appropriate.

[0061] In step S11, the prediction unit 11 predicts the presence of an obstacle located on the route indicated by the route information acquired from the mobile object. In step S12, the output unit 12 outputs information regarding the presence of the obstacle predicted by the prediction unit 11 to the mobile object.

[0062] In addition, in response to acquiring a plurality of pieces of route information from a mobile object, the mobile object management system 10 may execute the process shown in FIG. 2 for at least one of the plurality of pieces of route information.

[0063] 2 for each of a plurality of moving objects. Here, the mobile object management system 10 executes the process shown in Fig. 2, and the moving object A for which information regarding the presence of an obstacle is output may be regarded as an obstacle in the prediction process of step S11 for another moving object B. In other words, upon acquiring route information from the moving object B, the prediction unit 11 can recognize the moving object A located on the route indicated by the route information as an obstacle and predict the presence of the moving object A.

[0064] [Explanation of Effect] As described above, the mobile object management system 10 can predict the presence of an obstacle located on the path of a mobile object and output information regarding the presence of the obstacle to the mobile object. When communicating with the mobile object to safely move the mobile object, the mobile object management system 10 does not need to output information regarding the virtual space reproduced by the digital twin to the mobile object. Therefore, the mobile object management system 10 can reduce the communication load with the mobile object when determining the path of the mobile object.

[0065] The prediction unit 11 may also predict the probability of an obstacle being present in the target area on the route at the time the mobile object moves into the target area. The output unit 12 outputs information on the presence probability predicted by the prediction unit 11 to the mobile object. Compared to when the output unit 12 outputs binary information indicating whether an obstacle is present or not, this can provide the mobile object with highly accurate information regarding the presence of an obstacle. Therefore, the mobile object can use the highly accurate information to determine a safer route.

[0066] Furthermore, when the prediction unit 11 predicts the presence probability for each of a plurality of obstacles, the output unit 12 may output information on the integrated presence probability for the plurality of obstacles to the mobile body. When the integrated presence probability information is output, the mobile body management system 10 can reduce the amount of information output to the mobile body and further reduce the communication load with the mobile body, compared to when information on the presence probability for each of a plurality of obstacles is output.

[0067] Furthermore, the prediction unit 11 may predict the probability of an obstacle's existence using a probabilistic digital twin that treats information on whether an object exists in a certain area as an existence probability.

[0068] As described in the "Definition" section above, the mobile object management system 10 constructs a virtual space using a probabilistic digital twin technique, enabling accurate prediction of the future position of an object. However, when the probabilistic digital twin technique is adopted, the amount of data required to construct the virtual space may be larger than when a conventional digital twin is adopted. Therefore, if the mobile object management system 10 outputs information about the virtual space reproduced by the probabilistic digital twin to a mobile object, it is expected that the communication load with the mobile object will be large. However, in order to safely move the mobile object, the mobile object management system 10 only needs to output information about the existence probability regarding the path of the mobile object to the mobile object; it is not necessary to output all information about the virtual space reproduced by the probabilistic digital twin to the mobile object. Therefore, when using the probabilistic digital twin technique, the mobile object management system 10 can effectively reduce the communication load with the mobile object.

[0069] In addition, the prediction unit 11 may change the size of the area on the route in which the presence of an obstacle is predicted, depending on, for example, at least one of the characteristics of the obstacle, the characteristics of the moving body, the width of the route passed by the moving body, or the distance between the moving body and the obstacle.

[0070] There are various types of obstacles, ranging from small mobile robots to large unmanned forklifts or unmanned cranes. When the obstacle is large, the possibility that the mobile object will come into contact with or approach within a predetermined distance from the obstacle is higher than when the obstacle is small. Therefore, when the obstacle is large, the prediction unit 11 increases the size of the target movement area compared to when the obstacle is small, thereby enabling prediction of the presence of the obstacle with greater consideration for safety. The mobile object can then determine a safer route using the information about the presence of the obstacle output from the output unit 12.

[0071] Conversely, when an obstacle is small, the mobile body is more likely to be able to avoid the obstacle by temporarily stopping or taking a detour on the route compared to when the obstacle is large. Therefore, the prediction unit 11 reduces the size of the target movement area when the obstacle is small compared to when the obstacle is large. If an obstacle is present in a position that the mobile body can avoid while moving, if the size of the target movement area remains large, the prediction unit 11 may predict that the obstacle is present throughout the entire target movement area. In this case, it is expected that the mobile body will change a route that was originally safe using information about the presence of the obstacle output from the output unit 12. This may result in a longer travel time for the mobile body compared to the original route. However, by reducing the size of the target movement area, the prediction unit 11 is less likely to predict that the obstacle is present throughout the entire area before changing the size of the target movement area. This may increase the likelihood that the mobile body can move along the originally set route or reduce the extent to which the mobile body will change the originally set route. In other words, this may lead to a shorter travel time for the mobile body.

[0072] Furthermore, the width of the route through which the moving body travels may change depending on at least one of the factors of the size of the moving body itself and the environment in which the moving body is located. When the width of the route through which the moving body travels is large, the possibility that the moving body will come into contact with an obstacle or come within a predetermined distance of the obstacle is higher than when the width of the route is small. When the distance between the moving body and the obstacle is short, or when the speed or acceleration of the obstacle is high, the possibility that the moving body will come into contact with the obstacle or come within a predetermined distance of the obstacle is higher than when this is not the case.

[0073] Therefore, the prediction unit 11 can predict the presence of obstacles in the vicinity area as well depending on the situation, thereby making it possible to predict the presence of obstacles with greater consideration given to safety.The mobile object can then determine a safer route using the information about the presence of obstacles output from the output unit 12.

[0074] Furthermore, the prediction unit 11 may predict the presence of both moving and stationary objects as obstacles. By predicting the presence of stationary objects as well, the prediction unit 11 can make predictions that take into account objects that the moving body may come into contact with more precisely. The moving body can then determine a safer route using the information about the presence of obstacles output from the output unit 12.

[0075] The information on the various thresholds used by the prediction unit 11 may be changed as appropriate. The mobile object management system 10 may autonomously change the threshold information. Alternatively, when the mobile object management system 10 receives an instruction to change the threshold from the mobile object 21, the mobile object management system 10 may change the threshold information in accordance with the received instruction. When there are multiple mobile objects managed by the mobile object management system 10, the threshold information may be different for each mobile object, or the same threshold information may be used for multiple mobile objects.

[0076] When each unit of the mobile object management system 10 is realized by multiple computers, the method of distributing each unit of the mobile object management system 10 among the multiple computers is arbitrary. As an example, the mobile object management system 10 may be configured by installing the prediction unit 11 in a first computer and the output unit 12 in a second computer, and connecting the first and second computers. However, the mobile object management system 10 may also be realized by a single computer by installing the prediction unit 11 and the output unit 12 in a single computer.

[0077] Some or all of the components of the mobile object management system 10 may be provided on a cloud server built on a cloud, or on other types of virtualized servers generated using virtualization technology, etc. Functions other than those provided on servers such as cloud servers or virtualized servers are placed on edges. For example, in a system that monitors video captured near a mobile object via a network, edges are devices placed at or near the site, and are also devices that are close to the terminal in terms of the network hierarchy.

[0078] (1B) A variation of the first embodiment will be described below. In (1B), a mobile object control system will be described. The mobile object control system of (1B) is realized by a plurality of computers including mobile objects.

[0079] 3 includes a mobile object 21 in addition to a prediction unit 11 and an output unit 12. The processes executed by the prediction unit 11 and the output unit 12 are the same as those in (1A), and therefore will not be described here. The prediction unit 11 and the output unit 12 may also constitute a mobile object management system as shown in (1A).

[0080] The moving body 21 acquires information about the presence of an obstacle predicted by the prediction unit 11 from the output unit 12. Then, the moving body 21 determines the route of the moving body 21 based on the information about the presence of an obstacle.

[0081] For example, the mobile object 21 transmits route information including information indicating the moving position of the mobile object 21 as a query request to a mobile object management system including the prediction unit 11 and the output unit 12. Then, when the mobile object 21 acquires information regarding the presence of an obstacle predicted by the prediction unit 11, the mobile object 21 determines whether or not there is a moving position on the route of the route information where an obstacle exists or where the probability of the existence of an obstacle is equal to or greater than a predetermined threshold.

[0082] If there is one or more travel positions on the route in the route information where an obstacle exists or where the probability of an obstacle existing is equal to or greater than a predetermined threshold, the mobile body 21 can change the route indicated in the route information. For example, the mobile body 21 may reconfigure the route indicated in the route information so that the route does not include a travel position where the probability of an obstacle existing is equal to or greater than a predetermined threshold. Alternatively, the mobile body 21 may reconfigure the route so as to change the time at which the mobile body 21 arrives at a travel position where the probability of an obstacle existing is equal to or greater than a predetermined threshold.

[0083] On the other hand, if there is no movement position on the route indicated in the route information where an obstacle exists or the probability of an obstacle existing is greater than or equal to a predetermined threshold, the mobile body 21 does not need to change the route indicated in the route information.

[0084] The moving body 21 is mounted on a computer different from the computer on which the prediction unit 11 is mounted and the computer on which the output unit 12 is mounted. The moving body 21 is controlled by a hardware controller (not shown) in the moving body 21.

[0085] The effects of the mobile object control system 20 shown in (1B) are similar to those of the mobile object management system 10 shown in (1A), and therefore will not be described further.

[0086] In the following embodiment 2, a specific example of the mobile object control system described in embodiment 1 will be disclosed. However, the specific example of the mobile object control system described in embodiment 1 is not limited to the one shown below. Furthermore, the configurations and processes described below are examples and are not limited to these.

[0087] (2A) [Configuration] The robot control system 30 shown in Fig. 4 includes robots 100A to 110C and a management server 200. The robots 100A to 110C have the same configuration, and hereinafter the robots 100A to 110C will be collectively referred to as robot 100. The robot 100 is wirelessly connected to the management server 200, and as described below, communicates with the management server 200 regarding inquiries including information on route candidates and responses to the inquiries. The management server 200 is a server installed in a cloud environment and located at a location separate from the robot 100.

[0088] In the second embodiment, the robot 100 is an automated guided vehicle that moves within a work area in the real world to transport luggage, but the example of the robot 100 is not limited to this. Also, although three robots 100 are illustrated in Fig. 4, any number of robots 100 may be provided. The configurations of the robot 100 and the management server 200 will be described below.

[0089] 5 is a block diagram showing an example of a robot 100. The robot 100 includes a route candidate generating unit 101, a transmitting / receiving unit 102, a possibility evaluating unit 103, and a route determining unit 104.

[0090] The route candidate generation unit 101 generates candidates for movement routes of the robot 100. Hereinafter, the candidate movement routes will also be simply referred to as route candidates. The route candidate generation unit 101 may generate route candidates using, for example, pre-set route information. Here, the route candidates include information indicating each movement position to which the robot 100 will move and information indicating the time when the robot 100 will arrive at each movement position. The route candidate generation unit 101 may include information on any number of movement positions and information indicating the corresponding time in the route candidates.

[0091] However, the route candidate generation unit 101 may generate route candidates using environmental information around the robot 100 acquired from a sensor of the robot 100. In either case, the route candidate generation unit 101 can generate travel route candidates using any known method. The route candidate generation unit 101 may also generate multiple route candidates. The route candidate generation unit 101 outputs information on the generated route candidates for the robot 100 to the transmission / reception unit 102.

[0092] The transmitting / receiving unit 102 includes information about the route candidates output from the route candidate generation unit 101 in an area information request and transmits the area information request to the management server 200. The transmitting / receiving unit 102 may include information about multiple route candidates in the area information request. The area information request queries the management server 200 about the probability that an obstacle exists in an area that includes a route candidate. Hereinafter, the probability that an obstacle exists will also be simply referred to as the existence probability. Furthermore, an area that includes a route candidate refers to an area that includes each of the travel positions that are destinations in the route candidate, and is composed of one or more small areas. The small areas are set by the route division unit 202, and will be described in detail below.

[0093] The route candidate generation unit 101 may represent information about route candidates using the same map as the map of virtual space used in the management server 200. In this case, the management server 200 can use the area information transmitted from the transmission / reception unit 102 as is to execute the processing described below. However, the route candidate generation unit 101 may represent information about route candidates using a map different from the map of virtual space used in the management server 200. In this case, the management server 200 converts the map format of the transmitted area information into the map format of virtual space used in the management server 200. Thereafter, the management server 200 executes processing using the area information.

[0094] The transmitting / receiving unit 102 also receives area information transmitted by the management server 200 in response to an area information request. The area information includes information on route candidates and information on the probability of obstacles being present in the area including the route candidates, which information is derived by the management server 200. The transmitting / receiving unit 102 may receive information on route candidates and information on the probability of obstacles being present for a plurality of route candidates. The transmitting / receiving unit 102 outputs the received area information to the possibility evaluation unit 103.

[0095] The transmitting / receiving unit 102 is configured as an interface used for wireless communication with the management server 200. Note that the information that the transmitting / receiving unit 102 transmits and receives with the management server 200 is not limited to the area information request and the area information. The transmitting / receiving unit 102 corresponds to the output unit 12 according to the first embodiment.

[0096] The possibility evaluation unit 103 evaluates at least one of the possibility that the robot 100 will come into contact with an obstacle or the possibility that the robot 100 will approach within a predetermined distance when the robot 100 moves along a candidate route by referring to the area information. For example, the possibility evaluation unit 103 may evaluate at least one of the possibility that the robot 100 will come into contact with an obstacle or the possibility that the robot 100 will approach within a predetermined distance by comparing the probability of an obstacle existing on the candidate route with a predetermined threshold. Alternatively, the possibility evaluation unit 103 may determine the number of obstacles existing on the candidate route with a probability equal to or greater than a predetermined threshold by referring to the area information. The possibility evaluation unit 103 may evaluate the possibility that the robot 100 will come into contact with an obstacle or approach within a predetermined distance using the probability of the obstacle existing and the number of obstacles. For example, the possibility evaluation unit 103 may evaluate the possibility that the robot 100 will come into contact with an obstacle or approach within a predetermined distance as either "high" or "low."

[0097] When the area information includes multiple route candidates, the possibility evaluation unit 103 may use the method described above to evaluate, for each route candidate, the possibility that the robot 100 will come into contact with or approach within a predetermined distance from an obstacle. The possibility evaluation unit 103 outputs the evaluation result described above to the route determination unit 104.

[0098] If the area information includes multiple route candidates, the possibility evaluation unit 103 may use the following method to evaluate the possibility that the robot 100 will come into contact with an obstacle or approach within a predetermined distance. For example, the possibility evaluation unit 103 may prioritize the multiple route candidates in descending order of the probability that the robot 100 will come into contact with an obstacle or approach within a predetermined distance. Alternatively, the possibility evaluation unit 103 may prioritize the multiple route candidates based not only on the probability that the robot 100 will come into contact with an obstacle or approach within a predetermined distance, but also on the number of obstacles that the robot 100 will come into contact with or approach within a predetermined distance. In either case, the possibility evaluation unit 103 can prioritize the multiple route candidates using any known method. The possibility evaluation unit 103 outputs information on the prioritized multiple route candidates to the route determination unit 104 as an evaluation result.

[0099] The path determination unit 104 determines a movement path for the robot 100 based on the evaluation result output by the possibility evaluation unit 103. For example, if the possibility evaluation unit 103 evaluates that the possibility that the robot 100 will come into contact with an obstacle or approach within a predetermined distance for a predetermined path candidate is "high," the path determination unit 104 does not adopt the predetermined path candidate as the movement path for the robot 100. By not adopting the predetermined path candidate as the movement path for the robot 100, the path determination unit 104 can determine a safer path.

[0100] On the other hand, when the possibility evaluation unit 103 evaluates that the possibility of the robot 100 coming into contact with an obstacle or approaching within a predetermined distance is "low" for the predetermined route candidate, the route determination unit 104 adopts the predetermined route candidate as the movement route of the robot 100. The predetermined route candidate is, for example, a route determined in consideration of shortening the movement time. Therefore, the route determination unit 104 can determine a route that is not only safe but also allows efficient movement.

[0101] It is also possible that the possibility evaluation unit 103 evaluates the possibility that the robot 100 will come into contact with or approach within a predetermined distance from an obstacle as "low" for multiple route candidates. In this case, the route determination unit 104 can adopt any route candidate from the multiple route candidates as the movement route of the robot 100.

[0102] It is also assumed that the possibility evaluation unit 103 may assign higher priorities to multiple route candidates in order of the lowest probability that the robot 100 will come into contact with or approach within a predetermined distance from an obstacle. Here, the route determination unit 104 may adopt, for example, the route candidate assigned the highest priority as the movement route of the robot 100. Alternatively, the route determination unit 104 may select one route candidate from multiple route candidates assigned priorities within a predetermined order and adopt it as the movement route of the robot 100. In either case, the route determination unit 104 can execute a process of adopting or not adopting the evaluated route candidate as the movement route of the robot 100 using any known method.

[0103] 6 is a block diagram showing an example of a management server 200. The management server 200 includes a transmitter / receiver 201, a route dividing unit 202, a database 203, an object position predictor 204, and a target area determiner 205.

[0104] The transmitting / receiving unit 201 receives the area information request transmitted from the transmitting / receiving unit 102 of the robot 100. The transmitting / receiving unit 201 outputs the received area information request to the route dividing unit 202. As described above, the area information request may include information on multiple route candidates.

[0105] Furthermore, the transmitting / receiving unit 201 acquires obstacle presence probability information in an area including a route candidate predicted by the target area determination unit 205 in response to the area information request. The transmitting / receiving unit 201 transmits area information including information on the route candidate and obstacle presence probability information in the area including the route candidate to the robot 100. As described above, the area information may include information on multiple route candidates.

[0106] The transmitting / receiving unit 201 is configured as an interface used for wireless communication with the robot 100. Note that the information transmitted and received by the transmitting / receiving unit 201 to and from the robot 100 is not limited to the area information request and area information.

[0107] The route division unit 202 projects information about the route candidates indicated in the area information request onto a map of virtual space. Then, the route division unit 202 sets small areas on the map that include each of the movement positions that will be the movement destinations on the route candidates. When the route candidates are indicated on a map of the virtual space in the probabilistic digital twin, the small areas indicate areas into which the map is divided, and correspond to the movement target areas described in the first embodiment.

[0108] FIG. 7 shows a case where the path dividing unit 202 has set small regions for a certain path candidate K1. In FIG. 7, the small region is a square with a side length of F1. On the path candidate K1 on the map M, which is a two-dimensional map, the robot 100 starts from the small region D1, passes through the small region D5, and reaches the small region D15 within the period from time t=0 to t=15. The robot 100 is located in the small region D1 at time t=0, in the small region D5 at time t=5, and in the small region D15 at time t=15. Each small region on the map M is a region where the presence of an obstacle is predicted. The path dividing unit 202 can set any shape for the small region. The illustrated times are examples of time points.

[0109] The route dividing unit 202 can change the size of each small area on the map M depending on the size of the obstacle to be predicted. For example, the route dividing unit 202 may set a first threshold Th1 and a second threshold Th2 for the size of the target movement area, where Th1>Th2. The route dividing unit 202 compares the size Sc of the obstacle, which is information acquired from the database 203 (described later), with Th1 and Th2 to determine which is larger.

[0110] If Sc > Th1, the path dividing unit 202 sets the small region to a square with a side length of F1. If Th1 > Sc > Th2, the path dividing unit 202 sets the side length of the small region to F1 / 2. That is, the path dividing unit 202 sets the area of ​​the small region to 1 / 4. If Th2 > Sc, the path dividing unit 202 sets the side length of the small region to F1 / 3. That is, the path dividing unit 202 sets the area of ​​the small region to 1 / 9. However, variations in the method by which the path dividing unit 202 changes the size of each small region are not limited to this. Examples of variations are as described in the first embodiment.

[0111] When the area information request includes information on multiple route candidates, the route dividing unit 202 may project two or more route candidates onto the map of the same virtual space. However, the route dividing unit 202 may also project each of the route candidates onto a map of a different virtual space.

[0112] The route dividing unit 202 may also acquire information on multiple obstacles from the database 203. Here, since the obstacles vary in size, it is assumed that the size of each small area on the map corresponding to each obstacle may vary depending on the obstacle. Here, the route dividing unit 202 may be configured to handle multiple obstacles with small areas of the same size on the same map onto which one or more route candidates are projected. Furthermore, the route dividing unit 202 may set small areas on different maps for multiple obstacles with small areas of different sizes. The route dividing unit 202 projects one or more route candidates onto each of the different maps.

[0113] The database 203 stores information on not only moving objects but also stationary objects as obstacles.

[0114] The route dividing unit 202 outputs map information including the route candidates and information on the small areas set as described above to the target area determining unit 205 .

[0115] Information about obstacles that exist in the real space is stored in association with the virtual space in the probabilistic digital twin in the database 203. Here, the information about the obstacles includes information about the movement of the obstacles as well as information about the shapes and sizes of the obstacles.

[0116] Specifically, information on the position, direction of travel, and speed of an obstacle in real space is stored as information on the movement of the obstacle in database 203. Here, the information on the movement of the obstacle stored in database 203 is represented as the position where the obstacle exists on a map in virtual space, and the direction of travel and speed on the map. In addition, information such as the shape and size of the obstacle in real space is also stored in database 203 as information on the shape and size of the obstacle on the map in virtual space.

[0117] The information about the obstacles described above is acquired by devices such as cameras or sensors installed in the real space, and is stored in the database 203. Note that the database 203 may store information about a plurality of obstacles.

[0118] The object position prediction unit 204 uses obstacle information acquired from the database 203 to predict, as position prediction information, the position of an obstacle at each time point indicated by the route candidate. Specifically, the object position prediction unit 204 uses a probabilistic digital twin to predict, as position prediction information, the existence probability of an obstacle being present on the map of the virtual space. The existence probability is as described in the first embodiment. Note that the object position prediction unit 204 may generate position prediction information for multiple obstacles.

[0119] Note that the object position prediction unit 204 may also predict, as position prediction information, the position where an obstacle exists at each time point indicated by the route candidate, for the information on stationary objects stored in the database 203. Here, the object position prediction unit 204 sets the existence probability of a stationary object at a position where the stationary object exists to a value of 1 or close to 1, and sets the existence probability of a stationary object at other positions to a value of 0 or close to 0. In other words, the object position prediction unit 204 sets the variance of the probability distribution at the position of a stationary object to a value of 0 or close to 0.

[0120] The target area determination unit 205 acquires map information from the route division unit 202 and acquires obstacle position prediction information from the object position prediction unit 204. The target area determination unit 205 projects the obstacle position prediction information onto the map information. Then, the target area determination unit 205 derives the presence probability that an obstacle exists in the small area in which the robot 100 is located at each time point indicated by the route candidate. The route division unit 202, the object position prediction unit 204, and the target area determination unit 205 correspond to the prediction unit 11 according to the first embodiment.

[0121] 8A shows a situation in which the target area determination unit 205 projects position prediction information for a certain obstacle A onto the map M shown in FIG. 7. Obstacle A is a moving obstacle. In FIG. 8A, an area where the probability of the obstacle's existence is greater than 0 is shown as area A1. Note that the object position prediction unit 204 generates position prediction information for obstacle A such that the probability of the robot 100's existence increases from the edge to the center of area A1.

[0122] For example, assume that the target area determination unit 205 derives the probability of the presence of obstacle A at time t = 5. Here, the target area determination unit 205 evaluates the degree of overlap between small area D5, in which robot 100 exists at time t = 5, and area A1. The target area determination unit 205 refers to the map shown in FIG. 8A and determines that part of small area D5 is included in the edge of area A1. Then, the target area determination unit 205 determines that the probability that obstacle A exists in small area D5 at time t = 5 is 0.1.

[0123] 8B shows a situation in which the target area determination unit 205 projects position prediction information for an obstacle B onto the map M shown in FIG. 7. Obstacle B is a moving obstacle. The path division unit 202 also sets the size of the small area corresponding to obstacle B to be the same as the size of the small area corresponding to obstacle A. In other words, the small area corresponding to obstacle A and the small area corresponding to obstacle B are both squares with one side measuring F1.

[0124] 8B, the region where the probability of an obstacle's existence is greater than 0 is shown as region B1. The object position prediction unit 204 generates the position prediction information for obstacle B so that the probability of the robot 100's existence increases from the edge to the center of region B1. Here, the area of ​​region B1 is smaller than the area of ​​region A1.

[0125] The target area determination unit 205 evaluates the degree of overlap between small area D5, where the robot 100 exists, and area B1 at time t = 5. The target area determination unit 205 refers to the map shown in Figure 8B and determines that part of small area D5 is included in the edge of area B1. The target area determination unit 205 then determines that the probability that obstacle B exists in small area D5 at time t = 5 is 0.2.

[0126] 8A and 8B, the area of ​​region B1 is smaller than the area of ​​region A1, and therefore region B1 has a higher probability density per area than region A1. Therefore, the probability that obstacle B exists in small region D5 at time t=5 is higher than the probability that obstacle A exists in small region D5.

[0127] In this way, the target area determination unit 205 can derive the presence probabilities of obstacles A and B at time t = 5. The target area determination unit 205 derives the presence probabilities of obstacles A and B in the small area in which the robot 100 is located, using a similar method, at each time within the period from time t = 0 to t = 15 in route candidate K1. Note that because obstacles A and B are moving, area A1 and area B1 are located in different positions at each time. The target area determination unit 205 derives the presence probabilities of obstacles A and B in the small area in which the robot 100 is located at each time, thereby deriving the presence probabilities of obstacles for an area including the entire route of route candidate K1.

[0128] The target area determination unit 205 associates the derived prediction result information with information indicating the position of the small area being determined, and outputs the information to the transmission / reception unit 201. The prediction result information includes information indicating the presence probability at each point in time for multiple obstacles A and B. The transmission / reception unit 201 transmits area information to the robot 100, including information on the route candidates and information on the presence probability of obstacles in the area including the route candidates.

[0129] If the target area determination unit 205 derives the presence probability for multiple obstacles A and B at each time point, the transceiver unit 201 may transmit the presence probability for each obstacle individually. Alternatively, the target area determination unit 205 may integrate the presence probabilities of obstacles A and B to calculate a single presence probability value. For example, in the example shown in FIGS. 8A and 8B , the presence probability of obstacle A in the small area D5 at time t = 5 is derived as 0.1, and the presence probability of obstacle B is derived as 0.2. In this case, the target area determination unit 205 may integrate the presence probabilities of 0.1 and 0.2 to calculate the higher presence probability value of 0.2. However, the example of the integration process performed by the target area determination unit 205 is not limited to this. The transceiver unit 201 transmits the integrated obstacle presence probability information to the robot 100.

[0130] In the example described above, the path dividing unit 202 sets the size of the small areas to be determined to be the same for obstacle A and obstacle B. However, since obstacle A and obstacle B are different in size, it is conceivable that the size of the small areas to be determined for each obstacle will be different.

[0131] 8C shows an example in which the path dividing unit 202 sets the small region corresponding to the obstacle B as a square with a side length of F1 / 2. In FIG. 8C, the region D5 in FIG. 8B is divided and shown.

[0132] For example, if the size of obstacle A is AL and the size of obstacle B is BL, the path dividing unit 202 determines the magnitude relationship between AL, BL, and the first threshold value Th1 and the second threshold value Th2 as follows: AL > Th1 Th1 > BL > Th2 In this case, the path dividing unit 202 sets the small area corresponding to obstacle A as a square with one side measuring F1, and sets the small area corresponding to obstacle B as a square with one side measuring F1 / 2.

[0133] In the example shown in FIG. 8C , the target area determination unit 205 evaluates the degree of overlap between small regions D51 to D54 and region B1 at time t=5. The target area determination unit 205 references the map shown in FIG. 8C and determines that portions of small regions D51 and D53 are included at the edges of region B1. The target area determination unit 205 then determines that the probabilities of obstacle B being present in small regions D51 and D53 at time t=5 are 0.15 and 0.05, respectively. Meanwhile, small regions D52 and D54 do not overlap with region B1. Therefore, the target area determination unit 205 determines that the probability of obstacle B being present in small regions D52 and D54 at time t=5 is 0. The target area determination unit 205 can perform a similar presence probability determination at each time during the period from time t=0 to t=15 for route candidate K1.

[0134] It is also possible to assume that an obstacle C exists in addition to obstacles A and B. If the size of obstacle C is CL, the path dividing unit 202 determines the magnitude relationship between CL and the first and second thresholds Th1 and Th2 as follows: Th1 > CL > Th2. In this case, the target area determination unit 205 performs a presence probability determination for obstacle C, similar to that for obstacle B. The presence probability information derived for obstacles B and C may be output separately by the transmitting / receiving unit 201. Alternatively, the target area determination unit 205 may combine the presence probability for obstacle B and the presence probability for obstacle C to calculate a single presence probability value. In this way, the target area determination unit 205 can perform a process of combining the presence probabilities of multiple obstacles that occupy the same small area to be determined. Therefore, for example, the target area determination unit 205 can perform a process of combining the presence probabilities for obstacles of similar size or the same type.

[0135] The target area determination unit 205 outputs the existence probability information derived as described above to the transmitting / receiving unit 201. The transmitting / receiving unit 201 transmits area information including the existence probability information to the robot 100.

[0136] If there are multiple small areas that are to be subjected to obstacle detection at a predetermined time point indicated by a route candidate, the target area determination unit 205 may set the area information to include information on the presence probability for each small area. However, the target area determination unit 205 may set the area information to include information on the presence probability for one or more small areas, but not to include information on the presence probability for other small areas.

[0137] For example, in the example shown in FIG. 8C , the target area determination unit 205 may set the area information to include information on the existence probability of each of the small regions D51 to D54. Alternatively, the target area determination unit 205 may set the area information to include information on the existence probability of each of the small regions D51 to D54 that is equal to or greater than a predetermined threshold. As another example, the target area determination unit 205 may set the area information to include information on the existence probability that falls within a predetermined ranking when the existence probabilities of each of the small regions D51 to D54 are sorted in descending order. For example, the target area determination unit 205 may set the area information to include information on the existence probability of each of the small regions D51 to D54 that is the highest, 0.15.

[0138] When the set area information is transmitted to the robot 100, the path determination unit 104 of the robot 100 refers to the output area information. The path determination unit 104 then determines that the probability of an obstacle being present in the small area D52 or D54 at time t=5 is low. Therefore, the path determination unit 104 can control the movement of the robot 100 so that the robot 100 is located in either the small area D52 or D54. By performing the above process, the robot 100 can move along the initially set path candidate K1, since the size of the obstacle B is small.

[0139] In the example described above, the target area determination unit 205 derives the presence probability of moving obstacles A and B on the route of the route candidate K1. However, the target area determination unit 205 may also derive the presence probability of stationary objects on the route of the route candidate K1 using a similar method.

[0140] Furthermore, the target area determination unit 205 may derive the probability of an obstacle existing in each small area where the robot 100 exists at one or more arbitrary points in time on the route, rather than for the entire route of the route candidate K1.

[0141] Furthermore, when the area information request includes information on a plurality of route candidates, the target area determination unit 205 may derive, for each route candidate, the probability of an obstacle existing at each time point indicated by the route candidate.

[0142] [Explanation of Processing] Fig. 9 is a sequence diagram showing an example of a typical process of the robot control system 30. By referring to Fig. 9, the process of each segment of the robot control system 30 will be explained. Note that the details of each process are as described above, and therefore the explanation will be omitted as appropriate.

[0143] First, in step S21, the route candidate generation unit 101 generates route candidates for the robot 100. In step S22, the transmission / reception unit 102 includes information on the route candidates output from the route candidate generation unit 101 in an area information request, and transmits the area information request to the management server 200. The transmission / reception unit 201 receives the transmitted area information request.

[0144] In step S23, the route dividing unit 202 sets small areas on the map that include each of the destination positions on the route candidates, using the information on the route candidates indicated in the area information request. In step S24, the object position predicting unit 204 uses the information on the obstacles to predict the positions of the obstacles at each time point indicated by the route candidates. Note that the processes of steps S23 and S24 may be executed either first, or may be executed in parallel.

[0145] In step S25, the target area determination unit 205 derives the probability of an obstacle being present in a small area on the route candidate. In step S26, the transmitting / receiving unit 201 transmits area information including the obstacle presence probability information derived by the target area determination unit 205 to the robot 100.

[0146] In step S27, the possibility evaluation unit 103 refers to the area information to evaluate the possibility that the robot 100 will come into contact with an obstacle when the robot 100 moves along the route candidate. In step S28, the route determination unit 104 determines the movement route of the robot 100 based on the evaluation result output by the possibility evaluation unit 103.

[0147] [Description of Effect] In recent years, the application of autonomously mobile robots has progressed. A robot plans its own travel route from a starting point to a destination point and moves along the planned route. In this case, the robot may acquire information such as the position and speed of at least one of moving or stationary obstacles to plan the travel route. The robot uses the acquired obstacle information to predict the future position of the obstacle. By performing the prediction process, the robot generates a route that avoids contact with the obstacle.

[0148] To generate a path, the robot needs to know the presence of each obstacle. However, there is a possibility that the robot will not be able to detect the presence of each obstacle using sensors or the like provided on the robot. In particular, it is considered difficult for the robot to accurately detect information such as the presence and speed of a moving obstacle. Furthermore, even if the robot can accurately detect information about the obstacles, the internal processing of the robot for detecting the obstacle information may be complicated.

[0149] As another example, IoT (Internet of Things) technology may be applied to allow a robot to collect information from multiple sensors installed in the robot's movement space. Using information from the sensors, the robot can generate a path in real time while moving. Furthermore, since multiple sensors are installed separately from the robot, the robot can accurately detect information such as the presence and speed of obstacles. However, the robot must send and receive information to and from the multiple sensors installed in the movement space. Therefore, if the number of robots that autonomously generate paths increases, the amount of communication between the robots and the sensors may become enormous, potentially preventing normal communication between the robots and the sensors. If communication is not performed normally, the robot may not be able to collect necessary information from the sensors in real time.

[0150] For these reasons, one possible method is to install sensors in the space where the robot moves, and have a server installed on the cloud side use the information acquired from the sensors to detect obstacles around the robot. With this method, the amount of communication between the sensors and the server does not change even if the number of robots increases. Therefore, the server can collect necessary information from the sensors in real time.

[0151] As an example, a server installed on the cloud side could determine the robot's route based on information about detected obstacles and send the determined route information to the robot. In other words, the server controls all of the robot's movements, and the robot moves according to the control instructions from the server. However, the server may only be able to directly control a limited number of robots. This means that in a system consisting of a server and a robot, compatibility between the robots may be low.

[0152] To improve compatibility between robots, it is expected that when a server installed on the cloud side detects an obstacle, it will transmit the detected obstacle information to the robot. However, if the robot frequently queries the server for information on obstacles, etc., in order to identify obstacles in the surrounding environment, the amount of communication between the robot and the server may increase.

[0153] In contrast, the robot control system 30 disclosed in the present disclosure performs the following processing. The management server 200 provided on the cloud side predicts the future position of the obstacle in the virtual space of the digital twin by using current information about the obstacle collected from sensors. Then, in response to receiving an area information request from the robot 100, the management server 200 provides the robot 100 with information on the existence probability of areas around the route, regarding the route indicated by the route candidate included in the area information request. In other words, the management server 200 does not need to provide the robot 100 with information about areas other than those around the route. Therefore, the management server 200 can significantly reduce the amount of communication between the robot 100 and the management server 200.

[0154] In particular, when operating multiple robots 100, the effect of reducing the amount of communication between the robots 100 and the management server 200 is significant. If the management server 200 were to transmit the entire digital twin data to each robot 100, the amount of communication between the management server 200 and the robots 100 would increase, and communication delays could occur. However, in the robot control system 30 disclosed herein, the amount of data transmitted from the management server 200 to each robot 100 is extremely small, making it possible to significantly reduce the amount of communication.

[0155] Furthermore, the management server 200 stores information on the current and future positions of obstacles and the like, and is specialized in the function of providing the robot 100 with information on obstacles in response to an inquiry from the robot 100. Therefore, even when inquiries are made from various types of robots 100, the management server 200 can provide the robot 100 with information that is effective for movement control. This increases the compatibility of robots that can be used in the robot control system 30.

[0156] Furthermore, the robot 100 can determine an optimal movement route that takes safety and movement time into consideration by receiving area information from the management server 200. Furthermore, the robot 100 itself does not need to detect other objects in the surrounding environment. Therefore, there is no need to attach devices such as sensors or cameras to the robot 100, which reduces the cost of the robot 100.

[0157] Below, (2B) and (2C) will be described variations of the configuration and processing shown in (2A). It goes without saying that the configurations and processing shown in (2B) and (2C) can be combined in part or in whole.

[0158] (2B: Changing the notification using a threshold value) In (2A), the management server 200 provides the robot 100 with information on the existence probability of areas around the route indicated by the route candidate. However, depending on the situation, the management server 200 may not need to provide the robot 100 with information on the existence probability of areas around the route.

[0159] 10 is a block diagram showing another example of a management server 200. The management server 200 includes a transmitting / receiving unit 201 to a target area determining unit 205, as well as a notification availability determining unit 206.

[0160] The notification possibility determination unit 206 acquires obstacle presence probability information in an area including a route candidate, which is output by the target area determination unit 205 to the transmission / reception unit 201. Then, the notification possibility determination unit 206 uses the presence probability information and a predetermined threshold value stored in the database 203 to determine whether the presence probability information should be provided to the robot 100.

[0161] Specifically, the notification determination unit 206 determines whether the probability of an obstacle existing in the small area where the robot 100 is located is greater than or equal to a predetermined threshold value at each time point indicated by the route candidate. If the probability of the obstacle existing is equal to or less than the predetermined threshold value, the notification determination unit 206 controls the transmission / reception unit 201 not to output information on the probability of the obstacle existing to the robot 100. On the other hand, if the probability of the obstacle existing is greater than the predetermined threshold value, the notification determination unit 206 controls the transmission / reception unit 201 to output information on the probability of the obstacle existing to the robot 100.

[0162] For example, if the presence probability is equal to or lower than a predetermined threshold at each time point indicated by the route candidate, the notification determination unit 206 may control the transmission / reception unit 201 not to output presence probability information regarding the entire route candidate to the robot 100. Specifically, the notification determination unit 206 may control the transmission / reception unit 201 to transmit area information that does not indicate any presence probability to the robot 100. As another example, the notification determination unit 206 may control the transmission / reception unit 201 to transmit area information including information indicating that no obstacles exist in the entire route candidate to the robot 100. Alternatively, the notification determination unit 206 may control the transmission / reception unit 201 to stop transmitting area information to the robot 100.

[0163] It is also possible that, among the multiple time points indicated by the route candidate, the presence probability is equal to or less than a predetermined threshold at some time points and is greater than the predetermined threshold at other time points. In this case, the notification possibility determination unit 206 may cause the robot 100 to transmit area information that does not indicate the presence probability for time points where the presence probability is equal to or less than the predetermined threshold, but indicates the presence probability for time points where the presence probability is greater than the predetermined threshold.

[0164] When the target area determination unit 205 derives presence probability information for multiple route candidates, the notification determination unit 206 can execute the above process for the multiple route candidates. As described above, the notification determination unit 206 determines whether to notify the presence probability information using threshold information.

[0165] Furthermore, the database 203 may store a plurality of different thresholds according to the characteristics of the obstacle as the predetermined threshold. The characteristics of the obstacle refer to classification of the obstacle based on at least one of the following viewpoints, for example: (5) whether the obstacle is a human being; (6) whether the obstacle is an object smaller than a predetermined size; (7) whether the obstacle is a predetermined type of machine; and (8) the speed of the obstacle predicted by the prediction unit 11 at the time when the obstacle reaches the target movement area. Regarding (5) to (7), it can also be defined that a plurality of different thresholds are stored according to the type of obstacle.

[0166] For example, with regard to (5), the predetermined threshold may be set to Th3 if the obstacle is a human, and Th4 if the obstacle is something other than a human. If the obstacle is a human, it is preferable to output the presence probability information to the robot 100 as much as possible, taking into consideration the safety of the human. Therefore, the magnitude relationship between Th3 and Th4 may be set to Th4>Th3, for example.

[0167] Regarding (6), the predetermined threshold may be set to Th5 when the obstacle is smaller than a predetermined size, and set to Th6 when the obstacle is equal to or larger than the predetermined size. When the obstacle is equal to or larger than the predetermined size, the possibility of contact with the robot 100 increases compared to when the obstacle is smaller than the predetermined size. Therefore, when the obstacle is equal to or larger than the predetermined size, it is preferable to output presence probability information to the robot 100 as much as possible, taking into consideration the safety of the robot 100. Therefore, the magnitude relationship between Th6 and Th5 may be set to, for example, Th5>Th6.

[0168] Regarding (7), the predetermined threshold may be set to Th7 if the obstacle is a predetermined type of machine, and set to Th8 if the obstacle is not a predetermined type of machine. Here, the magnitude relationship between Th7 and Th8 may be set to Th8>Th7, for example. A machine being a predetermined type of machine means, for example, that the machine is expensive or important.

[0169] Regarding (8), if the speed of the obstacle when it reaches the target area is less than a predetermined speed, Th9 may be set as the predetermined threshold. If the speed of the obstacle when it reaches the target area is equal to or greater than a predetermined speed, Th10 may be set as the predetermined threshold. If the speed of the obstacle is fast, the possibility of unexpected contact with the robot 100 increases. Therefore, in consideration of the safety of the robot 100, it is preferable to output presence probability information to the robot 100 as much as possible. Therefore, the magnitude relationship between Th9 and Th10 may be set as Th9 > Th10, for example.

[0170] Furthermore, the database 203 may store a plurality of different thresholds according to the characteristics of the robot 100 as the predetermined threshold. The characteristics of the robot 100 refer to, for example, categorizing the robot 100 based on at least one of the following viewpoints: (9) Type of robot 100 (10) Speed ​​of the robot 100 at the time when the robot 100 reaches the target movement area (11) Type of cargo carried by the robot 100 The type of robot 100 is classified according to the size, use, etc. of the robot 100. Information on the speed of the robot 100 at the time when the robot 100 reaches the target movement area may be included in the information on the route candidates in the area information request, or may be calculated by the target area determination unit 205 based on the information on the route candidates.

[0171] For example, with regard to (9), Th11 may be set when the robot 100 is equal to or larger than a predetermined size, and Th12 may be set when the robot 100 is smaller than the predetermined size. When the robot 100 is equal to or larger than the predetermined size, the robot 100 is more likely to come into contact with an obstacle than when the robot 100 is smaller than the predetermined size, so it is preferable to output as much presence probability information as possible to the robot 100. Therefore, the magnitude relationship between Th11 and Th12 may be set as Th12>Th11, for example.

[0172] Regarding (10), if the speed of the robot 100 when it reaches the target movement area is less than a predetermined speed, Th13 may be set as the predetermined threshold. If the speed of the robot 100 when it reaches the target movement area is equal to or greater than a predetermined speed, Th14 may be set as the predetermined threshold. If the speed of the robot 100 is high, the possibility of unexpected contact with an obstacle increases. Therefore, in consideration of the safety of the robot 100, it is preferable to output presence probability information to the robot 100 as much as possible. Therefore, the magnitude relationship between Th13 and Th14 may be set as Th13 > Th14, for example.

[0173] Regarding (11), if the load carried by the robot 100 is a predetermined type of load, Th15 may be set as the predetermined threshold, and if the load carried is a load other than the predetermined type, Th16 may be set as the predetermined threshold. A load being a predetermined type indicates, for example, that the load is important. If the load carried by the robot 100 is a predetermined type of load, it is preferable to output presence probability information to the robot 100 as much as possible, taking into consideration the safety of the load. Therefore, the magnitude relationship between Th15 and Th16 may be set as Th16 > Th15, for example.

[0174] The above-described threshold setting methods are merely examples and are not limited to these. For example, thresholds may be set by combining two or more of the viewpoints (5) to (11). For example, when (5) and (6) are combined, the following thresholds may be set as the predetermined thresholds: - When the obstacle is a human: Th17 - When the obstacle is a non-human and is smaller than a predetermined size: Th18 - When the obstacle is a non-human and is equal to or larger than a predetermined size: Th19 Here, the magnitude relationship between Th17 to Th19 may be set as, for example, Th18 > Th19 > Th17 or Th18 > Th17 > Th19.

[0175] When the target area determination unit 205 derives the presence probabilities of multiple obstacles, the notification determination unit 206 can determine whether the presence probability of each obstacle is greater than or equal to a predetermined threshold. When the target area determination unit 205 integrates the presence probabilities of multiple obstacles, the notification determination unit 206 can determine whether the integrated presence probability is greater than or equal to a predetermined threshold.

[0176] For example, assume that the target area determination unit 205 calculates a presence probability P1 by aggregating the presence probabilities of multiple obstacles that are humans, and calculates a presence probability P2 by aggregating the presence probabilities of multiple obstacles that are not humans. In this case, the notification determination unit 206 determines whether to notify the presence probability information for obstacles that are humans by determining the magnitude relationship between the presence probability P1 and Th3. The notification determination unit 206 also determines whether to notify the presence probability information for obstacles that are not humans by determining the magnitude relationship between the presence probability P2 and Th4. Note that Th4 is set to be greater than Th3. As described above, when multiple obstacles are classified into two or more types of obstacles, the target area determination unit 205 can determine whether to notify the presence probability information for each type of obstacle by comparing the aggregated presence probability for obstacles of the same type with a threshold.

[0177] When the management server 200 transmits area information to multiple robots 100 in response to an area information request, the predetermined threshold may be the same or different for each robot 100. When different thresholds are set depending on the type of obstacle, different thresholds may be set for each robot 100 for the same type of obstacle, or the same threshold may be set for multiple robots 100.

[0178] Furthermore, the threshold information may be stored in advance in the database 203, or may be transmitted from the robot 100 to the management server 200. For example, the path candidate generation unit 101 of the robot 100 may determine a plurality of different thresholds according to the type of obstacle, and transmit information on the determined thresholds to the management server 200 via the transmission / reception unit 102. Therefore, when there are a plurality of robots 100, different threshold information may be transmitted to the management server 200 according to the type of robot 100, etc.

[0179] The transmitting and receiving unit 102 may include information about the determined threshold in the area information request and transmit it to the management server 200. However, the transmitting and receiving unit 201 may transmit information about the determined threshold to the management server 200 separately from the area information request. Furthermore, when the path candidate generation unit 101 updates the threshold information, the transmitting and receiving unit 102 can transmit the updated threshold information. For example, when the path candidate generation unit 101 detects that the environment in which the robot 100 exists has changed, the transmitting and receiving unit 102 can update the threshold information.

[0180] Other processes executed by the transmitter / receiver 201 to the target area determination unit 205 are as described in (2A), and therefore will not be described here.

[0181] [Explanation of Effect] As described above, the notification possibility determination unit 206 can be set not to transmit information regarding the presence of an obstacle to the robot 100 when the probability of an obstacle's presence is equal to or less than a predetermined threshold. In other words, when the probability of an obstacle's presence is low and it is considered that there is little need to provide the robot 100 with information regarding the presence probability, the transmission / reception unit 201 does not transmit information regarding the presence of an obstacle to the robot 100. By not transmitting information regarding the presence of an obstacle to the robot 100, the management server 200 can further reduce the amount of communication between the robot 100 and the management server 200 while ensuring that the robot 100 moves safely.

[0182] The predetermined threshold used by the notification possibility determination unit 206 for determination may be transmitted from the robot 100. In other words, the robot 100 can autonomously determine information on the threshold that determines its own movement path according to conditions such as the characteristics of the robot 100 or the environment in which the robot 100 exists. Therefore, the robot control system 30 can ensure the safety and efficiency of movement of the robot 100 according to the individual circumstances of the robot 100, such as the characteristics of the robot 100 and the environment in which the robot 100 exists.

[0183] Furthermore, the predetermined threshold value may have a different value depending on the characteristics of at least one of the obstacle or the robot 100. In other words, even if the probability of an obstacle being present is the same, the management server 200 may change whether or not to transmit information about the presence of an obstacle depending on the type of the obstacle or the robot 100. This ensures the safety of the movement of the robot 100 by transmitting information about the presence of an obstacle when necessary, and reduces the amount of communication by not transmitting information about the presence of an obstacle when not necessary.

[0184] (2C: Changing the area to be judged) The target area judgment unit 205 of the management server 200 may change the small area around the candidate route that is the target for predicting the probability of an obstacle's presence, depending on at least one of the characteristics of the obstacle, the characteristics of the robot 100, the width of the route the robot 100 will take, or the distance between the robot 100 and the obstacle.

[0185] An example will be described below in which the target area determination unit 205 changes the small area depending on the width of the route or the distance between the robot 100 and an obstacle. In the example shown in FIG. 8A or 8B , the width of the route along which the robot 100 travels is set to W1, and a predetermined threshold value to be compared with the route width is set to Th13. Here, if W1 is less than Th13, as described in (2A), the target area determination unit 205 sets a small area that is a square with a side length of F1 and is located on the route indicated by the route candidate as the area for which the probability of an obstacle presence is predicted. For example, at time t=5, small area D5 is set as the area for which the probability of presence is predicted.

[0186] On the other hand, if W1 is equal to or greater than Th13, the target area determination unit 205 predicts the probability of an obstacle's presence not only in the small square area with a side length of F1 located on the route indicated by the route candidate, but also in surrounding small areas. For example, at time t=5, the target area determination unit 205 may predict the probability of an obstacle's presence not only in the small area D5 but also in the four small areas adjacent to the small area D5 on all four sides. Alternatively, the target area determination unit 205 may predict the probability of an obstacle's presence in a square area with a side length of 3F1 and including the small area D5 at its center. As described above, the target area determination unit 205 can set an area larger than the small area D5 and including the small area D5 as the target for predicting the probability of an obstacle's presence.

[0187] Similarly, the target area determination unit 205 may compare the distance between the robot 100 and the obstacle at the time of prediction with a predetermined threshold. If the distance is greater than the predetermined threshold, the target area determination unit 205 sets small region D5 as the target for predicting the presence probability at time t = 5. On the other hand, if the distance is equal to or less than the predetermined threshold, the target area determination unit 205 can set an area that is larger than and includes small region D5 as the target for predicting the presence probability at time t = 5.

[0188] The characteristics of the obstacle and the characteristics of the robot 100 are as shown in (5) to (11) above. As an example, the target area determination unit 205 may predict the presence probability for small region D5 at time t=5 in any of the following cases: - When the obstacle is not a human - When the obstacle is an object smaller than a predetermined size - When the obstacle is a predetermined type of machine - When the speed or acceleration of the robot 100 at the time of prediction is less than a predetermined threshold - When the robot 100 is a predetermined type of robot - When the speed of the robot 100 at the time when the robot 100 reaches the movement target area is less than a predetermined threshold - When the luggage carried by the robot 100 is a predetermined type of luggage

[0189] On the other hand, the target area determination unit 205 may set an area that is larger than the small area D5 and that includes the small area D5 as the target for predicting the presence probability in any of the following cases at time t=5: - When the obstacle is a human - When the obstacle is an object of a predetermined size or larger - When the obstacle is not a predetermined type of machine - When the speed or acceleration of the robot 100 at the time of prediction is equal to or greater than a predetermined threshold - When the robot 100 is not a predetermined type of robot - When the speed of the robot 100 at the time when the robot 100 reaches the target movement area is equal to or greater than a predetermined threshold - When the luggage carried by the robot 100 is not a predetermined type of luggage

[0190] Note that the target area determination unit 205 can execute the above-described process even when the size of the small area changes as shown in Fig. 8C. In other words, the target area determination unit 205 can execute the above-described process for each obstacle even when the type of obstacle changes.

[0191] As described above, the target area determination unit 205 can predict the presence of obstacles with greater consideration given to safety by changing the area set as the target for predicting the presence probability depending on the situation. The robot 100 can then determine a safer route by acquiring the presence probability information derived by the target area determination unit 205.

[0192] In the above-described embodiments, the present disclosure has been described as a hardware configuration, but the present disclosure is not limited to this. The present disclosure can also be realized by causing a processor in a computer to execute a computer program to perform the processing of the devices constituting the mobile object management system, the mobile object control system, and the robot control system described in the above-described embodiments.

[0193] 11 is a block diagram showing an example of the hardware configuration of an information processing device 90 that executes the processing of the system or device described in each embodiment. Referring to FIG. 11, the information processing device 90 includes a signal processing circuit 91, a processor 92, and a memory 93.

[0194] The signal processing circuit 91 is a circuit for processing signals in accordance with the control of the processor 92. The signal processing circuit 91 may include a communication circuit for receiving signals from a transmitting device.

[0195] The processor 92 is connected to the memory 93, and performs the processing of the device described in the above embodiment by reading and executing a computer program from the memory 93. As an example of the processor 92, one of a CPU (Central Processing Unit), an MPU (Micro Processing Unit), an FPGA (Field-Programmable Gate Array), a DSP (Demand-Side Platform), and an ASIC (Application Specific Integrated Circuit) may be used, or a plurality of these may be used in parallel.

[0196] The memory 93 may be a volatile memory, a nonvolatile memory, or a combination thereof. The memory 93 is not limited to one, and may be provided in multiple units. The volatile memory may be, for example, a random access memory (RAM) such as a dynamic random access memory (DRAM) or a static random access memory (SRAM). The nonvolatile memory may be, for example, a read only memory (ROM) such as a programmable random only memory (PROM) or an erasable programmable read only memory (EPROM), a flash memory, or a solid state drive (SSD).

[0197] The memory 93 is used to store one or more instructions. Here, the one or more instructions are stored as programs in the memory 93. The processor 92 can perform the processes described in the above embodiments by reading and executing these programs from the memory 93.

[0198] The memory 93 may include a memory provided outside the processor 92, as well as a memory built into the processor 92. The memory 93 may also include a storage device located away from the processors constituting the processor 92. In this case, the processor 92 can access the memory 93 via an I / O (Input / Output) interface.

[0199] As described above, one or more processors included in each device in the above-described embodiments execute one or more programs including instructions for causing a computer to execute the algorithms described using the drawings. Execution of the programs enables the information processing described in each embodiment to be realized.

[0200] The program includes instructions or software code that, when loaded into a computer, causes the computer to perform one or more functions described in the embodiments. The program may be stored in a non-transitory computer-readable medium or a tangible storage medium. By way of example and not limitation, computer-readable media or tangible storage media include random-access memory (RAM), read-only memory (ROM), flash memory, solid-state drive (SSD) or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disk (DVD), Blu-ray disc or other optical disk storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices. The program may also be transmitted on a transitory computer-readable medium or communication medium. By way of example and not limitation, transitory computer-readable media or communication media include electrical, optical, acoustic, or other forms of propagated signals. The transitory computer-readable medium or communication medium may provide the program to the computer via a wired communication path, such as an electric wire or optical fiber, or via a wireless communication path.

[0201] Some or all of the above embodiments may be described as, but are not limited to, the following supplements. Furthermore, some or all of the elements described in Supplementary Notes 2 to 9 that are dependent on Supplementary Note 1 may also be dependent on Supplementary Notes 10, 16, and 21 in the same dependency relationship as Supplementary Notes 2 to 9. In this manner, some or all of the elements described in any supplementary note may be applied to various hardware, software, recording means for recording software, systems, and methods. (Supplementary Note 1) A mobile object management system comprising: a prediction means for predicting the presence of an obstacle located on a route indicated by route information of a mobile object in response to acquisition of the route information from the mobile object; and an output means for outputting, to the mobile object, information regarding the presence of the obstacle predicted by the prediction means. (Supplementary Note 2) The mobile object management system described in Supplementary Note 1, wherein the prediction means predicts a probability that the obstacle will be present in an area on the route indicated by the route information at the time the mobile object moves into the area; and the output means outputs the probability information as information regarding the presence of the obstacle. (Supplementary Note 3) The mobile object management system according to Supplementary Note 2, wherein the prediction means predicts the probability for each of the plurality of obstacles, and the output means outputs information on the probability integrated for the plurality of obstacles as information on the presence of the obstacle. (Supplementary Note 4) The mobile object management system according to Supplementary Note 2 or 3, wherein the output means does not output the information on the presence of the obstacle to the mobile object if the probability indicated by the information on the presence of the obstacle is equal to or lower than a predetermined threshold. (Supplementary Note 5) The mobile object management system according to Supplementary Note 4, wherein the predetermined threshold is output from the mobile object. (Supplementary Note 6) The mobile object management system according to Supplementary Note 4 or 5, wherein the predetermined threshold has a different value depending on characteristics of at least one of the obstacle or the mobile object. (Supplementary Note 7) The mobile object management system according to any one of Supplements 2 to 6, wherein the prediction means predicts the probability that the obstacle is present in a certain area using a digital twin that treats information on whether an object is present in the area as a presence probability.(Supplementary Note 8) The mobile object management system according to any one of Supplementary Notes 1 to 7, wherein the prediction means changes the size of an area on the route for which the presence of the obstacle is predicted, depending on at least one of a characteristic of the obstacle, a characteristic of the mobile object, a width of a route along which the mobile object will pass, or a distance between the mobile object and the obstacle. (Supplementary Note 9) The mobile object management system according to any one of Supplementary Notes 1 to 8, wherein the prediction means predicts the presence of moving objects and stationary objects as the obstacles. (Supplementary Note 10) A mobile object control system comprising: a prediction means for predicting the presence of an obstacle located on a route indicated by route information of the mobile object in response to acquiring the route information of the mobile object from the mobile object; an output means for outputting, to the mobile object, information regarding the presence of the obstacle predicted by the prediction means; and the mobile object that determines the route of the mobile object based on the information regarding the presence of the obstacle. (Supplementary Note 11) The mobile body control system according to Supplementary Note 10, wherein the prediction means predicts a probability that the obstacle will be present in an area on the route indicated by the route information at the time the mobile body moves into the area, and the output means outputs information on the probability as information related to the presence of the obstacle. (Supplementary Note 12) The mobile body control system according to Supplementary Note 11, wherein the prediction means predicts the probability for each of a plurality of the obstacles, and the output means outputs information on the probability integrated for the plurality of obstacles as information related to the presence of the obstacle. (Supplementary Note 13) The mobile body control system according to Supplementary Note 11 or 12, wherein the output means does not output information related to the presence of the obstacle to the mobile body if the probability indicated by the information related to the presence of the obstacle is equal to or less than a predetermined threshold. (Supplementary Note 14) The mobile body control system according to Supplementary Note 13, wherein the predetermined threshold is output from the mobile body. (Supplementary Note 15) The mobile object control system according to Supplementary Note 13 or 14, wherein the predetermined threshold has a value that varies depending on characteristics of at least one of the obstacle or the mobile object. (Supplementary Note 16) A mobile object management method, comprising: in response to acquiring route information of the mobile object from the mobile object, predicting the presence of an obstacle located on the route indicated by the route information; and outputting information relating to the predicted presence of the obstacle to the mobile object.(Supplementary Note 17) The mobile object management method according to Supplementary Note 16, comprising predicting a probability that the obstacle will be present in an area on a route indicated by the route information at the time the mobile object moves into the area, and outputting information on the probability as information related to the presence of the obstacle. (Supplementary Note 18) The mobile object management method according to Supplementary Note 17, comprising predicting the probability for each of a plurality of the obstacles, and outputting information on the probability integrated for the plurality of obstacles as information related to the presence of the obstacle. (Supplementary Note 19) The mobile object management method according to Supplementary Note 17 or 18, comprising not outputting information on the presence of the obstacle to the mobile object if the probability indicated by the information on the presence of the obstacle is equal to or less than a predetermined threshold. (Supplementary Note 20) The mobile object management method according to Supplementary Note 19, wherein the predetermined threshold is output from the mobile object. (Supplementary Note 21) A program causing a computer to execute the following steps: predicting the presence of an obstacle located on the route indicated by the route information of the mobile object in response to acquiring the route information of the mobile object from the mobile object; and outputting information on the presence of the obstacle predicted by the prediction means to the mobile object.

[0202] Although the present disclosure has been described above with reference to the embodiments, the present disclosure is not limited to the above-described embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present disclosure within the scope of the present disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.

[0203] REFERENCE SIGNS LIST 10 Mobile object management system 11 Prediction unit 12 Output unit 20 Mobile object control system 21 Mobile object 30 Robot control system 100 Robot 101 Route candidate generation unit 102 Transmitting / receiving unit 103 Possibility evaluation unit 104 Route determination unit 200 Management server 201 Transmitting / receiving unit 202 Route division unit 203 Database 204 Object position prediction unit 205 Target area determination unit 206 Notification possibility determination unit

Claims

1. A prediction means that predicts the presence of obstacles located along the path indicated by the path information obtained from the moving object, The system includes an output means that outputs information about the presence of an obstacle predicted by the prediction means to the moving body. Mobile device management system.

2. The prediction means predicts the probability that the obstacle is present in the region on the path indicated by the path information at the time the moving object moves into the region on the path indicated by the path information. The output means outputs the probability information as information regarding the presence of the obstacle. The mobile device management system according to claim 1.

3. The prediction means predicts the probability for each of the multiple obstacles, The output means outputs the integrated probability information for the plurality of obstacles as information regarding the presence of the obstacles. The mobile device management system according to claim 2.

4. The output means does not output information regarding the presence of the obstacle to the moving body if the probability indicated by the information regarding the presence of the obstacle is below a predetermined threshold. The mobile device management system according to claim 2 or 3.

5. The predetermined threshold is output from the mobile body. The mobile device management system according to claim 4.

6. The predetermined threshold has different values ​​depending on the characteristics of at least one of the obstacle or the moving object. The mobile device management system according to claim 4.

7. The prediction means uses a digital twin, which treats information about whether or not an object exists in a certain region as a probability of existence, to predict the probability that the obstacle exists in the region. The mobile device management system according to claim 2 or 3.

8. A prediction means that predicts the presence of obstacles located along the path indicated by the path information obtained from the moving object, An output means that outputs information regarding the presence of an obstacle predicted by the prediction means to the moving body, The mobile body determines the path of the mobile body based on information regarding the presence of the aforementioned obstacles. Mobile control system.

9. In response to obtaining the path information of the moving object from the moving object, the presence of obstacles located along the path indicated by the path information is predicted. The mobile body outputs information regarding the presence of predicted obstacles. Mobile management method.

10. In response to obtaining path information of a moving object from the moving object, predict the presence of an obstacle located on the path indicated by the path information, The mobile body outputs information regarding the presence of predicted obstacles. A program that causes a computer to perform a task.