Information processing system

Abstract

To provide an information processing system which can obtain an inference result in consideration of states of respective places of an object of inference when the state of the object is different depending on the place.SOLUTION: An information processing system includes: a plurality of inference devices 10 which acquire pieces of input information different from each other and infer a state of an object; and a device. One inference device 10 out of the plurality of inference devices 10 includes: an acquisition unit 11 which acquires input information; an inference unit 12 which generates a first inference result being a result obtained by inferring a state of the object on the basis of the input information; a receiving unit 13 which receives first result information including a second inference result in another inference device 10 from another inference device 10; a determination unit 15 which generates a determination result on the basis of the first inference result and the second inference result; and a control unit 16 which controls the device on the basis of the determination result.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present invention relates to an information processing system.

Background Art

[0002] The information processing system described in Patent Document 1 is a distributed cooperative inference device having a plurality of inference devices. In this information processing system, solutions are generated by cooperative inference of a plurality of inference device units for problems that cannot be handled by a single inference device unit. In this information processing system, the generation order of information is commonly managed for the information input to the plurality of inference devices.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Patent Document 3

Patent Document 4

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the information processing system as described above, common inputs are processed in parallel by a plurality of inference devices. Therefore, when the state of the inference object varies depending on the location, information regarding the state of the object at each location cannot be processed. As a result, inference results cannot be obtained in consideration of the state for each location of the object.

[0005] Therefore, one aspect of the embodiment was made in view of the above problems, and aims to provide an information processing system that can obtain inference results by taking into account the state of the object at each location when the state of the object to be inferred differs depending on the location. [Means for solving the problem]

[0006] An information processing system according to one aspect of the embodiment comprises a plurality of inference devices connected to each other so as to be able to send and receive information, each acquiring different input information and inferring the state of an object; at least one device controllably connected to an inference device; and a plurality of sensors that perform sensing to generate input information and output the generated input information to a corresponding inference device among the plurality of inference devices. One of the plurality of inference devices has an acquisition unit that acquires input information; an inference unit that generates a first inference result which is the result of inferring the state of an object based on the input information acquired by the acquisition unit; a receiving unit that receives first result information including a second inference result from another inference device; a determination unit that generates a determination result based at least on the first inference result and the second inference result; and a control unit that controls a device based at least on the determination result.

[0007] In this information processing system, different input information is input to multiple inference devices. Each inference device generates a first inference result in which the state of the object is inferred based on the input information, and a determination result is generated based at least on the first inference result and a second inference result from another inference device. With this configuration, if the state of the object to be inferred differs depending on the location, the state of the object at each location is inferred by each inference device to generate the first and second inference results. This makes it possible to obtain an inference result that takes into account the state of the object at each location. Furthermore, since the equipment is controlled based on the first and second inference results, the equipment can be controlled while taking into account the state of the object at each location.

[0008] Furthermore, in one aspect of the above, the input information may be an image of the object, and the inference unit may generate a first inference result, which is the result of inferring the state of the object based on the appearance of the object in the image. With such a configuration, the state of the object is inferred based on the appearance of the object at each location. This makes it possible to generate the first inference result with greater accuracy.

[0009] Furthermore, in one aspect of the above, one of the multiple inference devices may further include a transmission unit that transmits a second result information, which is the first result information with the first inference result added to it, to another inference device. With such a configuration, another inference device can generate a judgment result with greater accuracy based on the inference results generated by the inference device and the other inference device.

[0010] Furthermore, in one aspect of the above, the object may be a moving body flowing from upstream to downstream along a predetermined flow path. Multiple devices may be arranged along the flow path and operate to increase or decrease the amount of flow of the object. Multiple sensors may be arranged along the flow path. Multiple inference devices may be provided so that their operation can be controlled individually. The inference unit may generate a first inference result, which is the result of inferring the amount of flow based on the input information. The receiving unit may receive first result information from another adjacent inference device on the upstream side of the flow path. The transmitting unit may transmit second result information to yet another adjacent inference device on the downstream side of the flow path. The determination unit may perform a first comparison process that calculates the difference between the first inference result and the second inference result and generates a first comparison result, which is the result of comparing the difference with a first threshold. The control unit may control the device corresponding to either the inference device corresponding to the control unit or another inference device based on the first comparison result.

[0011] In this configuration, the device is controlled based on a first comparison result, which is obtained by comparing the difference between the amount of flow of the object at the location corresponding to the inference device and the amount of flow of the object at the location corresponding to another adjacent inference device upstream, with a first threshold. This makes it possible to control the device in response to changes in the amount of flow of the object.

[0012] Furthermore, in one aspect of the above, the determination unit may perform a second comparison process to generate a second comparison result, which is the result of comparing the first inference result with a second threshold. The control unit may control the equipment corresponding to the inference device having the control unit based on the second comparison result. With this configuration, the equipment corresponding to the inference device is controlled based on the result of comparing the amount of flow of the object at the position corresponding to the inference device with the second threshold. This makes it possible to control the equipment more appropriately in response to changes in the amount of flow of the object.

[0013] Furthermore, in the above aspect, the object may be a moving object and a path that the moving object can travel. Multiple inference devices and multiple sensors may be provided and arranged corresponding to each of the multiple paths. Multiple devices may be provided for each inference device and may be devices that notify the moving object of permission information that permits it to pass. The inference unit may generate a first inference result that infers the presence of a moving object on the path based on the input information. The determination unit may generate a determination result that determines whether or not to permit the moving object to pass on the path based on the multiple first inference results generated by the multiple inference devices. The control unit may control whether or not to notify the device corresponding to the path of permission information based on the determination result generated by the determination unit of any of the multiple inference devices.

[0014] With this configuration, if a moving object is present on the path, the device controls whether or not to notify the moving object of information that permits it to pass. With this configuration, the presence of moving objects on multiple paths is taken into consideration, and the inference device determines whether or not to permit the moving object on the path to pass. This makes it possible to limit the number of moving objects passing on a path or at points where multiple paths intersect, thereby enabling smoother movement of moving objects along the path.

[0015] Furthermore, in one aspect of the above, if the first inference result infers that a moving object exists on the path, a type inference result inferring the type of the moving object based on the input information may be generated and added to the first inference result. The determination unit may generate a determination result in which it determines whether or not to permit the passage of the moving object on the path based on the multiple type inference results included in the multiple first inference results. With such a configuration, the type of moving object is taken into consideration when determining whether or not to permit the passage of the moving object on the path. For example, it may be determined to prioritize the passage of smaller moving objects. This makes it possible to allow moving objects to pass more smoothly on the path based on their type.

[0016] Furthermore, in the above aspect, the object may be a movable machine. Multiple inference devices and multiple sensors may be located around the movable machine. The equipment may be installed on the movable machine and capable of notifying information regarding the determination result. The inference unit may generate a first inference result, which is the result of inferring the presence or absence of an obstacle in a part of the periphery of the movable machine based on the input information. The determination unit may generate a determination result, which is the result of determining the presence or absence of an obstacle around the movable machine based on the multiple first inference results generated by the multiple inference devices. If the determination unit generates a determination result indicating the presence of an obstacle around the movable machine, the control unit may control the equipment to notify information indicating the presence of an obstacle around the movable machine.

[0017] According to such a configuration, when an obstacle exists around the movable machine, the device is controlled to notify the presence of the obstacle. Thereby, when an obstacle exists around the movable machine, it becomes possible to stop the operation of the movable machine.

[0018] Also, in the above aspect, the receiving unit may receive detection information indicating that it has been detected that the movable machine is operating. The inference unit may start generating a first inference result when detection information or first result information is received by the receiving unit. According to such a configuration, when the movable machine is operating, the inference unit starts generating a first inference result. Thereby, when the movable machine is operating, it is possible to determine whether or not an obstacle exists around the movable machine. As a result, when an obstacle exists around the movable machine, the operation of the movable machine can be more reliably stopped.

Effect of the Invention

[0019] According to one aspect of the present invention, it is possible to provide an information processing system capable of obtaining an inference result in consideration of the state of each location of the object when the state of the object to be inferred differs depending on the location.

Brief Description of the Drawings

[0020] [Figure 1] It is a schematic configuration diagram of an information processing system according to the first to third embodiments. [Figure 2] It is a schematic diagram showing an information processing system according to the first embodiment. [Figure 3] It is a block diagram showing a functional configuration of an inference device according to the first to third embodiments. [Figure 4] It is a flowchart showing a processing procedure according to the first to third embodiments. [Figure 5] It is a schematic diagram showing an information processing system according to the second embodiment. [Figure 6] It is a schematic diagram showing an information processing system according to the third embodiment. [Modes for carrying out the invention]

[0021] Embodiments of the present invention will be described in detail below with reference to the attached drawings. In this description, the same reference numerals will be used for the same element or element having the same function, and redundant explanations will be omitted.

[0022] Figure 1 is a schematic diagram of the information processing system 1 according to the first to third embodiments. The information processing system 1 is a system for inferring the state of an object T at each location and controlling a control device based on the inferred results. The information processing system 1 is composed of a plurality of sensors 2, at least one control device (device) 3, and a plurality of inference devices 10. The plurality of inference devices 10 are spatially distributed and connected to each other so as to be able to send and receive information. A sensor 2 is provided for each of the plurality of inference devices 10. The control device 3 is connected to at least one of the plurality of inference devices 10 so as to be able to send and receive information.

[0023] Multiple inference devices 10 each acquire different input information. Each inference device 10 infers the state of object T based on the input information according to a pre-built algorithm. At least one of the multiple inference devices 10 controls the control device 3 based on the inference result. Specifically, object T moves moment by moment, and the state of object T changes. Sensor 2 performs sensing on this object T and generates input information. Sensor 2 outputs the generated input information to the inference device 10 corresponding to Sensor 2. One of the multiple inference devices 10 controls the control device 3 based on the inference result.

[0024] The inference device 10 generates a first inference result that infers the state of object T based on the input information output from sensor 2. For example, the input information is an image of object T. In this case, the inference device 10 generates a first inference result that infers the state of object T based on the appearance of object T in the image. One of the multiple inference devices 10 acquires a second inference result from another inference device 10, generates a determination result based on the first and second inference results, and controls the control device 3 based on the determination result. In this way, the state of object T at each location is inferred by each inference device 10, and the generation of a determination result based on multiple inference results is performed sequentially by each inference device 10.

[0025] [First Embodiment] The information processing system 1A according to the first embodiment will be described with reference to Figure 2. Figure 2 is a schematic diagram showing the information processing system 1A according to the first embodiment. As shown in Figure 2, in the information processing system 1A according to the first embodiment, the object T is a moving body that flows from upstream to downstream on a predetermined channel F, for example, a river T1 (object, moving body). In the example shown in Figure 2, multiple tributaries F1, F2, and F3 branch off from points A, B, and C of the river T1, and sluice gates G1 to G3 are provided at positions corresponding to the branching points between the channel F and the tributaries F1 to F3. In the example shown in Figure 2, the water volume at point A is assumed to be 100, the water volume at point B is 300, and the water volume at point C is 100. The river T1 flows from point A to point C. Furthermore, the object T in the first embodiment may be any moving body that flows in one direction, for example, multiple items moved on a conveyor belt in a factory, or multiple people located on a moving walkway.

[0026] Multiple control devices 3 are arranged along the flow path F. In the example shown in Figure 2, multiple control devices 3(1), 3(2), and 3(3) are installed at sluice gates G1 to G3, respectively. Each control device 3 is, for example, an actuator capable of operating each of the sluice gates G1 to G3, and increases or decreases the water volume (flow rate) of the river T1 by operating to open or close each of the sluice gates G1 to G3. In other words, each control device 3 is a device that operates to increase or decrease the water volume (flow rate) of the river T1. Note that the multiple control devices 3 only need to be arranged in a way that allows control of the sluice gates corresponding to each control device 3. For example, control devices 3(1), 3(2), and 3(3) may be wirelessly connected to each of the sluice gates G1 to G3, respectively, and arranged in a way that allows remote control of the sluice gates G1 to G3.

[0027] Multiple sensors 2 are arranged along the channel F. In the example shown in Figure 2, the multiple sensors 2 are cameras that image the river T1 at points A, B, and C. The multiple sensors 2 generate images of the river T1 at each point along the channel F as input information. The multiple sensors 2 output the generated input information to the inference device 10.

[0028] Multiple inference devices 10 are distributed along the flow path F. The multiple inference devices 10 monitor the water volume of the river T1 via sensors 2 and control the control device 3. The multiple inference devices 10 pass the inference results from each inference device 10 from the upstream inference device 10 to the downstream inference device 10 and calculate the difference between the inference results. The multiple inference devices 10 control each control device 3 so that this difference is minimized. The operation of each of the multiple inference devices 10 is controlled independently of the control device 3.

[0029] In the example shown in Figure 2, multiple inference devices 10(1), 10(2), and 10(3) are provided in correspondence to three sensors 2 located at points A, B, and C. Each inference device 10 acquires input information from each sensor 2 and infers the water volume of the river T1 based on the acquired input information. The multiple inference devices 10(1), 10(2), and 10(3) control multiple control devices 3(1), 3(2), and 3(3) respectively based on the inference results. This controls the opening and closing of multiple sluice gates G1 to G3.

[0030] The specific functions of the inference device 10 will be described with reference to Figures 2 and 3. Figure 3 is a block diagram showing the functional configuration of the inference device 10 according to the first to third embodiments. The inference device 10 consists of an acquisition unit 11, an inference unit 12, a receiving unit 13, a transmitting unit 14, a determination unit 15, and a control unit 16. The acquisition unit 11 acquires input information from the sensor 2. Specifically, the acquisition unit 11 acquires images of the river T1 captured by the sensor 2 at points A, B, and C, respectively.

[0031] The inference unit 12 outputs a first inference result, which is the result of inferring the state of object T based on the input information acquired by the acquisition unit 11. In the example shown in Figure 2, the state of object T is the water volume of river T1. The inference unit 12 generates a first inference result, which is the result of inferring the water volume of river T1 based on the image acquired from sensor 2, according to a pre-built algorithm. For example, the inference unit 12 inputs the image into a machine learning model that has been built in advance by supervised learning or the like, and takes the value output from the machine learning model as the first inference result.

[0032] The receiving unit 13 receives first result information from another inference device 10, which includes the second inference result from that other inference device 10. In the first embodiment, the receiving unit 13 receives first result information from another adjacent inference device 10 located upstream of the flow path F. The transmitting unit 14 transmits second result information, which is the first result information with the first inference result added to it, to yet another inference device 10. Specifically, the transmitting unit 14 transmits second result information to yet another adjacent inference device 10 located downstream of the flow path F.

[0033] In the example shown in Figure 2, the transmitter 14 of the inference device 10(1) transmits information indicating the water volume at point A, inferred by the inference unit 12, to the inference device 10(2). The receiver 13 of the inference device 10(2) receives the information indicating the water volume at point A from the inference device 10(1). The transmitter 14 of the inference device 10(2) transmits the information indicating the water volume at point A and the information indicating the water volume at point B, generated by the inference unit 12 of the inference device 10(2), to the inference device 10(3). The inference device 10(3) receives the information indicating the water volume at point A and the information indicating the water volume at point B.

[0034] In other words, the inference device 10(1) generates information indicating the water volume at point A as the first inference result and transmits it to the inference device 10(2) as the second result information. The inference device 10(2) receives the information indicating the water volume at point A (second inference result) as the first result information. The inference device 10(2) generates information indicating the water volume at point B as the first inference result and transmits the information indicating the water volume at point B, which is the sum of the information indicating the water volume at point A, to the inference device 10(3) as the second result information. The inference device 10(3) receives the information indicating the water volume at point A and the information indicating the water volume at point B (second inference result) as the first result information.

[0035] The determination unit 15 generates a determination result based on the first inference result generated by the inference unit 12 and the second inference result included in the first result information received by the receiving unit 13. The control unit 16 controls the control device 3 based at least on the determination result.

[0036] In the first embodiment, the determination unit 15 calculates the difference between the first inference result and the second inference result and performs a first comparison process to generate a first comparison result, which is the result of comparing the first threshold with the difference. Based on the first comparison result, the control unit 16 controls the control device 3, which corresponds to either the inference device 10 corresponding to the control unit 16 or another inference device 10.

[0037] For example, if the first threshold is a positive number, the determination unit 15, in the first comparison process, derives the difference obtained by subtracting the second inference result from the first inference result, and generates a first comparison result that determines whether the difference exceeds the first threshold, which is a positive number. If the difference in the first comparison result exceeds the first threshold, the control unit 16 controls the control device 3 corresponding to another inference device 10 via the other inference device 10 to reduce the water volume of the river T1. The first threshold, which is a positive number, may be, for example, a value between 70% and 80% of the maximum water volume at each point in the river T1. The first threshold may also be a value that has been set in advance, or a value that has been derived by the inference device 10.

[0038] In the example shown in Figure 2, if the difference obtained by subtracting the water volume at point A from the water volume at point B exceeds a first threshold value which is a positive number, the sluice gate G1 corresponding to point A is opened and the water at point A is discharged into the tributary F1. Specifically, in the first comparison process, the determination unit 15 of the inference device 10(2) derives a difference of 200 obtained by subtracting the water volume of river T1 at point A (100) from the water volume of river T1 at point B (300), and generates a first comparison result which is the result of determining that the difference of 200 exceeds the first threshold value (150). Since the difference in the first comparison result exceeds the first threshold value, the control unit 16 controls the control device 3(1) via the inference device 10(1) to open the sluice gate G1 corresponding to point A. As a result, the amount of water flowing from point A to point B is reduced, and flooding of river T1 from channel F at point B is suppressed.

[0039] Furthermore, for example, if the first threshold is a negative number, the determination unit 15 generates a first comparison result in the first comparison process, which is the result of determining whether the difference obtained by subtracting the second inference result from the first inference result is lower than the first threshold, which is a negative number. If the difference in the first comparison result is lower than the first threshold, the control unit 16 controls the control device 3 corresponding to the inference device 10 having the control unit 16 to reduce the water volume of the river T1. Note that the first threshold, which is a negative number, may be, for example, a value obtained by multiplying the value of 70% to 80% of the maximum water volume at each point in the river T1 by -1.

[0040] In the example shown in Figure 2, if the difference obtained by subtracting the water volume at point B from the water volume at point C falls below a first threshold value, which is a negative number, the sluice gate G3 corresponding to point C is opened and the water at point C is discharged into the tributary F3. Specifically, in the first comparison process, the determination unit 15 of the inference device 10(3) generates a first comparison result in which it determines that the difference obtained by subtracting the water volume at point B (300) from the water volume at point C (100) is -200, which is below the first threshold value of -150. The control unit 16 determines that the difference in the first comparison result is below the first threshold value, and therefore controls the control device 3(3) to open the sluice gate G3 corresponding to point C. As a result, the water volume at point C decreases, and flooding of the river T1 from the channel F at point C is suppressed.

[0041] In the first embodiment, it is sufficient that either a positive or negative number is used in the first comparison process. For example, both a positive and a negative number may be used as the first threshold in the first comparison process, or only a positive number or only a negative number may be used as the first threshold in the first comparison process.

[0042] In the first embodiment, the determination unit 15 performs a second comparison process to generate a second comparison result, which is the result of comparing the first inference result with a second threshold. The control unit 16 controls the control device 3, which corresponds to the inference device 10 having the control unit 16, based on the second comparison result. Specifically, the determination unit 15 performs a second comparison process to generate a second comparison result, which is the result of determining whether the first inference result exceeds the second threshold. If the first inference result exceeds the second threshold in the second comparison result, the control unit 16 controls the control device 3, which corresponds to the inference device 10 having the control unit 16, to reduce the water volume of the river T1. The second threshold may be, for example, a value of 70% or more and 80% or less of the maximum water volume at each point in the river T1. The second threshold may also be a preset value or a value derived in the inference device 10.

[0043] In the example shown in Figure 2, when the water volume at point B exceeds the second threshold, the sluice gate G2 corresponding to point B is opened and the water at point B is discharged into the tributary F2. Specifically, the determination unit 15 of the inference device 10(2) generates a second comparison result in the second comparison process, which determines that the water volume 300 at point B exceeds the second threshold 200. The control unit 16 determines that the water volume at point B exceeds the second threshold in the second comparison result, and therefore controls the control device 3(2) to open the sluice gate G2 corresponding to point B. As a result, the water volume at point B decreases, and flooding of the river T1 from channel F at point B is suppressed.

[0044] The inference device 10 may generate a determination result based on the first inference results from multiple inference devices 10. In the example shown in Figure 2, the determination unit 15 of the inference device 10(3) may determine the flood risk of the entire river T1 based on first result information including information indicating the water volume at point A and information indicating the water volume at point B, and the first inference result which is information indicating the water volume at point C. In this case, the inference device 10 located further downstream among the multiple inference devices 10 has more information regarding the water volume of the river T1, so it is possible to determine the flood risk with greater accuracy. For example, the determination unit 15 of the inference device 10(3) calculates the total or average value of the water volumes at points A, B, and C, and determines that the flood risk is high if the calculated total or average value exceeds a predetermined threshold. Furthermore, if the determination unit 15 determines that the flood risk of the entire river T1 is high, the control unit 16 may control multiple control devices 3(1), 3(2), and 3(3) to open all sluice gates G1 to G3 in the river T1.

[0045] The processing procedure of the information processing system 1A according to the first embodiment will be explained with reference to Figure 4. Figure 4 is a flowchart showing the processing procedure according to the first to third embodiments. First, input information is acquired by the acquisition unit 11 (step S1). Next, the inference unit 12 generates a first inference result, which is the result of inferring the state of the object T based on the input information (step S2). Subsequently, the receiving unit 13 receives first result information, including a second inference result from another inference device 10, from another inference device 10 (step S3). Subsequently, the transmitting unit 14 transmits second result information, which is the first result information with the first inference result added to it, to yet another inference device 10 (step S4). Subsequently, the determination unit 15 generates a determination result based on the first result information and the second result information (step S5). Finally, the control unit 16 controls the control device (device) 3 based on the determination result (step S6).

[0046] The effects and advantages of the information processing system 1A according to the first embodiment described above will now be explained. In the information processing system 1A according to the first embodiment, different input information is input to each of the multiple inference devices 10. Each inference device 10 generates a first inference result in which the state of the object T is inferred based on the input information. With this configuration, if the state of the object T differs depending on the location, the state of the object T is inferred by each inference device 10 and a first inference result and a second inference result are generated. This makes it possible to obtain the first inference result and the second inference result while considering the state of the object T at each location. Furthermore, since the control device 3 is controlled based on the first inference result and the second inference result, the control device 3 can be controlled while considering the state of the object T at each location.

[0047] Furthermore, in the information processing system 1A according to the first embodiment, determination results are generated sequentially in each inference device 10. Compared to the case where multiple inference devices operate in cooperation, there is no waiting time until the processing of other inference devices 10 is completed, and no time is required for the mutual exchange of inference results with other inference devices 10. Therefore, determination results regarding the state of the object T can be generated in real time in response to changes in the state of the object T.

[0048] Furthermore, in the information processing system 1A according to the first embodiment, the input information is an image of the object T, and the inference unit 12 generates a first inference result, which is the result of inferring the state of the object T based on the appearance of the object T in the image. With this configuration, the state of the object T is inferred based on the appearance of the object T at each location. This makes it possible to generate the first inference result with greater accuracy.

[0049] Furthermore, in the information processing system 1A according to the first embodiment, one of the multiple inference devices 10 is equipped with a transmission unit 14 that transmits a second result information, which is the first result information with the first inference result added to it, to another inference device 10. With this configuration, the determination unit 15 of the other inference device 10 can generate a determination result with greater accuracy based on the inference results generated by the inference device 10 and the other inference device 10.

[0050] Furthermore, in the information processing system 1A according to the first embodiment, the control device 3 is controlled based on a first comparison result, which is obtained by comparing the difference between the amount of flow of object T at a position corresponding to the inference device 10 and the amount of flow of object T at a position corresponding to another adjacent inference device 10 upstream, with a first threshold. This makes it possible to control the control device 3 in response to changes in the amount of flow of object T.

[0051] Specifically, when the first threshold is a positive value, if the difference obtained by subtracting the water volume of river T1 at the location corresponding to another adjacent inference device 10 upstream from the water volume of river T1 at the location corresponding to the inference device 10 exceeds the first threshold, the control device 3 corresponding to the other inference device 10 is controlled via the other inference device 10 to reduce the water volume of river T1. This makes it possible to suppress a further increase in the water volume of river T1 at the location corresponding to the inference device 10 when the water from river T1 at the location corresponding to the other inference device 10 moves downstream, especially when the water volume of river T1 at the location corresponding to the inference device 10 has increased significantly. As a result, it is possible to suppress the river T1 at the location corresponding to the inference device 10 from overflowing from the channel F.

[0052] In the example shown in Figure 2, if the water volume at point B increases significantly, the difference obtained by subtracting the water volume at point A from the water volume at point B increases. Here, since this difference exceeds the first threshold, which is a positive value, the sluice gate G1 is opened and the water at point A is released into the tributary F1. As a result, the amount of water moving from point A to point B decreases, and the further increase in the water volume at point B is suppressed. Consequently, it is possible to prevent the river T1 from overflowing from channel F at point B.

[0053] Specifically, when the first threshold is a negative value, if the difference obtained by subtracting the amount of water in the river T1 at an adjacent upstream inference device 10 from the amount of water in the river T1 at the location corresponding to the inference device 10 falls below the negative first threshold, the control device 3 is controlled to reduce the amount of water in the river T1 at the location corresponding to the inference device 10. This makes it possible to suppress a further increase in the amount of water in the river T1 at the location corresponding to the inference device 10 when the water in the river T1 at the location corresponding to another inference device 10 moves downstream. As a result, it is possible to suppress the river T1 at the location corresponding to the inference device 10 from overflowing from the channel F.

[0054] For example, if the water volume at point B increases significantly, the difference between the water volume at point C and the water volume at point B becomes negative, and the absolute value of the difference increases. When this difference falls below the first threshold, which is a negative number, the sluice gate G3 corresponding to point C is opened. This prevents the river T1 from overflowing from the channel F at point C when water moves from point B to point C.

[0055] Furthermore, in the information processing system 1A according to the first embodiment, the determination unit 15 executes a second comparison process to generate a second comparison result, which is the result of determining whether the first inference result exceeds a second threshold. If the first inference result exceeds the second threshold in the second comparison result, the control unit 16 controls the control device 3 corresponding to the inference device 10 having the control unit 16 to reduce the water volume of the river T1. With this configuration, if the water volume of the river T1 at the location corresponding to the inference device 10 exceeds the second threshold, the control device 3 is controlled to reduce the water volume of the river T1 at the location corresponding to the inference device 10. This makes it possible to suppress the river T1 at the location corresponding to the inference device 10 from overflowing from the channel F when the water volume of the river T1 at the location corresponding to the inference device 10 increases significantly. In the example shown in Figure 2, if the water volume at point B increases significantly, the sluice gate G2 is opened and the water at point B is released into the tributary F2, thereby suppressing the river T1 from overflowing from the channel F at point B.

[0056] [Second Embodiment] The information processing system 1B according to the second embodiment will be described with reference to Figures 3 and 5. Figure 5 is a schematic diagram showing the information processing system 1B according to the second embodiment. The information processing system 1B is a system that monitors traffic by placing inference devices 10, etc., along multiple target routes. As shown in Figure 5, the target object T is a moving body and a route that the moving body can travel on, for example, an automobile (target object, moving body) T2 and a road (target object, route) T3 on which the automobile T2 travels. In Figure 5, multiple roads T31, T32 that intersect at point P and multiple automobiles T21, T22 located on the multiple roads T31, T32 are shown. Note that the target object T in the second embodiment is not limited to the above, and may be, for example, a pedestrian and a sidewalk that pedestrians can travel on.

[0057] Multiple sensors 2, multiple control devices 3, and multiple inference devices 10 are arranged for each road T3. In the example shown in Figure 5, sensor 2(1), control device 3(1), and inference device 10(1) are located on road T31, and sensor 2(2), control device 3(2), and inference device 10(2) are located on road T32.

[0058] Multiple sensors 2 generate images of road T3 and output the generated images as input information to multiple inference devices 10. In the example shown in Figure 5, multiple sensors 2(1) and 2(2) are positioned on multiple roads T31 and T32, respectively. Multiple sensors 2(1) and 2(2) are cameras that generate images of multiple roads T31 and T32, and output the generated images to multiple inference devices 10(1) and 10(2), respectively.

[0059] The control device 3 is a device that notifies vehicle T2 of permission information to proceed, and is, for example, a display or electronic signboard that displays information granting permission to proceed. In the example shown in Figure 5, the control device 3(1) notifies vehicle T21 that it is permitted to proceed by displaying the string "Proceed". The control device 3(2) notifies vehicle T22 that it is not permitted to proceed by displaying the string "Stop".

[0060] The inference device 10 determines whether or not to permit the passage of vehicle T2 on road T3 based on the input information acquired from sensor 2, and controls the control device 3 based on the determination result. The functional configuration of the inference device 10 will be described in detail below with reference to Figure 3.

[0061] The acquisition unit 11 acquires input information from the sensor 2. The inference unit 12 generates a first inference result that infers the presence of automobile T2 on road T3 based on the input information. If the inference unit 12 generates a first inference result, it also generates a type inference result that infers the type of automobile T2 and adds it to the first inference result. In the example shown in Figure 5, the acquisition unit 11 of the inference device 10(1) acquires images of automobile T21 and road T31 from the sensor 2(1). The inference unit 12 of the inference device 10(1) infers that automobile T21 is on road T31 and that automobile T21 is a light vehicle. Similarly, the inference unit 12 of the inference device 10(2) infers that automobile T22 is on road T32 and that automobile T22 is a regular passenger car.

[0062] Another inference device 10 transmits first result information, including the second inference result from the other inference device 10, to another inference device 10. The receiving unit 13 receives the first result information from the other inference device 10. In the example shown in Figure 5, the transmitting unit 14 of inference device 10(1) transmits first result information to inference device 10(2) that includes information indicating that automobile T21 is present on road T31 and information indicating that automobile T21 is a light vehicle. The receiving unit 13 of inference device 10(2) receives the first result information.

[0063] The determination unit 15 generates a determination result that determines whether or not to permit the passage of automobile T2 on road T3, based on the multiple first inference results generated by the multiple inference devices 10. In the example shown in Figure 5, the determination unit 15 of the inference device 10(2) generates a determination result that permits the passage of automobile T21 on road T31 and does not permit the passage of automobile T22 on road T32, based on information that multiple automobiles T21 and T22 exist on multiple roads T31 and T32, respectively, generated by the multiple inference devices 10(1) and 10(2).

[0064] The determination unit 15 generates a determination result that determines whether or not to permit vehicle T2 to pass on road T3, based on the multiple type inference results included in the multiple first inference results. In the example shown in Figure 5, the determination unit 15 of the inference device 10(2) decides to prioritize the passage of vehicle T21, which is a light vehicle, based on the fact that the types of multiple vehicles T21 and T22 are a light vehicle and a regular vehicle, respectively. Based on this decision, the determination unit 15 generates a determination result that permits vehicle T21 to pass on road T31 and does not permit vehicle T22 to pass on road T32.

[0065] The control unit 16 controls whether to not notify or notify the control device 3 corresponding to the road T3, based on the determination result generated by the determination unit 15 of any of the multiple inference devices 10. In the example shown in Figure 5, the control unit 16 of the inference device 10(2) controls the control device 3(2) so that the control device 3(2) corresponding to the road T32 does not notify the permit information, based on the determination result by the determination unit 15. The control unit 16 of the inference device 10(2) controls the control device 3(1) via the inference device 10(1) so that the control device 3(1) corresponding to the road T31 notifies the permit information, based on the determination result by the determination unit 15 of the inference device 10(2).

[0066] In the information processing system 1B according to the second embodiment, processing is performed in the same procedure as in the first embodiment, as shown in Figure 4.

[0067] The effects of the information processing system 1B according to the second embodiment described above will now be explained. In the information processing system 1B according to the second embodiment, if an automobile T2 is present on road T3, the control device 3 controls whether or not to notify the automobile T2 of information that permits it to pass. With this configuration, the presence of automobiles T2 on multiple roads T3 is taken into consideration, and the inference device 10 determines whether or not to permit the automobile T2 to pass on road T3. This makes it possible to limit the number of automobiles T2 traveling on road T3 or at points where multiple roads T3 intersect, thereby enabling automobiles T2 to travel more smoothly on road T3. In the example shown in Figure 5, the number of automobiles traveling at the point where road T31 and road T32 intersect can be limited to automobile T21 only.

[0068] Furthermore, in the information processing system 1B according to the second embodiment, if the first inference result infers that an automobile T2 exists on road T3, a type inference result is generated by inferring the type of automobile T2 based on the input information and added to the first inference result. The determination unit 15 generates a determination result that determines whether or not to permit automobile T2 to pass on road T3 based on the multiple type inference results included in the multiple first inference results. With this configuration, the type of automobile T2 is taken into consideration when determining whether or not to permit automobile T2 to pass on road T3. In the example shown in Figure 5, it is determined that automobile T21, which is smaller in size, will be given priority in passing. This makes it possible to allow automobile T2 to pass more smoothly on road T3 based on the type of automobile T2.

[0069] The information processing system 1B according to the second embodiment also provides the same effects and advantages as the first embodiment.

[0070] [Third Embodiment] The information processing system 1C according to the third embodiment will be described with reference to Figures 3 and 6. Figure 6 is a schematic diagram showing the information processing system 1C according to the third embodiment. The information processing system 1C is a system that detects the presence or absence of obstacles around an object T and notifies a control device installed on a movable machine. The object T is a movable machine such as a port cargo handling machine, for example, a crane (object) T4. An obstacle is an object that hinders the transfer of cargo by the crane T4, for example, other cargo handling machines, workers, and animals.

[0071] Multiple inference devices 10 and multiple sensors 2 are located around the crane T4. The sensors 2 are capable of sensing the crane T4 and its surroundings. The multiple inference devices 10 are configured to send and receive information from each other. A control device 3 is installed on the crane T4 and can notify it of information regarding the determination results generated by the inference devices 10.

[0072] In the example shown in Figure 6, the inference device 10(1) and sensor 2(1), and the inference device 10(2) and sensor 2(2) are installed around the crane T4. The inference device 10(3) and sensor 2(3) are installed on a drone D1 that flies around the crane T4. The inference device 10(4) and sensor 2(4) are installed on a drone D2 that flies around the crane T4. Each sensor 2 is a camera that images the crane T4 and its surroundings. The multiple inference devices 10(1) to 10(4) are connected to each other via wireless communication. The control device 3 is installed on the crane T4 and is connected to be controllable by any of the multiple inference devices 10.

[0073] Multiple sensors 2 output images of the crane T4 and its surroundings as input information to multiple inference devices 10. When the multiple inference devices 10 receive detection information indicating that the operation of the crane T4 has been detected, or first result information from another inference device 10, they generate a first inference result, which is the result of inferring the presence or absence of obstacles around the crane T4 based on the input information. The multiple inference devices 10 determine the presence or absence of obstacles around the crane T4 and control the control device 3 based on the determination result. The control device 3 notifies the equipment that controls the crane T4 or the operator of the crane T4 of the information regarding the presence or absence of obstacles around the crane T4.

[0074] The functional configuration of the inference device 10 will be described in detail below with reference to Figures 3 and 6.

[0075] The acquisition unit 11 acquires input information from the sensor 2. When the receiving unit 13 receives detection information or first result information (described later), the inference unit 12 generates a first inference result, which is the result of inferring the presence or absence of obstacles in a part of the area surrounding the crane T4 based on the input information. In the example shown in Figure 6, when the crane T4 is operating, a part of the space R through which the movable part T41 (object) of the crane T4 passes is imaged by the sensor 2(1). The inference unit 12 generates a first inference result, which is the result of inferring the presence or absence of obstacles in a part of the space R through which the movable part T41 passes, based on the image captured by the sensor 2(1). The inference unit 12 may also generate a first inference result at times other than when the receiving unit 13 receives detection information or first result information.

[0076] The receiving unit 13 receives first result information from another inference device 10, which includes the second inference result from that other inference device 10. The transmitting unit 14 transmits second result information to yet another inference device 10, which is the first result information with the first inference result added to it. In the example shown in Figure 6, the receiving unit 13 of inference device 10(2) receives first result information from inference device 10(1), which includes the second inference result generated by the inference unit 12 of inference device 10(1). The transmitting unit 14 of inference device 10(2) transmits second result information to inference device 10(3), which is the first result information with the first inference result generated by the inference unit 12 added to it.

[0077] The receiving unit 13 receives detection information indicating that the crane T4 is operating. In this case, the inference unit 12 starts generating a first inference result when it receives detection information or first result information from the receiving unit 13. In the example shown in Figure 6, detection information is transmitted from the crane T4 to the inference device 10(1), and the first inference result is generated in the inference device 10(1). Next, the first result information is passed from the inference device 10(1) to the inference device 10(2), and the first inference result is generated in the inference device 10(2). In this way, multiple first inference results are sequentially generated in multiple inference devices 10(1) to (4). The detection information may be generated in any of the multiple inference devices 10, or it may be generated in the crane T4.

[0078] The determination unit 15 generates a determination result, which is the result of determining whether or not there are obstacles around the crane T4, based on a plurality of first inference results generated by a plurality of inference devices 10. If the determination unit 15 generates a determination result indicating that an obstacle exists around the crane T4, the control unit 16 controls the control device 3 to notify it of information indicating that an obstacle exists around the crane T4.

[0079] In the example shown in Figure 6, the determination unit 15 of the inference device 10(4) generates a determination result, which is the result of determining whether or not there is an obstacle in the space R through which the movable part T41 passes, based on a plurality of first inference results generated by a plurality of inference devices 10(1) to (4). The control unit 16 of the inference device 10(4) controls the control device 3 to notify information indicating that an obstacle exists in the space R through which the movable part T41 passes.

[0080] In the information processing system 1C according to the third embodiment, processing is performed in the same procedure as in the first embodiment, as shown in Figure 4. However, in the information processing system 1C according to the third embodiment, the process (step S2) in which the inference unit 12 generates a first inference result, which is the result of inferring the state of the object T based on the input information, is started when the receiving unit 13 receives detection information or first result information.

[0081] The effects and benefits of the information processing system 1C according to the third embodiment described above will now be explained. In the information processing system 1C according to the third embodiment, the control device 3 is controlled to notify the system of the presence of an obstacle when an obstacle exists around the crane T4. This makes it possible to stop the operation of the crane T4 when an obstacle exists around the crane T4, for example, when an obstacle exists in the space R through which the movable part T41 passes.

[0082] Furthermore, in the information processing system 1C according to the third embodiment, the receiving unit 13 receives detection information indicating that the crane T4 is operating. The inference unit 12 starts generating a first inference result when the receiving unit 13 receives the detection information or the first result information. With this configuration, the inference unit 12 starts generating the first inference result when the crane T4 is operating. This makes it possible to determine whether or not there are obstacles around the crane T4 when the crane T4 is operating. As a result, the operation of the movable machine can be stopped more reliably when there are obstacles around the movable machine. In addition, since the inference processing and determination processing in the inference device 10 are executed only when the crane T4 is operating, the information processing system 1C can be operated more efficiently.

[0083] The information processing system 1C according to the third embodiment also provides the same effects and advantages as the first embodiment.

[0084] Although various embodiments of the present invention have been described above, the present invention is not limited to the above embodiments, and may be modified or applied to other things without changing the gist of each claim.

[0085] In the first to third embodiments described above, the number of inference devices 10 was between two and four, but is not limited to this. For example, the number of inference devices 10 may be five or more. More specifically, in the first embodiment, inference devices 10 and sensors 2 may be provided at points other than points A to C. In the second embodiment, inference devices 10, sensors 2 and control devices 3 may be provided at each of three or more roads T3. In the third embodiment, a drone may be provided in addition, or inference devices 10 and sensors 2 may be provided in addition.

[0086] Furthermore, while the order in which the first result information is passed is determined in the second and third embodiments, the system is not limited to this. In the second embodiment, it is sufficient for one of the multiple inference devices 10 to have the first inference results from all the inference devices 10, so the first result information may be passed in any order. In this case, when the receiving unit 13 of the inference device 10 receives the first result information, the transmitting unit 14 may transmit to another inference device 10 a second result information, which is the first result information with the first inference result added to it.

[0087] In the third embodiment, it is sufficient for one of the multiple inference devices 10 to have the first inference results from multiple inference devices 10 capable of monitoring space R, so the first result information may be passed in any order. In addition, an inference device 10 that monitors the portion of space R through which the movable part 41 passes faster may prioritize transmitting the first result information or detection information. This makes it possible to prioritize determining whether or not there is an obstacle in the portion of space R through which the movable part T41 passes faster. As a result, it is possible to more reliably suppress the crane T4 from interfering with an obstacle.

[0088] The constituent elements of the present invention are described below. <Invention 1> Multiple inference devices are connected to each other in a way that allows them to send and receive information, and each device acquires different input information from the others to infer the state of an object. At least one device controllably connected to the inference device, Multiple sensors that perform sensing to generate the input information and output the generated input information to the corresponding inference device among the multiple inference devices, Equipped with, One of the multiple inference devices, An acquisition unit that acquires the aforementioned input information, An inference unit that generates a first inference result, which is the result of inferring the state of the object based on the input information acquired by the acquisition unit, A receiving unit that receives first result information, including a second inference result from another inference device, A determination unit that generates a determination result based at least on the first inference result and the second inference result, An information processing system comprising: a control unit that controls the device based at least on the determination result. <Invention 2> The input information is an image of the object, The information processing system according to Invention 1, wherein the inference unit generates the first inference result, which is the result of inferring the state of the object based on the appearance of the object in the image. <Invention 3> The information processing system according to invention 1 or 2, wherein one of the plurality of inference devices further comprises a transmission unit that transmits a second result information, which is the first result information with the first inference result added to it, to another inference device. <Invention 4> The aforementioned object is a moving body that flows from upstream to downstream along a predetermined flow path. Multiple of the aforementioned devices are arranged along the flow path and operate to increase or decrease the amount of the flow of the object, The plurality of sensors are arranged along the flow path, The plurality of inference devices are provided so that their operation can be controlled for each device. The inference unit generates the first inference result, which is the result of inferring the amount of flow based on the input information. The receiving unit receives the first result information from the other inference device adjacent to it on the upstream side of the flow path. The transmitting unit transmits the second result information to yet another inference device located downstream of the flow path. The determination unit calculates the difference between the first inference result and the second inference result and performs a first comparison process to generate a first comparison result which is the result of comparing the first threshold with the difference. The information processing system according to Invention 3, wherein the control unit controls the device corresponding to either the inference device corresponding to the control unit or the other inference device based on the first comparison result. <Invention 5> The determination unit executes a second comparison process to generate a second comparison result, which is the result of comparing the first inference result with the second threshold. The information processing system according to Invention 4, wherein the control unit controls the device corresponding to the inference device having the control unit based on the second comparison result. <Invention 6> The aforementioned object is a moving body and a path through which the moving body can travel. The plurality of inference devices and the plurality of sensors are provided in correspondence for each of the plurality of paths. Multiple of the aforementioned devices are provided for each of the inference devices and are devices that notify the moving object of permission information that permits it to pass. The inference unit generates the first inference result, which infers the presence of the moving object on the path, based on the input information. The determination unit generates a determination result that determines whether or not to permit the passage of the moving object on the path, based on the plurality of first inference results generated by the plurality of inference devices. The control unit controls the notification / non-notification of the permission information to the device corresponding to the route, based on the determination result generated by the determination unit of one of the plurality of inference devices, in the information processing system according to any one of claims 1 to 3 of the invention. <Invention 7> If the inference unit infers in the first inference result that the moving object exists on the path, it further generates a type inference result that infers the type of the moving object based on the input information and adds it to the first inference result. The information processing system according to Invention 6, wherein the determination unit generates a determination result that determines whether or not to permit the passage of the moving object on the path, based on the plurality of type inference results included in the plurality of first inference results. <Invention 8> The aforementioned object is a movable machine, The plurality of inference devices and the plurality of sensors are located around the movable machine, The aforementioned device is installed on the movable machine and is capable of notifying information regarding the determination result. The inference unit generates the first inference result, which is the result of inferring the presence or absence of obstacles in a part of the periphery of the movable machine based on the input information. The determination unit generates the determination result, which is the result of determining whether or not there is an obstacle around the movable machine, based on the plurality of first inference results generated by the plurality of inference devices. The information processing system according to any one of inventions 1 to 3, wherein the control unit controls the device to notify the information indicating the presence of an obstacle around the movable machine when the determination unit generates a determination result indicating the presence of an obstacle around the movable machine. <Invention 9> The receiving unit receives detection information indicating that the movable machine is operating. The information processing system according to Invention 8, wherein the inference unit starts generating the first inference result when the receiving unit receives the detection information or the first result information. [Explanation of Symbols]

[0089] 1,1A,1B,1C Information Processing System 2,2(1),2(2),2(3),2(4) Sensors 3,3(1),3(2),3(3) Control device 10,10(1),10(2),10(3),10(4) Inference device 11 Acquisition Department 12 Reasoning part 13 Receiving Unit 14. Transmitter 15 Judgment section 16 Control Unit F channel T object T1 River (object, moving body) T2, T21, T22: Automobile (object, moving object) T3, T31, T32 Roads (Objects, Routes) T4 crane (object) T41 Movable part (object).

Claims

1. Multiple inference devices are connected to each other in a way that allows them to send and receive information, and each device acquires different input information from the others to infer the state of an object. At least one device controllably connected to the inference device, Multiple sensors that perform sensing to generate the input information and output the generated input information to the corresponding inference device among the multiple inference devices, Equipped with, One of the plurality of inference devices, An acquisition unit that acquires the aforementioned input information, An inference unit that generates a first inference result, which is the result of inferring the state of the object based on the input information acquired by the acquisition unit, A receiving unit that receives first result information, including a second inference result from another inference device, A determination unit that generates a determination result based at least on the first inference result and the second inference result, A control unit that controls the device based at least on the determination result, An information processing system further comprising a transmission unit that transmits to another inference device a second result information obtained by adding the first inference result to the first result information.

2. The aforementioned object is a moving body that flows from upstream to downstream along a predetermined flow path. Multiple of the aforementioned devices are arranged along the flow path and operate to increase or decrease the amount of the flow of the object, The plurality of sensors are arranged along the flow path, The plurality of inference devices are provided so that their operation can be controlled for each device. The inference unit generates the first inference result, which is the result of inferring the amount of flow based on the input information. The receiving unit receives the first result information from the other inference device adjacent to it on the upstream side of the flow path. The transmitting unit transmits the second result information to yet another inference device located downstream of the flow path. The determination unit calculates the difference between the first inference result and the second inference result and performs a first comparison process to generate a first comparison result which is the result of comparing the first threshold with the difference. The information processing system according to claim 1, wherein the control unit controls the device corresponding to either the inference device corresponding to the control unit or the other inference device based on the first comparison result.

3. The determination unit executes a second comparison process to generate a second comparison result, which is the result of comparing the first inference result with the second threshold. The information processing system according to claim 2, wherein the control unit controls the device corresponding to the inference device having the control unit based on the second comparison result.

4. The aforementioned object is a moving body and a path through which the moving body can travel. The plurality of inference devices and the plurality of sensors are provided in correspondence for each of the plurality of paths. Multiple of the aforementioned devices are provided for each of the inference devices and are devices that notify the moving object of permission information that permits it to pass. The inference unit generates the first inference result, which infers the presence of the moving object on the path, based on the input information. The determination unit generates a determination result that determines whether or not to permit the passage of the moving object on the path, based on the plurality of first inference results generated by the plurality of inference devices. The information processing system according to claim 1, wherein the control unit controls the notification / non-notification of the permission information to the device corresponding to the route, based on the determination result generated by the determination unit of one of the plurality of inference devices.

5. If the inference unit infers in the first inference result that the moving object exists on the path, it further generates a type inference result that infers the type of the moving object based on the input information and adds it to the first inference result. The information processing system according to claim 4, wherein the determination unit generates a determination result that determines whether or not to permit the passage of the moving object on the path, based on the plurality of type inference results included in the plurality of first inference results.

6. A plurality of inference devices that are connected to each other so as to be able to send and receive information from each other, and each acquires different input information from each other to infer the state of an object, At least one device controllably connected to the inference device, Multiple sensors that perform sensing to generate the input information and output the generated input information to the corresponding inference device among the multiple inference devices, Equipped with, One of the plurality of inference devices, An acquisition unit that acquires the aforementioned input information, An inference unit that generates a first inference result, which is the result of inferring the state of the object based on the input information acquired by the acquisition unit, A receiving unit that receives first result information, including a second inference result from another inference device, A determination unit that generates a determination result based at least on the first inference result and the second inference result, The device includes a control unit that controls the device based at least on the determination result, The aforementioned object is a movable machine, The plurality of inference devices and the plurality of sensors are located around the movable machine, The aforementioned device is installed on the movable machine and is capable of notifying information regarding the determination result. The inference unit generates the first inference result, which is the result of inferring the presence or absence of obstacles in a part of the periphery of the movable machine based on the input information. The determination unit generates the determination result, which is the result of determining whether or not there is an obstacle around the movable machine, based on the plurality of first inference results generated by the plurality of inference devices. The control unit is an information processing system that, when the determination unit generates a determination result indicating that an obstacle exists around the movable machine, controls the device to notify the information indicating that an obstacle exists around the movable machine.

7. The receiving unit receives detection information indicating that the movable machine is operating. The information processing system according to claim 6, wherein the inference unit starts generating the first inference result when the receiving unit receives the detection information or the first result information.

8. The information processing system according to claim 6, wherein one of the plurality of inference devices further comprises a transmission unit that transmits to another inference device a second result information obtained by adding the first inference result to the first result information.

9. The input information is an image of the object, The information processing system according to any one of claims 1 to 8, wherein the inference unit generates the first inference result, which is the result of inferring the state of the object based on the appearance of the object in the image.

Images