Autonomous body task management device and autonomous body task management method

Abstract

Achieving both optimal scheduling and online control. [Solution] An autonomous body task management device that manages tasks for multiple autonomous bodies in an automated system in which the autonomous bodies perform tasks is characterized by comprising: a state determination unit that determines whether the autonomous body is executing an independent task that is not affected by other autonomous bodies; a task planning unit that, if the autonomous body is executing the independent task, plans a task to be executed after the independent task is completed; and a task allocation unit that assigns the task planned by the task planning unit to the autonomous body.

Description

[Technical Field]

[0001] The present invention relates to an autonomous body task management device and an autonomous body task management method. [Background technology]

[0002] Japanese Patent Application Laid-Open No. 2023-72447 (Patent Document 1) describes a technology for managing the movement of autonomous objects. This publication states, "We provide a mobile object control device that can optimize overall efficiency while avoiding interference between mobile objects." "In the mobile control system 100, two mobile objects cannot pass each other on the same route. Therefore, when the routes intersect, in order to avoid collisions between the AGVs 30, one AGV 30 must wait and then move the other AGV 30 before moving the other AGV 30. Therefore, a collision prohibition condition is used as a constraint for the mixed integer programming problem to calculate a transportation schedule that minimizes the evaluation value while avoiding collisions between the mobile objects. Furthermore, a transportation process is a process in which if one AGV 30 is present on a specific route, other AGVs 30 cannot enter the specific route. Since the transportation schedule is determined by solving such a mixed integer programming problem, it is possible to optimize overall efficiency while avoiding interference between the AGVs 30." [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2023-72447 Summary of the Invention [Problem to be solved by the invention]

[0004] In automated systems where multiple autonomous entities such as robots and guided vehicles operate, a control method (rolling horizon method) that optimizes task scheduling while predicting the future behavior of each autonomous entity is effective. The optimality of scheduling improves the longer the prediction period for the system behavior being considered is set. However, the longer the prediction period is set, the higher the calculation cost for scheduling becomes, and there is a concern that the online nature of control may be impaired.

[0005] Therefore, the present invention aims to achieve both optimal scheduling and online control. [Means for solving the problem]

[0006] In order to achieve the above-mentioned object, one representative autonomous body task management device of the present invention is an autonomous body task management device that manages tasks of an autonomous body in an automated system in which multiple autonomous bodies perform tasks, and is characterized by comprising: a state determination unit that determines whether the autonomous body is executing an independent task that is not affected by other autonomous bodies; a task planning unit that, if the autonomous body is executing the independent task, plans a task to be executed after the independent task is completed; and a task allocation unit that assigns the task planned by the task planning unit to the autonomous body. Furthermore, one representative autonomous body task management method of the present invention is an autonomous body task management method for managing tasks of an autonomous body in an automated system in which a plurality of autonomous bodies perform tasks, and is characterized by including a state determination step for determining whether the autonomous body is executing an independent task that is not affected by other autonomous bodies, a task planning step for planning a task to be executed after completing the independent task if the autonomous body is executing the independent task, and a task allocation step for assigning the task planned in the task planning step to the autonomous body. [Effects of the Invention]

[0007] According to the present invention, both optimal scheduling and online control can be achieved. Problems, configurations, and effects other than those described above will become clear from the following description of the embodiments. [Brief explanation of the drawings]

[0008] [Figure 1] An explanatory diagram of a configuration in which an autonomous body is managed by one management device. [Figure 2] An explanatory diagram of a configuration in which an autonomous body is managed by multiple management devices. [Figure 3] Configuration diagram of the autonomous task management device [Figure 4] A flowchart showing the process of planning a task [Figure 5] Flowchart detailing successor task selection (part 1) [Figure 6] Flowchart detailing successor task selection (part 2) [Figure 7] An example of unmanned truck operation (part 1) [Figure 8] Illustration of an example of unmanned truck operation (part 2) [Figure 9] An explanatory diagram showing the correspondence between states and tasks [Figure 10] Order picking system diagram DETAILED DESCRIPTION OF THE INVENTION

[0009] Hereinafter, an embodiment will be described with reference to the drawings. [Example]

[0010] In the first embodiment, operation management in a mine is exemplified as an automated system in which multiple autonomous entities perform tasks. An autonomous haulage system (AHS) loads mining excavated materials onto multiple unmanned trucks and repeatedly transports them to a storage location. The travel time from the loading location or storage location to the next target point until the branch point is reached is used for scheduling, and a calculation time of several tens of seconds to several minutes is secured to optimize the transportation task (selection of route and destination).

[0011] A solo task is a task that is not affected by other autonomous entities until it reaches a branching point. An autonomous entity is an unmanned truck or other work machine. A solo task is a task that is performed only by the autonomous entity being scheduled, and since the uncertainty of the work is small, the completion time can be easily predicted with high accuracy. One example of a solo task is traveling along a fixed route. Note that tasks that are affected by work by workers or tasks that involve fluctuations in the weight of the load have low accuracy in predicting the completion time, so it is preferable to exclude them from solo tasks.

[0012] A task that can be completed with the help of other autonomous entities is called a collaborative task. Collaborative tasks have a large degree of uncertainty in the tasks, making it difficult to predict the completion time.

[0013] In the system of the first embodiment, when an independent task is being executed, a task to be executed after the independent task is completed is planned. The system can use the time until the independent task is completed, thereby improving the optimality of scheduling. Furthermore, if the planning of the next task is completed at the time the independent task is completed, online control can be ensured.

[0014] FIG. 1 is an explanatory diagram of a configuration in which autonomous bodies are managed by one management device. Autonomous bodies A1 to AN are unmanned trucks and can communicate with a central management device 10 via a wireless network. Autonomous bodies B1 to BM are unmanned excavators and can communicate with the central management device 10 via a wireless network. In the configuration shown in FIG. 1, the central management device 10 plans tasks for multiple types of autonomous bodies (autonomous bodies A1 to AN and autonomous bodies B1 to BM) that have different functions.

[0015] FIG. 2 is an explanatory diagram of a configuration in which autonomous bodies are managed by multiple management devices. The autonomous bodies A1 to AN are unmanned trucks and can communicate with the autonomous body A management device 20 via a wireless network. The autonomous bodies B1 to BM are unmanned excavators and can communicate with the autonomous body B management device 30 via a wireless network. The autonomous body A management device 20 and the autonomous body B management device 30 can communicate with the central management device 10, for example, via a wireless network. In the configuration shown in FIG. 2, the autonomous body A management device 20 plans tasks for the autonomous bodies A1 to AN. The autonomous body A management device 20 acquires tasks for the autonomous bodies B1 to BM via the central management device 10 and plans tasks for the autonomous bodies A1 to AN after taking into account the tasks of the autonomous bodies B1 to BM. Furthermore, the autonomous body B management device 30 plans tasks for the autonomous bodies B1 to BM. The autonomous body B management device 30 acquires the tasks of the autonomous bodies A1 to AN via the central management device 10, and plans the tasks of the autonomous bodies B1 to BM after taking into consideration the tasks of the autonomous bodies B1 to AN. The central management device 10 receives and manages the tasks created by the autonomous body A management device 20 and the tasks created by the autonomous body B management device 30.

[0016] FIG. 3 is a configuration diagram of an autonomous body task management device 40. The autonomous body task management device 40 executes a process of planning and managing tasks for an autonomous body. If the central management device 10 plans tasks for an autonomous body, the central management device 10 operates as the autonomous body task management device 40. If the autonomous body A management device 20 plans tasks for an autonomous body, the autonomous body A management device 20 operates as the autonomous body task management device 40. If the autonomous body B management device 30 plans tasks for an autonomous body, the autonomous body B management device 30 operates as the autonomous body task management device 40. If an autonomous body plans its own tasks, the autonomous body operates as the autonomous body task management device 40.

[0017] The autonomous body task management device 40 is a computer having a processor 41, a memory 42, a storage unit 43, and a communication unit 444. Specifically, the processor 41 reads out a predetermined program from the storage unit 43, loads it into the memory 42, and executes it sequentially, thereby realizing various functions of the autonomous body task management device 40. The communication unit 44 is an interface for communicating with external management devices and autonomous bodies.

[0018] The functional units of the autonomous unit task management device 40 include a state determination unit 51, a task planning unit 52, and a task allocation unit 53. The state determination unit 51 determines whether or not an autonomous unit is executing an independent task that is not affected by other autonomous units.

[0019] When an autonomous body is executing an independent task, the task planning unit 52 plans a task to be executed after the independent task is completed. Specifically, when an autonomous body is executing an independent task and there are multiple candidate tasks to be executed after the independent task is completed, the task planning unit 52 plans a task to be executed after the independent task is completed. The candidate tasks are defined in advance in association with the completion event of the independent task.

[0020] Furthermore, the task planning unit 52 can accept termination of the task planning process during the task planning process. As one example, the task planning unit 52 terminates the task planning process when the task planning process is not completed within a time limit set corresponding to the content of the single task. As another example, the task planning unit 52 terminates the task planning process when the autonomous body passes a specific position due to its movement. When the task planning process is terminated, the task planning unit 52 executes the task planning process using a different algorithm. Specifically, until a predetermined condition is met, the task planning unit 52 executes the process of planning an optimal task by setting a long prediction period for the system behavior to be considered, and then employs a suboptimal calculation with a low calculation cost after the predetermined condition is met.

[0021] The task allocation unit 53 allocates the tasks planned by the task planning unit 52 to the autonomous bodies. The autonomous bodies execute the allocated tasks sequentially. Therefore, the task allocation unit 53 allocates the next task before the autonomous body completes the task currently being executed, thereby eliminating waiting time between tasks and ensuring online control of the autonomous bodies.

[0022] 4 is a flowchart showing the processing procedure for planning a task. When an autonomous object k reaches a state s, the system of the first embodiment sequentially executes the following steps S101 to S106.

[0023] In step S101, the task allocation unit 53 assigns a task associated with the current state s. Here, the assigned task is an independent task T s I Then, the process proceeds to step S102.

[0024] Step S102: The autonomous entity k executes the independent task T s I Then, the process proceeds to step S103. In step S103, the task planning unit 52 s I It is determined whether there are multiple candidates for the succeeding task after completion. If there are multiple succeeding tasks, the process proceeds to step S104. If there is only one candidate for the succeeding task, the process proceeds to step S106.

[0025] In step S104, the task planning unit 52 selects a succeeding task T' from among multiple candidates. Then, the process proceeds to step S105. Details of this step will be described later. In step S105, the task allocation unit 53 adds the selected succeeding task T' to the schedule of the autonomous body, and ends the process. In step S106, the task allocation unit 53 adds the succeeding task to the schedule of the autonomous body, and the process ends.

[0026] 5 is a flowchart showing the details of the selection of the succeeding task shown in step S 104. When switching the method for obtaining a task based on a time limit, the autonomous task management device 40 sequentially executes steps S201 to S207.

[0027] In step S201, the task planning unit 52 calculates the independent task T s I The corresponding calculation time t s I Then, the process proceeds to step S202. In step S202, the task planning unit 52 sets a condition C for selecting a succeeding task, and then the process proceeds to step S203.

[0028] In step S203, the task planning unit 52 calculates the calculation time t s I If the time has passed, the process proceeds to step S207. If the time has not passed, the process proceeds to step S204. In step S204, the task planning unit 52 calculates the calculation result R for selecting the succeeding task, and then the process proceeds to step S205. In step S205, the task planning unit 52 determines whether or not the calculation result R satisfies the condition C. If not, the process returns to step S203. If satisfied, the process proceeds to step S206.

[0029] In step S206, the task allocation unit 53 adds the succeeding task T' based on the calculation result R to the schedule of the autonomous body, and then ends the process. This step corresponds to step S106 in FIG. In step S207, the task planning unit 52 selects a succeeding task T' from among the multiple candidates based on the alternative method, and the task allocation unit 53 adds the succeeding task T' to the schedule, and the process ends.

[0030] 6 is a flowchart showing the details of the selection of the succeeding task shown in step S 104. When switching the method for determining a task based on the passage of a specific position, the autonomous task management device 40 sequentially executes steps S301 to S307.

[0031] In step S301, the task planning unit 52 calculates the independent task T s I The corresponding calculation continuation point p s I Then, the process proceeds to step S302. In step S302, the task planning unit 52 sets a condition C for selecting a succeeding task, and then the process proceeds to step S303.

[0032] In step S303, the task planning unit 52 s I If it has passed, the process proceeds to step S307. If it has not passed, the process proceeds to step S304. In step S304, the task planning unit 52 calculates the calculation result R for selecting the succeeding task, and then the process proceeds to step S305. In step S305, the task planning unit 52 determines whether or not the calculation result R satisfies the condition C. If not, the process returns to step S303. If satisfied, the process proceeds to step S306.

[0033] In step S306, the task allocation unit 53 adds the succeeding task T' based on the calculation result R to the schedule of the autonomous body, and then ends the process. This step corresponds to step S106 in FIG. In step S307, the task planning unit 52 selects a succeeding task T' from among the multiple candidates based on the alternative method, and the task allocation unit 53 adds the succeeding task T' to the schedule, and the process ends.

[0034] 7 and 8 are explanatory diagrams of an example of the operation of an unmanned truck. An unmanned truck 60 repeats the task of loading earth and sand at point 2 or point 3 and unloading it at point 1 or point 4. For convenience, points 1 to 4 are assumed to be reachable by a single road from one branch point.

[0035] Figure 7 shows an unmanned truck 60 that has unloaded earth and sand at point 1 traveling to a branch point. Traveling from point 1 to the branch point is not affected by other autonomous bodies, is not affected by workers, and the weight of the load does not change. Therefore, traveling from point 1 to the branch point is a single task. After the branch point, the candidate tasks are to head to point 2 and to head to point 3.

[0036] Figure 8 shows an unmanned truck 60 loaded with earth and sand traveling from point 2 to a branch point. Traveling from point 2 to the branch point is not affected by other autonomous bodies, is not affected by workers, and the weight of the load does not change. Therefore, traveling from point 2 to the branch point is a single task. After the branch point, the candidate tasks are a task heading to point 1 and a task heading to point 4.

[0037] 9 is an explanatory diagram showing the correspondence between states and tasks. At a location where the unmanned truck 60 performs work, an in-progress trigger that starts an individual task, an individual task that is executed by the in-progress trigger, and a candidate destination are linked and managed.

[0038] Specifically, at point 1, there is a work-in-progress trigger "Loading of goods onto unmanned truck is completed" and a single task T1 I , and the movement candidates "Point 2, Point 3" are linked. Point 2 has a work-in-progress trigger "Unloading of transported goods from unmanned truck is completed" and an independent task T2 I , and the movement candidates "Point 1, Point 4" are linked. Point 3 has a work-in-progress trigger "Unloading of transported goods from unmanned truck is completed" and a single task T3 I , and the movement candidates "Point 1, Point 4" are linked. Point 4 has a work-in-progress trigger "Loading of transported goods onto unmanned truck is completed" and a single task T4 I , and the travel candidates "Point 2, Point 3" are linked. [Example]

[0039] In Example 2, an order picking system for a logistics warehouse is used as an example of an automated system in which multiple autonomous entities perform tasks. In an order picking system, an automated guided vehicle (AGV) lifts shelves filled with stored items and transports them to a work station where workers pick and sort the items. When determining the destination of the AGV after the picking work is completed, if transport scheduling takes too long, the work station will be blocked and picking work from subsequent shelves will not be possible. Therefore, the travel time from the station to the route network of the transport area is used in the scheduling calculation.

[0040] Figure 2 is an explanatory diagram of an order picking system. Stored items are filled on inventory shelves in the inventory area. An AGV crawls under the inventory shelves, lifts them up, and carries them to the work station. A picker, a worker, removes items from the inventory shelves, sorts them, and places them in shipping boxes on the sorting shelf. The shipping boxes are then transported by conveyor to the inspection area. An exit path 70 is provided next to the work station. After completing the picking operation, the AGV first leaves the work station via the exit path 70. While traveling along the exit path 70, the AGV plans its next destination. This allows the AGV to immediately vacate the work station without waiting for the next destination to be determined. Furthermore, the time spent traveling along the exit path 70 can be used to calculate the destination, ensuring sufficient calculation time.

[0041] As described above, the device disclosed in the embodiment is an autonomous body task management device 40 that manages tasks of an autonomous body in an automated system in which a plurality of autonomous bodies perform tasks, and is characterized by including a state determination unit 51 that determines whether the autonomous body is executing an independent task that is not affected by other autonomous bodies, a task planning unit 52 that, if the autonomous body is executing the independent task, plans a task to be executed after the independent task is completed, and a task allocation unit 53 that assigns the task planned by the task planning unit to the autonomous body. According to this configuration and operation, by taking into consideration the timing for creating a plan for the tasks of the autonomous body and creating the plan at an appropriate timing, it is possible to achieve both optimal scheduling and online control.

[0042] Furthermore, when the single task is being executed and there are multiple candidate tasks to be executed after the single task is completed, the task planning unit 52 plans the task to be executed after the single task is completed. This allows the optimum task to be selected from multiple task candidates.

[0043] Furthermore, the task planning unit 52 can accept a request to stop the task planning process during the task planning process. This makes it possible to avoid situations where the autonomous body is put into a waiting state without completing the task planning process.

[0044] Furthermore, the automated system involves multiple types of autonomous entities with different functions. This configuration allows for the creation of plans for tasks of autonomous entities in a system that includes a variety of autonomous entities.

[0045] Furthermore, the task planning section 52 stops the task planning process if the task planning process is not completed within a time limit set in accordance with the content of the single task. Furthermore, the task planning unit 52 stops the process of planning the task when the autonomous body passes a specific position due to its movement. Furthermore, when the task planning section 52 stops the process of planning the task, it executes a process of planning the task using an algorithm different from that of the process. According to this configuration and operation, if an optimal schedule cannot be calculated by the time an individual task is completed, a suboptimal schedule is calculated, thereby ensuring online control of the autonomous body.

[0046] Furthermore, the tasks of the autonomous body include tasks in which the weight of the payload changes and tasks in which the weight of the payload does not change, and the independent task is a task in which the weight of the payload does not change. In this way, by not using the execution of a task that includes factors that reduce the accuracy of the prediction in the schedule calculation, the schedule can be calculated stably.

[0047] The task candidates are predefined in association with the completion event of the single task. Therefore, the task to be executed after the single task can be appropriately selected from the plurality of tasks.

[0048] The present invention is not limited to the above-described embodiments, but includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to those including all of the described configurations. Furthermore, not only can the configurations be deleted, but also replacements and additions of configurations are possible. For example, in the above embodiment, unmanned trucks and AGVs are shown as examples of autonomous bodies, but the autonomous bodies and tasks can be set arbitrarily. [Explanation of symbols]

[0049] 10: central control unit, 20: autonomous body A control unit, 30: autonomous body B control unit, 40: autonomous body task control unit, 41: processor, 42: memory, 43: storage unit, 44: communication unit, 51: state determination unit, 52: task planning unit, 53: task allocation unit, 60: unmanned truck, 70: exit road

Claims

1. An autonomous body task management device that manages tasks of a plurality of autonomous bodies in an automated system in which the autonomous bodies perform tasks, comprising: a state determination unit that determines whether the autonomous entity is executing an independent task that is not affected by other autonomous entities; a task planning unit that, when the single task is being executed, plans a task to be executed after the single task is completed; a task allocation unit that allocates the task planned by the task planning unit to the autonomous body; An autonomous body task management device comprising:

2. 2. The autonomous body task management device according to claim 1, wherein the task planning unit plans a task to be executed after the completion of the single task when the single task is being executed and there are multiple candidate tasks to be executed after the single task is completed.

3. 2. The autonomous body task management device according to claim 1, The autonomous body task management device is characterized in that the task planning unit is capable of accepting a stop of the process of planning the task during the process of planning the task.

4. 2. The autonomous body task management device according to claim 1, The autonomous body task management device is characterized in that the automation system involves a plurality of types of autonomous bodies having different functions.

5. 4. The autonomous body task management device according to claim 3, The task planning unit is configured to stop the process of planning the task if the process of planning the task is not completed within a time limit set corresponding to the content of the single task.

6. 4. The autonomous body task management device according to claim 3, The autonomous body task management device is characterized in that the task planning unit suspends the process of planning the task when the autonomous body passes a specific position due to movement of the autonomous body.

7. 4. The autonomous body task management device according to claim 3, The autonomous body task management device is characterized in that, when the task planning process is stopped, the task planning unit executes a process of planning a task using an algorithm different from that of the process.

8. 2. The autonomous body task management device according to claim 1, wherein the tasks of the autonomous body include a task in which the weight of the payload changes and a task in which the weight of the payload does not change, and the independent task is a task in which the weight of the payload does not change.

9. 3. The autonomous body task management device according to claim 2, The task management device according to the present invention is characterized in that the task candidates are defined in advance in association with a completion event of the single task.

10. 1. An autonomous body task management method for managing tasks of a plurality of autonomous bodies in an automated system in which the autonomous bodies perform tasks, comprising: a state determination step of determining whether the autonomous entity is executing an independent task that is not affected by other autonomous entities; a task planning step of planning a task to be executed after the single task is completed when the single task is being executed; a task allocation step of allocating the task planned in the task planning step to the autonomous body; 1. A method for managing tasks of an autonomous body, comprising:

Images