KR20240029952A - Apparatus and method for planning tasks in which parts transportation is automated based on reinforcement learning through interworking with production management systems - Google Patents

Abstract

A planning device is provided. The planning device includes an acquisition unit that obtains order information; a generating unit that generates a first task necessary to achieve the order information; a distribution unit that distributes the second task, in which the process of the first task is initially subdivided, to the working machine; It may include a planning unit that plans a third task in which the process of the second task is divided into two parts for each work machine.

Description

Apparatus and method for planning tasks in which parts transportation is automated based on reinforcement learning through interworking with production management systems}

The present invention relates to an apparatus and method for scheduling automated parts transport.

A plan can be developed to realize logistics automation using autonomous robots.

As an example, a remote control system that generates a command, selects a robot, calculates a movement path, and generates a work command may be proposed.

However, there is no specific technology to create optimal movement routes, and as a result, distribution of logistics automation using self-driving robots is limited.

Korean Patent Publication No. 2019-0146945 discloses a self-driving robot technology that transports objects from a random point to another point while autonomously driving along a moving line upon receiving a work command.

Korea Patent Publication No. 2019-0146945

The present invention is a plan to establish a work plan to quickly supply various parts to the process line by creating work according to the parts order, distributing the created work to each work machine, and planning the detailed actions of the work machine. It is intended to provide a device and method.

The planning device of the present invention includes an acquisition unit that acquires order information; a generating unit that generates a first task necessary to achieve the order information; a distribution unit that distributes the second task, in which the process of the first task is initially subdivided, to the working machine; It may include a planning unit that plans a third task in which the process of the second task is divided into two parts for each work machine.

The distribution unit may perform reinforcement learning in a direction that minimizes the move action of the work machine and generates the second task to have the minimized move action.

The planning method of the present invention includes a creation step of identifying parts needed for each process line; It may include a distribution step of creating a transport job for the part and assigning or distributing it to a working machine.

When the number of pallets that the working machine can load at one time is defined as the upper limit number, and the pallet order corresponding to the order of loading or unloading the pallets is defined, the distribution step is performed by selecting the parts within the range that satisfies the upper limit number. The pallet order can be set in a way that minimizes the number of transports.

According to the planning device and planning method of the present invention, required parts in each process line can be identified through linkage with a production management system at the manufacturing site such as ERP/MES. Additionally, according to the present invention, transportation tasks, etc. can be created and assigned based on the information obtained in this way and reinforcement learning.

According to the present invention, a system can be provided in which workers can plan optimized operations in which workers can transport parts to each process line and retrieve empty pallets with minimal movement.

1 is a block diagram showing a planning device of the present invention.
Figure 2 is a table showing order information.
Figure 3 is a table showing remaining parts and pallet information.
Figure 4 is a table showing manufacturing information.
Figure 5 is a diagram showing the first task created in the generation unit.
Figure 6 is a flow chart showing the operation of the distribution unit.
Figure 7 is a flowchart showing the planning method of the present invention.
8 is a diagram illustrating a computing device according to an embodiment of the present invention.

Below, with reference to the attached drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily implement the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein. In order to clearly explain the present invention in the drawings, parts not related to the description are omitted, and similar parts are given similar reference numerals throughout the specification.

In this specification, duplicate descriptions of the same components are omitted.

Also, in this specification, when a component is mentioned as being 'connected' or 'connected' to another component, it may be directly connected or connected to the other component, but may be connected to the other component in the middle. It should be understood that may exist. On the other hand, in this specification, when it is mentioned that a component is 'directly connected' or 'directly connected' to another component, it should be understood that there are no other components in between.

Additionally, the terms used in this specification are merely used to describe specific embodiments and are not intended to limit the present invention.

Also, in this specification, singular expressions may include plural expressions, unless the context clearly dictates otherwise.

In addition, in this specification, terms such as 'include' or 'have' are only intended to designate the presence of features, numbers, steps, operations, components, parts, or combinations thereof described in the specification, and one or more It should be understood that this does not exclude in advance the presence or addition of other features, numbers, steps, operations, components, parts, or combinations thereof.

Also, in this specification, the term 'and/or' includes a combination of a plurality of listed items or any of the plurality of listed items. In this specification, 'A or B' may include 'A', 'B', or 'both A and B'.

Additionally, in this specification, detailed descriptions of well-known functions and configurations that may obscure the gist of the present invention will be omitted.

Figure 1 is a block diagram showing a planning device 100 of the present invention.

The planning device 100 shown in FIG. 1 may include an acquisition unit 110, a generation unit 130, a distribution unit 150, and a planning unit 170. Work machines, such as self-driving robots, can be provided to carry the parts necessary for manufacturing the products included in the order information to each process line. The planning device 100 of FIG. 1 can plan the work of the corresponding work machine. The work machine can transport parts as a unit of a pallet on which a plurality of parts are mounted. The work plan of the work machine may include not only the process of transferring pallets loaded with parts, but also the process of retrieving empty pallets placed in each process line.

The acquisition unit 110 may acquire order information i1. Order information i1 may include delivery date, number of products, type of product, product specifications, etc.

The generator 130 may generate the first task necessary to achieve order information i1.

As an example, the first task may be defined as only a part of the process line-centered task among the entire process required to produce a product, such as 'bring part number 1 to a process line.'

In order to assemble a product, a process line may require only one work machine to bring part number 1. This may correspond to the first task. However, the work machine cannot operate normally with only the first task. For example, this is because the work machine does not know where to load part 1.

In order to operate the work machine normally, a second task in which the first task is subdivided into primary tasks is required.

The distribution unit 150 may create a second job in which the process of the first job is divided into primary parts and distribute it to the work machine.

The second task may include additional work required by the worker to accomplish the first task. The second task may include a plurality of subdivided additional tasks, and one of the additional tasks may include the first task. Alternatively, the plurality of additional tasks may be pieces of the first task. In this case, when all of the plurality of additional tasks are performed, the first task can be naturally achieved.

For example, a second task can be arranged to go to the warehouse and load part 1, and when the loading of part 1 is completed, move to process line a and unload part 1 when it arrives at process line a. there is. When all of the above two second operations are performed, the first operation of placing part 1 on process line a can be completed.

As another example, the second task is to go to the warehouse and load two parts No. 1, and when the loading of part No. 1 is completed, move to the b process line, and when it arrives at the b process line, the second task is to unload one part No. 1. It may include a second operation of moving from process line b to process line a and unloading one part 1 when arriving at process line a. When all of the above three second operations are performed, the first operation of placing component No. 1 on process line a can be completed, although a process of placing part No. 1 on process line b is added in the middle.

The distribution unit 150 may perform reinforcement learning in a direction that minimizes the move action of the work machine and generates the second task to have the minimized move action. A move action may mean an action in which a work machine travels from its current location to another location.

As an example, a first task may be generated by the generation unit 130 to place part 1 in process line A and place part 2 in process line B.

In this case, the distribution unit 150 may create a second job and distribute it to the work machine as in the first embodiment below.

1) Loading part 1 in the warehouse (pallet unit)

2) Move to a process line

3) Unloading part 1 into process line a, recovering the empty first pallet present in process line a.

4) Return to warehouse

5) Unload the first empty pallet into the warehouse and load the second part into the warehouse.

6) Move to process line b

7) Unloading part 2 in process line b, recovering the empty second pallet present in process line b.

8) Return to warehouse

9) Unloading the empty second pallet into the warehouse

Among the above second tasks, 2) moving to process line a, 4) returning to the warehouse, 6) moving to process line b, and 8) returning back to the warehouse may correspond to move actions.

In the above example, there are a total of 4 move actions.

The distribution unit 150 determines whether the first pallet and the second pallet can be loaded on the work machine at the same time using the information on the first pallet on which part a of the work machine is mounted and the information on the second pallet on which part b is mounted. You can.

If it is determined that the first pallet and the second pallet can be loaded together on the work machine, the distribution unit 150 may create a second job as in the second embodiment below.

1) Load part 1 and part 2 in the warehouse (pallet unit)

2) Move to a process line

3) Unloading part 1 into process line a, recovering the empty first pallet present in process line a.

4) Move to process line b

5) Unloading part 2 into process line b, recovering the empty second pallet from process line b.

6) Return to warehouse

7) Unloading empty first pallet and empty second pallet into warehouse

Among the above second tasks, 2) moving to process line a, 4) moving to process line b, and 6) returning to the warehouse may correspond to move actions.

Among the above second tasks, 1) loading part 1 and part 2 in the warehouse can be expressed as one task as (Pick, warehouse, part 1, 1 piece, part 2, 1 piece). However, in reality, it may contain two operations, such as (Pick, warehouse, part 1) and (Pick, warehouse, part 2). This is because one task unconditionally targets one palette. At this time, a problem may arise as to which of (Pick, warehouse, part 1), which indicates the loading of part 1, or (Pick, warehouse, part 2), which indicates the loading of part 2, is performed first. The order may change depending on the loading and unloading method of the robot. For example, in the case of a method of unloading a pallet loaded later, it may be desirable for part 1, which is unloaded first, to be loaded after part 2 among parts 1 and 2.

In the above example, there are a total of three move actions. It can be seen that the number of move actions has been reduced by one compared to the previous example. Although one move action may only be removed, it is obvious that the difference will be noticeable if the first task is accumulated multiple times. Increasing the movement time of the work machine means that the overall working time increases. Additional work machines may be added to reduce the overall work time. In this case, schedule work needs to be added to prevent interference and collisions between multiple work machines.

In order to shorten the overall work time and minimize the input of additional work machines and the preparation of separate schedule work, it is better to minimize the movement time of the work machine as a result. To minimize the movement time of the work machine, the fewer the number of move actions, the better.

The planning unit 170 may create or plan a third task in which the process of the second task is subdivided for each work machine. The third task may include detailed control items of the work machine. For example, in order to perform the second task of unloading part 1 into a process line, the work machine needs to realistically perform a plurality of third tasks. For example, the third task at this time is moving the work tool to the unloading position of the preset process line a, rotating the a set angle based on the process line, unloading the first part, and placing the empty pallet in the process line a. This can include turning towards, loading empty pallets, etc.

In order to minimize the number of move actions, additional information in addition to order information i1 is required. To obtain the information, the acquisition unit 110 may communicate with an order server equipped with at least one of ERP (Enterprise Resource Planning) or MES (Manufacturing Execution System) that manages order information i1.

The acquisition unit 110 may acquire order information i1, remaining part quantity i2 in the current process line, and pallet information i3 requiring recovery from the order server.

The distribution unit 150 may create a second job using all of the order information i1, the remaining part amount i2, and the pallet information i3 and distribute it to the work machine.

The acquisition unit 110 may acquire order information i1, remaining part quantity i2 in the current process line, and pallet information i3 requiring recovery through various paths as well as the order server.

The distribution unit 150 can determine the upper limit number corresponding to the number of pallets that the work machine can load at one time through analysis of pallet information i3. In order to determine the upper limit number, the distribution unit 150 may obtain the loading capacity of the work machine in advance.

The distribution unit 150 may set a pallet order corresponding to the order of loading or unloading pallets through analysis of the remaining part quantity i2 and order information i1 of each process line. At this time, according to the palette order, the number of move actions of the work machine and the order of move actions can be determined as in the first embodiment above.

The distribution unit 150 can determine the number of move actions in which the work machine travels within the work area through analysis of the upper limit number and pallet order.

The distribution unit 150 may correct the palette order to minimize the number of move actions.

The distribution unit 150 may create a second job containing the corrected pallet order. As an example, the distribution unit 150 may create a second job as in the second embodiment described above.

The distribution unit 150 may integrate a plurality of move actions using the same driving path into a single common action among the identified plurality of move actions.

For example, in the first embodiment above, the two move actions '4) return to warehouse' and '8) return to warehouse' use the same driving path. In this case, the distribution unit 150 may integrate the two move actions into a single common action '6) Return to warehouse', as in the second embodiment.

In order to integrate multiple move actions into a single common action, the palette order scheduled to be performed in the step before or after the multiple move actions integrated into the common action needs to be changed.

At this time, the distribution unit 150 may correct the palette order so that a common action can be achieved.

A common action may be integrated into the last move action performed among a plurality of move actions using the same driving path. Accordingly, the palette order for the preceding move action can be adjusted by the distribution unit 150 so as not to use the same travel path. Reinforcement learning can be used to control and correct the palette order.

Planning device 100 may operate as follows.

The generation unit 130 may create a first task of transporting the required quantity of parts to the process line and recovering a pallet of used parts.

In order to create the first job, the creation unit 130 includes order information i1 entered through linkage with a production management system at the manufacturing site such as ERP/MES and an order server, and a parts remaining quantity including the number of remaining parts in the current factory process lines. You can receive palette information i2, i3, which includes the number of empty palettes.

Ordering information i1 may additionally include manufacturing information such as the type and number of parts needed to produce a product and the processing line that must be operated.

The generator 130 outputs several unit operations (first operations), and one unit operation may be expressed as (line A, part 3). (Line a, part 3) may refer to the operation of transferring one pallet loaded with part 3 to line a.

The distribution unit 150 may generate an action list of a work machine corresponding to the second task based on the first task.

The distribution unit 150 is configured based on a reinforcement learning algorithm and can receive a reward based on whether the action is successful or not and the maximum number of move actions at each stage (state).

The operation of the distribution unit 150 will be explained using tasks performed by a plurality of robots as an example.

For example, 1 pallet of Part 1 and 1 pallet of Part 3 are supplied to process A & 1 pallet of Part 1 and 1 pallet of Part 4 are supplied to process B, and 2 pallets of process A and 1 pallet of process B are recovered. A task can be created as follows:

[(A,1), (A,3), (B,1), (B,4), (A,P), (A,P), (B,P)]

(A,1) may mean supplying one pallet of Part 1 to process A.

(A,3) may mean supplying one pallet of Part 3 to process A.

(B,1) may mean supplying one pallet of Part 1 to process B.

(B,4) may mean supplying one pallet of part 4 to process B.

(A,P) may mean to retrieve one empty pallet from process A.

Since each job created by the generation unit targets only one pallet, two identical recovery jobs, such as (A,P) and (A,P), can be created to recover two pallets in process A. there is.

(B,P) may mean to retrieve one empty pallet from process B.

If the distribution unit 150 is excluded, the robot can perform one task and, upon completion, proceed to perform the next task.

Robot 1: (A,1), Robot 2: (A,3) -> Robot 1: (B,1), Robot 2: (B,4) -> Robot 1: (A,P), Robot 2: (A,P) -> Robot 1: (B,P)

In this case, the number of move actions is Robot 1 = 8, Robot 2 = 6 (total = 14).

When the distribution unit 150 is applied, a second task may be created on an action basis.

(Pick,G,1),(Place,A,1),(Pick,G,3),(Place,A,3),(Pick,G,1),(Place,B,1),(Pick ,G,4),(Place,B,4),(Pick,A,P),(Place,G,P),(Pick,A,P),(Place,G,P),(Pick,B ,P),(Place,G,P)

Here, A and B represent the process, G represents the warehouse, and P represents an empty pallet (recovery).

Robot 1: (Pick,G,1),(Pick,G,3),(Place,A,3),(Place,A,1),(Pick,A,P),(Pick,A,P) ,(Place,G,P),(Place,G,P)

To interpret, this means loading part 1 and part 3 in the warehouse, unloading them at A, recovering the pallet, and transporting it to the warehouse.

Robot 2: (Pick,G,1),(Pick,G,4),(Place,B,4),(Place,B,1),(Pick,B,P),(Place,G,P)

To interpret, this means loading part 1 and part 4 in the warehouse, unloading them at B, recovering the pallet, and transporting it to the warehouse.

At this time, the number of move actions is Robot 1 = 2 times, Robot 2 = 2 times (total = 4 times).

Looking at it, it can be seen that the number of move actions is significantly reduced compared to the comparative example in which tasks are processed one by one in order.

The action list (second task) generated by the distribution unit 150 needs to be subdivided into actions required to perform the task of the actual robot corresponding to the work machine.

In order for an actual robot to perform even the single action of picking a pallet loaded with parts, various detailed actions such as movement, rotation, and approach for picking are required (corresponding to a third task). This is needed. Such segmentation work may be performed by the planning unit 170. Detailed actions and third tasks generated by the planning unit 170 may be transferred to the work machine.

Figure 2 is a table showing order information. Figure 3 is a table showing the remaining parts quantity and pallet information. Figure 4 is a table showing manufacturing information.

Order information 'order' may include product type and quantity.

Part remaining quantity and pallet information 'state' may include the type and remaining amount of parts used in each process line, and the number of pallets that need to be recovered.

Product production-related information corresponding to manufacturing information may include information on the type and number of process lines and parts required to produce each product. Manufacturing information may include information on the quantity of each part that can be mounted per pallet.

Figure 5 is a diagram showing the first task generated by the generation unit 130.

The generation unit 130 can calculate the total quantity of parts required for the order using the order quantity for each product in the 'Order' information and the number of parts required for production for each product in the 'product production-related information'.

The generator 130 can calculate the number of parts required for each line using 'state' information.

The generation unit 130 can calculate the parts that need to be transported on a pallet basis using the calculated number of parts required for each line and the number of parts per pallet information in product production-related information.

The creation unit 130 may create the first task in the form of (position, part) and store it in the form of a list.

The creation unit 130 may create and store the first task related to recovering empty pallets in the form of (position, pallet) per pallet.

As an example, the process of creating a job using order information i1, remaining part quantity i2 in the production line, recovery pallet information i3, and product production-related information (manufacturing information) may be as follows.

i1: Produce 30 units of product 1

The generation unit 130 can calculate and set the parts required for production using product production-related information. For example, the generation unit 130 can be set to 450 parts 1, 30 parts 2, and 60 parts 3 in process A and 300 parts 1 and 60 parts 4 in process B.

The generator 130 can calculate and set the parts that need to be transported using the remaining part quantity i2. For example, the generation unit 130 can set 238 parts 1 and 35 parts 3 in process A, and 162 parts 1 and 38 parts 4 in process B.

The generation unit 130 can convert the number of parts into pallet units using product production-related information. For example, the creation unit 130 can set one palette of Part 1 and one palette of Part 3 in process A, and one palette of Part 1 and one palette of Part 4 in process B. The setting value at this time can be expressed as ‘(A,1), (A,3), (B,1), (B,4)’.

The generator 130 may add a recovery task using recovery pallet information i3. For example, the generator 130 can be set to retrieve two A process pallets and one B process pallet.

Combine the previous set value and number of times to get '(A,1), (A,3), (B,1), (B,4), (A,P), (A,P), (B,P )'. 'P' may mean pallet recovery. Figure 6 is a flowchart showing the operation of the distribution unit 150.

FIG. 6 may show the structure of a reinforcement learning algorithm of the distribution unit 150.

The distribution unit 150 may receive the allocation result of the first task to the robot from the generation unit 130 (401). The assignment result may include a list in the form of (Position, Object), such as [(A,1), (A,3), (B,1), (D,2), (D,P)]. You can.

The distribution unit 150 may generate a pick action (loading action) and a place action (unloading action) corresponding to the second task from the allocation result (402). Pick action may mean picking up and loading a pallet. Place action may mean putting down a loaded pallet.

The generated actions are [(Pick, G, 1), (Place, A, 1), (Pick, G, 3), (Place, A, 3), (Pick, G, 1), (Place A list in the form of (action, position, object) such as , B, 1), (Pick, G, 2), (Place, D, 2), (Pick, D, P), (Place, G, P)] may include.

The distribution unit 150 repeats learning as many times as the specified number of epochs, and when the specified number of epochs is reached (403), it can output an action list with the maximum reward (404).

If the specified number of epochs is not reached, the distribution unit 150 may check whether the current reward value is less than a specified random value t (405).

If the current reward value is smaller, the distribution unit 150 initializes the currently learning action list and reward (406) and performs a plurality of pick actions (Pick actions).

The initial action of each robot can be set randomly (407). An example setting is as follows. R1, R2, R3, and R4 may each represent a second task.

R1: [(Pick, G, 1)], R2: [(Pick, G, 3)], R3: [(Pick, G, 2)], R4: [(Pick, D, P)]

Remaining actions: [(Place, A, 1), (Place, A, 3), (Pick, G, 1), (Place, B, 1), (Place, D, 2), (Place, G, P )]

The distribution unit 150 may randomly select the next action of the robot from the remaining actions (408). An example is as follows.

Case 1

R1: [(Pick, G, 1), (Place, A, 1)], R2: [(Pick, G, 3), (Pick, G, 1)], R3: [(Pick, G, 2) , (Place, D, 2)], R4: [(Pick, D, P), (Place, G, P)]

Remaining actions: [(Place, A, 3), (Place, B, 1)]

Case 2

R1: [(Pick, G, 1), (Place, A, 3)], R2: [(Pick, G, 3), (Pick, G, 1)], R3: [(Pick, G, 2) , (Place, D, 2)], R4: [(Pick, D, P), (Place, B, 1)]

Remaining Action: [(Place, A, 1), (Place, G, P)]

The distribution unit 150 may update the currently learning reward and action list from the selected action (409).

The distribution unit 150 sets the maximum value of the number of move actions of each robot generated in the case of the order of executable actions as the reward, and at this time, the reward is a negative value. Can be updated with value. In case of failure, the specified failure penalty value can be updated as a reward.

Case 1 >> Success

R1: [(Pick, G, 1), (Move, G, A), (Place, A, 1)]

R2: [(Pick, G, 3), (Pick, G, 1)]

R3: [(Pick, G, 2), (Move, G, D), (Place, D, 2)]

R4: [(Pick, D, P),(Move, D, G), (Place, G, P)]

Reward: Number of move actions: [1, 0, 1, 1] >> Maximum 1 >> -1

Case 2 >> Failure

R1: [(Pick, G, 1), (Place, A, 3)] >> Parts not picked cannot be placed

Reward: Failure penalty value

Examples of failure include placing, unloading, and unloading parts that have not been picked, loaded, or loaded, or not placed in the reverse order of the picked order, or the robot fails once. It is possible to pick more pallets than can be loaded.

If there are remaining actions (410), the distribution unit 150 can go back and re-execute from the reward < t conditional statement (405).

If there are no remaining actions, the distribution unit 150 determines whether the current reward is greater than the stored maximum reward value (411), and if it is greater, the reward and action list can be updated ( 412).

As discussed above, the distribution unit 150 can perform reinforcement learning by using the maximum value of the number of move actions rather than the movement distance of each work machine as a reward.

Figure 7 is a flowchart showing the planning method of the present invention.

The planning method of FIG. 7 may be performed by the planning device 100 shown in FIG. 1 .

The planning method may include an acquisition step (S 510), a creation step (S 520), a distribution step (S 530), and a planning step (S 540).

In the acquisition step (S510), order information, remaining part quantity, pallet information, and manufacturing information can be obtained. The acquisition step (S510) may be performed by the acquisition unit 110.

The creation step (S 512) can identify the parts needed for each process line. The generation step (S520) may be performed by the generation unit 130. The creation step (S512) may create a first task of the work machine related to transport of the part.

In the distribution step (S 530), the first task may be subdivided to create a second task corresponding to the transport task of the parts and assigned or distributed to the work machine. The distribution step (S530) may be performed by the distribution unit 150.

When the number of pallets that a worker can load at once is defined as the upper limit, and the pallet order corresponding to the order of loading or unloading the pallets is defined, the distribution step (S 530) is to distribute parts within the range that satisfies the upper limit. The pallet order can be set in a way that minimizes the number of transports. In the distribution step (S530), the pallet order can be set in a way that minimizes the number of parts transported by using reinforcement learning.

In the planning step (S 540), a third task in which the process of the second task is subdivided for each work machine can be planned and assigned to each work machine.

8 is a diagram illustrating a computing device according to an embodiment of the present invention. Computing device TN100 of FIG. 8 may be a device described herein (e.g., planning device 100, etc.).

In the embodiment of FIG. 8, the computing device TN100 may include at least one processor TN110, a transceiver device TN120, and a memory TN130. Additionally, the computing device TN100 may further include a storage device TN140, an input interface device TN150, an output interface device TN160, etc. Components included in the computing device TN100 may be connected by a bus TN170 and communicate with each other.

The processor TN110 may execute a program command stored in at least one of the memory TN130 and the storage device TN140. The processor TN110 may refer to a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor on which methods according to embodiments of the present invention are performed. Processor TN110 may be configured to implement procedures, functions, and methods described in connection with embodiments of the present invention. The processor TN110 may control each component of the computing device TN100.

Each of the memory TN130 and the storage device TN140 can store various information related to the operation of the processor TN110. Each of the memory TN130 and the storage device TN140 may be comprised of at least one of a volatile storage medium and a non-volatile storage medium. For example, the memory TN130 may be comprised of at least one of read only memory (ROM) and random access memory (RAM).

The transceiving device TN120 can transmit or receive wired signals or wireless signals. The transmitting and receiving device (TN120) can be connected to a network and perform communication.

Meanwhile, the embodiments of the present invention are not only implemented through the apparatus and/or method described so far, but may also be implemented through a program that realizes the function corresponding to the configuration of the embodiment of the present invention or a recording medium on which the program is recorded. This implementation can be easily implemented by anyone skilled in the art from the description of the above-described embodiments.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements made by those skilled in the art using the basic concept of the present invention defined in the following claims are also possible. It falls within the scope of invention rights.

100...Planning Device 110...Acquisition Department
130...generation part 150...distribution part
170...Planning Department

Claims (5)

An acquisition unit that obtains order information;
a generating unit that generates a first task necessary to achieve the order information;
a distribution unit that distributes the second task, in which the process of the first task is initially subdivided, to the working machine;
It includes a planning unit that plans a third operation in which the process of the second operation is divided into two parts for each work machine,
The distribution unit performs reinforcement learning in a direction to minimize the move action of the work machine, and generates the second task to have the minimized move action.
According to paragraph 1,
The acquisition unit communicates with an order server equipped with at least one of ERP (Enterprise Resource Planning) or MES (Manufacturing Execution System) that manages the order information,
The acquisition unit acquires the order information, the remaining amount of parts in the current process line, and pallet information requiring recovery from the order server,
The distribution unit is a planning device that creates the second job using all of the order information, the remaining part quantity, and the pallet information and distributes it to the work machine.
According to paragraph 1,
The acquisition unit acquires the order information, the remaining amount of parts in the current process line, and the pallet information requiring recovery,
The distribution unit analyzes the pallet information to determine the upper limit of the number of pallets that the work machine can load at one time,
The distribution unit sets a pallet order corresponding to the order of loading or unloading the pallet through analysis of the remaining part quantity of each process line and the order information,
The distribution unit determines the number of move actions in which the work machine travels within the work area through analysis of the upper limit number and the pallet order,
The distribution unit corrects the palette order to minimize the number of move actions,
A planning device wherein the distribution unit generates the second task containing a corrected pallet sequence.
According to paragraph 3,
The distribution unit integrates a plurality of move actions using the same driving path among the identified plurality of move actions into a single common action,
The distribution unit corrects the palette order so that the common action is achievable.
In a planning method performed by a planning device,
A creation stage that identifies the parts needed for each process line;
A distribution step of creating a transport job for the part and assigning or distributing it to a work machine,
When the number of pallets that the working machine can load at once is defined as the upper limit number, and the pallet order corresponding to the order of loading or unloading the pallets is defined,
The distribution step is a planning method of setting the pallet order in a direction that minimizes the number of transportation of the parts within a range that satisfies the upper limit number.

Images