Lacquering robot system

Abstract

To provide a lacquering robot system which self-automatically performs lacquered production by a robot arm.SOLUTION: The lacquering robot system includes: a plurality of robot work stations (RBT_WS) for controlling a plurality of lacquering robots; and a lacquer transaction computer 3 which allocates a lacquering task accepted from a customer terminal to one lacquering robot in response to the lacquering task and causes this lacquering robot to perform the lacquering task. The lacquering robots are arranged in a plurality of locations, and each robot work station RBT_WS generates a lacquering task database having stroke feature data which includes locus data of strokes for an image to be lacquered by the lacquering task and material data of materials including paints required for painting loci of the strokes or tools including brushes, and orders a purchase request for a material or a tool in the material data to a material market via a material transaction network on the basis of the lacquering task database.SELECTED DRAWING: Figure 18

Description

[Technical Field]

[0001] The present invention relates to a robot system for lacquering. [Background technology]

[0002] Lacquerware is a traditional craft in which lacquer is applied with a brush to the surface of various materials such as wood and acrylic. In recent years, the aging of lacquerware artisans and the decline in the number of successors have led to a decline in the production of lacquerware, and the decline of the traditional lacquerware industry has become a problem.

[0003] Non-Patent Document 1 reports an attempt to use a robotic arm to apply lacquer with a brush. According to this document, a lacquer brush is attached to the tip of a six-axis general-purpose robotic arm, the trajectory of the arm is programmed, and the robotic arm applies lacquer to the surface of a vessel or other object to decorate it. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] "Urushi Brush Painting Using a Robot Arm" 2018 Research Presentation Meeting of the 4th Branch of the Japanese Society for the Science of Design Summary of the Invention [Problem to be solved by the invention]

[0005] However, simply using a robotic arm to brush lacquer is not enough to revitalize the lacquering industry.

[0006] Therefore, an object of this embodiment is to provide a lacquering robot system that performs lacquering production using a robot arm automatically, or preferably autonomously. [Means for solving the problem]

[0007] The first aspect of this embodiment is a robot control and material transaction computer that controls each of the plurality of lacquering robots; a lacquer trading computer that, in response to a lacquering task received from a customer terminal, assigns the lacquering task to one of the lacquering robots and causes the lacquering robot to execute the lacquering task, and, after the lacquering task is completed by the lacquering robot, executes a sales process between the customer terminal and the lacquer item for which the lacquering task has been performed via a lacquer trading network; the plurality of lacquering robots are disposed at a plurality of locations; 5. Robot control and material transaction computer a lacquering image of the lacquering task is decomposed into a plurality of strokes, and a lacquering task database is generated having stroke feature data corresponding to each of the plurality of strokes, the stroke feature data having stroke trajectory data and material data having a material including paint or a tool including a brush required to paint the stroke trajectory; This is a lacquering robot system that, based on the lacquering task database, places an order to a material market via a material trading network for a purchase request for the materials or tools of the material data corresponding to the received lacquering task. [Effects of the Invention]

[0008] According to the first aspect, the lacquering robot system can automatically carry out lacquering production using a robot arm. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a diagram showing a schematic configuration example of a lacquering robot system according to an embodiment of the present invention; [Figure 2] FIG. 1 is a diagram showing an example of the configuration of a first lacquering robot system according to the present embodiment. [Figure 3] FIG. 10 is a diagram showing a configuration example of a second lacquering robot system in the present embodiment. [Figure 4]FIG. 10 is a flowchart of a first example of a request acceptance process and a task generation process by the lacquer trading computer 3 in this embodiment. [Figure 5] FIG. 10 is a flowchart of a second example of the request acceptance process and task generation process by the lacquer trading computer 3 in this embodiment. [Figure 6] FIG. 10 is a flowchart showing a first example of a task allocation process to a lacquering robot RBT by a lacquer trading computer 3. [Figure 7] FIG. 10 is a flowchart showing a second example of the task allocation process to the lacquering robot RBT by the lacquer trading computer 3. [Figure 8] FIG. 8 is a flowchart showing details of the estimation processes S44 and S45 of the first and second total times TT_A1 and TT_A2 in FIG. 7. [Figure 9] This is a flowchart showing an example of the lacquering process of the robot workstation (robot control / material transaction computer) RBT_WS. [Figure 10] FIG. 10 is a flowchart showing a first example of the lacquer trading process executed by the lacquer trading computer 3 after the lacquer item is completed. [Figure 11] FIG. 10 is a flowchart showing a first example of a material trading process of the material trading computer MAT_C. [Figure 12] FIG. 10 is a flowchart showing a second example of the material trading process of the material trading computer MAT_C. [Figure 13] 1 is a diagram showing an image of a lacquering task design, a lacquering task database, and an outline of the lacquering process. [Figure 14] 1 is a diagram showing an image of a lacquering task design, a lacquering task database, and an outline of the lacquering process. [Figure 15] FIG. 10 is a diagram showing a process for generating a lacquering task database. [Figure 16] FIG. 10 is a diagram showing a process for generating a lacquering task database. [Figure 17] FIG. 10 is a diagram showing an example of one stroke and example of feature data associated with the stroke. [Figure 18] FIG. 2 is a diagram illustrating an example of the hardware configuration of a robot workstation RBT_WS. DETAILED DESCRIPTION OF THE INVENTION

[0010] FIG. 1 is a diagram showing an example of the schematic configuration of a lacquering robot system in this embodiment. In the lacquering robot system, a lacquer trading computer 3 and multiple customer terminals CST_1 to CST_N are able to communicate with each other via a lacquer trading blockchain (or lacquer trading network) 1. Furthermore, multiple lacquering robot hands RBT_1 to RBT_N and the material market MAT_M are able to communicate with each other via a material trading blockchain (or material trading network) 2. The lacquer trading computer 3 is a lacquer trading center that controls the buying and selling of lacquer products (lacquer items) with customers. Hereinafter, the robot hand will be simply referred to as a robot.

[0011] The basic workflow of the lacquering robot system is as follows: The customer terminal CST accesses the lacquer trading computer 3 via the lacquer trading network 1, browses lacquer items, which are lacquered goods such as lacquerware, and sends a purchase request for the selected lacquer item to the lacquer trading computer 3. The lacquer trading computer 3 receives the purchase request and assigns the lacquering task of producing the desired lacquer item to one of multiple lacquering robots RBT_1 to RBT_N deployed in different locations. The lacquering robot RBT assigned the task produces the lacquer item (lacquerware) requested for purchase and notifies the lacquer trading computer 3 of the completion of production upon completion. The lacquer trading computer 3 registers the completed lacquer item in the lacquer trading blockchain 1, and the customer terminal CST completes the purchase procedure for the lacquer item. Finally, the lacquering robot RBT assigned the lacquering task delivers the produced lacquer item to the customer.

[0012] A robot workstation (not shown) that controls the lacquering robot RBT procures lacquering materials, lacquering tools, etc. as needed from the material market MAT_M via the material trading blockchain 2. Optionally, each lacquering robot RBT is associated with a warehouse (not shown) that stores lacquering materials, tools, etc., and the procured materials and tools are stored in the warehouse.

[0013] Figure 2 is a diagram showing an example of the configuration of a first lacquering robot system in this embodiment. In the first lacquering robot system, each lacquering robot is assigned a warehouse WHS that stores lacquering materials (materials and / or consumables such as tools (hereinafter referred to as materials or tools for short)). As in Figure 1, the lacquer trading blockchain 1 is configured so that the lacquer trading computer 3 and customer terminal CST can communicate with each other. Meanwhile, the material trading blockchain 2 is configured so that the material market MAT_M, multiple warehouses WHS for material inventory management, and multiple robot workstations RBT_WS can communicate with each other.

[0014] Multiple robot workstations RBT_WS are arranged in different areas and locations. Each robot workstation RBT_WS has a lacquering robot RBT, a robot control computer RBT_C that controls the lacquering robot RBT, and a material transaction computer MAT_C that executes transactions of lacquering materials, tools, etc. Each warehouse WHS is associated with the lacquering robot RBT of each robot workstation RBT_WS.

[0015] The lacquer trading computer performs lacquering trend analysis and generates production tasks (lacquering tasks) for popular lacquer items. The customer terminal CST accesses the lacquer trading computer 3 and sends a purchase request for a lacquer item. The lacquer trading computer assigns the lacquering task for the lacquer item in the purchase request to a lacquering robot RBT at one of the robot workstations. The lacquering robot RBT is controlled by the robot control computer RBT_C to perform the lacquering task for the lacquer item.

[0016] When the robot workstation RBT_WS completes the assigned lacquering task, it sends a completion notification to the lacquer trading computer 3. The lacquer trading computer 3 registers the completed lacquer item in the lacquer trading blockchain 1. The customer terminal CST that made the purchase request carries out the purchase procedure for the completed lacquer item via the lacquer trading blockchain 1. Once the purchase procedure is complete, the robot workstation RBT_WS delivers the completed lacquer item to the customer. In addition, the profit from the sale of the lacquer item is returned to the lacquering robot RBT via the lacquer trading blockchain 1.

[0017] In the process of performing the lacquering task, the material transaction computer MAT_C in each robot workstation RBT_WS checks the inventory of materials or tools in the warehouse and purchases the necessary materials or tools from the material market MAT_M via the material transaction blockchain 2. The material transaction computer MAT_C may also purchase the necessary materials or tools in cooperation with the warehouse. In this case, the material transaction computer MAT_C requests the necessary materials or tools from the warehouse. In response to this, the corresponding warehouse WHS sends a material purchase request to the material market MAT_M via the material transaction blockchain 2. The purchased materials or tools are shipped from the material market MAT_M to the warehouse and stored in the warehouse. The robot control computer RBT_C obtains the necessary materials or tools from the warehouse WHS and controls the lacquering robot RBT to perform the lacquering task.

[0018] Figure 3 is a diagram showing an example of the configuration of a second lacquering robot system in this embodiment. Unlike the first lacquering robot system in Figure 2, the second lacquering robot system does not have a warehouse assigned to the lacquering robot. Otherwise, as in Figure 2, the lacquer trading blockchain 1 is configured so that the lacquer trading computer 3 and the customer terminal CST can communicate with each other. Meanwhile, the material trading blockchain 2 is configured so that the material market MAT_M and multiple robot workstations RBT_WS can communicate with each other.

[0019] The multiple robot workstations RBT_WS are the same as in Figure 2. The process from requesting a lacquer item to completing the transaction is also the same as in Figure 2, except for the process related to the warehouse.

[0020] Unlike Figure 2, in the process of executing the lacquering task in Figure 3, the material trading computer MAT_C in each robot workstation RBT_WS checks the inventory of materials or tools and purchases the necessary materials or tools from the material market MAT_M via the material trading blockchain 2. The purchased materials or tools are shipped from the material market to the lacquering robot RBT.

[0021] The lacquer trading computer 3, the robot control computer RBT_C, and the material trading computer MAT_C shown in Figures 2 and 3 may be integrated into a virtual server of a server located in a server center. Alternatively, the above three computers may be distributed and configured as virtual servers of servers in server centers at different locations.

[0022] [Processing by lacquer trading computer 3] The request reception process, task generation process, task allocation process, etc., performed by the lacquer trading computer 3 will be described below.

[0023] [Request acceptance processing, task creation processing] 4 is a flowchart of a first example of the request reception process and task generation process by the lacquer trading computer 3 in this embodiment. In the figure, the process of the lacquer trading computer 3 is shown on the left, and the process of the customer terminal CST is shown on the right. The leftmost column shows the task generation process S10 to S13, and the rest show the request reception process S15 to S27.

[0024] In the task generation process, the lacquer trading computer 3 collects data on lacquering to analyze popular themes of lacquer design (S10), analyzes the data (S11), sets a lacquering theme (S12), and designs a lacquering task to be performed by a lacquering robot and its specifications for the set lacquering theme (S13).

[0025] In the request reception process, first, the customer terminal CST sends a purchase request for the desired lacquer item to the lacquer trading computer 3 (S14). When the lacquer trading computer 3 receives the purchase request (S15), it searches a library that covers all the themes and topics necessary for lacquer production to check whether the requested theme is available for order (S16). If the order cannot be accepted, the lacquer trading computer 3 replies to the customer terminal CST that the request is canceled (S17). On the other hand, if the order can be accepted, it replies to the customer terminal CST that the request is valid (S18).

[0026] Furthermore, the lacquer trading computer 3 interprets the contents of the request (S19), estimates a price (S20), and transmits the estimate to the customer terminal CST (S21). If the request is valid (YES in S22), the customer terminal CST receives the estimate (S23), and if it does not accept the estimate (NO in S24), it returns a cancellation to the lacquer trading computer 3 (S25). If it accepts the estimate (YES in S24), the customer terminal CST returns acceptance of the estimate (S26).

[0027] When the lacquer trading computer 3 receives an estimate acceptance from the customer terminal (YES in S27), it proceeds to task allocation, in which the lacquering task corresponding to the request is assigned to one of the lacquering robots (S28). The task assignment process will be explained in Figures 6, 7, and 8.

[0028] Figure 5 is a flowchart of a second example of the request reception process and task generation process by the lacquer trading computer 3 in this embodiment. In the second example, the customer terminal CST searches the library in the lacquer trading computer 3 and checks for a purchase request for the desired lacquer lacquer (S30). If the desired lacquer lacquer theme is in the library (YES in S31), the customer terminal CST sends a purchase request to the lacquer trading computer 3 (S14). The rest of the process is the same as the first example in Figure 4, so a description thereof will be omitted. Figures 4 and 5 have the same reference number S** for the same process.

[0029] [Laquering task allocation processing] FIG. 6 is a flowchart showing a first example of the task assignment process to the lacquering robots RBT by the lacquer trading computer 3. In this first example, the lacquer trading computer 3 checks whether each lacquering robot of the multiple lacquering robots RBT is currently available for assignment to the lacquering task (S40, S41). Whether a task is available for assignment means whether the lacquering robot is currently performing the production of a lacquering task. If the lacquering robot is currently performing the production of a lacquering task, the task is unavailable; if the lacquering robot is not currently performing the production of a lacquering task, the task is available for assignment. This check is performed, for example, by querying the robot control computer RBT_C of the multiple robot workstations RBT_WS. The lacquer trading computer 3 then selects the lacquering robot closest to the location of the customer who sent the purchase request from the list of available lacquering robots and assigns the task to that robot (S42). If the multiple lacquering robots RBT have approximately the same functions and performance, the time it takes each lacquering robot to complete the lacquering task will be approximately the same. Therefore, the lacquering robot closest to the location of the customer who issued the purchase request is predicted to have the shortest delivery time for the completed lacquer item. The robot workstation then proceeds to the lacquer production process (S43).

[0030] 7 is a flowchart showing a second example of the task allocation process to the lacquering robots RBT by the lacquer trading computer 3. In the second example, the lacquer trading computer 3 checks whether each of the multiple lacquering robots RBT is currently able to be assigned a task (S40, S41). If task allocation is possible, a first total time TT_A1 is estimated for that lacquering robot, which is the sum of the time it takes for the lacquering robot to complete lacquering and the time it takes to deliver the lacquer item to the customer (S44).

[0031] On the other hand, if task assignment is not possible, a second total time TT_A2 is estimated for that lacquering robot, which is the sum of the time until the lacquering robot becomes available for task assignment, the time until the lacquering robot completes lacquering, and the time until the lacquering item is delivered to the customer (S45). A lacquering robot that cannot be assigned a task is currently lacquering another task, and cannot be assigned a lacquering task. Therefore, the time until the currently executing task or a waiting task is completed is the time until the lacquering robot becomes available for task assignment. Then, based on the first total time TT_A1 or the second total time TT_A2 of all lacquering robots RBT, the lacquering robot with the shortest total time is selected and assigned the task (S46).

[0032] Figure 8 is a flowchart showing details of the estimation processes S44 and S45 for the first and second total times TT_A1 and TT_A2 in Figure 7. In the estimation process S44, the lacquer trading computer 3 estimates the time TT_1 required for the lacquering robot to complete lacquering and the time TT_2 required for the delivery of the lacquer item to the customer for a lacquering robot that can be assigned a task, and adds these together to calculate the first total time TT_A1. Furthermore, in the estimation process S45, for a lacquering robot that cannot be assigned a task, the second total time TT_A2 is calculated by adding the time TT_3 required for the lacquering robot to become available for task assignment to the first total time TT_A1.

[0033] It is also possible to estimate the time required to procure the materials or tools necessary for executing the lacquering task, and include this estimated time in the total time TT_A1 or TT_A2.

[0034] [Configuration and processing overview of the robot workstation RBT_WS] [Robot Workstation RBT_WS Configuration] 18 is a diagram showing an example of the hardware configuration of the robot workstation RBT_WS. The diagram shows the robot workstation RBT_WS, the customer terminal CST that can communicate via a network 37, the lacquer trading computer 3, the material market MAT_M, and the warehouse WHS.

[0035] As shown in Figure 2, the robot workstation RBT_WS has a robot control computer RBT_C that controls the operation of the lacquering robot, and a material transaction computer MAT_C that procures materials or tools necessary for lacquering. In this specification, the robot workstation RBT_WS is also referred to as the robot control / material transaction computer. This robot control / material transaction computer is connected to the lacquering robot RBT via a bus 29.

[0036] The robot workstation RBT_WS has a processor 11, a main memory 12, an interface 13, and a bus 29 connecting them. The robot workstation RBT_WS stores in its storage a robot control program and a material transaction program 20, an image-to-stroke model program 22, a stroke-to-feature model program 24, image-to-stroke training data and stroke-to-feature training data 26, and a lacquering task database 27.

[0037] First, the processor 11 executes a robot control program to control the lacquering operation of the lacquering robot RBT. Second, the processor executes the robot control program to perform training (parameter adjustment) of the image-to-stroke model 22 and the stroke-to-feature model 24 using training data 26. Third, the processor executes the robot control program to generate a lacquering task database 27 from the lacquering tasks using the trained image-to-stroke model 22 and the stroke-to-feature model 24. This database 27 is described in detail in FIG. 17.

[0038] [Lacquer coating process using a robotic workstation] Figure 9 is a flowchart showing an example of the lacquering process of the robot workstation (robot control / material transaction computer) RBT_WS. The robot control computer RBT_C receives a new lacquering task (S50) and checks whether the feature data of the image of the lacquering task has been registered in the lacquering task database (S51). If it has been registered (Yes in S51), it accesses the feature data of the lacquering task in the lacquering task database (S52_1). If it has not been registered (No in S51), the robot control computer uses the trained models 22, 24 to generate feature data of the strokes of the image of the lacquering task (S52_2).

[0039] The generation of the above-mentioned characteristic data of the lacquering task may be executed by the lacquer trading computer 3 when the lacquering task is generated.

[0040] Next, the robot control computer RBT_C checks whether there is a shortage of materials or tools required to execute the task (S53), and if there is a shortage (S54_Yes), requests the material trading computer MAT_C to procure the missing materials or tools. As will be described in Figures 11 and 12, the checks (S53, S54) whether there is a shortage of materials or tools required to execute the task may be performed by the material trading computer MAT_C.

[0041] In response to the material or tool procurement request, the material transaction computer MAT_C initiates a blockchain transaction for the material transaction via the material transaction blockchain 2 (S55) and purchases the missing materials or tools (S56). In the material transaction for procuring the missing materials or tools, the materials or tools to be procured are extracted based on the lacquering database of the lacquering task, and the extracted materials or tools are purchased. The materials or tools to be procured are calculated based on the lacquering database of the lacquering task, taking into account the inventory status of the warehouse.

[0042] Next, the robot control computer RBT_C controls the lacquering robot RBT to perform the lacquering work based on the database of lacquering tasks (S57_C, S57). In this lacquering work, the robot control computer causes the lacquering robot to use materials or tools procured based on the characteristic data of the multiple strokes of the lacquering task, and perform the lacquering work of each stroke. When the lacquering work is completed, the robot control computer RBT_C notifies the lacquer trading computer 3 of the completion.

[0043] Finally, the lacquer transaction computer 3 initiates blockchain processing for the lacquerware transaction on the lacquer transaction blockchain 1 (S58), resulting in the transaction of the lacquer item being executed (S59).

[0044] As described above, material transactions for procuring materials or tools and the production of lacquerware are carried out based on the database of lacquerware tasks shown in Figure 17.

[0045] In the above explanation, after receiving a new lacquering task, the robot control computer RBT_C checks whether there is a shortage of materials or tools required to perform the task (S53). If there is a shortage, the material trading computer MAT_C purchases the missing materials or tools (S56). However, the material trading computer MAT_C may speculatively purchase materials or tools required for the predicted lacquering task from the material market MAT_M based on a demand forecast for the lacquering task. Such speculative purchases can improve the accuracy of the total time, even if the total times TT_A1 and TT_A2 calculated in Figures 7 and 8 do not include the time for purchasing materials.

[0046] [Lacquer trading process after lacquer item completion executed by lacquer trading computer 3] 10 is a flowchart showing a first example of the lacquer trading process after a lacquer item is completed, which is executed by the lacquer trading computer 3. In the figure, the left column shows the processing of the lacquer trading computer 3, and the right column shows the processing of the customer terminal CST.

[0047] First, the lacquer trading computer 3 receives a notification of the completion of lacquerware production from the robot workstation RBT_WS (S60). In response to the completion notification, the lacquer trading computer 3 generates various data to register information about the lacquer item in the lacquer trading blockchain 1. For example, this includes the lacquer item's unique identification number ID and a digital certificate to verify the lacquer item's origin and quality. The identification number ID and digital certificate, along with important information such as price, materials used, and production date, are registered in the lacquer trading blockchain 1 (S61). This allows the customer terminal CST to obtain all information about the lacquer item and a complete and transparent record of the purchase. Once the lacquer item information is registered in the blockchain, the lacquer trading computer 3 notifies the customer terminal CST that the lacquer item is now available for purchase (S62).

[0048] In response to the notification that the lacquer item is available for purchase, the customer terminal CST starts a transaction on the lacquer transaction blockchain (S68) and transfers the cryptocurrency to the specified address (S69). As a result, the purchase process is completed (S70), and the customer terminal CST sends a purchase procedure completion notification to the lacquer transaction computer 3 (S71).

[0049] When the lacquer transaction computer 3 receives the notification that the purchase procedure has been completed, it confirms the completion of the purchase process (S64), coordinates the shipping of the lacquer item (S64), generates appropriate shipping and tracking information (S66), and sends it to the customer terminal CST (S67). At this time, a shipping instruction for the lacquer item is sent to the robot workstation RBT_WS, and the lacquer item manufactured by the lacquering robot is shipped to the customer.

[0050] The customer terminal CST receives the shipping and tracking information (S72), and the lacquer items are delivered to the customer (S73), and the receipt of the lacquer items is confirmed on the lacquer transaction blockchain (S74).

[0051] [Material transaction processing by the material transaction computer MAT_C in the robot workstation RBT_WS] Figure 11 is a flowchart showing a first example of material transaction processing by the material transaction computer MAT_C. The first example is material transaction processing when using a warehouse associated with a lacquering robot. The first example corresponds to the lacquering robot system shown in Figure 2. In the figure, the left column shows the processing of the material transaction computer MAT_C, and the right column shows the processing of the warehouse WHS terminal.

[0052] When the material transaction computer MAT_C receives a lacquering task, it checks whether the materials (materials or tools) required for the lacquering task are in sufficient stock in the warehouse (S80). This inventory check may be performed periodically, separate from when the task is received. For example, based on a demand forecast for lacquering tasks, periodic inventory checks may be performed speculatively in preparation for lacquering tasks that are likely to occur in the future. If the material inventory is insufficient, or if the material inventory is predicted to be insufficient (NO in S81), the material transaction computer MAT_C sends a material request to the warehouse WHS (S82) and waits for the confirmation result (S83).

[0053] The warehouse terminal receives the material request (S87), and if there is insufficient material in stock in the warehouse (NO in S88), it replies with a material shortage notification (S89). At the same time, the warehouse terminal starts the material purchase process on the material transaction blockchain 2 (S90, S91). When the warehouse terminal receives the material (S92), it updates the inventory data (S93), sends a material preparation completion notification to the material transaction computer MAT_C (S94), and simultaneously dispatches the material to the material transaction computer MAT_C in the robot workstation RBT_WS (S95), and updates the inventory information (S96).

[0054] Meanwhile, when the materials transaction computer MAT_C receives the materials (S84), it sends a receipt confirmation to the warehouse terminal (S85), and the lacquer robot's material inventory is updated (S86).

[0055] FIG. 12 is a flowchart showing a second example of the material transaction processing of the material transaction computer MAT_C. The second example is a material transaction processing when a warehouse is not used. The second example corresponds to the lacquering robot system shown in FIG. 3. As in FIG. 11, when the material transaction computer MAT_C receives a lacquering task, it performs an inventory check to determine whether the lacquering robot has enough materials (materials or tools) in stock to complete the lacquering task (S80). If the material inventory is insufficient (NO in S81), the material transaction computer MAT_C starts a transaction on the lacquer transaction blockchain (S90_1), purchases the materials (S91_1), and when the materials arrive (S92_1), updates the inventory information (S93_1). That is, the material transaction computer MAT_C executes the processes S90 to S93 by the warehouse terminal in FIG. 11.

[0056] In the second example as well, the material trading computer MAT_C may speculatively purchase materials or tools that may be in short supply in the future based on a demand forecast for the lacquering task.

[0057] [The lacquering task database generation process executed by the robot control computer RBT_C] In a self-controlled lacquering robot system, the robot workstation RBT_WS receives a lacquering task and self-controls the procurement of materials (materials or tools) necessary to complete the lacquering task, and the lacquering robot lacquers the design of the lacquering task. To accomplish this, a lacquering task database is constructed from an image or figure (hereinafter referred to as "image") of the design of the lacquering task, and the procurement of the materials and the lacquering of the design are carried out based on this database. The process of generating the lacquering task database is described in detail below.

[0058] Figures 13 and 14 show an image of a lacquering task design, a lacquering task database, and an overview of the lacquering process. In Figure 13, image IMG_1 of the lacquering task design is an image of a crescent moon. The robot control computer RBT_C decomposes image IMG_1 into multiple strokes drawn by the lacquering robot. As a result, image IMG_1 is decomposed into outer and inner curved strokes STR_1 and STR_2. Then, for each stroke, stroke trajectory data TRJ_1 and TRJ_2 and material data MAT_1 and MAT_2 containing the material or tool are generated. The material is, for example, the type and color of lacquer, and the tool is a lacquering brush. The trajectory data is, for example, a set of multiple control points on the stroke, or a function of the stroke's curve or line. The trajectory data TRJ and material data MAT are stroke feature data.

[0059] The material transaction computer MAT_C calculates the materials or tools required for the lacquering task based on the material data in the lacquering task database and carries out the procedures for purchasing the materials or tools. The robot control computer RBT_C controls the lacquering by the lacquering robot based on the trajectory data and material data in the database. In the example of Figure 13, the lacquering robot performs two strokes of lacquering (Step 1, 2) (see Paint in the figure).

[0060] In FIG. 14, image IMG_2 of the picture for the lacquering task is an image of a house. The robot control computer RBT_C breaks down image IMG_2 into a plurality of drawing strokes to be made by the lacquering robot, i.e., 20 strokes STR_1 to STR_20. Then, for each stroke, stroke trajectory data TRJ_1 to TRJ_20 and material data MAT_1 to MAT_20 including materials or tools are generated. Then, the lacquering robot performs lacquering of the 20 strokes (Step 1 to Step 20).

[0061] When applying lacquer to the surface of an image IMG, the surface of the image is divided into multiple strokes. Applying lacquer with multiple strokes makes it possible to apply lacquer to the surface of the image.

[0062] Next, we will explain the process of generating a lacquer painting task database. Figures 15 and 16 are diagrams showing the process of generating a lacquer painting task database. Figure 15 is a diagram showing the training process of the image-to-stroke model 22 and the stroke-to-feature model 24. Figure 16 is a diagram showing the inference process of both models.

[0063] 15, first, there is a preparation step Step 1(a) of preparing a data set in which image data, which is input data, is associated with stroke data, which is teacher data, as training data TRN_D_1, and a training step Step 2(a) of using the training data TRN_D to train the image-to-stroke model 22. This enables the image-to-stroke model 22 to convert an image into a set of strokes.

[0064] In the figure, the training data TRN_D_1 has a set of multiple strokes for each image data Image of multiple image data. The image-to-stroke model 22 is, for example, a deep neural network model. In the training process, one image data item of the training data TRN_D_1 is input to the image-to-stroke model 22, and the parameters of the neural network are adjusted so as to reduce the difference between the model output corresponding to the input and the stroke data (teacher data) of the training data TRN_D_1.

[0065] 15, secondly, there is a preparation step Step 1(b) of preparing a data set in which stroke data, which is input data, is associated with feature data, which is teacher data, as training data TRN_D_2, and a training step Step 2(b) of using the training data TRN_D to train the stroke-to-feature model 24. This enables the stroke-to-feature model 24 to convert the stroke data into feature data.

[0066] In the figure, the training data TRN_D_2 has a set of features for each stroke data Stroke of multiple stroke data. The stroke-to-feature model 24 is also, for example, a deep neural network model. In the training process, one stroke data of the training data TRN_D_2 is input to the stroke-to-feature model 24, and the parameters of the neural network are adjusted so as to reduce the difference between the model output corresponding to the input and the feature data (teacher data) of the training data TRN_D_2.

[0067] The feature data associated with a stroke is a comprehensive set of features including, for example, brush type, ink type and color, stroke thickness size, speed, pressure, drying time, brush cleaning time, trajectory, etc. To identify a function that most accurately represents the trajectory of a stroke, curve fitting may be used to identify a function from the control source of the stroke.

[0068] Figure 16 shows the inference process of the image-to-stroke model 22 and the stroke-to-feature model 24. The robot control computer RBT_C executes the program of the image-to-stroke model 22 to convert one image data IMG into an intermediate representation including a set of strokes STR into which the image is decomposed. Furthermore, the robot control computer RBT_C executes the program of the stroke-to-feature model 24 to convert each stroke data STR into stroke feature data (including trajectory TRJ and material MAT). The process of converting into feature data is performed for all strokes in the set of strokes.

[0069] 17 is a diagram showing an example of a stroke and example feature data associated with the stroke. On the left side of the diagram, an example of two strokes in two-dimensional coordinates is shown. This example has a stroke from point C to point A and a stroke from point A to point B.

[0070] In the figure, an example of feature data associated with a certain stroke is shown on the right side. Brush type: Round brush (specialized for drawing thin lines that highlight the details of an image) Ink color: Forest Green (RGB value: (72, 119, 75), HEX color code: #48774B) Ink type: Wiped lacquer (a technique in which raw lacquer is repeatedly applied and then wiped off.) Stroke size: 0.1 (Stroke thickness is 10% of the coordinate scale. This ensures proportional accuracy.) Speed: 0.1 per second (The stroke speed is 0.1 per second, which means it travels at a speed that covers 10% of the coordinate scale in 1 second.) Pressure: 2 Newtons (The robot hand applies 2 Newtons of pressure to the brush during painting.) Drying time: 20 seconds (time for paint to dry before starting the next stroke) Brush cleaning time: 0 seconds (The time to clean the brush before starting the next stroke. If the next stroke is the same lacquer, the cleaning time is 0 seconds.) Trajectory: Functions F1(x), F2(x) (The brush moves from point C to point A according to the function F1(x), then moves from point A to point B according to the function F2(x). As the value of x is updated, the brush moves continuously.)

[0071] The stroke trajectory is calculated in an elaborate way. The function F1(x) of the first stroke (line segment C-to-A) from point C (3.17, 5.75) to point A (2.33, 4.58) is determined by curve fitting as follows: F1 (x)=2.54×10^(-3) e^2x+4.34 When drawing a stroke, the brush moves along the value of the function F1(x) from x=3.17 to 2.33.

[0072] The function F2(x) for the next stroke from point A to point B is determined as follows: F_2 (x)=4.58 When drawing a stroke, the brush moves along the value of the function F2(x) from x=2.33 to 2.83.

[0073] The robot control computer RBT_C or the material transaction computer MAT_C calculates the number and quantity of materials or tools required for the lacquering task based on any of the parameters of the characteristic data of the lacquering task, such as the type of brush, the color and type of ink, the stroke size, the speed, the pressure, or the drying time.

[0074] The robot control computer RBT_C controls the lacquering robot to perform the lacquering task based on any of the parameters of the characteristic data of the lacquering task, namely, brush type, ink color, ink type, stroke size, speed, pressure, drying time, and trajectory.

[0075] The above embodiment is a lacquering robot system. However, the embodiment can also be applied to a painting robot system that applies glazes (or paints or coating materials) called painting on ceramics, a traditional craft other than lacquering, or a painting robot system. Therefore, in the embodiment, lacquering should be understood to include painting or painting.

[0076] As described above, the lacquering robot system of this embodiment generates a database of lacquering tasks from images of the lacquering tasks, automatically purchases lacquering materials such as ingredients or tools required for the lacquering tasks based on the database, and has the lacquering robot perform the lacquering process based on the database. This allows the lacquering robot system to perform various lacquering tasks automatically, and preferably autonomously. [Explanation of symbols]

[0077] 1: Lacquer trading network, lacquer trading blockchain 2: Material trading network, material trading blockchain 3: Lacquer trading computer 20: Robot control computer, material trading computer 22: Image vs. Stroke Model 24: Stroke-to-feature model 26: Image vs. stroke data, stroke vs. feature data 27: Lacquerware task database CST: Customer terminal RBT: Lacquer-painted robot, Lacquer-painted robot hand RBT_WS: Robot Workstation, Robot Control and Material Transaction Computer RBT_C: Robot control computer MAT_C: Material Trading Computer MAT_M: Materials Market WHS: Warehouse

Claims

1. a plurality of robot control and material transaction computers that respectively control a plurality of lacquering robots; a lacquer trading computer that, in response to a lacquering task received from a customer terminal, assigns the lacquering task to one of the lacquering robots and causes the lacquering robot to execute the lacquering task, and, after the lacquering task is completed by the lacquering robot, executes a sales process between the customer terminal and the lacquer item for which the lacquering task has been performed via a lacquer trading network; the plurality of lacquering robots are disposed at a plurality of locations; The robot control and material transaction computer a lacquering image of the lacquering task is decomposed into a plurality of strokes, and a lacquering task database is generated having stroke feature data corresponding to each of the plurality of strokes, the stroke feature data having stroke trajectory data and material data having a material including paint or a tool including a brush required to paint the stroke trajectory; A lacquering robot system that places an order for the purchase of materials or tools of the material data corresponding to the received lacquering task on the basis of the lacquering task database with a material market via a material trading network.

2. The lacquering robot system according to claim 1, wherein the robot control and material transaction computer causes the lacquering robot to which the lacquering task has been assigned to perform the lacquering process of the lacquering task based on the lacquering task database corresponding to the assigned lacquering task.

3. The robot control and material transaction computer generating the strokes corresponding to the image of the lacquering task using an image-to-stroke model trained based on first training data having the lacquered image and the strokes corresponding to the lacquered image; The lacquering robot system according to claim 1, wherein the stroke feature data corresponding to each of the plurality of strokes is generated by a stroke-to-feature model trained based on second training data having each of the plurality of strokes and the stroke feature data corresponding to each stroke.

4. The lacquering robot system of claim 1, wherein the robot control / material trading computer speculatively places purchase requests for materials or tools required for the material data of the lacquering task predicted in the future, based on the demand forecast for the lacquering task, with the material market via the material trading network.

5. 5. The lacquering robot system according to claim 1, wherein the lacquer trading computer assigns the lacquering task to a lacquering robot that is closest to the customer of the customer terminal from among the plurality of lacquering robots.

6. The lacquering robot system according to claim 1 , wherein the lacquering includes painting ceramics and painting with paint.

Images